(Publications On Ocean Development) Marc Weller, Barbara Metzger, Niall Johnson - Settling Self-Determination Disputes - Complex Power-Sharing in Theory and Practice-Hotei Publishing (2008)

Settling Self-determination Disputes
Settling Self-determination
Disputes:
Complex Power-sharing in Theory and Practice
Edited by
Marc Weller and Barbara Metzger
Assistant Editor
Niall Johnson
Published under the Auspices of the

Centre for International Constitutional Studies at the
University of Cambridge
leiden • boston
2008
Printed on acid-free paper.
A c.i.p. record for this book is available from the Library of Congress.
isbn: 978 9004 16482 6
© 2008 by Koninklijke Brill nv, Leiden, The Netherlands.

Koninklijke Brill nv incorporates the imprints Brill, Hotei Publishers, idc Publishers, Mar-
tinus Nijhoff Publishers and vsp.
http://www.brill.nl
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronic, mechanical, photocopying,
microfilming, recording or otherwise, without written permission from the Publisher.
Authorization to photocopy items for internal or personal use is granted by Brill provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood
Drive, Suite 910, Danvers ma 01923, usa.
Fees are subject to change.
Printed and bound in The Netherlands.

Table of Contents
Preface ix
Settling Self-determination Conflicts: An Introduction xi

Marc Weller
Section One Framework and Case Studies 1
Part One Historical and Theoretical Framework 3

Chapter 1 Nationalism, Self-determination, and the Doctrine of
Territorial Unity 5
James Mayall
Chapter 2 Why the Legal Rules on Self-determination Do Not Resolve

Self-determination Disputes 17
Marc Weller
Chapter 3 The Logics of Power-sharing, Consociation and Pluralist

Federations 47
Brendan O’Leary
Part Two Case Studies 59

Chapter 4 Complex Power-sharing in and over Northern Ireland:
A Self-determination Agreement, a Treaty, a Consociation,
a Federacy, Matching Confederal Institutions,
Intergovernmentalism, and a Peace Process 61
Brendan O’Leary
Chapter 5 Resolving the Bougainville Self-determination Dispute:

Autonomy or Complex Power-sharing? 125
Anthony J. Regan
vi Table of Contents
Chapter 6 Resolving Self-determination Disputes Through Complex

Power-sharing Arrangements: The Case of Mindanao,
Southern Philippines 161
Mark Turner
Chapter 7 Power-sharing and International Intervention: Overcoming

the Post-conflict Legacy in Bosnia and Herzegovina 193
Florian Bieber
Chapter 8 Interim-governance for Kosovo: The Rambouillet

Agreement and the Constitutional Framework Developed
under UN Administration 243
Marc Weller
Chapter 9 Power-sharing in Macedonia? 265

Farimah Daftary and Eben Friedman
Chapter 10 Gagauzia and Moldova: Experiences in Power-sharing 307

Priit Järve
Chapter 11 Case Study of the Conflict in South Ossetia 345

Ketevan Tsikhelashvili and Natasha Ubilava
Section Two Cross-cutting Analysis 383
Part Three Vertical Power-sharing 385

Chapter 12 Addressing the Self-determination Dispute 387
Marc Weller
Chapter 13 Power-sharing and the Vertical Layering of Authority: A

Review of Current Practices 407
Stefan Wolff
Chapter 14 Electoral Arrangements in Systems of Complex Power-

sharing 451
Andrew Reynolds
Chapter 15 Third-party Involvement in Self-determination Conflicts 467

Ulrich Schneckener
Part Four Functional Power-sharing 501

Chapter 16 Education 503
Mark Turner
Table of Contents vii
Chapter 17 Resolving Self-determination Disputes Using Complex

Power-sharing: The Role of Economic Policies 531
John Bradley
Chapter 18 Policing Territories Previously Subject to Civil War and

Ethnic Violence 559
Philip Towle
Chapter 19 The Military Dimension of Security Sector Governance in

Complex Power-sharing Arrangements 571
Paul Cornish
Chapter 20 Transforming Justice, Reclaiming the Rule of Law: Legal

Transition in Complex Power-sharing Agreements 599
Angela Hegarty
Chapter 21 The Role of Human and Minority Rights in Complex Power-

sharing 627
Jennifer Jackson-Preece
Chapter 22 ‘Bridges’ in Self-determination Disputes? External Relations

of Sub-national Entities and Minority Groups 667
Francesco Palermo
Part Five Conclusions 689

Chapter 23 Conclusion
Power-sharing Theory: Lessons from the Complex Power-
sharing Project 691
John McGarry
Biographies 721
Bibliography 729
Index 765
Preface
This study is the result of an international collaborative project supported and

funded by the Carnegie Corporation of New York. This multi-year venture has
involved a research team of some forty chapter authors and commentators. The
research has been accompanied by three major workshops on project methodol-
ogy, initial chapter reviews and final discussions. We made a point of including
both scholars and practitioners involved in power-sharing settlements in the re-
view process, hoping to learn more about the actual implementation of the settle-
ments under investigation.
The project team was united in its wish to explore whether long-standing se-
cessionist conflicts have been addressed effectively through the significant num-
ber of self-determination settlements that were generated in response to the wave
of internal conflicts of the 1990s. We were also committed to testing whether
consociationalist and integrative techniques of conflict settlement really are as
mutually exclusive as is sometimes supposed, or whether they can in fact be mu-
tually reinforcing. Finally, the project derives its impetus from the necessity to
critically rethink the doctrine of self-determination. One may question whether
its traditional, restrictive interpretation will be adequate in confronting the wide
variety of future challenges to the territorial integrity of states.
This volume represents an initial consideration of these issues in the light of an
exploration of eight cases of attempted settlement of self-determination conflicts.
By self-determination conflict, we understand to mean those cases where a unit
within an existing state is struggling, at least at some stage of its campaign, for
independence. We somewhat stretched this definition by including the Republic
of Macedonia. There, an internal movement had launched an armed campaign
in favour of autonomy. However, the regional environment was precarious and a
possible disruption of the territorial integrity of Macedonia was generally feared
as the ultimate consequence of this campaign.
A second criterion for the selection of our cases concerns the doctrine of ‘com-
plex’ power-sharing. We attempted to investigate a mix of cases, featuring ei-
ther the full range of elements of complex power-sharing, or at least a significant
number of them. We understand complex power-sharing settlements to be those
that concurrently deploy autonomy, governmental, and executive power-sharing,
veto and voting mechanisms, human and minority rights regimes, dispute settle-
Marc Weller
ment mechanisms, and the involvement of a layer of international involvement

or even international governance. Again, we stretched this definition somewhat
in considering the case of Georgia, which features a more pronounced autonomy
dimension. However, it was felt prudent to try and cover a fair spectrum of at-
tempted solutions within our overall area of interest.
Of course, the work presented here only partly reflects the outcomes of our
work. It is to be seen in connection with a study of autonomy settlements al-
ready published under the leadership of Professor Stefan Wolff (Weller and Wolff
2005) and a further enquiry into asymmetrical solutions to state construction
that is being concluded now. A further output of this work will be a review of
mechanisms aiming to enhance political participation of non-dominant groups
in situations where conflict appears to be looming or has just been provisionally
concluded. In addition, the Cambridge-Carnegie team has become active as advi-
sor in a number of ongoing settlement attempts, ranging from Kosovo, Georgia,
Moldova, and Sudan to Sri Lanka.
In addition to our contributors and commentators, we have many individuals
to thank. First and foremost, there is Stephen del Rosso of the Carnegie Corpora-
tion of New York, whose steady support has allowed us to develop this project
over time. We have also had the benefit and advice of Professor Sir Elihu Laut-
erpacht, QC, CBE and Professor James Crawford, QC in the University of Cam-
bridge. In Cambridge, the project was very ably supported by Ms Jessie Fyfe, Ms
Carmen Semenescou, and Mr Niall Johnson, who also took on a significant role
as assistant editor. Moreover, the project was carried out in close cooperation
with the European Centre for Minority Issues. At ECMI, we had the benefit of
the support of Josie Lavoie, Matthew Ward, Colleen French, Adriana Nikolova,
Janina Dill, Vladislav Michalcik, and many others.*
Finally, it is necessary to record our gratitude to Ms Lindy Melman of Brill
Publishers, who managed the publications process with her customary calm ef-
ficiency.
Marc Weller
Cambridge
* Ms Katherine Nobbs was instrumental in bringing the editorial process to its con-
clusion
Settling Self-determination Conflicts:
An Introduction
Marc Weller
The claim to self-determination encapsulates the hopes of ethnic peoples and

other groups for freedom and independence. It provides a powerful focus for
nationalist fervour, and it provides a convenient tool for ethnic entrepreneurs
seeking to mobilize populations and fighters in pursuit of a secessionist cause.
Indeed, self-determination conflicts are among the most persistent and destruc-
tive forms of warfare. Unless the aim of secession is achieved, self-determination
conflicts are likely to drag on for decades, in some instances for half a century
(Burma/Myanmar). Given the structural inequality between an armed self-deter-
mination movement and the opposing central government, the ‘national libera-
tion movement’ will often resort to irregular methods of warfare, possibly includ-
ing terrorist tactics. This may lead to profound destabilization of societies placed
at risk of disintegration, as could be seen in Sri Lanka or Sudan. Also, due to the
doctrine of non-intervention, international actors are traditionally hesitant to in-
volve themselves in attempts to bring about a settlement of the conflict.
Of course, self-determination conflicts will not go away by virtue of being ig-
nored. Unless defeated by the government, secessionist campaigns will persist
with varying degrees of violence. At present, there are about thirty ongoing armed
self-determination conflicts. Some are simmering at a lower level of terrorist vio-
lence, others amount to more regular internal armed conflicts, with secessionist
groups maintaining control over significant swaths of territory to the exclusion
of the central government. In addition to these active conflicts, it is estimated
that there are another fifty or so campaigns for self-determination at present that
might turn violent if left unaddressed. Indeed, some analysts predict that state
collapse and disintegration may be a defining feature of African politics over the
next decades, drawing an ominous arch of crisis across northern Africa, and then
extending it south to reach as far as Zimbabwe. Given the difficulties encountered
in that region in terminating non-ethnically motivated internal armed conflicts
that are lacking in the secessionist dimension, one can easily imagine the effects
for regional stability if self-determination were to be added to the equation.
The explosive nature of self-determination claims is not only explained by the
powerful force of nationalism or ethnic entrepreneurship. At the structural level,
the very doctrine of self-determination contributes to the fact that, traditionally,
few existing conflicts have been addressed, and those that have emerged anew
xii Marc Weller
appeared to be unsolvable. The doctrine of self-determination has traditionally

been seen as an all-or-nothing proposition. True, self-determination has numer-
ous layers of meaning. This includes a right to democratic participation for in-
dividuals that can be derived from the doctrine of self-determination, certain
human rights entitlements for minorities and additional benefits for indigenous
peoples. But at the sharp end, where opposed unilateral secession is concerned,
the doctrine has contributed to conflict, rather than helping to resolve it.
International legal rules are made by governments. Governments do have an
interest in perpetuating the legitimating myth of statehood based on an exercise
of the free will of the constituents of the state – their own legitimacy depends on
it. But while embracing the rhetoric of free will and self-constituting states, gov-
ernments have at the same time ensured that the legal right to self-determination
is strictly rationed and cannot ever be invoked against the state they represent.
Traditionally, self-determination in the sense of secession, accordingly, only ap-
plies to colonial entities in the classical sense, and closely analogous cases. While
Chechnya might want to argue that it was occupied forcibly by a metropolitan
power during the age of imperialism for the purpose of economic exploitation
(a lay-man’s working definition of colonialism perhaps), it nevertheless did not
qualify for self-determination. Unsurprisingly, Russia, and many other states like
it, have made sure that the doctrine of self-determination can only be invoked in
the classical and narrowly defined circumstances of colonialism, which practi-
cally no longer exist anywhere.
Even in relation to such traditional colonies, the right can only be exercised
within the boundaries established by the colonial power – in that way it does
not overcome the effects of colonialism, but the self-determination entity itself is
defined through it. Furthermore, the right is of singular application. As soon as
a colony has gained independence, it will itself start defending its own territorial
integrity with utmost vigour. There is no secession from secession. And, when
armed self-determination conflicts break out outside of the colonial context,
a legal inequality with profound practical consequences comes into existence.
Colonial self-determination movements are entitled to establish national libera-
tion movements, and the international system is twisted in their favour, to help
overcome the last vestiges of colonialism. Other rebel movements hiding in the
deserts and jungles of the world also lay claim to the label of ‘national liberation’.
However, in their case, the self-determination privilege does not apply; instead,
the international system is structured in such a way as to help the central state
ensure their defeat. However just the cause of groups fighting on behalf of sup-
pressed and tortured peoples may also be outside of the colonial context, they
are classified as secessionist rebels and terrorists. Hence, they can be engaged
with minimum international legal restraint, under the very legal order of the state
from which they seek to escape.
Overall, the all-or-nothing game of self-determination has helped to sustain
conflicts, rather than resolving them. Self-styled self-determination movements
see no alternative to a strategy of fight and win in order to achieve their aims.
Central governments see little alternative to violent repression.
Settling Self-determination Conflicts: an Introduction xiii
Of course, there is one alternative. This has been the attempt of buying off se-
cessionists by offering them autonomy within the state, instead of independence
outside of the state. Classical cases include the Aaland Islands within Finland,
South Tyrol in Italy, and the accommodation of various groups within India. But
over the period of 1945 to the end of the Cold War around 1988, only a handful
of settlements of this kind were obtained. With the termination of the Cold War,
however, the situation has changed significantly. Since 1988, we have had a prolif-
eration of self-determination settlements. This is, in part, a result of the fact that
the end of the Cold War was, itself, a trigger for the outbreak of new conflicts
requiring settlement, including those affiliated with the collapse of the former
Yugoslavia and the Soviet Union. In view of the enormously destructive force of
these conflicts, especially in Bosnia and Herzegovina, international actors have
finally become involved and at times imposed settlements.
In other instances, the actors involved have recognized that they have been
locked in an unsolvable situation of mutually hurting stalemate and, often with
international assistance, they have come to consensual agreements. These in-
clude settlements in Northern Ireland, Aceh, Bougainville, Sudan (at least the
South and East), to name but the most prominent from among a clutch of about
twenty-two cases. Suddenly, it seems, self-determination conflicts can be settled,
although it remains a difficult process and one where implementation is generally
no less complicated than the process of reaching agreement.
The new wave of settlements can be roughly divided into three principal cat-
egories. First, there are autonomy settlements that trade self-government for
a cessation of violence. Such settlements will generate a special status for the
formerly secessionist entity or entities while the overall state remains basically
unchanged. But asymmetrical autonomy settlements of this kind have certain
risks. They tend to be concluded in circumstances where the government sees no
prospect even of forcible reincorporation of the entity in question. Instead, the
unit will be granted a special status within the state, offering comprehensive self-
governance with minimum involvement by the central authorities. Settlements
concluded under such pressure will tend to neglect the need to assure good gov-
ernance, including human rights, within the newly autonomous body. Too often,
power is simply handed on to the resistance fighters, who will feel entitled to rule
forevermore and who have learnt that they gain little by playing according to the
rules of the international system. Also, some settlements fail to add a sufficiently
strong integrative element, to balance the autonomy. These are, technically, poor
and incomplete settlements, concluded because a bad peace is better than per-
petual conflict.
This study considers the designs for the Philippines and the proposed settle-
ments concerning Moldova and Georgia as examples of this category. It asks
whether these designs have advanced upon traditional autonomy solutions and
manage to avoid these pitfalls by starting to deploy complex power-sharing tech-
niques in addition to simple autonomy.
In the course of our study, we have developed the concept of complex power-
sharing. Complex power-sharing describes a solution where there is a complex
xiv Marc Weller
layering of public authority, both horizontally and vertically. This will include
autonomous structures. This is matched by the application of consociationalist
techniques, such as governmental power-sharing, guaranteed parliamentary rep-
resentation for the minority, veto rights for ethnic communities or ethno-ter-
ritorial entities, the granting of minority rights and agreements on the transfer
of economic resources. In addition, there is an element of international involve-
ment in the negotiation and implementation of the settlement, and in post-con-
flict governance.
If public power is allocated to various points, from the local level to regions to
the central, and perhaps even the international level, it is no longer easily possible
to determine where sovereignty actually lies. Contested sovereignty therefore is
no longer an all-or-nothing, zero-sum game. Power is exercised at the layer where
public services can be delivered best. Where power is retained at the centre, this
is balanced by the consociationalist techniques mentioned above, such as guar-
anteed representation of regions in the overall government and legislature, soft
blocking powers relating to the vital interests of the respective ethnic communi-
ties, and international corrective mechanisms helping to ensure compliance by all
actors. A first attempt to adopt this method was made in the Good Friday Agree-
ment on Northern Ireland, which is studied at length in the present volume.
This technique was deployed in its most pronounced form in relation to Bosnia
and Herzegovina. There, an extreme amount of public power was assigned to
the constituent units, to the exclusion of the centre. Moreover, the functioning
of the overall state was mortgaged through the application of wide-ranging veto
powers on the part of of the constituent units. Secessionist pressure remained
strong. This situation could only be contained through the long-term presence of
a sizeable military force. Moreover, the inability of the veto-bound institutions to
generate decisions was compensated for by the Office of the High Representative.
This body was intended to have advisory powers, but gradually turned itself into
an institution of co-governance.
Many observers foresaw a collapse of Bosnia and Herzegovina. However, some
ten years after the Dayton agreements, it appears as if this design, and the pas-
sage of time, have managed to keep in place the unity of the state. Gradually, the
Bosnian institutions are taking over more of the functions of government inde-
pendently and the international security presence is reducing.
Since Dayton, power-sharing designs have improved. Macedonia, for instance,
managed to address the concerns of the sizeable ethnic Albanian community
without introducing changes that would block the functioning of the overall state.
While Bosnia and Macedonia are still at some risk – a risk that may increase in
the light of further developments relating to Kosovo – complex power-sharing is
generally seen as a success.
In addition to more advanced autonomy settlements and complex power-shar-
ing, a third approach would focus on tackling the self-determination dimension
in a more direct way. As was already noted, autonomy and complex power-shar-
ing will generally trade claims to independence for self-government within the
state and a greater share in the decision-making of the overall state. The renun-
Settling Self-determination Conflicts: an Introduction xv
ciation of the self-determination claim is meant to be permanent. However, in a

number of instances, such an approach was not feasible. Instead, the possibility
of eventual independence is provided for in various forms.
First, there can be what is called constitutional self-determination. In these
cases, the central state adds a provision to its constitutional arrangements ex-
pressly providing for the possibility of secession. For instance, the constitution
of Ethiopia now provides that every nation, nationality, and people in the coun-
try “has an unconditional right to self-determination and secession”. Lichtenstein
even allows every municipality to “remove itself ” from the principality if it so
wishes. Canada has recognized in the Clarity Act that it would act in accordance
with a clear and unambiguous outcome of a referendum on secession. The UK’s
unwritten constitution, too, would accept the outcome of a genuine exercise of
popular will in Scotland or Wales.
This approach attempts to forestall self-determination conflicts by clarifying
that independence can be an option, provided it is brought about by a transpar-
ent and democratic campaign within the entity in question, and provided that a
potential divorce is the result of negotiation between the centre and the entity.
Where self-determination is, in principle, available, pressure for the exercise of
the independence option may lessen.
A somewhat more limited application of this approach is evident in deferrals
of the self-determination issue, by promising mechanisms to address it in the
future. This tack was taken in the 1999 Rambouillet agreement on Kosovo, which
provided for the calling of an international meeting to determine mechanisms for
a final settlement for Kosovo after a period of three years. Security Council Reso-
lution 1244 (1999) also provided for a process towards a final settlement without
adding any specifics. It took the renewed outbreak of ethnic violence in Kosovo
in March 2004 to remind the organized international community of this commit-
ment, and to organize the present process of negotiations on the future status of
Kosovo in Vienna.
Another approach consists of an even balancing of self-determination claims
where conflict has already happened. This is the secret of the Good Friday Agree-
ment on Northern Ireland. On the one hand, it confirms that the island of Ireland
as a whole is a self-determination unit. On the other hand, it also establishes
that no change in the present status can be obtained without the joint consent
of the populations of Northern Ireland and the South, respectively. In that sense,
the North, too, can claim to be a self-determination entity, given that no deci-
sion about status can be taken without its consent. Both sides, it is hoped, can
defend the agreement in relation to their respective constituents without having
to admit to having given up the essential self-determination claim. In the mean-
time, power-sharing mechanisms are meant to provide for stable and effective
governance and a gradual building of confidence among the populations that will
isolate those bent on violence ever further.
Where the balancing of self-determination claims is not possible, recent prac-
tice has generated the concept of external and internal conditionality of self-
determination. An example of external conditionality is furnished by the 1994
xvi Marc Weller
agreement on the status of the autonomous territory of Gagauzia within Mol-

dova. The agreement proclaims that the entity has a special status as a form of
self-determination of the Gagauzes. Should the Republic of Moldova undergo
a significant modification of its own status, for instance by way of a merger of
its ethnically Romanian-inhabited territories with Romania, then Gagauzia too
would have the right of external self-determination.
Internal conditionality was established in relation to Bougainville. The people
of that island have the right to hold a referendum on independence from Papua
New Guinea no earlier than ten years and no later than fifteen years after the
election of the first autonomous government under the terms of the settlement.
However, the decision on the holding of the referendum is also dependent on
whether or not weapons have been effectively disposed of on the island by that
date, and whether good governance prevails. It is therefore up to the authorities
in Bougainville internally to create the conditions necessary for triggering the
referendum.
The Bougainville settlement is inspired by the hope of the central government
that ten to fifteen years of genuine autonomy may lessen the pressure for inde-
pendence. The population might realize, it is hoped, that it has more to gain from
a continued association with a larger state. A similar hope is expressed in the
2005 settlement on Southern Sudan. Both North and South are supposed to cam-
paign for continued unity during an interim period of six years of self-governance
for the South. Then the population in the South can decide on independence in
a referendum. In this instance, the emphasis on seeking to maintain continued
unity is window dressing, given the general expectation that the South will ex-
press itself with one voice in favour of independence. Conditionality serves as
a face-saving device to help the central government disguise the fact that it has,
after prolonged and bloody conflict, agreed to eventual independence.
At present, the organized international community is still desperately seeking
a way of finding a justification or disguising of the inevitable independence of
Kosovo. Belgrade is hardly likely to consent to any settlement that would lead
to independence. The international actors would not wish to admit that an en-
tity outside of the colonial context, such as Kosovo, does have a right to self-
determination in the sense of secession. Accordingly, it may be determined that
independence can also occur as a result of an international settlement process
driven by the need to maintain stability in a certain region. Moreover, Kosovo’s
unique status under the former Yugoslav constitution can be invoked, given that
it granted federal powers to Kosovo much like those of the full republics that
gained independence as a result of the dissolution of Yugoslavia.
What, then, are we to make of this profusion of settlements and of differing
approaches? The good news is that there is a realization that the old, restrictive
doctrine of self-determination does not help resolve the issue in the post colo-
nial world. Neither does ignoring the problem. Accordingly, enhanced autonomy
settlements and complex power-sharing arrangements have been adopted to of-
fer alternatives short of secession if it can be avoided or as a step towards in-
dependence where it is inevitable. This study reviews in its first volume many
Settling Self-determination Conflicts: an Introduction xvii
of the most prominent examples. The second part of this work seeks to draw
cross-cutting lessons from this experience, addressing issues ranging from state
construction to security sector reform, human rights, economic issues, etc. The
project as a whole, however, also pursues a more theoretical aim. It seeks to ad-
dress the consociationalist – integrationalist divide that has long dominated the
literature in this field.
This study asks whether these approaches really are irreconcilable, or whether
the concept of complex power-sharing manages to build on that which is best in
both of them. Our hypothesis is that complex power-sharing is a useful and pow-
erful tool to address secessionist conflicts if the self-determination claim itself is
denied or deferred. Our studies show that complex power-sharing involves au-
tonomy and consociationist techniques in the first settlement stage. Once ethnic
stabilization has occurred, integrative techniques can be increasingly applied, in
order to transform an interim situation into a permanent solution. Throughout,
international involvement can offer a crucial contribution to dissolving tensions
and lessening the effect of the zero-sum game of ethnic politics.
Section One
Framework and Case Studies
Part One
Historical and Theoretical Framework
Chapter 1
Nationalism, Self-determination, and the Doctrine of
Territorial Unity
James Mayall
The grumbling dispute between the United States of America and Turkey in the
run-up to the Iraq war over the future status of the Kurdish-occupied region of
Northern Iraq provides a powerful contemporary illustration of both the need
for complex power-sharing as a solution to self-determination disputes and the
continuing political difficulties in achieving it. Many Kurds believe that they were
robbed of their right of self-determination, which was first conceded at the time
of the Versailles Peace Conference and then withdrawn following Ataturk’s revo-
lution in post-Ottoman Turkey. It would be surprising if some of them at least
would not see in the overthrow of Saddam Hussein’s regime in Iraq a chance to
redeem this pledge. It would be equally surprising if Turkey – which has only
recently abandoned its attempt to forcibly assimilate its own Kurdish popula-
tion and still regards the main Kurdish party as a terrorist organization – did not
view this prospect with grave concern. The need to reassure Turkey led those
charged with framing a new Iraqi constitution towards a de-centralized power-
sharing solution that would guarantee continued autonomy for the Kurds while
meeting the international commitment to maintain Iraq’s territorial unity. Be-
yond recognition of their cultural and language rights – which some claim have
been granted only to give Turkey’s application to join the European Union some
chance of future success – it seems unlikely that Turkey’s Kurdish minority will
be equally fortunate.
If we are to understand the problems, as well as the prospects for diverting
such disputes away from secessionist and other forms of civil conflict, we need
to ask first why self-determination claims more often than not become part of
the problem than of the solution. The answer lies in two deeply etched features
of the international landscape: the central role of sovereignty and the contested
nature of nationalism. Sovereignty is a doctrine of political supremacy over a
population living within the internationally recognized boundaries of a particular
state. Internally, the principle is upheld by the government’s claim to monopolize
both legitimate force and jurisdiction within the territory; externally, sovereigns
look to the recognition of their peers. From this point of view, international soci-
Marc Weller and Barbara Metzger (eds.), Settling Self-Determination Disputes: Complex Power-Sharing in Theory and Practice
© Koninklijke Brill NV. Printed in The Netherlands. isbn 978 9004 16482 6. pp. 5-16
James Mayall
ety is composed of solipsists, whose cooperation is confined to recognizing each

other’s narcissism. Just as turkeys don’t vote for Christmas, states don’t favour a
carve-up except when it favours them.
Traditional international society was largely composed of dynastic sovereign
states. The patrimony of the rulers – and with it the borders of their states – could
be changed as a result of the fortunes of war or the construction of dynastic alli-
ances through marriage, and consequently also by the acquisition of title through
inheritance. The members of international society were thus the sovereign states,
not their populations. This conception was challenged by the American and
French Revolutions, but the United States was only peripherally involved in in-
ternational relations during its early years and, after the defeat of Napoleon in
1815, the old real-estate system was restored in Europe. It survived, dented but
more or less intact, until the First World War.
Since 1919, international society has ostensibly been based on a principle of
popular sovereignty, namely national self-determination. The collapse of the
Hapsburg, Hohenzollern, Romanov and Ottoman empires dealt a mortal blow to
the dynastic principle. It was no longer possible to defend the state as a private
possession of particular individuals or families. But if prescription was out, con-
sent had to be in; ownership of the state, in other words, had to be transferred to
the people. The difficulty in effecting this transfer arose because, in the last analy-
sis, individuals alone can give or withhold consent. Yet individuals do not and
cannot live alone. Which, therefore, are the appropriate collective selves, whose
right to self-determination must be recognized as the basis of the new political
order?
This question would be straightforward if the identity of the nation were self-
evident. For most nationalists, this is indeed the case. They almost invariably
invoke particular historical myths and theories to justify their own claims and
to denigrate those of their opponents. Once in power, they habitually use the
school curriculum to perpetuate their favoured national myths and to construct
a national culture that will both justify and run congruently with state boundar-
ies. Whether they succeed in this endeavour or not, the reality is that while the
doctrine of nationalism is clear – i.e., that the world should be divided into nation
states – the question of national identity is contested.
Two broad theoretical answers have been advanced in response to the question
of national identity, although within each of them there are a number of varia-
tions, some of which overlap. Primordialists maintain that the national map of
the world was laid down a very long time ago, even if very few these days cling
to the belief that it accurately reflects the natural world, which can therefore be
assumed to have remained essentially unaltered since the beginning of human
history (Smith 1985). By contrast, modernists see the nation as a recent invention,
dating, except for a few somewhat anomalous, or at least unexplained cases, only
from the American and French revolutions (Gellner 1983, 1997).
For primordialist writers, assigning the right to self-determination is, in prin-
ciple, a soluble problem, however difficult it may be in practice. Perhaps, just
because it is so difficult, they do not often address the matter directly. The recipe
1 Nationalism, Self-determination, and the Doctrine of Territorial Unity
is deceptively simple. First find your ethnic group or ethnicity. This is done by
identifying a group of people who share one or more of a list of ‘objective’ char-
acteristics (the ‘one or more’ is normally added to accommodate Switzerland)
– a name, a common language, a homeland in which they generally although not
invariably reside, common symbols, a common myth of origin or ancestry and a
sense of themselves as a people with a shared history of triumphs and disasters
and, on the basis of these, shared hopes and aspirations.
Modernists have paid almost as little attention to the international implica-
tions of their theories as the primordialists have. For the most part, their interest
is in the historical and sociological conditions that ushered in the nationalist era
rather than in the justification offered for a state’s entry into international soci-
ety. Thus, for example, Gellner (1983, 1997) argued that a national culture was a
necessary accompaniment to the transition from agricultural to modern society,
largely because the division of labour on which industrialism depends, and the
competition to which it leads, requires occupational mobility and therefore a lit-
erate and trainable labour force. Peasants, who mostly stay in one place, do not
need to read and write in order to function and tend to regard whoever governs
them with deep suspicion. On the other hand, modern states require educated
citizens whose loyalty they can command. In attempting to explain why citizens
themselves identify with the state, the modernist author Benedict Anderson has
traced the rise of the nation to the development of print capitalism; the profit mo-
tive requiring a market of readers, which in turn put a premium on the produc-
tion of literary works in the vernacular and allow for an imagined community of
people who do not know one another directly (Anderson 1983).
Anderson and Gellner – and indeed most modernist writers on nationalism
– adopt a broadly realist approach to international relations, to the extent that
they consider them, which is not often. They seldom address legal or normative
questions at all. At the same time, implicit in their arguments is the recognition
that political identity – like political boundaries – is a contingent matter. This is
the crucial point. What is contingent cannot be settled by rational argument or
a democratic vote. For political argument to take place, boundaries must be in
place, but they lie behind or beyond such argument all the same.
This reality was not immediately grasped in 1919 and many democrats are re-
luctant to accept it even now. The original Wilsonian solution to the problem
was the plebiscite. It failed, not merely because of the irreconcilable territorial
claims in Central and Eastern Europe after the First World War, nor because the
great powers had no intention of testing their legitimate title in their own pos-
sessions by this method, but also because it too regarded the identity question as
self-evident. As Ivor Jennings famously put it in 1956, “on the surface it seemed
reasonable: let the people decide. It was in practice ridiculous because the people
cannot decide until someone decides who are the people.” (Jennings 1956: 56).
The attempt to democratize international society deepened rather than miti-
gated the problem. Popular sovereignty, if it meant anything, implied consent,
but Jennings was right, it could not tell you whose consent was to count. When
the drafters of the United Nations Charter and the Universal Declaration of Hu-
James Mayall
man Rights turned their attention to the right of self-determination, they referred
to peoples rather than nations, presumably in an attempt to avoid the destructive
confusion that had accompanied the reconstruction of Europe after 1918. Eventu-
ally, the territorial claims and counter-claims of the European successor states
had been settled politically, broadly along national lines. But while this was better
than allowing free rein to forced population movements and or communal mas-
sacres, it nonetheless left plenty of room for conflict. Nor did the minority rights
treaties that were imposed, and consequently much resented, as the price of ad-
mission to the League of Nations provide the necessary reassurance.
The wording of the UN Charter was not a huge improvement. Finding ob-
jective criteria to define ‘a people’ is no easier than – indeed, no different from
– defining a nation, unless that is, the right is assigned to pre-existing states or
territorial units and the people, whoever they may be, are simply assumed to be
identified with and represented by state governments. Whether or not those who
drew up these documents were already clear about what they were doing, the
principle was in practice interpreted as applying – ex-post facto – to all existing
states and to the overseas colonies of the European imperial powers. It was also
generally accepted, in a move that made greater psychological than logical or le-
gal sense, that self-determination required the end of apartheid in South Africa.
Despite a rearguard action by defenders of the imperial idea (for example, Burns
1957; Seton Watson 1960) at the time it was not seriously advanced in relation to
the Soviet Union’s imperial legacy, on the eastern fringes of Europe, around the
Baltic, in Central Asia and beyond.
Nor were most governments willing to insist on a democratic test of opinion
before extending international recognition to states that underwent a revolu-
tion. It is true that the United States attempted – for more than twenty years
successfully – to blackball the People’s Republic of China from the United Na-
tions, but even Washington’s closest allies were unimpressed by this attempt to
enforce ideological conformity across international borders and in obvious defi-
ance of Article 2(7) of the Charter. Irredentism got equally short shrift. Around
the edges of their inheritance, some colonial successor states consolidated their
territory without suffering serious international consequences: thus, India swal-
lowed Goa; Indonesia, first West Irian and then, in 1974, East Timor; and China,
Tibet – a decidedly pre-modern form of conquest, which the outside world was
nonetheless able to digest because the country had never enjoyed formal sover-
eignty or international recognition. But, in general, during the Cold War there
was widespread antipathy to opening up the domestic political arrangements of
sovereign states to outside scrutiny, and no indication that the forceful pursuit of
irredentist claims – outside the immediate context of European imperial with-
drawal – would be tolerated.
This conventional interpretation was accepted by the Soviet Union when it put
itself into ‘voluntary’ liquidation after 1989. The different Soviet Socialist Repub-
lics were treated as colonies with a right to self-determination, a right that was
not extended to autonomous regions such as Chechnya within the Russian Fed-
eration itself. Nothing in the conventional interpretation of self-determination
1 Nationalism, Self-determination, and the Doctrine of Territorial Unity
prevented secession where it could be agreed. There were precedents – Norway’s

secession from Sweden, Ireland’s from the UK, Singapore’s from Malaysia, the
separation of the Czech and Slovak republics – before and after decolonization
and before and after the Cold War. But if it did happen, it did not happen very
often and it did not dispose state authorities to recognize any general right of
secessionist self-determination.
After 1989, some observers concluded that the apparent stability of the con-
ventional interpretation – only Bangladesh fought its way to independence and
international recognition and then only with Indian help – meant that it was a
function of the Cold War. There are two obvious weaknesses with this view. First,
the conventional interpretation was constantly challenged – in Burma, Canada,
Nigeria, Sri Lanka, Sudan and elsewhere – it merely did not engage the full at-
tention of the strongest powers, which were preoccupied with the Cold War con-
frontation. Secondly, the evidence suggests that its adoption had more to do with
the territorial imperative than with the Cold War. It is a depressing fact that na-
tional sovereignty, unlike its dynastic predecessor, seems to transform land into
a positional good.
Irredentist claims are seldom abandoned altogether, but the ambitions of
governments that harbour them – such as Spain to Gibraltar, the Philippines to
Sabah, Morocco to Mauritania, the Republic of Ireland to Ulster, Argentina to
the Falklands, and Taiwan to the Chinese mainland, face formidable practical
constraints. Despite their disagreement on most other issues, the superpowers
would not openly support such territorial change. On this issue at least, they also
supported the international legal prohibition on the use of force as an instrument
of foreign policy. With the partial exception of the Spanish claim to Gibraltar,
which was sympathetically viewed by African states on anti-colonial grounds,
most other states were similarly hostile to proposals to redraw the political map.
The extent to which the Wilsonian vision of a world made safe for democ-
racy and national self-determination would challenge rather than support the
international order became evident immediately after the First World War. The
gruesome consequences of the demands for organic democracy in much of Eu-
rope were submerged by the Second World War and the territorial stabilization
imposed on Europe by the Cold War division and the re-introduction, in the east,
of authoritarian rather than democratic rule. The end of the Cold War put the
clock back, revealing that little had been done in Europe to resolve the underlying
problem in the meantime.
The post-colonial attempt to avoid a similar outcome looked back to the nine-
teenth century. The irredentist wars that followed the withdrawal of Spain and
Portugal from South America – most famously the war between Paraguay and
the Triple Alliance, which killed off a major part of Paraguay’s population – pro-
vided a warning of the dreadful consequences of treating territory as a positional
good and therefore as non-negotiable. The principle of uti possidetis juris was
intended as a practical solution to an otherwise unresolvable ethical problem.
Since appealing to the principle of self-determination could not settle the issue of
rival territorial claims, what criteria were to be used to decide who had title? The
10 James Mayall
answer was that, in the absence of a negotiated boundary adjustment, successor

states would accept the borders that they had inherited at independence.
The principle was revived after 1960 when African successor states, whose
leaders had often previously called for the redrawing of African boundaries in
line with African social and cultural realities, became fearful of opening a Pan-
dora’s box of ethnic claims and counter-claims. As already noted, the principle
did not originally enjoy the same status in post-colonial Asia, where several coer-
cive consolidations occurred in the immediate aftermath of decolonization, and
in the case of Indonesia’s annexation of East Timor, as late as 1974. However,
this annexation was increasingly referred to as an aberration, and in 1999 the
conventional interpretation of self-determination as decolonization was used to
justify the territory’s claim to independence and to distinguish it from separatist
demands elsewhere in the Indonesian archipelago (Mayall 2000). Uti possidetis
was also imported into the former Soviet Union and Eastern Europe where it
was used to transform the internal administrative boundaries of the Soviet Union
and, after much bloodshed, of the former Yugoslav Republic, into internationally
recognized borders.
Secession seldom solves the human problem to which it is addressed, namely
the need to buttress the identity and enhance the security of the breakaway pop-
ulation. The reason is that it creates disenfranchised minority groups trapped
within the new states. To that extent, the application of uti possidetis is a useful
exercise in damage limitation. In a recent legal study, Steven Ratner concluded,
somewhat optimistically, that the principle need not trump self-determination
because the circumstances that had made it good law in the colonial context no
longer applied. He is no doubt right to argue that “decolonization did not have
to entail adoption of uti possidetis” but, strictly speaking he is wrong to maintain
that it was adopted “because it kept decolonization – a development regarded
almost universally as imperative – orderly” (Ratner 1996: 610). It was a post-colo-
nial not a colonial law. It was revived not by the departing colonial powers but by
successor African governments, although it certainly appealed to state authori-
ties everywhere as a principle of order.
It is also true that sometimes there are possibilities for satisfying demands for
self-determination – and hence strengthening security – without creating a new
independent state. Referenda were used successfully, although not without of-
fering future hostages to fortune, in settling the national identity of the divided
Trust territories of Cameroon and Togo at the time of independence from Britain
and France. South Tyrol is a more recent example. Such cases, however, are more
the exceptions that prove the rule that under conditions of democratic govern-
ment, there is nothing so difficult to change as an international border.
What is to happen, moreover, when such rational solutions are not accepted
because a previously internal border that was regarded as tolerable within the
framework of a Federation, as in Tito’s Yugoslavia, is not accepted as legitimate
within the context of its dissolution and independent statehood? Or, as in Sudan,
when those who win control of a successor state seek to replace the indirect rule
of the former imperial power by forced assimilation and/or centralized nation-
1 Nationalism, Self-determination, and the Doctrine of Territorial Unity 11
building? These symptoms generally presage a slide towards a form of patho-

logical ultra-nationalism. When they are present, the flip side of the uti possidetis
doctrine becomes apparent. It ceases to operate as an incentive to rational prob-
lem-solving, and serves instead to legitimize savagery and the militarization of
society to a point where the cure is often worse than the disease.
It is not hard to think of examples from most parts of the world. In Africa,
for example, it was unfortunate that the model that was inherited at indepen-
dence was that of the national security state, although, in the majority of cases,
there was no external threat. This inheritance paved the way for the widespread
military hijacking of the state apparatus, which was regarded by the new politico-
military class as an exploitable resource for their own enrichment. In many coun-
tries, the government itself was the major source of insecurity. Nor is it clear that
the return to civilian, and ostensibly democratic, rule in many African countries
during the 1990s has done much to improve either the self-determination or se-
curity of the mass of the population.
Uti possidetis has three main weaknesses as a mechanism for conflict reso-
lution, as opposed to an ordering principle for an established society of states.
First, popular passions are more likely to be involved in the dispute. This may
be, as in the Serbian attachment to Kosovo, because a territory, or even an entire
landscape, is imbued with symbolic significance within a particular nationalist
mythology. Alternatively, it may be that changes in the regional and wider inter-
national climate provide an opportunity for a majority community to overturn
the traditional dominance of a minority, as with the Kosovo Albanians or the
Rwandan Hutu. Where political life is organized along lines of communal con-
frontation, the fact that there is no local or international provision for territorial
adjustment can lead society to implode. The case of Somalia, where the state dis-
integrated once the safety valve of irredentism in the Ogaden was closed in 1978,
provides a dramatic illustration of this possibility.
The second reason why the uti possidetis principle may backfire is simply that
there is no appeal beyond it. Admittedly, it was the central aim of the Westphalia
system to remove such an appeal in inter-state relations. But, in the period be-
tween 1648 and the French Revolution, the fact that territory could change hands
without it having a cathartic effect on the lives of ordinary people allowed the
system to operate more or less as intended. Once politics were nationalized, mili-
tary defeat in contested territory such as Alsace-Lorraine began to be followed
by mass population transfers. The twentieth century attempts to outlaw war as
an instrument of foreign policy amounted to a recognition that, in a world of
popular sovereignties, territory could no longer be treated as so much real estate,
to be traded on the battlefield, brought under new ownership by inclusion in the
dowry of royal brides, or literally bought and sold as in the cases of the Louisiana
and Oregon purchases and Alaska.
The final unfortunate side-effect of the doctrine is that it raises the symbolic
value of holding territory at a time when changes in the nature of the world econ-
omy are undermining both its economic value and the ability of many govern-
ments to resolve internal political problems by economic means. Many of the
12 James Mayall
mercantilist moorings of the international economy have been sheared, with the
result that it can no longer be accurately described as primarily a set of inter-state
relationships and transactions.
The main challenge to the conventional interpretation of self-determination as
de-colonialization came from secessionists: these were, after all, precisely those
who took the principle seriously, and who understandably drew the conclusion
that, if self-determination was a fundamental human right, then it should ap-
ply to them. Of the three secessionist crises that spilled onto the world stage
during the Cold War – Katanga, Biafra and Bangladesh – only the Biafran case
was debated seriously in terms of the substantive meaning of self-determination.
The reintegration of Katanga into the Congo was the price the United States was
prepared to pay to marginalize Soviet influence within the United Nations peace-
keeping operation. The rebellion in East Bengal was often explained by academic
analysts in terms of a theory of internal colonialism, but it was the Indian army
that expelled Pakistan, not the Bangladeshis themselves.
Biafra’s bid for independence collapsed because, unlike Bangladesh, the Bi-
afrans failed to secure a powerful external patron who was prepared to defy the
international consensus in favour of the territorial status quo. France came close,
but, in the end, de Gaulle indicated that he would be guided by African opinion.
By 1969, four African states – Ivory Coast, Gabon, Tanzania and Zambia – had
broken ranks and recognized the Biafran government. A number of others were
rumoured to be sympathetic to its cause. At the annual OAU Summit in 1969,
President Nyerere of Tanzania circulated a memorandum to his fellow African
Heads of State. Nyerere had himself proposed the 1964 OAU resolution, which
committed African countries to accept the boundaries inherited at independence,
but he now urged that, in this case, they should abandon the commitment. His ar-
gument was straightforward, and in its own terms, compelling. Colonial borders,
Nyerere suggested, had been accepted for practical reasons – to facilitate inter-
state cooperation, minimize opportunities for conflict and to release energies that
could be better devoted to development and improving the lot of Africa’s peoples.
Nonetheless, the right of governments to rule rested on their ability to serve the
population as a whole. When a government could no longer protect the lives of
all of its citizens, and when a particular group believed itself to be threatened by
genocide, it forfeited its legitimacy. In these circumstances, the same political
considerations that had earlier led him to accept existing territorial arrangements
could now (and in his view, should) be advanced in support of partition.
Nyerere’s attempt to establish internal standards of accountability and good
government as relevant criteria for international recognition failed. His argu-
ment was revived, however, after the Cold War by the then UN Secretary-Gen-
eral Boutros Boutros-Ghali, at least implicitly. In his Agenda For Peace, the docu-
ment that was commissioned following the first ever Security Council Summit
in January 1992, he made three statements that seemed to hint at a process of
managed constitutional reform for international society.
Tanzania’s memorandum on Biafra’s case. For text, see Kirk-Greene (1971: 429–439).
First, he insisted that the state must remain as the foundation stone, but that its
authority was not absolute. “Respect for its fundamental sovereignty and integ-
rity are crucial to any common international progress. The time of absolute and
exclusive sovereignty, however, has passed; its theory was never matched by real-
ity.” Secondly, he argued that while the United Nations had not closed its doors to
new members, “if every ethnic, religious or linguistic group claimed statehood,
there would be no limit to fragmentation, and peace, security and economic well-
being for all would become ever more difficult to achieve.” Finally, he suggested
that the way to resolve the rival claims of sovereignty and self-determination was
through respect for human rights, particularly the rights of minorities, on the one
hand, and democratization, on the other. “Respect for democratic principles at all
levels of social existence is crucial: in communities, within states and within the
community of states.”
The admission that the United Nations had not closed its doors to new mem-
bers offered a hostage to fortune, since the Secretary General did not specify
the circumstances under which the principle of uti possidetis might legitimately
be breached in the future. At the same time, because the democratic side had
won the Cold War, if only by default, its victory ensured that claims for self-
determination would henceforth have to be cast in democratic form. The two
international experts consulted by the Canadian government for an opinion on
whether Quebec might have a unilateral right of secession provided some indica-
tion of the way legal opinion on the subject was evolving. They were clear that
Quebec did not have such a right but nonetheless concluded that “there may be
developments in the principle of self-determination according to which not only
colonialism but also flagrant violations of human rights or undemocratic regimes
could lead to a right of unilateral secession” (Department of Justice 1997).
Since the underlying problem of political identity remains unresolved, it is prob-
ably inevitable that, despite such rulings, secessionists will continue to press their
democratic right to self-determination. The odds against success are still long,
however. The two successful cases of enforced secession since the end of the Cold
War – Eritrea’s separation from Ethiopia in 1993 and East Timor’s from Indonesia
in 1999 – are better explained by reference to the conventional interpretation of
self-determination as decolonization in accordance with uti possidetis, than in
terms of the evolution of new criteria. If at least some potential secessionists are
to be persuaded that complex power-sharing represents a more rational way of
resolving self-determination disputes, it will be necessary to convince them first
that democracy and secession are likely to prove incompatible in practice.
There are three main arguments. The first is that secession must be ruled out
both to avoid anarchy and in the interests of public welfare. This was the posi-
Boutros Boutros-Ghali, Agenda for Peace, paragraphs 17 and 18. For text, see Roberts
and Kingsbury (1993: Appendix A, 468–498).
The two experts were Professor James Crawford, Whewell Professor of International
Law, University of Cambridge; and Professor Luzius Wildhaber, University of Basel.
Professor Wildhaber is also a judge on the European Court of Human Rights.
14 James Mayall
tion adopted by Abraham Lincoln during the American Civil War and, in a much
diluted form, it survives in Boutros Boutros-Ghali’s Agenda for Peace. All citizens
have the same fundamental rights, grounded in the constitution and protected by
law. The only way the minority can become the majority, on this view, is by per-
suading the majority to change their allegiance at the next election. Guaranteed
rights of free speech and free association allow them, in principle, to compete on
level terms with the government for the affections of the people.
The theory is attractive: there is no ground for holding that opposition will be
regarded as treachery, and the government itself will be regularly held to account
and will be changed following electoral defeat. The trouble is that, at the end of
the twentieth century, as in the middle of the nineteenth, this is not an accurate
description of social reality in many parts of the world. In particular, as Harry
Beran has pointed out, it cannot deal with societies that are structurally deeply
divided (Beran 1987: 39–42). The argument that Jinnah advanced on behalf of the
Muslim League fifty years ago was that Indian Muslims were a separate people,
in a sense that would inevitably define and limit their political destiny. They could
not see themselves as other than Muslims, and they could not hope to become
the majority under any foreseeable circumstances. On the basis of this argument,
the British were eventually persuaded to partition the country.
Whether this was a wise, or indeed a necessary decision, will no doubt continue
to be debated amongst historians and nationalists in both India and Pakistan. The
relevant point is that until power was transferred at midnight on August 15 1947,
the British remained in control and were, therefore, in a position to make the
decision. International society has no such authority to settle conflicts in societ-
ies where people insist on identifying themselves in ethnic or communal terms
rather than as citizens. After the Croatian Declaration of Independence in 1991,
the government wooed international recognition by passing legislation to protect
Serbian rights. The trouble was that Serbs were not reassured and continued to
resist rule from Zagreb. A blind eye had to be turned toward the ethnic cleans-
ing of the Krajina before Croatia would accept the American-brokered settle-
ment in Bosnia. Similarly, when in 1994 the RPF began to close in on Kigali, the
Hutu-dominated government of Rwanda launched a genocide rather than sur-
render power. In such cases, Lincoln’s splendidly civic argument fails to address
not merely the aspirations but, much more importantly, the fears of the people
on both sides of the conflict.
The second argument is that democracy will be subverted and political free-
dom destroyed where two or more powerful national groups compete through the
ballot box to capture the state. This is the instrumental, as distinct from rights-
based, defence of partition advanced by Mill in Representative Government (Mill
1865: Chapter XVI). Written in the 1860s, it can be read as an almost uncanny
prediction of what happened in Yugoslavia in the early 1990s. Mill’s argument is
implicitly historicist: a non-national or civic national democracy will only emerge
if peoples of different nationality have grown accustomed to living together – and
being governed under a single authority – prior to the era of both nationalism
and democracy. However, if national self-consciousness and a desire for ‘free in-
stitutions’ emerge together, the absence of an overarching political culture will

ensure that democracy will act as a source of conflict rather than of legitimacy.
The merit of Mill’s argument is its concern with practical outcomes rather than
collective rights in any abstract sense. His primary concern is how to protect hu-
man freedom; his advocacy of partition under certain circumstances is directed
to the same end, not the conservation of particular national cultures or forms of
life as ends in themselves. For those committed to the latter objectives, it is Mill’s
indifference to minorities that are not sufficiently numerous to strike at the dem-
ocratic foundations of the state that most weakens his argument’s appeal. Except
in this limiting case, Mill was as opposed to secession as Lincoln, and his assimi-
lationist assumption that minority discontent could be overcome by education,
while benign in intention and possibly true in some cases and over the very long
term, has been falsified by events in many parts of the world.
The third argument is that, while the existence of a nation is a prerequisite for
democracy, nations will come into being, in the future as in the past, only through
a process of struggle and self-assertion against other forms of imposed imperial
sovereignty. This bleak argument rests on a compelling but not necessarily ethnic
logic. The essence of representative democracy is majority rule. No matter how
consociational the constitutional arrangements, there will always be some party
or interest that is left out of the ruling coalition. So what is it that persuades the
minority to put up with rule by the majority? Lincoln had suggested that it was
the opportunity to change the government by swinging public opinion behind
the opposition. But why should they be prepared to wait, particularly if the odds
on success at the polls are long and there seems a better chance of influencing
events by taking the law into their own hands? A possible answer is that the mi-
nority will only put up with majority rule if there is some overarching community
sentiment, that is stronger than the conflicts of ideological, economic or political
interest that customarily divide people on a daily basis.
It is not clear that a community of this kind can be engineered, at least from
the outside. But we know that most, if not all, well-established nation-states de-
velop democratic institutions only after a prolonged period of power struggles
and feuding at home and abroad. Even the United States had to fight a civil war
to prove to themselves that they were one people and not two. Collective am-
nesia, as Renan famously remarked, as much as shared memories or invented
traditions, is a defining characteristic of most nations (Renan 1882). Whether it is
true that behind every great fortune there is a great crime, it is difficult to think
of a great nation (or even a small one for that matter) that does not rest on great
cruelties and/or injustices in its past. The difficulty posed by this argument is
only too apparent: on the one hand, as an account of how the democratic world
came into being and maintains itself it is plausible; on the other hand, insofar as
it condemns us to live – indeed to define ourselves – by the sword, it is morally
repellent. Moreover, governments of all ideological persuasions have accepted
the legal prohibition on the use of force except in self-defence, a view that is chal-
lenged but not fundamentally repudiated by the new American doctrine of pre-
emption. According to the argument in question, the twentieth century world
16 James Mayall
wars were fought, as least partly, to replace the rule of the sword with that of the
constitution, internationally as well as nationally. This process was interrupted
by the Cold War, but there is no evidence to suggest that the popular hunger for
self-government that erupted after 1989 was an aberration.
None of these arguments for and against secession in the name of secession is
satisfactory. The first fails on empirical and the second on moral grounds, while
the third endorses violence as the midwife of change. If, as I have argued, it is im-
possible to devise rational rules for redrawing the political map to accommodate
the aspirations of all self-conscious groups, it remains to be asked whether there
is an alternative way of interpreting self-determination within international soci-
ety. The evidence reviewed in the case studies in this book suggests that complex
power-sharing, sometimes of a federal or confederal kind, sometimes of a con-
sociational kind, may indeed provide such a way forward. It seems unlikely that
many groups will be initially willing to forgo the temptation of framing their po-
litical programmes in romantic and/or essentialist and uncompromising terms.
The very complexity that is necessary to keep those who have been long involved
in deadly conflict from returning to the battlefield is also itself something of a
deterrent – the culture of nationalist insurgency will be familiar to those who
have done the fighting and whose education seldom fits them to benefit from the
peace. Nonetheless, when the battle cannot be won, a window of opportunity will
eventually be opened for rational argument. The cases reviewed here may not add
up to a trend, but they do suggest that strategies exist that are capable of being
adapted to different circumstances and that may prove acceptable to all parties
– in the first case to cover the transition from war to peace – and possibly for the
longer term as well.
Chapter 2
Why the Legal Rules on Self-determination Do Not
Resolve Self-determination Disputes*
Marc Weller
I Introduction
“It is for the people to determine the destiny of the territory and not the territory
the destiny of the people.” This renowned definition of the right to self-determi-
nation, expressed by Judge Dillard in his Individual Opinion in the 1975 Western
Sahara case, proves two things. First, it confirms that lawyers, too, can manage a
pretty turn of phrase. Second, this statement, perhaps like no other famous dic-
tum in international law, demonstrates the dangers of well-intentioned judicial
activism, for there has rarely been a pronouncement more dangerously mistaken
than this one.
Judge Dillard proposes, most sensibly it seems at first sight, that people must
triumph over the accidents of geography and of historically established territo-
rial divisions. People act according to their free will and must therefore be able to
shape their destiny through collective decisions. Since 1945, this view underpins
the international system as a legitimizing myth. The legitimacy of its most basic
building block, the ‘sovereign’ state, is derived from the assumption that the state
is nothing other than a machine to form and implement an aggregated common
will of its people. Accordingly, the state itself is supposed to have been formed by
an act of will of its citizens. If the creation of the state is the product of an act of
will, then a further collective decision should also suffice to undo it. Moreover,
human beings do not surrender their free will by deciding to join into, or form, a
collectivity. Hence, one would presume that groups within an existing state must
also be able to assert their will by deciding to leave an existing state and form a
new sovereign unit. Such a decision would be made manifest by an expression of
will of the population concerned, for instance through a referendum.
While this sounds logical, reality is, of course, very different. The international
system has ‘balanced’ the competing aims of accommodating the ideology of free
* An earlier version of this chapter appeared in 4(2) Ethnopolitics (2005).

Western Sahara, 1975 ICJ 12, 122.
18 Marc Weller
will against the purported aim of maintaining stability, order and peace. Accord-
ingly, the international system has developed in a way that can, on the one hand,
draw legitimacy from a doctrine of popular will. The political principle, and legal
rule, of self-determination is the most potent expression of this concept. How-
ever, on the other hand, those who operate the system have ensured that the very
doctrine of self-determination that purports to enfranchise people actually serves
to disenfranchise them, in the interest of maintaining peace and stability. Rather
than offering citizens a choice, the doctrine of self-determination has been con-
structed in a way to limit or deny choice. In fact, generally self-determination is a
rule that empowers those who oppose choice, even by violent means, where the
territorial definition of the state is concerned.
Rather than preventing conflict, however, the rule of self-determination has
generated a dynamic that sustains conflict. For, those who seek to assert their
identity are forced into an absolute position. As the international system only
privileges the interest of the state over the position of groups challenging its ter-
ritorial unity, most central governments feel under no pressure to accommodate
demands for change. Constitutional settlements that might maintain the unity of
the state while permitting a greater expression of diverse identities through the
legal and political system have therefore often been denied. Instead, the state will
tend to label groups that agitate for a more pronounced identity as secessionists
and rebels. Under cover of international rules, it will attempt to suppress these.
However, in most instances, this has not led to a disappearance of the problem.
Instead of giving up, in many instances resistance groups have responded by radi-
calizing their demands, claiming that only outright statehood can preserve the in-
terests of their constituents – as is demonstrated by the very repression that was
launched by the central state in response to their original campaign. A vicious
circle ensues. Given the absence of any sort of international remedy – states hav-
ing protected their freedom to engage ‘rebellions’ of this kind in an undisturbed
manner through the doctrine of non-intervention – the opposition movements
will often see an armed struggle as the only way of furthering their aims.
These types of conflict have been among the most damaging and protracted that
have bedevilled states and the international system since 1945. Marshall and Gurr
list seventy-two self-determination conflicts that have been conducted over that
period, with only twenty-four of these having been settled or concluded through
victory by one side (Marshall and Gurr 2003: 30). They have been sustained,
rather than resolved, through the doctrine of self-determination. The doctrine
has energized secessionist movements, on the one hand, in their mistaken belief
that they are engaged in a just struggle that must ultimately be recognized by the
international system. On the other hand, the governments have taken comfort in
the fact that they have constructed the self-determination rule in a way that does
not in any way affect their ability to quash separatist groups.
Accordingly, virtually all of the instances of opposed unilateral secession (a)
resulted in violent conflict and were (b) either brought to a close through a deci-
sive victory of the government, or they festered for decades. A classical example
is furnished by the extremely destructive conflict over Biafra that resulted in a
2 Why the Legal Rules on Self-determination Do Not Resolve: Self-determination Disputes 19
humanitarian catastrophe. The United Nations refused to intervene during the

conflict and only mounted a humanitarian operation after Nigeria had defeated
the secession. In another instance, Katanga seceded from the Congo at the point
of decolonization. A UN peacekeeping force actually defeated the secessionists
under a slightly ambiguous UN mandate. Only East Pakistan managed to obtain
statehood through unilateral opposed secession, although under rather unique
circumstances, and in consequence of an armed action from neighbouring In-
dia.
Where there was no decisive military result, conflicts have continued at times
over decades. For instance, some of the ethnic self-determination insurgencies in
relation to Burma/Myanmar have persisted literally since independence over half
a century ago, others have been concluded only very recently.
Of course, the rigidity of the classical doctrine of self-determination has been
subjected to numerous challenges since it consolidated during the 1960s. In par-
ticular, the unfreezing of the Cold War certainties since 1989 has brought with it
significant challenges to the doctrine of territorial unity. However, as this chapter
will argue, even these challenges were addressed in a way that has left the restric-
tive doctrine of self-determination in place. The result is the fragile insistence
on the continued existence or territorial unity of threatened states in Eastern
and Central Europe. In Bosnia and Herzegovina, a massive international military
presence has been deployed for a full ten years now with this end in mind. In
relation to Moldova and Georgia, the OSCE has been attempting for a similar
period to negotiate settlements to the Transdniestria, Abkhaz and South Ossetia
conflicts that somehow maintain the unity of both states. Similarly, initial pledges
were made to maintain the territorial integrity of the former Yugoslav Federation
(Serbia and Montenegro) in relation to Kosovo, however impossible that might
ultimately appear to be.
Nevertheless, after the disasters of the destructive ethnopolitical and ethnoter-
ritorial conflicts of the Balkans, the Caucasus and other areas of the first half of
the 1990s, a ray of hope has emerged. We can now see the emergence of some
self-determination settlements that appear to indicate a willingness of central
governments, the self-styled self-determination movements and international
actors to escape from the trap imposed by the self-determination rule. Hence,
new settlements have been emerging that do not always necessarily preclude self-
determination in the sense of secession, while at the same time offering a new
relationship between the central state and the secessionist unit that could make
continued territorial unity possible.
This chapter dissects the main strands of classical discourse about self-deter-
mination. It then turns to consider the development of the doctrine of constitu-
tional self-determination – an innovation of the 1990s – and asks whether it of-
fers a solution to self-determination conflicts. It then briefly introduces the most
recent practice of complex power-sharing settlements and asks whether these
will indeed point a way out of the self-determination trap – an issue pursued
further in Chapter 12.
20 Marc Weller
II The Classical Right to Self-determination

Self-determination disenfranchises populations. This process of disenfranchise-
ment has traditionally proceeded in five steps. First, self-determination is intrin-
sically linked with, and deployed to justify, the disenfranchizing doctrine of terri-
torial unity. Second, there is the issue of the definition of the object of protection
of the right to self-determination – that is to say, the definition of the types of
‘people’ entitled to exercise this right. Third, there is the scope of application of
the right to self-determination. That is to say, even if a ‘people’ is designated as
a right holder, does this right trump previously existing territorial definitions, or
is it exercised within these confines? Then, there is the issue of the singularity of
implementation of the right – is it a continuous process, or is it a one-time-only
event? Finally, there is the problem of the modalities of achieving the point of
self-determination.
Before turning to each of these features of classical self-determination dis-
course in turn, it might be useful first to distinguish the concept from other con-
texts in which it is used.
A A Concept with Multiple Meanings

This book addresses self-determination as the right of all peoples freely to deter-
mine their political, economic and social status – the formulation used in virtu-
ally all relevant UN documents addressing the issue. However, this definition is
broad and can be taken to encompass both external and internal self-determina-
tion. External self-determination will normally be taken to include the right to
secession. Internal self-determination concerns the choice of a system of gover-
nance and the administration of the functions of governance according to the will
of the governed. The following are examples of the different layers of meaning of
self-determination in a legal sense:
• Self-determination as an individual right: self-determination is not only a
right exercised by peoples or groups. It is also a human right of individuals.
Hence, individuals are entitled to participation in the political, economic or
cultural system of their state. In that sense, the individual right to self-de-
termination might be seen to be co-extensive with the right to some form
of democratic governance. However, for a long time, this ‘right’ has been
reduced to an underlying political doctrine that was not actionable. It is only
now, albeit somewhat hesitantly, surfacing as a firm legal entitlement.
• Self-determination as a right appertaining to members of groups and perhaps
groups themselves: self-determination is also a right that can be invoked by
members of certain groups, such as national, religious, ethnic or linguis-
tic minorities. In this sense, self-determination is congruent with minority
rights. Minority rights protect the existence of national, religious, linguistic
or ethnic groups, facilitate the development of their identity and ensure that
they can fully and effectively participate in all aspects of public life within
the state. While it was previously argued that minority rights are only held
by members of minorities individually, it is clear that they can exercise these

rights in community with others. There may also be emerging a recognition
of a group identity as an object of legal protection, although this remains
controversial. This includes entitlements to cultural autonomy. Some would
argue that there may also be an entitlement to territorial autonomy where
national minorities constitute a local majority, but this is not accepted as yet
in general practice.
• Self-determination and indigenous peoples: in addition to their cultural iden-
tity, indigenous populations tend to claim a historic and particularly strong
bond with certain territories they have occupied since time immemorial.
Indigenous rights, therefore, not only seek to enhance the maintenance of
the cultural identities of indigenous peoples, but they may also extend to
land rights and political/territorial autonomies. While the technical term
‘people’ is applied to indigenous populations in ILO Convention 169, the
Convention immediately clarifies that this is not meant to imply a people’s
right to external self-determination in the sense of international law.
• Self-determination in case of a limited territorial change: where a significant
tranche of territory is moved from one sovereign to another, the popula-
tion of that territory may be entitled to express and subsequently exercise
its preferences through a plebiscite. As opposed to the self-determination
of peoples, this entitlement does not extend to a free determination of the
international legal status of the territory – for instance to opt for independ-
ence or association with a third state. Instead, it is limited to an endorsement
or rejection of the change that is proposed by the governments concerned.
This doctrine is, however, displaced in certain circumstances, for instance
in cases of territorial change that is anticipated in historical arrangements,
such as the handover of Hong Kong. At times, it may be contested whether
the inhabitants of the territory in question are a ‘people’ entitled to self-de-
termination of peoples, or merely a population attached to a stretch of terri-
tory and hence only entitled to a plebiscite. For instance, the population of
Gibraltar might argue that it is a people entitled to full self-determination,
while Spain and the United Kingdom take somewhat differing views in rela-
tion to a more limited form of self-determination that may apply subject to
the provisions of the Treaty of Utrecht of 1713. That treaty provides that the
territory must revert to Spain should the UK ever withdraw from it.
• External self-determination of peoples: self-determination of peoples implies
a right unilaterally to initiate a change in the status of a territory through
an act of will of the population of that entire territory. In this way, self-de-
termination of peoples differs from the right of a population to co-deter-
mine the future of a portion of territory through a plebiscite that was just
noted above. This latter kind of ‘self-determination’ is ancillary to a decision
of states to effect a transfer of territory. A population rejects or ratifies the
decision of the states involved. Self-determination of peoples, on the other
hand, is an original right that is vested in ‘a people’, merely by virtue of the
fact that the technical label ‘people’ attaches to a specific population and
22 Marc Weller
territory. Whether the state involved favours any sort of territorial change is
inconsequential; the exercise of the will of the ‘people’ so nominated is alone
decisive.
Manifestly, the doctrine of self-determination has different legal consequences

in these different contexts. Within the confines of this discussion, the principal
focus must lie on self-determination as an entitlement of ‘peoples’ freely to deter-
mine the international legal status of a territory.
B The Issue of State Consent

Virtually all inhabitable portions of the globe are subject to territorial jurisdiction
by one state or another. Virtually all human beings also find attached to them-
selves the claim to jurisdiction of at least one state. Hence, if people wish to form
a new state, this can only occur at the expense of an existing one, both in terms
of human and territorial resources. This can occur either with the consent of
the central government concerned or, more likely, against the opposition of the
government. In the former case, it is of course not necessary to rely on a right to
self-determination.
A divorce by agreement has occurred in a few instances (e.g., Malaysia/Singa-
pore). Where this consent from the central government is lacking, the interna-
tional system will tend to deny legal personality to those seeking separation. This
may appear illogical, as the relevance of the ‘sovereign’ acts, such as the granting
of consent, of the central government in relation to the entity seeking secession
constitutes the essence of any self-determination dispute. However, the legal sys-
tem protects the claims of governments and will normally only offer status if the
government concerned is content.
Changes of status by consent occur in a number of instances. These include:
• Instances where one state joins another. For instance, when the new German
Federal states of the former German Democratic Republic joined the Fed-
eral Republic of Germany, the legal personality of the latter persisted, with
the former being extinguished. There can also be state unions, where a new
composite state is formed, with both constituent entities relinquishing their
international legal personality.
• Instances of dissolution of composite states. The division of Czechoslovakia
into the Czech and Slovak Republics serves as an example. Czechoslovakia
disappeared as a sovereign entity. In contrast, the constitution of the Union
of Serbia and Montenegro provided for the continuation of legal personal-
ity of the overall state by Serbia, once Montenegro opted to leave. Similarly,
when the Soviet Union dissolved, all its successor states agreed that the Rus-
sian Federation would continue the legal personality of the former Union.
• Instances of secession. In such cases, it is clear that only one element of a
composite state splits off, without putting into question the legal personality
of the state. An example is furnished by the secession by agreement of Erit-
rea from Ethiopia.
The manifestation of an act of will of the population is necessary even where a

government agrees to the separation of certain territories. Hence, the agreement
on the possible secession of Eritrea required the holding of a referendum after
an interim period, to confirm that this change in status is indeed in accordance
with popular will. Again, however, there remains a crucial difference to cases of
opposed secession. The exercise of the will of the population followed on from a
previous agreement by the central government that a referendum could be held
and that its results would be respected. An international legal entitlement to self-
determination was not necessarily the trigger for this process at the outset. In-
stead, the exercise of self-determination flowed from a previous, voluntary deci-
sion of the newly constituted central government that consisted of the victorious
former rebel movements. Subsequently, Ethiopia entered into its constitution a
provision permitting in advance the secession of its remaining constituent units
– a case of constitutional self-determination that will be considered later.
In contrast, the essence of the traditional right of self-determination of peoples
is that it in itself constitutes a valid basis for a claim to secede, irrespective of the
wishes of the central government. Therefore, one is really talking about a right
to unilateral and mostly opposed secession. Naturally, such a right is perceived
to be very dangerous by governments, as it can be exercised autonomously from
their consent and control. Hence, it is not surprising that the right to self-deter-
mination in the sense of unilateral and opposed secession has been defined very
restrictively. After all, it is the governments that make the law in the international
sphere, and they can be expected to do so according to their shared perception of
central state interests.
C Self-determination as an Exceptional Right

The right to opposed unilateral secession stands in obvious tension with the claim
to territorial integrity and unity of existing states. Governments have enshrined
the doctrine of territorial unity in countless international declarations and other
instruments, often tied to, or twinned with declarations concerning self-determi-
nation. The first element of disenfranchisement that obtains lies in the very ex-
istence of a right to self-determination. While this right purports to enfranchise
populations wishing to exercise their will, it does the opposite. In generating what
is an exceptional entitlement to secession, self-determination appears to confirm
that secession is not otherwise available in circumstances where the central gov-
ernment refuses to consent to a separation. This strengthens the view that a se-
cession that is not covered by the exceptional right to (colonial) self-determina-
tion amounts to an internationally unlawful act. This, for example, was the view
(wrongly) taken by the rump-Yugoslavia in relation to Croatia, Slovenia, Bosnia
and Herzegovina, and Montenegro. The consequence of this – mistaken – view
would be that an entity that succeeds in secession would be an unlawful entity.
Unlawful entities are well known in international law. These are entities that
have come into being in violation of essential rules of the international commu-
24 Marc Weller
nity as a whole, such as the prohibition of the use of force by states (Northern
Cyprus), the right to self-determination (Southern Rhodesia after its Unilateral
Declaration of Independence by the white minority government), the prohibi-
tion of apartheid (the so-called Banthustans), or the prohibition of genocide and
ethnic cleansing (the Republika Srpska and Herzeg Bosna). Such entities may dis-
play the objective criteria of statehood of territory, population and government.
However, this will not trigger the consequence of statehood that would ordinar-
ily obtain. Instead, these entities are non-states and all states may be under an
obligation not to recognize them as states or to assist them in maintaining their
illegal status.
Practice shows that opposed unilateral secession that does not involve the
unlawful use of external force, genocide, apartheid, etc., is not in itself interna-
tionally unlawful. Hence, an entity that manages to secede and to maintain itself
effectively can over time obtain statehood, and have this fact confirmed through
international recognition, even if the central government objects. However, in
its attempts to obtain statehood, the entity is not legally privileged – it enjoys no
right as a legal subject in and of itself. Therefore, the central government will con-
tinue to claim an entitlement to incorporate the seceding entity through fire and
sword if necessary. This entitlement would persist until the point in time when
the entity has demonstrated its effectiveness to the extent necessary for state-
hood. However, in the absence of external recognition, it is difficult to identify
this point in time. After all, the central government (or former central govern-
ment) can argue that the entity is not effective, and will never be effective, as it
only exists so long as it is not forcibly re-incorporated. And such an act can occur
at any moment chosen by the central government. One might say that Somalil-
and is at present in such a state of legal uncertainty. As was already noted above,
Biafra and Katanga are examples of entities that were forcibly re-incorporated,
without much international opposition. Chechnya, too, was subjected to armed
re-incorporation, despite assurances to the contrary that had been given by Mos-
cow in a series of peace settlements.
It is by way of a lack of international legal protection of its status that an ‘ef-
fective entity’ differs greatly from a ‘self-determination entity’. The latter is inter-
nationally privileged long before it obtains effective independence. Indeed, it is
the essence of the right to self-determination to ensure that a self-determination
entity can freely exercise the option of independence if it so wishes. The former
has to fight the threat or attempt of forcible reincorporation and will only mature
into a state if it wins decisively and with a prospect of permanency of its new
status emerging. An armed contest is therefore a structural element of discourse
about self-determination outside of the context of self-determination entities as
they have been classically defined.
D The Definition of the Entity

The classical right of colonial self-determination is now a core part of interna-
tional law and enjoys a status that is legally superior to other international norms
that do not enjoy this elevated position (ius cogens). However, it is applied only
to colonial and non-self-governing territories, of which there are practically none
left. This is the second disenfranchizing aspect of the doctrine of self-determina-
tion: it is established as an exception to the doctrine of territorial unity (above),
but the exception is framed so narrowly that it does not apply to many or any
situations of struggle for independence outside of the colonial context.
There is no formal definition of what constitutes a colonial territory. However,
as a rule of thumb, it only includes those territories that one would intuitively
recognize as such. These are territories that were forcibly acquired by a racially
distinct metropolitan power divided by an ocean during the time of imperialism
and subjected to a colonial regime for the purposes of economic exploitation. The
long list of qualifications contained in this sentence indicates the lengths govern-
ments have gone to in order to ensure that self-determination cannot ever be
invoked against themselves. Colonial self-determination only consolidated into a
firm legal rule in the early 1960s, when the only remaining colonial powers resist-
ing decolonization were international pariahs. These were, principally, Portugal
and Spain, both held in the grip of dictatorships. Analogous situations to which
the rule of colonial self-determination was also applied (Palestine and South Af-
rica) were similarly unique. Hence, it was safe for the rest of the governments of
the world to consecrate the doctrine of self-determination as a firm legal rule,
provided it could only be applied to these ‘others’. In relation to them, self-deter-
mination was framed as a very aggressive doctrine, in order to help address the
historic injustice that was, by then, clearly recognized in relation to these special
cases.
Of course, many populations in other circumstances claim to be disenfran-
chised or suppressed. They will argue that they, too, have been subjected to co-
lonialism. However, they are excluded from the application of the concept. For
example, Chechnya argued that it was forcibly incorporated into Russia during
the period of imperialism and colonially exploited. Nevertheless, its claims to
colonial self-determination were simply brushed aside on the international stage.
Some politicians in Kosovo were tempted to make a similar argument in rela-
tion to Serbia. Again, this argument would not have offered a chance of success.
Kosovo therefore instead opted for making an argument based on constitutional
self-determination.
As was already noted, in addition to genuine colonies, it is accepted that peo-
ples living under alien occupation (Palestine) and under racist regimes (formerly
South Africa) are entitled to the right of self-determination. The same applies
to ‘secondary’ colonies. These are entities that were entitled to colonial self-de-
termination in the first place. However, when they were at the very point of ad-
ministering the act of self-determination, they were forcibly incorporated into
another state. East Timor and Western Sahara are the two principal examples of
this phenomenon. The recent holding of a referendum in East Timor and its inde-
In the UN framework, there exists a technical identification of ‘non-self-governing

territories’. However, this definition is not that useful in new contexts.
26 Marc Weller
pendence are therefore an example of colonial self-determination in the classical

sense, although held with some delay, rather than constitutional self-determina-
tion.
E Scope of Application
While self-determination is an activist right that is intended to overcome the
evils of colonialism, it is in fact administered in a way that is consistent with the
territorial designs and administrative practices imposed by the colonizers. This is
the third level of disenfranchisement administered through the doctrine of self-
determination. For the definition of the entity that is entitled to exercise the right
of self-determination is in itself a product of colonial administration. Hence, self-
determination does not aim to restore ethnic or tribal links amongst populations
that were artificially divided by the colonizers. Instead, the ‘people’ entitled to
self-determination are those who happen to live within the colonial boundaries
drawn by the colonial powers.
Accordingly, the International Court of Justice confirmed in relation to the
Western Sahara that links may have existed between that territory and Morocco
before colonialism. However, these would not be restored through self-determi-
nation. The pre-existing links were not of a kind that could displace the right to
separate identity that was actually manufactured through the process of colo-
nial administration – the very evil the doctrine of self-determination purports to
overcome. Morocco had argued that the people of the Western Sahara had pre-
viously owed allegiance to its leadership. This relationship had been artificially
disrupted by the imposition of Spanish colonial rule in the Western Sahara. Now
that colonialism was in the process of being removed, the previous status should
be restored and the Western Sahara should fall to Morocco once more.
This view was rejected by the Court. While the Court confirmed that there
may have been some pre-existing legal links between the two territories, the very
act of colonialism is constitutive of a new legal status for the colonial entity. Co-
lonialism generates the self-determination entity and therefore defines the state
that may ensue. That entity holds original rights that displace legal ties that might
have existed before. Most strikingly this was also affirmed in relation to the is-
land of Timor. Once Portugal withdrew from East Timor, Indonesia claimed that
the island should be unified again and the Eastern part would naturally merge
with the Western section. However, the separate colonial administration of the
East by Portugal, as opposed to Dutch administration in the West, had rendered
it a separate self-determination entity. Indonesia’s occupation of the island was
therefore internationally opposed as an act undertaken in violation of the right
to self-determination. As was noted above, eventually Indonesia consented to the
holding of a referendum on genuine self-determination. That referendum was
strongly in favour of independence, which was subsequently implemented with
the assistance of an international peacekeeping/enforcement mission.
The aim of decolonization is therefore not the restoration of the situation that
may have existed before colonialism. Instead, action is taken in a way that does
not fully overcome, but merely reshapes, facts on the basis of the reality of co-
lonial administration. And it is the territorial shape of that administration that
defines the self-determination entity, not the will of the people. Herein lies the
third element of disenfranchisement. Contrary to the dictum of Judge Dillard
quoted at the outset, it is not the act of free will of populations that can fully
assert itself. Instead, it can only apply itself within boundaries that have been
colonially defined. For instance, different ethnic groups within a colonial terri-
tory would not be entitled to form separate states, or perhaps to associate in part
with neighboring ethnic kin states. Instead, the entire territory, as defined by the
colonial masters, must exercise the right to self-determination as one whole and
undivided entity.
While some might regard this practice of retaining ‘artificial’ colonial bound-
aries as reprehensible, it has been accepted by the African states upon indepen-
dence. In fact, it has been fiercely defended by them. This principle of uti posside-
tis has been described by the International Court of Justice as follows:
21. … The fact that the new African States have respected the administrative bounda-
ries and frontiers established by the colonial powers must be seen not as a mere prac-
tice contributing to the gradual emergence of a principle of customary international
law, limited in its impact to the African continent as it had previously been to Spanish
America, but as the application in Africa of a rule of general scope. …
23. … The essence of the principle lies in its primary aim of securing respect for the ter-
ritorial boundaries that moment when independence is achieved. …
24. … There is no doubt that the obligation to respect pre-existing international fron-
tiers in the event of a State succession derives from a general rule of international law,
whether or not the rule is expressed in the formula uti possidetis. Hence, the numerous
solemn affirmations of the intangibility of the frontiers existing at the time of the inde-
pendence of African States, whether made by senior African statesmen or by organs of
the Organization of African Unity itself, are evidently declaratory rather than constitu-
tive: they recognize and confirm an existing principle, and do not seek to consecrate
a new principle or the extension to Africa of a rule previously applied only in another
continent.
25. However, it may be wondered how the time-hallowed principle has been able to
withstand the new approaches to international law as expressed in Africa, where the
successive attainment of independence and the emergence of new States have been ac-
companied by a certain questioning of traditional international law. At first sight this
principle conflicts outright with another one, the right of peoples to self-determination.
In fact, however, the maintenance of the territorial status quo in Africa is often seen as
the wisest course, to preserve what has been achieved by peoples who have struggled
for their independence, and to avoid a disruption which would deprive the continent of
the gains achieved by much sacrifice. The essential requirement of stability in order to
survive, to develop and gradually to consolidate their independence in all fields, have
induced African States judiciously to consent to the respecting of colonial frontiers,
Burkina Faso/Mali Frontier Dispute, Merits, 1986 ICJ 564.

28 Marc Weller
and to take account of it in the interpretation of the principle of self-determination of

peoples.
Subsequently, this doctrine was endorsed by the Badinter Commission appointed

by the European Community to advise on international legal issues in the context
of the dissolution of Yugoslavia. According to this view, uti possidetis applies as a
universal principle in all regions of the world.
F Self-determination as a Singularity
The fourth layer of disenfranchisement relates to the fact that colonial self-deter-
mination is an ‘act’ that occurs only once, and not an ongoing process. Of course,
self-determination continues to occur in its internal sense, according to the doc-
trine of democratic governance. However, in its external sense, it is a once-in-a-
lifetime opportunity. Hence, once a colonial territory had exercized the option of
independence or integration (the exception being, theoretically, association), the
right to external self-determination would have expired.
Self-determination is not available to distinct ethnic entities within the self-
determination unit that may feel that they too should have had the option of
secession from secession. The doctrine of territorial unity protects the territorial
identity of the self-determination entity before, during and after the act of self-
determination. The present dispute involving Comores and Mayotte may serve
as an example, as does the following extract from a submission to the United
Nations by Sri Lanka:
2. It is the position of the Government of Sri Lanka that the words “the right to self-
determination” … apply only to people under alien and foreign domination and these
words do not apply to sovereign independent states or to a section of a people or nation.
It is well recognized in international law that the principle of self-determination cannot
be construed as authorizing any action which would dismember or impair totally or in
part the territorial integrity or political unity of sovereign and independent States. This
article of the Covenant cannot therefore be interpreted to connote the recognition of
the dismemberment and fragmentation on ethnic and religious grounds. Such an in-
terpretation would clearly be contrary, inter alia, to General Assembly Resolution 2526
(XXV) on the Declaration of Principles of International Law and incompatible with the
purposes and principles of the Charter.
An exception to this rule would relate to a self-determination entity that decides to

associate, but not integrate, with another state, instead of opting for independence.
Through association, the self-determination status of the entity is retained, or one
might say, it is transformed into a case of constitutional self-determination. However,
there is very little practice of this kind.
International Covenant on Civil and Political Rights, Third Periodic Reports of States
Parties due in 1991, Addendum, Report Submitted by Sri Lanka, 18 July 1994, CCPR/
C/70/Add.6, 27 September 1994.
In 1991, the doctrine of uti possidetis has been found to apply not only in the co-
lonial context. The Badinter Commission ruled that the constituent republics of
the Socialist Federal Republic of Yugoslavia might turn into states. However, enti-
ties within the republics – for instance, the mainly ethnic Serb-inhabited Krajina
region – could not make similar claims. According to the doctrine of uti posside-
tis, they might only claim territorial autonomy within the new state boundaries.
It will be convenient to return to this issue when considering the new practice
of constitutional self-determination. Before doing so, however, it is necessary to
consider the fifth step in the chain of disenfranchisement that lurks within the
doctrine of self-determination.
G Modalities of Reaching the Point of Self-determination

We have already noted that self-determination exceptionally enfranchises colo-
nial peoples struggling for the right to opt for a new status within colonial bound-
aries at one unique point of decision. All other cases are excluded and therefore
subjected to the negative, disenfranchising element of the doctrine of self-deter-
mination. These are:
• Cases that arise outside of a colonial context. As the colonial context is de-
fined very restrictively, this includes many instances of perceived colonialism
on the part of historically disenfranchised populations (Corsicans, Basques,
Chechens, Kosovars, etc).
• Cases that arise within the colonial context, but where populations want to
escape from the colonial definition of the self-determination entity and the
subsequent state (India-Pakistan (Kashmir), Komores and Mayotte, Bou-
gainville, Burma, Sri Lanka, Sudan, etc).
• Cases where the entity in question opposes the purported initial act of self-
determination in the form of association or integration and seeks to replay
its decision (formerly Eritrea and Somaliland, which violently questioned
their integration at the point of decolonization with Ethiopia and Somalia
respectively).
The treatment by the international system of classical colonial cases of self-deter-

mination is very different from the vast majority of self-determination conflicts
that remain. Those entities that qualify as classical self-determination entities are
legally entitled to mount a struggle. If the colonial state resists and represses the
struggle, there is a right to wage an armed struggle. While this struggle has the
form of a civil war as a matter of fact, on the legal plane the national liberation
movement representing the self-determination entity is entitled to turn it into an
international conflict in terms of international humanitarian law. That is to say,
national liberation fighters are to be treated according to the same privileges en-
joyed by combatants in international armed conflicts, including prisoner of war
status. Moreover, and controversially, the self-determination entity is entitled to
receive military support from abroad. This would probably not include the direct
support of foreign armed forces, but would include the basing of national libera-
30 Marc Weller
tion fighters in neighbouring territories, and training, equipping and supplying

them there. The government, on the other hand, is not entitled to receive inter-
national support. In short, the system has been arranged in a way to ensure that
the national liberation struggle will ultimately be a success. Given the ‘corrective’
nature of the doctrine of self-determination in relation to the evil of colonialism,
this should not be controversial. However, it is important not to confuse national
liberation warfare with an open licence to engage in acts of terrorism or other
grave violations. Such atrocities are never permissible, even in the cause of an
internationally lawful struggle of anti-colonial liberation.
In cases outside of the colonial context, the system is rigged in order to ensure
that the state prevails. However unjustly treated by history a self-styled ‘national
liberation movement’ may feel, their struggle is legally classified as a purely inter-
nal domestic rebellion. The central state can use its military or police power to re-
press and defeat such a movement. No external assistance may be given to those
struggling against the central government. Instead, traditionally, the government
was taken to be entitled to receive as much military support and assistance, in-
cluding probably the involvement of foreign forces it may invite into the country,
as it may deem necessary to crush the rebellion. The rebels themselves are not
elevated by humanitarian law to the status of combatants that enjoy the full pro-
tection of the law of international armed conflict. The government would claim
the right to treat them as traitors and bandits under the domestic law of the state.
Instead of being treated as prisoners of war, they can be criminally convicted and
shot. Their only hope lies in the minimal protection of the law of internal armed
conflict if the rebellion has taken on a significant territorial scope, as well as that
of general human rights.
The fifth element of disenfranchisement therefore relates to the imbalance in
the status of those struggling for independence outside of the colonial context
and the state. Unless another state is willing to break the rules and intervene (as
occurred when India invaded East Pakistan in an operation that led to the estab-
lishment of the state of Bangladesh), self-determination struggles could classi-
cally only result in a crushing defeat for the rebellion (Biafra) or an eternal stale-
mate, sustained over decades through low intensity fighting and perhaps terrorist
campaigns that cannot be decisively defeated by the central government.
The imbalance in status not only relates to the fact that the government can
treat those struggling for purported liberation as criminals in its domestic law.
Those engaged in the struggle are disenfranchised twice in this instance – domes-
tically, where their status is determined by their opponents, and internationally.
As was already noted, at the international level the doctrine of non-intervention
ensures that even those groups that control large slices of the territory and popu-
lation of a state cannot attract international support or recognition, or significant
international entitlements that would flow from some sort of international legal
personality. An outside government that offers support to a secessionist move-
ment is guilty of an act of unlawful intervention. External agencies have, in the
past, even been cautious about political initiatives aimed at settling a self-deter-
mination conflict, lest this be considered illegitimate interference. Accordingly,
the government seeking to oppose secession has classically also been largely im-
mune from diplomatic pressure or even external sanctions in relation to its at-
titude.
The doctrine of non-intervention has provided a cover for quite brutal uses
of force against secessionist entities, often at the cost of very significant civilian
suffering. While other governments may have, on occasion, feebly requested that
at least systematic and grave human rights abuses should cease (most recently in
relation to Chechnya), they have traditionally not felt able to insist that violence
cannot be a means of settling self-determination conflicts. Force being an ac-
ceptable option, or even the expected, routine response, threatened governments
have therefore generally done their utmost to achieve a military defeat of seces-
sion. Negotiations on a settlement were not foreseen in the international script
and international pressure for a negotiated settlement would have been deemed
intervention. Hence, it was victory for the state and crushing defeat for those that
claim an entitlement to self-determination, or, where a decisive result cannot be
achieved, a prolonged, mutually hurting stalemate.
This has only recently changed in two types of situations. In the first instance,
stalemate proved no longer acceptable domestically (Northern Ireland, Sudan,
perhaps Sri Lanka). In another kind of case, the humanitarian suffering result-
ing from the fighting, or the instability brought to neighbouring regions, have
been invoked to justify actual external armed intervention. As a result of such
intervention that may have been initially focused on humanitarian concerns, the
intervenors have found themselves constrained also to address the underlying
self-determination conflict. In Bosnia and Herzegovina, they found themselves
committed to the continued territorial integrity of that state, while initially ac-
cepting the reality of its internal division. In relation to Kosovo, an international-
ized status settlement is now being prepared. In relation to Northern Iraq, the
US government and others have attempted to generate a solution to the Kurdish
issue.
The presumption that force is the appropriate remedy to secessionist aspira-
tions has recently been put into question in the case of the dissolution of the
Socialist Federal Republic of Yugoslavia. However, as we shall see, this incident
has not established a general inhibition on the use of force as the principal form
of discourse in such instances. Instead, it has drawn attention to a new aspect of
the right to self-determination that had not been acted upon previously. This is
the doctrine of constitutional self-determination.
II Constitutional Self-determination
The crucial difference between colonial and constitutional self-determination
lies in the fact that, in the former case, the right to secession is based directly in
international law. In the latter, the claim to self-determination is derived from
a constitutional arrangement that establishes a separate legal personality for
component parts of the overall state. The constitution of a state is taken to be a
manifestation of the sovereign will of the state population. International law now
32 Marc Weller
appears to takes note of these features of domestic constitutional law and gives
effect to them. However, it is not constitutive of the claim to self-determination.
It is possible to distinguish three different types of constitutional self-determi-
nation:
• Express self-determination status.
• Effective dissolution of a federal-type state.
• Implied self-determination status.
It will be convenient to consider each of these in turn.
A Express Self-determination Status

A few constitutions will determine that certain nominated constituent entities
enjoy a right to external self-determination. One such clear case has been fur-
nished by the new Ethiopian constitution that was adopted after the final victory
of the internal opposition forces that displaced the central government. Article
39(5) of the new constitution of 8 December 1994 declares with the greatest clar-
ity that: “Every Nation, Nationality and People in Ethiopia has an unconditional
right to self-determination, including the right to secession.” Paragraph 5 adds an
unusual definition:
A ‘Nation, Nationality or People’ for the purpose of this Constitution, is a group of peo-
ple who have or share a large measure of a common culture or similar customs, mutual
intelligibility of language, belief in a common or related identities, a common psycho-
logical make-up, and who inhabit an identifiable, predominantly contiguous territory.
Another example is furnished by the Constitution of the Principality of Liech-

tenstein. Article 4(2) permits each municipality to “remove itself from the state-
community.” In both instances, the constitutions provide for a certain process
that must be gone through before secession.
An interesting sub-species of express constitutional self-determination is con-
ditional self-determination. For instance, the Law on the Special Legal Status of
Gagauzia describes that autonomous territorial unit within Moldova as an “inte-
gral part of the Republic” (Article 1(1)). Nevertheless:
In case of a change of the status of the Republic of Moldova as an independent state, the
people of Gagauzia shall have the right to external self-determination (Article 1(4)).
The change that is being contemplated is a possible division of Moldova, with its
larger segment possibly joining Romania.
The identity of this conditional constitutional self-determination unit is de-
fined in an unusual way too. Localities in which (ethnic) Gagauzes constitute less
than fifty per cent of the population may be included in the autonomous territo-
Author’s translation.
rial unit on “on the basis of the freely expressed will of a majority of the electorate
revealed during a local referendum” (Article 5(2)). Accordingly, this would be one
of the more recent examples, where the will of the people does, after all, triumph
over previous administrative/territorial arrangements. This is an interesting de-
parture from the classical colonial self-determination practice.
Ordinarily, constitutional self-determination will assign a right to secession
only to federal-type territorial units, such as constituent republics, that are clear-
ly defined in terms of territory. The classical example is furnished in the constitu-
tion of the former Union of Soviet Socialist Republics, the USSR. In accordance
with Leninist doctrine, Article 70 of the Constitution of 7 October 1977 provided
that the Union is an integral, federal, multinational state formed on the principle
of socialist federalism as a result of the free self-determination of nations and
the voluntary association of equal Soviet Socialist Republics. Article 72 simply
added that: “Each Union Republic shall retain the right freely to secede from the
USSR”.
Of course, it was probably not anticipated that any Union Republic would ever
dare to assert this constitutional right of self-determination. When, in 1989/1990,
the Baltic republics declared their intention to revive their full sovereignty and
move towards full independence, this was strongly resisted by Moscow. Given the
clear and unambiguous nature of Article 72 of the Constitution, it was not easy
to justify such a stance. However, in rather a strained argument, attention was
drawn to Article 78, which required ratification by the USSR of changes to the
boundaries of the Union Republics agreed among themselves, and to provisions
assigning competence in relation to the external boundaries of the federation
to the centre. This interpretation would render the unilateral right of secession
established in Article 72 meaningless, and a legal race developed on this issue be-
tween the Baltic republics and Moscow. The central Congress of People’s deputies
worked at high speed to prepare a “Law on Procedures for Resolving Questions
Related to the Secession of Union Republics from the USSR”. That instrument,
finally brought into effect on 3 April 1990, provided for a lengthy interim period
of at least five years, and left to the central Congress of USSR People’s Deputies
a final decision on giving effect to the will of the population of the republic con-
cerned. However, the previous month, on 11 March, Lithuania had declared the
renewal of its independence.
This event triggered a somewhat ambiguous response, especially on the part
of Western states. Many of these had never recognized de jure the incorporation
of the Baltic republics into the Soviet Union, arguing that this had occurred as
In the Philippines autonomy settlement (Final Peace Agreement between the Gov-
ernment of the Republic of the Philippines and the Moro National Liberation Front
of June 1996, at www.ecmi.de), the autonomous unit is also constituted through a
plebiscite, although no external self-determination is provided for.
Article 73(2) assigned jurisdiction to the USSR in relation to the determination of
the state boundaries of the USSR and also approval of changes in the boundaries
between Union Republics.
34 Marc Weller
the result of an unlawful use of force. Accordingly, it was difficult for them to
insist on the doctrine of territorial unity in this instance. On the other hand, very
few – other than heroic Iceland which did accord its recognition – were willing
to act on principle when confronting this fact. The issue was resolved when the
USSR dissolved entirely in the wake of an unsuccessful coup against President
Gorbachev. At that point, Western governments fell over themselves in seeking
to outdo one another in extending rapid recognition.
The case of the USSR therefore became one of outright dissolution of a federal
state, and the argument of express constitutional self-determination was not fully
tested in this instance. While, ultimately, the Socialist Federal Republic of Yugo-
slavia also dissolved (see below), there was nevertheless a strand of argument in
relation to the secession of Croatia and Slovenia that can be seen as the point of
discovery of this claim to self-determination in international relations.
The 1974 SFRY constitution provided that:
The nations of Yugoslavia, proceeding from the right of every nation to self-determina-
tion, including the right to secession, on the basis of their will freely expressed in the
common struggle of all nations and nationalities in the National Liberation War and So-
cialist Revolution, and in conformity with their historic aspirations, aware that further
consolidation of their brotherhood and unity is in the common interest, have, together
with the nationalities with which they live, united in a federal republic of free and equal
nations and nationalities and founded a socialist federal community of working people
– the Socialist Federal Republic of Yugoslavia ...10
This provision quite clearly assigned to all ‘nations’ contained in the federation
the right to self-determination, including expressly the right of secession. Each
of the federal republics was seen as the political expression of the constituent
nations. Hence, the republic had had assigned to them an express right to self-
determination and secession. This proposition was put to the test in 1991.
Under the Milosevic regime, Serbia had gained the ascendancy within the Yu-
goslav federation during the second half of the 1980s. With the support of some
other compliant republics, it was undoing the careful balancing act between the
different republics that was reflected in the design of Tito’s 1974 constitution.
Kosovo, in particular, suffered the virtual unilateral abolition of its autonomy.
Faced with this change in the balance of powers within the Federation, Croa-
tia and Slovenia attempted to protect their position by proposing a new federal
constitution that would enhance their status. Negotiations toward this end con-
ducted during 1990 and early 1991 were frustrated by Serbia. Croatia and Slo-
venia then unilaterally declared independence on 25/26 June 1991. The central
government in Belgrade had been unwilling to settle, as compromise would mean
a surrender of some of the very powers it had just captured. The republics – and
Latvia and Estonia had adopted declarations concerning full independence in May
and August 1990, respectively.
10 Constitution of the Socialist Federal Republic of Yugoslavia, Basic Principles, Sec I.
Kosovo – would have no option but to comply. After all, the international system
uniquely privileged the central government, permitting it to deploy the armed
forces of the state if necessary, in order to defend the central value of territorial
unity. Based on state practice over the past decades, it was clear that indepen-
dence was no option.
While international actors tried very hard to dissuade Croatia and Slovenia
from declaring independence, they rapidly acknowledged this fact once it oc-
curred. When Belgrade proceeded to answer the declarations of independence
of Slovenia and Croatia of 25 and 26 June 1991 with the use of force, the inter-
national community, led by the EU, took a somewhat ambivalent step. While it
failed to recognize the two entities until January 1992, it nevertheless adopted
the following unprecedented view only a few weeks after the declarations of in-
dependence:
The European Community and its member States are dismayed at the increasing vio-
lence in Croatia. They remind those responsible for the violence of their determination
never to recognize changes of frontiers which have not been brought about by peaceful
means and by agreement. ... The Community and its member States call on the Federal
Presidency to put an immediate end to this illegal use of the forces on its command.11
In short, while Belgrade continued to invoke the doctrine of territorial unity, the
EU took the view that both entities were either already states, or entitled to be-
come states and in possession of pre-state rights. These were the right to territo-
rial integrity and unity, and the protection against the use of force by the central
government. The EU then attempted unsuccessfully to negotiate an orderly se-
cession. When this failed, it declared, through the medium of its Badinter arbitra-
tion/advisory commission, that the entire Yugoslav federation was in a process
of dissolution. Obviously, this was not an agreed dissolution, which would have
been legally unproblematic. Instead, it would be an effective dissolution that was
still being opposed from the centre. Nevertheless, the Badinter Commission held,
and the EU government accepted, that the individual republics that wished inde-
pendence would obtain it unilaterally, provided that they complied with a num-
ber of requirements, including the holding of a free and fair referendum and the
acceptance of minority rights guarantees.
Of course, the thesis of a dissolution of the SFRY was somewhat daring. After
all, the federation was only dissolving because Croatia and Slovenia had seceded
in the first place. Hence, the argument of express constitutional self-determina-
tion of these two entities was deployed in addition.
The Yugoslav episode also pointed to the attempts of international actors to
ensure that no wide-ranging precedent would ensue that might encourage seces-
sion elsewhere. Hence, express constitutional self-determination was framed in a
very restrictive way. In relation to federal-type entities, there would be two con-
11 EPC Statement on Yugoslavia, 27 August 1991.

36 Marc Weller
ditions. First, the constitution would have to assign a right of self-determination

to constituent units in a clear and unambiguous way. Second, only the constitu-
ent federal republics themselves would be entitled to self-determination. The line
was drawn in relation to Kosovo. That territory enjoyed a dual status according
to the constitution of the Socialist Federal Republic of Yugoslavia. On the one
hand, it had full federal representation, along with the six constituent republics
of the SFRY and Vojvodina. Hence, it was represented equally in the rotating col-
lective federal presidency, it sent directly elected representatives into the federal
parliament, etc. Moreover, its substantive competences were similar to those of
constituent republics proper, including even the right to maintain its own central
bank. On the other hand, Kosovo was also a unit that was legally subordinated to
Serbia – a constituent republic in its own right. Hence, in relation to the federa-
tion, Kosovo was a federal unit; however, in relation to Serbia, it was an autono-
mous province.
Kosovo had been subjected to the unilateral abolition of its federal status from
1987 onwards, submerging it within Serbia which launched a campaign of consid-
erable repression in relation to it. Nevertheless, Kosovo claimed that its original
status was distinguishable from that of constituents in name only. Hence, if they
could leave, so should Kosovo.
Kosovo’s declaration of independence, and its initially peaceful campaign
for international acknowledgement, was internationally ignored. Once again, it
proved that only violence appeared to be a tool for forcing international attention
to be directed towards this situation. In the end, NATO felt constrained to mount
a massive armed intervention on the European continent in order to address
the Kosovo crisis when it finally spun out of control in 1999. However, the use
of military force was justified exclusively on humanitarian grounds and NATO
expressly committed itself to the territorial unity of the then rump Yugoslavia.
Upon conclusion of hostilities, the United Nations gave similar assurances when
launching an international governance mission for the territory, as is evidenced
in Security Council Resolution 1244 (1999).
Hence, it appears that not all federal units within a federation providing for
express self-determination status are entitled to self-determination. In addition,
contrary to the initial, rather loose formulations of the Badinter Commission, not
every federal-type state structure automatically implies a right of secession for
federal units. This was confirmed in relation to Chechnya. One may distinguish
between federal states where it is assumed that sovereignty continues to be ex-
clusively concentrated in the centre, while the exercize of authority is merely de-
volved to the federal units. This is the route Russia chose to go when adopting its
new federal constitution after the dissolution of the USSR. According to the 1993
Constitution of the Russian Federation, Chechnya had indeed been promoted to
the status of a Federal Republic in that newly independent state. Nevertheless,
this status was not one derived from the original sovereignty of the individual
federal entities. Instead, it was one derived from a limited grant of central au-
thority, which left sovereignty vested in the overall Federation. The Constitution
provided:
We, the multi-ethnic people of the Russian Federation, united by our common destiny
of our land, seeking to advance human rights and freedoms and promote civil peace
and accord, preserving a historically established state unity, guided by universally rec-
ognized principles of equality and self-determination of peoples, ... renewing the sover-
eign statehood of Russia, ...
Article 3.1. The multi-ethnic people of the Russian Federation shall be the bearer of its
sovereignty and the sole source of authority in the Russian Federation.
Article 4.1. The sovereignty of the Russian Federation shall extend to its entire terri-
tory.
Article 66.5. The status of a member of the Russian Federation may be altered by the
mutual consent of the Russian Federation and the members of the Russian Federation
in accordance with a federal constitutional law.
In this case, it is the “multi-ethnic people of the Russian Federation” that are the
“bearer of … sovereignty” as a collective entity. Federal entities such as Chechnya
are not assigned original sovereignty, which resides only in the centre. Hence, in
stark contrast to the response of the EU to the Yugoslav crisis, the EU confirmed
the territorial integrity of the Russian Federation, instead of the claimed rights
of Chechnya, for example, in the following Declaration of the EU presidency on
behalf of the Union of 18 January 1995:
The European Union would again urge strongly that there should be an immediate ces-
sation of hostilities in order to facilitate the bringing of humanitarian aid to the popula-
tion and allow negotiations to begin without delay. The European Union takes note in
this connection of the proposal made by the Prime Minister of the Russian Federation.
It calls for a peaceful settlement to the conflict which respects the territorial integrity
of the Russian Federation.
When Russia forcibly reincorporated Chechnya some years later, the internation-
al community condemned the excessive brutality of the venture and its human
rights abuses. But there was little or no support for the suggestion that Chechnya
had a claim to independence, notwithstanding the ceasefire agreement of 1996, to
which reference will be made below.
However limited the construction of the doctrine of express constitutional
self-determination may be, it has been regarded with considerable skepticism by
some governments. The Yugoslav precedents, in particular, have had some impact
on the attempts to negotiate settlements in a number of other secessionist con-
flicts. The option of a federal-type solution has proven unacceptable to a number
of central governments. They presume that the granting of federal status to an
entity in exchange for a cessation of the conflict will inevitably be the first step to
an invocation of constitutional self-determination and to eventual independence.
To overcome this inhibition to finding a solution, some federal or even confederal
settlements have been counter-balanced with ‘eternity’ clauses. Such clauses con-
firm that any right to self-determination that may have previously been held by
38 Marc Weller
the secessionist unity has been expended in the agreement on a settlement. The
federal union that has been achieved is there rendered ‘indissoluble’.12
B Effective Dissolution
As we have seen, some constitutions will confirm that the overall state is com-
posed of former sovereign entities that retain at least the seeds of original sov-
ereignty. They may even confirm a right to self-determination, and sometimes
refer expressly to secession, appertaining to sub-state units. This was the case in
the USSR and the SFRY. Nevertheless, the declarations concerning independence
of the Baltic Republics, and the secessions of Slovenia and Latvia, were opposed
by the centre. In both cases, it was convenient to the international agencies ad-
dressing this issue to refer to the doctrine of effective dissolution, in addition to
considering express constitutional self-determination.
The concept of dissolution assumes that federal units can obtain statehood
if the federation disappears. It is not clear whether a dissolution of a federation
will only result in automatic statehood for constituent units if the former fed-
eral constitution confirms that these are the holders of original sovereignty, or if
this might apply to any federation. Given the express self-determination status in
both cases of effective dissolution, this issue remains untested. In relation to the
USSR, the effective dissolution was followed by a regularization of the situation
through the Alma Ata Declaration. This document clarified that the USSR had
ceased to exist, confirmed statehood for all former Union Republics, and it nomi-
nated the Russian Federation as the universal successor to that state.13 In relation
to the SFRY, the dissolution remained opposed from the centre for a considerable
period of time. The dissolution thesis was initially offered by the Badinter Com-
mission, which had found that the SFRY was in a ‘process of dissolution’ during
the autumn of 1991. That process had been completed at the end of the year, ac-
cording to the Badinter Commision, which then recommended recognition of
statehood for some constituent republics. The concept assumed that constituent
units would gain full sovereignty if the collective institutions of the federation
were no longer functioning or representative of all federal units.
Of course, even according to the concept of effective dissolution, a sharp divid-
ing line was drawn. The entitlement was restricted to fully federal subjects only.
Hence, the claim of Kosovo to statehood was initially rejected. That entity had
indeed argued that it previously enjoyed a clear federal status. With the disap-
pearance of the federation, it too should be allowed to gain direct international
legal personality. Chechnya made a similar argument. The USSR being dissolved,
nothing would bar it from achieving statehood. However, Chechnya had not been
one of the full Union Republics. Instead, the constitution recognized a federal
12 See, for instance, the Annan-Plan on the Comprehensive Settlement of the Cyprus
Problem of 31 March 2004.
13 In the latter case, the new states regularized their new status, and succession issues,
through the Alma Ata Declaration of 21 December 2001, 31 ILM (1992) 177.
status for it, but within Russia. Hence, the entity that could emerge and form a
sovereign state in consequence was Russia, and not, in turn, one of its constituent
units. Chechnya’s claim that its legal identity was unconnected with that of the
new Russian Federation was therefore rejected. Instead, Chechnya was consid-
ered part and parcel of the new Russian federal system and was born into that
constitutional structure, even if it violently resisted such a result at that time.
C Implied Self-determination Status

It is also possible to envisage an implied constitutional self-determination status.
This would be the case where a distinct ‘nation’ or ‘people’ inhabit a clearly consti-
tutionally defined territory. Where the central government consents to the holding
of a referendum on the issue of secession, or where such provision exists accord-
ing to the constitution in the absence of an express reference to self-determina-
tion, there is an expectation that such a referendum would need to be respected
by the central authorities. An example is furnished with reference to Scotland.
There does not even exist a written UK constitution. Nevertheless, referenda on
independence might be possible with the agreement and cooperation of the cen-
tral authorities. Should the result be in favour of independence, it is likely that that
outcome would have attracted a significant element of international legitimacy.
This view was very strongly confirmed by the Canadian Supreme Court in
a reference concerning the possible secession of Quebec. Despite the fact that
there is no express constitutional self-determination status for Quebec in the Ca-
nadian constitution, the Court found that: “A clear majority vote in Quebec on
a clear question in favor of secession would confer democratic legitimacy on the
secession initiative which all of the other participants in Confederation would
have to recognize.” However, the Court also confirmed that independence is not
an automatic result. Instead, both sides would have to engage in good faith nego-
tiations about the implementation of the decision to secede.
D Self-determination Status Generated through Central Government

Conduct?
More recently, it has been asserted that a constitutional self-determination status
may also be generated through adverse conduct by the central authorities. There
are two variants of this argument. One is based in part on the rulings of the Bad-
inter Commission. The Commission appeared to have suggested that indepen-
dence should be possible for federal-type entities that are being denied effective
representation in the legal and political structure of the federation. If negotiations
to address this situation have been exhaustively attempted, and if these have been
frustrated by the central government, it is proposed that the entity in question
should be entitled to secede. However, it has to be admitted that this theory is as
yet untested. In the closest case that comes to mind – Kosovo – the organized
international community was generally reluctant to disrupt the continued ter-
ritorial unity of the overall state.
40 Marc Weller
A second variant of this thesis would address instances where an entity has
suffered actual genocide or ethnic cleansing, or a deliberate campaign to destroy
a population by denying to it that which is necessary for its survival (such as
emergency food aid) on the part of the central authorities. In such instances,
some scholars argue that the central government should no longer be entitled to
invoke the doctrine of territorial unity in relation to that territory. Again, actual
practice does not yet offer any evidence of the acceptance of this view. While
armed humanitarian action has been launched in response to such situations,
this has generally been coupled with a strong confirmation of the territorial in-
tegrity of the state against which the intervention was directed. Iraq furnishes an
example, where independence for the long-suffering Kurdish North has not been
contemplated.
Overall, therefore, constitutional self-determination has developed as a con-
servative doctrine. It does not purport to broaden the scope of the right to self-
determination in international law. Nevertheless, states will be increasingly mind-
ful of the constitutional situation in their response to self-determination claims.
Given the need for an express establishment of constitutional self-determination,
however, this situation will remain rare. Most instances are likely to concern what
may be best termed unprivileged entities seeking to establish statehood not on
the basis of a legal entitlement, but on the basis of effectiveness.
III Effective Entities

What happens where an entity cannot lay claim to any self-determination status,
be it colonial or constitutional, but it nevertheless manages to exhibit the facts
that fulfill the criteria of statehood (territory, population, effective government)?
Most cases of secession or dissolution have occurred with the consent of the
relevant central authorities (United Arab Republic, Mali Federation, Malaysia-
Singapore, Yemen), or in the colonial context. There are hardly any examples of
unilateral opposed secessions that were effective in the long term. As was noted
above, the case of Bangladesh appears to be truly exceptional, given the special
role played by India in that instance. In that case, state practice did initially main-
ly reject the result, as is evidenced in the records of the UN General Assembly,
but this position was not maintained over time.
It might be argued that Somaliland constitutes an effective entity. After some
fifteen years of effective independence, it is tempting to argue that this entity
should now at least enjoy pre-state rights. That it to say, it should be protected
through the doctrine of territorial integrity from forcible attempts at re-incor-
poration. However, that case is as yet unsettled and its status remains precari-
ous until it is confirmed through an internationalized agreement. The hesitancy
with which it has been addressed by the international community confirms the
caution that is still being exercised in favour of the maintenance of the doctrine
of territorial unity. Similarly, the outcome of the Kosovo issue is also still open,
although there are signs that the international community might be willing to ex-
ercise a greater degree of flexibility than may have been initially envisaged, given
the realities on the ground.
It is possible to analyze the situation of effective entities in two ways. One
would simply point to the objective criteria of statehood. If an entity manages to
exhibit a defined population and territory that is subjected to an effective govern-
ment, and if the entity also has made manifest the will to be a state, then it is a
state. Recognition is generally only seen as declaratory; that is to say, it reflects a
status that already exists, rather than creating it. However, with respect to effec-
tive entities, declaratory recognition by other states is particularly important. For,
until an entity has attracted such recognition, it is not easily possible to confirm
whether it has obtained sufficient effectiveness to merit statehood.
Prolonged effectiveness, even in the absence of recognition, might also yield
statehood according to a second view. If a population makes manifest its desire
to act together as an organized political community independent of others on
a defined territory, then this manifestation of popular will must ultimately be
reflected in some form of legal status. For instance, most governments have rec-
ognized the People’s Republic of China’s sovereignty also in relation to the terri-
tory effectively controlled by the Republic of China. The PRC claims the right to
extend its authority also to Taiwan, by use of force if necessary. However, were it
to do so without an invitation from Taiwan, for instance by way of an armed inva-
sion, one might argue this would amount to a violation of the rights of the popu-
lation. These rights would be derived from a new variant of the doctrine of self-
determination that is reflective of a separate legal identity that, in this instance,
has developed over a period of half a century. However, it has to be admitted that
this theory is as yet untested.
IV Escaping from the Self-determination Trap through

Internationalized Settlements
Recent practice has sought to address the self-determination deadlock in oth-
er innovative ways. The termination of the Cold War not only resulted in the
fighting of new wars of secession and purported liberation. A number of others
were concluded, either because the parties had fought themselves to a stage of
mutually hurting stalemate, or due to external political or even armed interven-
tion. Several new techniques have been deployed in this context. In Europe itself,
there has been an attempt to defend at least the principle of territorial integrity
of the successor states of the dissolution of the USSR and the former Yugosla-
via. This technique has attempted to trade often wide-ranging self-government
of secessionist units, coupled with power-sharing mechanisms, for the continued
existence of the threatened state. At Dayton, therefore, very extensive powers of
self-government were granted to the constituent entities in order to retain the
overall state of Bosnia and Herzegovina. Attempts are being made at present by
international agencies to advance settlements in Moldova (Trandniestria) and
Georgia (Abkhazia, or more likely Southern Ossetia) along similar lines. A less
pronounced, modest autonomy settlement appeared sufficient in the attempts to
42 Marc Weller
terminate the conflict in Macedonia through the Ohrid agreements of 2001. The
attempt to draw Kosovo back into the ambit of Serbia and Montenegro through
the passage of time, on the other hand, has failed and international actors are now
preparing to negotiate a solution to the status issue that would be more accom-
modating to the wishes of the ethnic Albanian population. Hence, Kosovo may
well become one of the new types of settlements that question the previously
unshakable view that the territorial integrity of the central state must be restored
under all circumstances that do not qualify under the doctrines of classical colo-
nial or constitutional self-determination.
Indeed, even the earlier attempt to address the Kosovo conflict through the
Rambouillet process also resulted in a formula that is not free of ambiguity. The
text, to which Resolution 1244 (1999) of the UN Security Council makes refer-
ence, foresees the establishment of an internationalized mechanism that will set
up a process for the settlement of the status issue. That process was intended to
take account of, inter alia, the ‘will of the people’.
In fact, a more flexible approach is evidenced in several recent settlements
around the world. It was already noted that a number of governments are willing,
often after prolonged conflict, to give in to demands for external self-determina-
tion. As was already mentioned, this situation obtained in relation to Eritrea,
after the change in government in Ethiopia (and before a formal new constitu-
tion was adopted, providing for the right to self-determination to all nations and
nationalities in Ethiopia). A formal agreement had been struck between rebel
forces prior to the overthrow of the Mengistu regime, which granted to Eritrea
self-determination status and the right to hold a referendum on independence
after the expiry of an interim period. After that referendum confirmed the over-
whelming wish of the population to secede, the central authorities implemented
this decision.
In terms of a new generation of self-determination settlements, the experience
of Chechnya, on the other hand, was less encouraging. In August 1996, Chechnya
concluded an agreement with the Russian Federation, restated in 1997, which ex-
pressly recognizes Chechnya’s status as a self-determination entity and foresaw a
resolution of the issue according to that principle, and according to international
law, by 31 December 2001. However, after alleged Chechen terrorist attacks in
Moscow, the Russian Federation unilaterally annulled that agreement and forci-
bly re-incorporated Chechnya. This example highlights the desirability of seeking
to internationalize any agreement granting self-determination status at the point
of the termination of a self-determination conflict. Of course, even if Chechnya
had taken greater care to entrench the settlement at the international level, this
might not have dissuaded Moscow from a forcible incorporation. Still, it would
have made it more difficult for international actors to remain silent.
Despite this negative experience, there are now a significant number of in-
struments, generally arrived at with international involvement, that address the
self-determination status of an area of conflict in innovative ways. Through such
internationalized settlements, an attempt is made to escape from the self-deter-
mination trap. One of the first examples was provided by the Northern Ireland
agreement. The agreement confirms in Article 1 that that territory is a self-deter-

mination entity, whose status can only be changed on the basis of the wishes of
the majority of its population:
The two Governments … (i) recognize the legitimacy of whatever choice is freely exer-
cised by a majority of the people of Northern Ireland with regard to its status, whether
they prefer to continue to support the Union with Great Britain or a sovereign united
Ireland.
At the same time, the Article also recognizes the legal identity of the island of
Ireland, confirming that it is “for the people of the island of Ireland alone, by
agreement between the two parts respectively and without external impediment,
to exercise their right of self-determination on the basis of consent, freely and
concurrently given, North and South …” In this way, a formula was found that
could potentially satisfy both sides. The majority population is reassured that no
territorial change can occur without its consent. The republican communities
on both sides of the border, on the other hand, are assured of the fact the United
Kingdom would accept a decision to change the status of the territory, if it is tak-
en concurrently by both units. Hence, there is an alternative to an armed struggle:
a change in status can be achieved through gradual demographic change coupled
with political persuasion and reassurance of reluctant constituencies.
The Bougainville Peace agreement of 30 August 2001 assigns wide-ranging au-
tonomy to Bougainville. However, it also provides for a referendum to be held
among Bougainvilleans on the future status of the territory, although, somewhat
confusingly, this decision is subject to the review of the central government. The
referendum is to be held no sooner than ten years and no later than fifteen years
after the elections of an autonomous Bougainville government. Interestingly,
the holding of the referendum is conditioned on a holding of the ceasefire and
disarmament of the armed formations. There are also requirements that Bou-
gainville must demonstrate its capacity to deliver good self-governance before a
referendum can be held. This example offers another way of replacing the former
monodimensional logic of the armed struggle. Here, the act of self-determina-
tion is conditioned on a genuine renunciation of violence, instead of being the
product of violence and victory of the one side in an armed struggle. In addition,
energy needs to be focused on achieving an actual capacity of self-government
according to standards of rule of law and genuine democracy if the referendum
is to come about.
Another example of conditional self-determination was already noted above.
This is provided by the Gagauz settlement within Moldova. According to the Law
on the Special Status of Gagauzia of 23 December 1994, that autonomous entity
“shall have the right of external self-determination” should Moldova cease being
an independent state, for instance through a merger of its larger part (exclud-
ing Gagauzia and Transdniestria with Romania. Present negotiations seeking
to establish a federal-type structure for Moldova, including Transdniestria and
44 Marc Weller
Gagauzia foresee a similar reserve for Gagauzia in a future comprehensive con-

stitutional settlement.
The Sudan settlement is of a different type again. In a radical departure from
classical practice, the settlement clearly determines that the Southern unit will be
entitled to the exercise of the right to self-determination, no conditions attached.
The settlement is contained in the Machakos Protocols of 2002, and a series of
further protocols adopted since then. However, in a new twist, the settlement
requires that both sides cooperate to their utmost capacity over an interim period
of six years to make the option of continued unity attractive in advance of the
holding of a referendum in the South. Hence, interim governance is to be used to
reduce the thirst for independence, however unlikely that may be in practice. In
this face-saving way, the central government can consent to an agreement that,
in the end, is likely to result in the independence of the South, without having to
admit it to its constituents.
It will be convenient to examine these attempts to address the underlying self-
determination conflict in greater detail in the second part of this book.
V Conclusion
The right of self-determination was established in the late 1950s and 1960s, when
the major instances of decolonization had in fact been completed. It therefore
became acceptable to transform this doctrine into a right, given that it would
only be enforceable in relation to a small number of governments that continued
to cling to colonial rule at that time. The broadening of the application of the
right to cases like South Africa also caused no problems, given the pariah status
of the targets.
The right to self-determination was conceived in a way that made it very po-
tent in relation to the few instances of colonialism or analogous situations that
remained. However, even where these cases were concerned, the right to self-de-
termination was constructed in a way to yield results that would favour stability
of the ensuing result over restorative justice in relation to colonialism. This is
particularly evident in the doctrine of uti possidetis and in the view that self-de-
termination in the sense discussed in this article is a one-time-only event.
The existence of the right of self-determination therefore served as a conve-
nient legitimizing myth for the existing state system. It made it possible to argue
that the economic, social and political status of all states other than colonies must
be reflective of the will of the people. The actual representativeness of those ex-
ercising a hold over the state structure was of course not questioned. Their status
could be retrenched through the application of the doctrine of non-intervention
and territorial integrity (preventing challenges from outside) and the doctrine of
territorial unity (preventing challenges from inside).
Others, struggling for what they perceived to be self-determination, were left
outside of the system. Only a very limited corpus of humanitarian and human
rights laws (never effectively enforced) would apply to them. In a sense, the state
was given a carte blanche in dealing with groups seeking to assert their separate
identity. This principle was only disrupted when the rump Yugoslavia was denied
the right to use force against Croatia and Slovenia, bringing about the discovery
of the doctrine of constitutional self-determination. However, this entitlement,
too, was construed in such narrow terms that it did not really cover a significant
number of other conflicts that have been raging around the world for decades.
Self-determination, therefore, can be seen as something of a curse. It appears
to offer a promise of independence to populations. However, governments have
ensured that this promise is a hollow one. Naturally, the system has been rigged
to ensure that central governments will prevail in self-determination conflicts.
Over sixty years after the establishment of the post-Second World War inter-
national system, self-determination conflicts have remained endemic. It has be-
come clear that this system is not likely to generate the peace and stability it is
meant to achieve. By privileging stability over ‘justice’ (at least as seen by those
struggling for ‘liberation’), peace has been sacrificed.
While the post-Cold War realignments have resulted in the outbreak of a num-
ber of additional catastrophic self-determination conflicts, there have also been
a number of positive results. In a number of instances of mutually hurting stale-
mates, settlements are being achieved. These either circumvent the underlying
self-determination issue by offering wide-ranging self-government (autonomy or
even federal status) and power-sharing, or they do provide a possibility of sepa-
ration, often after a prolonged interim period. The latter settlements tend to be
obtained with strong international involvement, sought by the conflict parties
or imposed on them. There is, therefore, emerging a sense that it is necessary to
escape from the current self-determination trap, either by engineering new forms
of co-governance within states, or by accepting that secession cannot in the end
be ruled out if other options do not suffice.
Chapter 3
The Logics of Power-sharing, Consociation and Pluralist
Federations*
Brendan O’Leary
Power-sharing, rightly, is a standard prescription for protracted national, eth-

nic and communal conflicts in deeply divided territories, especially ones focused
on antagonistic self-determination claims. But no sensible advocate of power-
sharing assumes it is a universal panacea. Commending power-sharing has to be
feasible as well as desirable. And power-sharing is not simple, obvious, or one-
dimensional. To avoid confusion, a distinction needs to be made between ‘the
division of power’, ‘competition for power’, and ‘power-sharing’.
In the liberal tradition, influenced by the arguments of Montesquieu and
Madison and strongly present in American federalism, dividing political power
is critical to preventing despotism. It commends separating executive, legisla-
tive and judicial institutions. Inhibiting a monopoly of power, especially in the
executive, avoids dictatorship. The separation of civilian from military power, of
nomination from appointment, of police powers to arrest and interrogate from
the judicial power to prosecute, of local governments from central governments,
are less recognized, but just as important parts of the same logic. To divide power
is to prevent its abuse; to check power with power controls public officials.
Some think that organizing the division of power is what really matters in
deeply divided territories. A well-structured division of power, they say, inhib-
its national, ethnic or communal majorities – or minorities – from dominating
others. Proponents of integration and assimilation make this claim. For them,
representative government is power-sharing as well as power-division, because
it obliges power-holders to work in anticipation of the checking and balancing
capacities of the others. They say that they are supporters of sharing power, but
not among communities.
The competition for power, by contrast, focuses on how officials ‘win’ posi-
tions, be they executive, legislative, judicial or bureaucratic. In the liberal tra-
dition, competition for executive and legislative power should occur through
elections; liberals are more doubtful about elections to judicial and administra-
* My thanks to John McGarry and Shane O’Neill.
48 Brendan O’Leary
tive positions, for which they favour competitive meritocratic appointment. The
minimal definition of representative government is a political system in which
officials compete for authoritative positions in free and fair elections for citizens’
votes; in which elected officials hold office for limited terms, make laws, and give
orders to unelected officials within constitutional norms that ensure accountabil-
ity – both through the ballot box and recourse to the courts.
The division of power, and competition for power, are intelligent principles.
But, on their own, they are unlikely to calm deeply divided territories. Indeed, the
combination of the division of power and the competition for power may be con-
ducive toward the oppression of national, ethnic and religious communities. The
competition for power expresses or creates majorities – and such majorities may
be constructed from national, ethnic or communal cleavages. Majorities from
the same community may win control over all major offices and governments
– even if the powers of those offices and governments are divided and checked
– and then propose discriminatory public policy and conceptions of merit. In-
tegrationists and assimilationists too easily presume that a nation of individuals
is in existence, or that one should be built. But most states are multi-ethnic and
multi-confessional, and many are pluri-national. For this reason, integrationist
and assimilationist politics, even when motivated by high-minded consider-
ations, and however historically vindicated elsewhere, are recipes for conflict in
deeply divided territories. To commend integration, and especially, assimilation
where there are rival national self-determination claims is partisan or, alterna-
tively, utopian. It is partisan when one community seeks to nationalize the state
or region in its image on no better claim than might (numbers) makes right, it is
utopian when (potentially or actually) antagonistic communities are instructed to
fuse. Partisans and utopians succeeded many times in the last two centuries, too
often after bloodbaths.
That is why many contemporary liberals commend power-sharing as a sup-
plementary approach to avoiding the worst in pluri-national, multi-ethnic and
multi-religious states. Power-sharing commends the sharing of power as well as
the division of power and the competition for power. It commends ‘coalition’ as a
considered way of doing things, but not as a substitute for the division of power
and competition for power. Advocates of power-sharing might personally favour
integration or assimilation, in which all citizens have a primary loyalty to the na-
tion-state, but have learned that they cannot guarantee that outcome, at least not
without risking ‘internal’ war. They might prefer a strong programmatic govern-
ment pushed by a single-minded party, but think the costs too high. They might
think it best if their state had ‘normal’ institutions, but know that potential con-
flict between national, ethnic and religious communities warrants systematically
sharing power as well as dividing it and subjecting it to competition.
Proponents of power-sharing start from the recognition that political or cul-
tural homogenization can lead to genocide, ethnic expulsion, imposed partition,
and coercive assimilation. Their best argument is therefore very simple: power-
Why we still call internal wars ‘civil’ is a testimony to irony.

3 The Logics of Power-sharing, Consociation and Pluralist Federations 49
holders and their challengers have often pursued extremist and exterminist strat-
egies with grim results, sometimes to their mutual ruin, and power-sharing is
preferable. If Rudolf Rummel’s (1994)calculations in Death by Government are
even approximately correct the last century was the most lethal in human history.
The argument for power-sharing is, however, more sophisticated than acknowl-
edging that what cannot be won on the battlefield is best allocated through a
common forum. Power-sharers follow Rousseau’s (1762) declared method in The
Social Contract that commends taking “men as they are, and laws as they might
be,” but, because they do not seek just one community they reject Rousseau’s par-
ticular proposals as disastrous: namely, inalienable, indivisible, and absolute sov-
ereignty, the rejection of partial associations, and one vigourous homogenizing
civic religion. Power-sharers seek social contracts between two or more peoples,
or between two or more territorial governments. The first of these possibilities
leads to what are called ‘consociational’ directions; the second toward territorial
or federative power-sharing. These two possibilities can be combined in complex
forms, where the complexity refers both to their institutional formats and their
contexts. Here the focus is on the prerequisites for consociations and pluralist
federations.
I Consociation
The concept of consociation has been traced to Johannes Althusius (1557-1638).
He coined the formula consociatio consociationum for his vision of a cooperative
commonwealth. The etymology is suggestive: ‘con’, from cum, Latin for ‘with’, and
‘societas’, society. A consociation therefore is a society of societies. The word’s
roots imply separate societies that nevertheless cooperate in peaceful coexis-
tence. A political consociation, it follows, exists in a state or region within which
two or more cultural or ethnic or national communities peaceably coexist, with
none being institutionally superior to the others, and in which the relevant com-
munities’ leaders cooperate politically through both self-government and shared
government. Equality between the consociational partners is supposed, and there
are elements of both autonomy and of integration. No caste-like, colonial or rac-
ist relations can exist between the partner peoples.
The consociational idea has been re-invented many times by philosophers and
politicians. Reformers in the Habsburg Empire thought each nationality should
be treated as a full cultural equal (as autonomous Kulturgemeinschaften), and that
each citizen, on the basis of ‘the personality principle,’ should be able to declare
to which nationality or linguistic community they belonged, and enjoy rights
(including voting rights) wherever they lived. “Nations,” Karl Renner argued,
should be constituted as “associations of persons instead of as areas of domina-
tion.” Arend Lijphart, the key contemporary consociational thinker, published
The Politics of Accommodation: Pluralism and Democracy in the Netherlands in
1968, defining accommodation as the “settlement of divisive issues and conflict,
where only a minimal consensus exists.” He thought that the secret of Dutch po-
litical stability – after a history of religious disputes – was a spirit of accommoda-
tion amongst its political leaders, who cooperated to avoid violent conflict in a
benign self-denying prophecy. They had developed four capacities: the ability to
accommodate the divergent interests and demands of their respective collective
communities; to transcend cleavages to create common interests; to commit to
maintain and improve the system; and, lastly, a prudent appreciation of the perils
of fragmentation.
Later Lijphart, reasoned that democracies could be differentiated by the style
of their political elites, which could be competitive or collaborative, and by their
overall political culture, (fragmented or divided). These two dimensions generate
four types of democracy:
1. “Centripetal democracy,” in which politicians compete within a homoge
neous national culture (the received image of how liberal democracies do
and should work in the Anglo-American world);
2. “Centrifugal democracy,” in which politicians compete within a fragmented
multi-cultural environment, threatening system-breakdown (as in Fiji);
3. “Cartel democracy,” in which politicians collaborate within a homogeneous
and depoliticized national culture; and
4. “Consociational democracy,” in which political elites cooperate within a het-
erogeneous political culture, and sustain a politics of accommodation.
The last type serves as a general definition of democratic consociation, subject

to two provisos: cooperation need not preclude political competition for power;
and the forms of cooperation encompass the division as well as the sharing of
power.
Lijphart, unfortunately, declared that “consociational democracy means gov-
ernment by an élite cartel designed to turn a democracy with a fragmented cul-
ture into a stable democracy.” A cartel is anti-competitive, and the outlawing of
competition (and opposition) cannot be part of any consociation that is demo-
cratic. There can, of course, be non-democratic consociations – when political
leaders cooperate and conduct themselves according to consociational but not
democratic practices. Power may be shared among elites with little or no refer-
ence to their communities. But the fact that there can be undemocratic conso-
ciations does not, however, mean that consociation is inherently undemocratic.
There can be democratic and undemocratic presidents; likewise there can be
democratic and undemocratic consociations.
In Democracy in Plural Societies (1977,) Lijphart analyzed places that he
thought had displayed some consociational characteristics: Belgium, Switzerland
and Austria in Europe; Canada in the Americas; Israel (among Jews), and Leba-
non in the Middle East; and Surinam and the Netherlands Antilles in the Dutch
post-colonial world. He identified four consociational practices: cross-commu-
nity executive power-sharing, proportionality, cultural self-government, and
blocking powers, each of which requires further analysis.
This wording amends Lijphart’s original – he has accepted it as friendly amend-

ment.
A Cross-community Executive Power-sharing

Executives can share power through ‘complete’, ‘concurrent’ or ‘minimal’ means.
When the elected political leaders of all significant communities are represented
in the executive (e.g., in a collective presidency) we have a complete consocia-
tional executive, or a grand coalition. But an ‘all-encompassing’ coalition is not
essential for consociation; jointness in decision-making is more critical than com-
prehensive inclusion. In a concurrent consociation, each significant community
has over half of its voters supporting parties in the government. An executive is
minimally democratically consociational if one or more community merely gives
its plurality assent to its leaders in the executive. (This distinction between com-
plete, concurrent, and minimally consociational executives is equally fruitfully
applied to legislative procedures, and constitutional amendment procedures).
This clarification of complete, concurrent and minimally consociational pow-
er-sharing resolves a recurrent misunderstanding. Consociation does not require
the absence of opposition (or competition), as some critics suggest.
A ‘catch-all’ and ‘cross-community’ party that wins control of the executive may
sometimes function consociationally: if it draws support from each major com-
munity of voters at plurality levels or above; if it is descriptively representative of
the state or region it governs; if its internal party organizational characteristics
are consociational (i.e., it practices internal executive power-sharing, autonomy,
proportionality and blocking rights); and, lastly, if it follows consociational prac-
tices to manage crises that have national, ethnic, linguistic or religious roots.
What type of executive is best for consociational power-sharing? Many fol-
low Lijphart in preferring parliamentary government, in which premiers and
cabinets comprise a collegial executive, believing that executive presidencies are
majoritarian rather than consociational institutions. That certainly seems true
of single-person executive presidencies directly elected by all the citizens under
winner-takes-all single ballot contests: ‘one’ person cannot be a ‘mixed’ or ‘grand’
coalition. But it is a fallacy to assume that executive presidencies must be single
person institutions, or necessarily directly elected by the citizens. There have
been a significant number of presidential executive systems, especially collective
and rotating presidencies, which are compatible with consociational principles:
Bosnia and Hercegovina, Switzerland, the Italian region of Trentino-Alto Adige,
and the European Union furnish examples.
Nothing about presidencies per se prevents them from being designed or run
in a consociational manner. Indeed a strong premiership may be problematic
from the perspective of consociational decision-making. Where a premier heads
a party with a legislative majority, he or she has the capacity to determine cabinet
positions, and their relative salience. If the parliament is elected by proportional
representation and multi-party government is more likely, then the likelihood
that a premier has control over the composition of the cabinet is reduced. Conso-
ciational cabinets are usually ‘over-sized’, i.e., there are either more parties in the
executive and/or more legislators supporting those parties than would be needed
to control the legislature with a minimum winning coalition.
B Proportionality
In fully-fledged consociations, each community expects to be represented (either
descriptively or through choice) in political bodies that govern its members, in at
least rough accordance with its share of the citizenry. The representation may be
either descriptive, appropriately mirroring shares of groups, or substantive, with
persons expected to act for (or at least not against) the interests of their groups.
Proportionality can be partially applied, just to formal political institutions, or it
can be applied to all common institutions in a state and civil society (excluding
those in which each community has self-governing autonomy – the appointment
of Protestants to Catholic school boards would be against the spirit of consocia-
tion if education is supposed to be a domain of community self-government).
Proportionality in political institutions is most appropriately underpinned by
well-designed electoral systems. Whole families of such systems ensure that leg-
islative bodies are composed so that parties are represented in proportion to their
vote-shares. The most commonly used are closed and open party list systems,
and hybrid or mixed systems (which generally combine winner-take-all systems
at the district level with compensatory proportional allocations of top-up seats
for parties). There cannot be any uncontroversial notion of proportionality,
given that human beings do not come in fractions, that voters are very unlikely
to divide their votes in neat easily convertible shares, and that each method for
achieving proportionality “minimizes disproportionality according to the way it
defines disproportionality” (Gallagher 1991: 49). All proportionality systems nec-
essarily require mathematical rules to deal with the necessary ‘rounding off ’ or
‘sequencing’ of votes into seats. Each rule will have an explicit or tacit notion of
what minimizing disproportionality involves. These rules or formulae, however,
may be rank-ordered as to whether, when bias is inevitable, they favour larger
parties over smaller ones.
The merits of using PR systems to achieve proportional outcomes are obvious.
Provided district magnitudes are not too small and relatively evenly sized, and
provided that there has not been significant gerrymandering of electoral districts,
PR methods will produce outcomes that are usually seen as technically fair and
consistent, even though each may each have distinct quirks of their own. Using
such systems on a common roll has the merit of enabling voters to decide wheth-
er they want to be represented by ethnic, trans-ethnic or non-ethnic parties, i.e.,
voters enjoy self-determination (rather then the pre-determination of corporate
systems). It is a common mis-belief that consociations invariably stop free elec-
toral competition by forcing voters to vote for ‘their’ national, ethnic, religious or
tribal candidates. Not so. Where a political system deliberately obliges voters to
vote only within their own community for their own leaders then, to that degree,
the system may be called corporately consociational: separate electoral registers
for each group do that. But in liberal consociational arrangements, all voters are
on a common electoral register for the key governmental posts and are not re-
quired to vote for leaders from their own community of origin. It is therefore
untrue that consociation necessarily privileges prior collective identities.
In countries which use non-proportional electoral systems, such as the win-

ner-takes-all method in single member districts, it may still be possible to achieve
proportional representation of communities, though it may be more difficult to
guarantee, and will usually involve separate rolls or corporate representation. Re-
served seats may be kept for certain groups – e.g., the Maoris of New Zealand
were reserved seats when New Zealand used plurality rule, broadly in line with
their share of the citizenry. The territorial concentration of communities may
facilitate proportional representation of legislators by ethnic, racial, religious or
linguistic origin. The Quebecois are proportionately represented in Canada’s par-
liament, although Canada uses a single member plurality electoral system. After
1943, seats in the Lebanese Chamber of Deputies were divided in the ratio of six
Christians to five Muslims, or 54.5:46.5, which supposedly reflected the shares of
the two communities in the 1932 census. In fact, the actual ratio of residents was
50:49 and of citizens 52:47, so parity or a ratio of 9:8 would have been more just.
A ratio of 6:5 was also inflexible. In Lebanese politics before its internal war, it
gradually became impossible to hold a fresh census. When Muslims, presumed to
be expanding demographically, demanded a fresh census, Christians riposted by
demanding that the Lebanese diaspora, presumed disproportionately Christian,
be included in any count. This led to a stalemate and was one of the grievances
that provoked the outbreak of the Lebanese civil war.
Systems of reserved seats, or informal norms, are less likely to achieve propor-
tionality smoothly than proportional electoral formulae. They may also involve
pre-empting people’s identities and preferences, or adopting corporate (pre-
determined) rather than liberal (self-determined) principles of representation.
One-shot fixed quotas create obvious difficulties. Much the least controversial
quota allocation is to give a community a guaranteed share of positions in the ex-
ecutive or legislature that is broadly proportional to its demographic or electoral
weight. By contrast, the over-representation of significant minorities automati-
cally creates serious tensions, especially if the over-represented minority already
has other protections or has historically been privileged. In one famous example,
the Cypriot constitution of 1960 guaranteed 30% of the seats in the Cypriot par-
liament, and three of the ten seats in the Council of Ministers to Turkish Cypriots
(Article 46), even though they comprised less than 20% of the population. This is
one reason why this constitutional arrangement collapsed in 1963.
Parity of representation amongst demographic unequals creates predictable
objections amongst majority communities who, correctly, argue that proportion-
ality is a different principle from parity. The new Belgian constitution requires that
the federal cabinet be comprised of equal numbers of Flemish and French speak-
ers, even though Flemish speakers are now a clear demographic and electoral
This situation should be distinguished from one where affirmative action is required
to rectify historical imbalances (usually the result of discrimination). Affirmative ac-
tion policies may temporarily over-represent the under-represented in new cohorts
of appointees, but are intended to achieve overall proportionality, not to deliver long-
run over-representation!
majority. Disproportional representation may well be conceived of as a method

of minority protection, and may be demanded by minorities on various grounds,
but such representation may be, and is legitimately, criticized as departing from
the norm of proportionality (to each according to their numbers) and therefore
from strict consociational logic. The conjunction of over-representation or par-
ity amongst unequals with blocking rights (see below) creates obvious problems
for majorities. Majorities may well accept blocking rights on matters of national,
ethnic, cultural or religious significance for minorities, but they do not appreciate
why they additionally have to accept under-representation or parity.
The idea of proportionality also serves as a standard for the disbursement of
public funds by a regional or sovereign consociation. Indeed, departures from
proportionality may occasion dissatisfaction, unless one community is demon-
strably poorer than another. The Constitution of Cyprus of 1960 guaranteed
funding for the Turkish communal chamber at 1/4 of the sums to be provided
for the Greek communal chamber, even though Turkish Cypriots comprised less
than 1/5 of the population. Proportional allocation of public expenditure is rela-
tively easy to achieve technically among groups that are sharply differentiated
and relatively equal in endowments, though the politics of such allocation cannot
be presumed to be unproblematic.
Lastly, in consociations, proportionality is used as a standard for the allocation
of public sector and sometimes private sector positions. In 1958, the Lebanese
president introduced a principle of strict parity between Christian and Muslim
appointees to the civil service. This calmed communal relations, though tensions
remained because of Maronite ‘preserves’ in the security sector; later, Muslims,
convinced of their increased numbers and educational uplift, started to demand
the end of the quota and the proportionality principle. Examples of the use of
proportionality rules or quotas to allocate administrative and judicial positions
abound: Northern Ireland, Canada, and Belgium are all regions or countries with
rigorous fair employment laws, including the use of affirmative action, and in
some cases quotas, to achieve proportionality in public life. The principle of pro-
portionality, in short, does not occasion any fundamental conceptual or opera-
tional difficulties.
C Cultural Self-government
Consociations share political power between communities on matters agreed to
be of common concern. They do so in executives and legislatures, and, in princi-
ple, in judiciaries. There is also policy integration. In independent sovereign con-
sociations, security institutions (the armed forces, intelligence organizations, and
central policing services) and economic institutions (such as central banks) are
usually integrated. But, equally, successful consociations delegate decision-mak-
ing or grant autonomy (self-government) to communities on matters deemed ap-
propriate. Tacitly, consociations work with a principle of ‘subsidiarity’: making it
inappropriate to seek autonomy over what are matters of common concern, and
equally inappropriate to try to make a matter of common concern what has been
decreed to be within a segment’s autonomous rights.
The core idea of self-government from the perspective of minorities is “rule by
the minority over itself in the area of the minority’s exclusive concern” (Lijphart
1977: 41). But the idea applies to all communities, including majorities that have
autonomy. The idea of autonomy is easy to state in principle but its institutional
and regulatory manifestations are complex and very varied. It is not the same as
independence or sovereignty (though there can be co-sovereign federative units
in a state that enjoy autonomy). The idea, however, requires minimally constitu-
tionalized (or statutory or normative) arrangements, which regulate the level of
discretion enjoyed by the authorities with autonomy.
The key contrast in self-governnment, for our purposes, differentiates territo-
rial and non-territorial autonomy. Territorial autonomy in federations will be dis-
cussed shortly. Non-territorial autonomy is distinctively consociational. Under
non-territorial, also known as personal or group, autonomy, members of com-
munities may exercise their community rights wherever they reside or work in a
state or region. So, for example, they may publicly profess their religious beliefs or
hold religious meetings in public no matter wherever they happen to be. Under
personal autonomy, each person may opt to be recognized, or to receive services,
in accordance with their group membership. So, for example, a person living in
Brussels, Belgium, may opt to receive information about public services in either
French or Flemish. The provision of separate broadcasting networks for different
linguistic users throughout the entirety of a state, especially if each network is
run by its own community’s political institutions or civil society, is a good exam-
ple of non-territorial autonomy. Striking examples of ‘corporate legal autonomy’
are the separate civil law and personal status arrangements of the communities of
Lebanon, and India’s separate personal laws for its Muslim and Christian minori-
ties. No less consociational, but more liberal, are the fully funded and separate
educational systems of the Netherlands, Belgium and Northern Ireland.
The feasibility of personal or corporate principles of autonomy is not in doubt,
but it is not without its difficulties. While ethnic, religious and linguistic asso-
ciational life is prima facie unproblematic, modern states cannot dispense with
common territorial jurisdiction, either within a state, or within a province. In
some matters – the usual examples given are criminal or business law – a single
code of behaviour and a single regime of sanctions will usually be rational. Nev-
ertheless, distinctively consociational self-government is administratively cogent
for territorially dispersed communities, and to communities that are extensively
geographically mixed, as has been demonstrated at various junctures in the his-
tories of Belgium, Cyprus, Lebanon, and Northern Ireland.
D Blocking Rights
The last element of consociational practice is partly implied by the existence of
the other three. If there is cross-community executive power-sharing, then each
community has at least weak and perhaps vigourous protection against majori-
tarian executive discretion, and legislative agenda-setting. If there is proportional

representation, then the capacity of each group to block dominance by others is
enhanced – though not guaranteed. If there is consociational self-government,
then minorities can stop other minorities or majorities from exercising executive
or legislative dominance over them within the functions and spaces where au-
tonomy applies. But formal blocking rights may be sought and institutionalized
to ensure the guaranteed protection of minorities. Veto rights may exist within
the constitution. Within the executive, collective presidents or dual premiers may
share agenda-setting and agenda-blocking powers, and vice-presidents may be
granted veto-rights. Within the legislature, the consent of all the affected com-
munities may be required before constitutional change can take place. This can
take the form of requiring unanimity within the executive, a concurrent majority
within the legislature, or a weighted or super-majority that ensures concurrent or
significant cross-community support – as in the ‘cross-community consent pro-
visions’ of the 1998 agreement in Northern Ireland. Similar restrictions to stan-
dard majority rule in the legislature may apply to ordinary legislation if minorities
have the right to petition that such matters be deemed of national, ethnic or com-
munal significance. Courts may be charged with protecting group autonomy in
bills of rights and charters that effectively place constitutional constraints on ma-
jorities that are equivalent to entrenched veto rights; ombudsmen may be given
similar quasi-judicial roles.
Veto-rights create obvious difficulties for majoritarian democratic theory: mi-
nority ‘tyrannies’ may block desirable change; deadlock, immobilism, or policy
stagnation can flow from the deployment of vetoes by all groups; and minori-
ties that over-use their veto rights may destabilize a consociational settlement.
That said, these difficulties should not be exaggerated. In consociational systems,
formal veto rights should mostly apply in the domains of the politics of identity
and security, i.e., in ethnic, religious or national domains, and not to every policy
sector. In effect, in these domains, groups require parity of power, rather than
proportional power.
II Pluralist Federations
There is less controversy about the meaning(s) of territorial power-sharing. It
involves a division of powers between territorial units, i.e., with spatially focused
self-government in at least some specific functions (executive, legislative, judicial,
or administrative), and shared rule between the territorially autonomous unit(s)
of government and the confederal, federal, or central government.
Two ways of organizing territorial power-sharing are well known: federations
and confederations. Another is less well known: constitutionally entrenched au-
tonomy, in which any change in the status and powers of the autonomous unit
requires joint consent and parallel procedures by both the autonomous and
the central (or federal) authorities – also known as a system of ‘federacy’ (Ela-
zar 1977). Each of these distinct modes of organizing territorial government has
multiple, and in principle infinite, variations (see Hannum 1996; Lapidoth 1997).
They vary, for example, over whether they grant rights of exit (secession), fusion
(unification with other units), or opt-outs, or over their degree of symmetry in
dividing powers.
The Forum of Federations sensibly advocates federation as the best means of
sharing power where there are deep national, ethnic and other cultural divisions.
But two polar types of federation are not always properly considered in this de-
bate. One type of federation is integrative. It builds on the liberal logics of divid-
ing power and of organizing competition for power. It recognizes one nation. It
is centralized: the federal government and judiciary have significant powers over
citizens and over the member-units of the federation. Lastly, it facilitates a strong
majoritarian federal government (in the executive or legislature or both). This
kind of federation may work with an already homogenized people, or a federa-
tion with an overwhelmingly dominant people, and may occasion no complaints
from immigrants willing to be integrated or assimilated. But it will not work with
multiple national peoples, where there is no overwhelmingly dominant people,
and where there are rival homeland (self-determination) claims. The USA and
the federations in Latin America resemble integrated federations. They tend to
recommend what they have to others.
The other model of federation, better suited to the circumstances of deep di-
visions, is pluralist (O’Leary, McGarry and Salih 2005). It is consistent with the
liberal logics of dividing power and organizing competition for power but also
insists on the prudence and merits of power-sharing. It recognizes more than one
nation or people. It is decentralized: the member-states have significant policy-
making and legal powers and can resist encroachments by the federal branches
of government. And, lastly, it tends to be consociational in the federal executive,
legislature and judiciary: presidencies and cabinets are broadly representative;
the decision-making and representation rules in the house of representatives and
the house of the member-states are consensual; the federal judiciary is represen-
tative of the member-states and cautious about interfering in the self-government
of the member-states. Belgium and Switzerland and Canada (in practice if not in
constitutional texts) resemble the pluralist model. They tend to recommend what
they have to others.
Some critics of pluralist federation recognize its appeal and the appeal of con-
sociational decision-making. But, regretfully or otherwise, they claim that plu-
ralist federations and consociations are precarious, recipes for the break-up of
states, and certainly insufficient to hold them together. The recognition of na-
tionality or ethnicity is but the prelude to secession or discriminatory conflict.
Decentralization facilitates local abuse of power. Consensual or consociational
decision-making is a recipe for gridlock.
Pluralist federations have indeed often broken down (and consociations have
had a limited track-record of success). But it is not clear that the critics have
better alternatives, other than letting one group win. And it is unclear that they
have understood what factors lead to the breakdown or success of pluralist fed-
erations. We have sought to explain what makes a pluralist federation work by
seeing what makes one fail (McGarry and O’Leary 2003; O’Leary 2001; O’Leary,
McGarry and Salih 2005). Careful analysis should not lead people to reject plu-
ralist federations.
Pluralist federations are more likely to break down if they are put together
coercively (e.g., the USSR and Yugoslavia). They need to be built as ‘voluntary
unions’, or rebuilt as voluntary pacts if they were put together coercively. It is usu-
ally centralizers who threaten the federation’s founding pact, and who endanger
stability – secession is less common as a first move than as a reaction to mal-
treatment. Pluralist federations that are undemocratic are vulnerable to break-
down at the moment of democratization, but if established as democracies and
maintained as democracies are more likely to hold together. The maltreatment of
minorities, especially compact minorities on the geographical perimeter of the
federation, is an incitement to break-up. Conversely, equal treatment and con-
sociational treatment give nationally compact and dispersed minorities reasons
to stay. Pluralist federations will break down if severe distributive conflicts arise
over natural and fiscal resources, but may survive if they manage those conflicts
equitably (proportionally). A generous dominant people can live with the conces-
sions made to national minorities and treat them as equal partners, knowing that
there is both prudence and self-interest in such conduct. A dominant people that
coerces others, maltreats them, or exploits them, will reap what it sows. Natu-
rally, the preconditions of a successful pluralist federation are not all domestic.
Pluralist federations will do better if they do not have interventionist neighbours,
or if they lack irredentist neighbours.
Conclusion
A last word on international norms and interventions is needed. There are inter-
national proscriptions against genocide and expulsion, and norms of some signif-
icance that reward states that are democratic – and make discriminatory control
regimes potential pariahs. There are additional proscriptions against coercive as-
similation, and strong biases against secession and partitions. The conjunction
of these norms leave international organizations and great powers, when they
intervene in national, ethnic, and communal conflicts, confined to promoting
one of three repertoires of democratic design: (i) federations (and variations on
territorial power-sharing); (ii) consociation; or (iii) integration. In some places,
to prescribe integration, in say Bosnia and Herzegovina, Macedonia, Northern
Ireland, Cyprus, or Iraq, is simply to prescribe the partisan victory of one com-
munity over another. In short, the exclusion of certain options that were once
available to interventionist powers has created additional leverage in favour of
pluralist federations and consociational arrangements. Since these logics of pow-
er-sharing may reduce the tides of national, ethnic, and communal blood that
flow in our times, exploring their productive possibilities has a pressing political
and intellectual urgency.
Part Two
Case Studies
Chapter 4
Complex Power-sharing in and over Northern Ireland:
A Self-determination Agreement, a Treaty,
a Consociation, a Federacy, Matching Confederal
Institutions, Intergovernmentalism,
and a Peace Process*
Brendan O’Leary
I The Sources of Conflict
A The Territory, Its Claimants, Its Peoples

Conflict in and over Northern Ireland had deep external and internal roots. It
has been matched by a ‘meta-conflict’, i.e., intellectual controversy over what the
conflict is fundamentally about (McGarry and O’Leary 1995a: 1). But the primary
historic antagonism is simply stated: Ulster unionists want Northern Ireland to
remain part of the United Kingdom; Irish nationalists want it to become part
of the sovereign Irish nation-state. The social cleavage underlying this political
antagonism is also clear: within Northern Ireland, Ulster unionists and Irish na-
tionalists are overwhelmingly the political expressions of largely endogamous
descent-groups, i.e., ethnic communities. Each of these respective communities
has been partially mobilized into the British and Irish nations; each seeks the
succour of its respective co-nationals and their sovereign states; each is partially
distanced from its co-nationals by historic, political and geographic separations.
The historical roots of the conflict are equally evident: “Northern Ireland was and
is a legacy of settler colonialism” (O’Leary and McGarry 1996: 101).
The precise political unit is of recent vintage, created in 1920 by the United
Kingdom parliament, through the Government of Ireland Act, unsupported by
any MPs from Ireland. Irish nationalists regarded its formation, and its borders, as
* The author thanks all the participants in the Resolving Self-determination Disputes
Through Complex Power-sharing project, but is especially grateful to Angela Hegarty,
John McGarry and Paul Mitchell. The Carnegie Corporation of New York’s funding,
the United States Institute of Peace, the London School of Economics, the University
of Pennsylvania and the Rockefeller Foundation’s Bellagio fellowship assisted in re-
search and composition. This chapter draws upon but extensively modifies previous
analyses (O’Leary and McGarry 1993; O’Leary 1999d; 1999e; 2001b; 2001c; 2001d;
2001g; 2002a; 2002b).
a coercive, fresh and unnatural cut across the homeland of the Irish people and the
historic province of Ulster. Ulster unionists, by comparison, regarded its creation
as a ‘second-best’ option. They would have preferred all of Ireland, or at least all of
historic Ulster, to have remained in the UK. Northern Ireland was forged from six
of the nine counties of historic Ulster, which in the seventeenth century had been
the site of extensive colonial plantations by Scots and English settlers encouraged
by King James I and his successors. These plantations were more successful than
others in the rest of the island (see map below). The settlers and their descendants
were originally extensively differentiated from the native Irish by their language,
culture and religion. Of these markers of difference, religion has been the most
durable.
ULSTER
Crown sponsored plantations
of James I
Private plantations of Lo Antrim

nd
Chichester, Hamilton on
and Montgomery Donegal de
rry
Tyrone
Fe
rm
an
ag Armagh D o w n
h
Cavan
CONNAUGHT
LEINSTER
MUNSTER
0 miles 50
0 km 50
Source: O’Leary and McGarry 1996: Ch 2, 58.

4 Complex Power-sharing in and over Northern Ireland 63
Northern Ireland was the by-product of both British and Irish nation-building
failures. In the sixteenth and seventeenth centuries, Ireland was re-conquered for
the English Crown. The Catholic religion of its natives and the ‘old English’ was
repressed, and outlawed. Massacres and mass starvation through the deliberate
destruction of crops accompanied the mid-seventeenth century Cromwellian re-
conquest. In two large-scale ‘settlements’, one under the aegis of Cromwell, the
other later in the century under William of Orange, the land of Irish Catholics,
native and old English, was almost entirely confiscated on behalf of the ‘New
English’, led by what became an Ascendancy caste of Protestant aristocrats who
controlled the Irish parliament. The Gaelic civilization of the native Irish was
crushed, its aristocrats were dispossessed, and many departed into exile (Car-
ty 1996). In the eighteenth century, the position of Irish natives and Catholics
gradually improved, though the century ended with significant peasant forces
participating in a Jacobin-inspired insurrection against British rule in 1798, which
prompted William Pitt to create the Union of Great Britain and Ireland in 1801.
British nation-building in the nineteenth century failed to integrate Ireland as
a co-equal unit of the Union. Its form of government combined integration and
colonial administration. The Westminster parliament failed to integrate the Irish,
especially the Catholic Irish, as co-nationals of the British: the formal emanci-
pation of Catholics did not take place until 1829. British nation-building failed
in significant measure because Irish Protestants, predominantly of settler ori-
gin, resisted the reversal of the seventeenth century conquests implicit in treat-
ing Irish Catholics as equals (Lustick 1993; 1985). In the ethnic memory of Irish
nationalists, British management of the catastrophic Great Famine of the 1840s
was proof of the indelibly inferior moral status of the Irish in English eyes. The
diaspora produced by the Famine scattered, and would include a new and large
Irish American stock (later an important resource for the nationalists of Ireland).
Subsequent land reforms initiated attempts to kill Irish nationalism with belated
kindness. But this history of conquest and catastrophe explains why Irish nation-
alist movements mobilized with considerable mass popularity, especially from
the late eighteenth century onwards. They failed, however, to mobilize uniformly
across the peoples of Ireland. Irish Protestants, especially Ulster Protestants,
were not universally regarded as full co-nationals of the re-emergent Irish nation.
Irish nationalism encompassed movements to reverse the conquest of the land,
to revive the ancient Gaelic culture, and to win sovereignty for the island. Though
it numbered some Protestants amongst its political and cultural elites, its mass
base was overwhelmingly Catholic (Hutchinson 1987).
Historic Ulster had the greatest spread and concentration of Protestants and
descendants of settlers in Ireland. They identified with the Crown and the British
state, given the Protestant nature of the kingdom and union with Great Britain
(Colley 1992). In Ulster, Protestants were found in all social classes, whereas in
the rest of the island, generally, they were disproportionately in the higher strata.
Not surprisingly, Ulster’s Protestants were most resistant to the demands of Irish
nationalists for the repeal of the Act of Union, for home rule, or for a sovereign
independent Ireland. When the democratization of the UK began in earnest in
the 1880s, Ulster unionists, in coalition with the Conservatives and their aristo-
cratic supporters in the House of Lords and the military, successfully resisted
proposed legislation to grant home rule or autonomy to Ireland, bringing the UK
to the brink of a regime breakdown on the eve of the First World War.
The partition of Ireland in 1920, creating ‘Northern’ Ireland, with a popula-
tion of one and a half million, and ‘Southern’ Ireland, with a population of less
than three million, was justified by British policy-makers as a way of dealing
with the rival identities, interests and demands of Irish nationalists and Ulster
unionists. Whatever the merits of these arguments, there was little normative
or cartographic merit to the chosen partition. Irish nationalists noted a double
standard: “to avoid imposing on Ireland a system rejected by a quarter of its pop-
ulation” British governments “created and defended an artificial Northern Ire-
land rejected by one third of its population” (Laffan 1983: 123). The Government
of Ireland Act was the result of the decisions of the UK cabinet, which wanted to
keep Ireland within the empire and to satisfy Conservative and unionist demands
on Ulster. The Act was not an act of self-determination on the part of any of the
peoples of Ireland (Gwynn 1950; Laffan 1983; Carty 1996; McGarry and O’Leary
1995a: Ch 1; O’Leary and McGarry 1996: Ch 2; Mansergh 1997). The UK govern-
ment acted from a mixture of imperialist and paternalist motivations, claiming
to be responding to majority and minority sections of Irish opinion. But, in the
preceding five centuries of British rule, Ireland had been regarded as an entity, in
earlier centuries with a parliament of its own, and since 1800 as one of the king-
doms joined in the United Kingdom of Great Britain and Ireland, with a separate
administration and legal system. Even considering matters solely from a British
constitutional perspective, the unit of self-determination should have been Ire-
land, since it was as such a unit that the Irish parliament had approved the Union
(Mansergh 1997). Irish nationalists, especially militant republicans, rejected the
Government of Ireland Act. It was, by contrast, accepted by Ulster Unionists as
the best available means of protecting their connection to Great Britain.
B The Scale and Duration of Conflict

Irish nationalists, North and South, resisted Northern Ireland’s formation. They
were, however, eventually forced to accept partition, albeit as a temporary fact. In
1921, a treaty, negotiated between Sinn Féin and the UK government, recognized
the domestic independence of the territory of what the British had initially called
‘Southern Ireland’, as the Irish Free State (Pakenham 1935; Lyons 1973; Fair 1972;
Laffan 1999; Lawlor 1980-1981). Its independence was circumscribed: the treaty
obliged the Free State to remain in the British Empire, according to the model of
dominion autonomy, and the British insisted on the recognition of their Crown,
an oath, and on provisions requiring Ireland to provide naval bases “in time of
war.” The newly established Northern Ireland parliament was given the right to
opt-out of the Irish Free State, a right promptly exercised. The treaty was accom-
panied by the promise of a Boundary Commission, because the UK’s negotia-
tors realized that the new border was scarcely politically or ethically defensible.
But, its terms of reference were ambiguous, and its composition (a nominee each
from Northern Ireland, the Irish Free State, and the UK government) gave Belfast
and London a controlling majority.
The Treaty prompted a civil war within the new Irish Free State. Anti-Trea-
tyites rejected the Crown and partition as denials of the Irish people’s right to
national self-determination. Pro-Treatyites believed the settlement was the best
available. They won the civil war. They hoped for a positive outcome from the
Boundary Commission, but when its proposals became evident and were plainly
tilted towards Ulster unionists, they pragmatically accepted the 1920 partition.
Within the Irish Free State political competition focused on establishing full
sovereignty and state-building. The anti-Treatyites, the losers of the Civil War,
eventually mobilized behind the Fianna Fáil party, which in the 1930s became the
dominant electoral and then governmental power in independent Ireland. Its po-
litical programme, including bilingualism and asserting Ireland’s freedom from
Britain, meant that Irish nation-building was regarded by Ulster unionists as in-
imical to their interests and identity – even though independent Ireland treated
its major cultural minorities, the Anglo-Irish and Irish Protestants, better than
Ulster unionists treated its much larger minority. Fianna Fáil designed a new Irish
Constitution, promulgated and popularly endorsed in 1937, which repudiated the
1920 Act and the 1921 Treaty, and asserted the right of the Irish state to govern
all of the island, and defined the entire island as national territory. But it also
committed the state to seeking Irish unification according to the principles of
international law, i.e., through peaceful diplomacy (Bunreacht na hÉireann 1937:
Articles 2, 3 and 29).
Within Northern Ireland, running a miniature version of the Westminster
model of parliamentary government, the Ulster Unionist Party (UUP) built a sys-
tem of majoritarian control, using political, legal, economic, cultural, policing
and discriminatory administrative mechanisms to organize the new majority and
to disorganize the new minority (O’Leary and McGarry 1996: Ch 3, modifying
Lustick 1979). The UUP governed without interruption from 1921 until 1972. The
system was built because Ulster Unionists feared the irredentism of Irish nation-
alists, North and South, and that British governments might betray them. It was
underwritten by the cultural legacy of settler colonialism, in which Irish national-
ists and Catholics were regarded as (and treated as) inferiors. The impact of the
system was palpable: Catholics emigrated in significantly disproportionate num-
bers. Having constituted approximately 35% of Northern Ireland’s population in
1921 they composed approximately 60% of its emigrants between 1921 and 1981
(O’Leary and McGarry 1996: 131). Emigration offset the higher Catholic birth-
rate and stabilized unionist dominance. The system proved durable until the late
1960s, successively defeating a series of efforts by the paramilitary organization,
the Irish Republican Army (IRA), to conduct guerrilla warfare as a national lib-
eration strategy, and successfully marginalizing constitutional Irish nationalist
opposition. But, in the late 1960s, the system broke down under the political and
moral pressure of a civil rights movement, modelled on that in the USA, and fo-
cused on achieving an end to discrimination.
The re-organization of most Northern nationalists was manifest in the for-

mation of the Social Democratic and Labour Party (SDLP), which recognized
Northern Ireland, but sought equality, reconciliation and Irish unification – in
that order. In the late 1960s the (then Labour) UK government intervened to pro-
mote reforms widely resisted within the dominant UUP, and in 1969 sent in the
British Army to uphold security because the Royal Ulster Constabulary (RUC)
had proved both highly partisan and incapable of maintaining order. In 1970 a
Conservative UK government, more sympathetic to the UUP, authorized a crack-
down, and in 1971 oversaw the re-introduction of internment without trial and
other measures targeted against a new breakaway and militant organization, the
Provisional IRA (PIRA). The latter argued that Northern Ireland could not be
reformed, and embarked upon an extensive campaign of political violence with
significant support amongst a minority of the minority. In 1972 the UK abolished
the local Northern Ireland Parliament and ruled the region directly from London
(Westminster and Whitehall). Subsequently it would combine direct rule both
with intermittent attempts to organize a power-sharing and political and military
efforts to repress republican violence.
There was extensive political violence within Northern Ireland from the late
1960s until 1994. A three-sided conflict was waged between the forces autho-
rized by the UK, namely, the British Army, including locally recruited regiments
(overwhelmingly comprised of Ulster Protestants), and the police, the RUC; and
opposing paramilitaries, republican and loyalist. The largest and most effective
republican organization was PIRA, now known simply as the IRA. Its political
party was Sinn Féin. The largest and most effective loyalists became the Ulster
Volunteer Force (UVF), and the Ulster Defence Association (UDA). The latter,
operating under various front organizations, remained legal until 1991. Their
political parties became the Progressive Unionist Party (PUP) and the Ulster
Democratic Party (UDP). The conflict’s epicentre was in Northern Ireland, but
republicans spread it to Great Britain, especially London, and also targeted Brit-
ish military bases on the European continent, while loyalists made occasional in-
cursions into Ireland. The death toll, over 3,600, was felt most deeply in Northern
Ireland. Republicans were responsible for the largest number of deaths, nearly
twice as many as loyalist organizations. Loyalist organizations were responsible
for killing more than twice as many people as the state’s security forces. Though
the state’s forces were restrained, especially by comparison with state agents in
some other ethno-national conflicts, there were extensive abuses of power, and
the law was not effectively applied to state-authorized killings (Ní Aoláin 2000).
The first full-scale audit of the human and economic costs of conflict between 1969
and 1989 was sketched by the author with others (O’Leary and McGarry 1993: Ch
1; first sketched in O’Duffy and O’Leary 1990). Subsequent research has confirmed
the relevant patterns, but with better confirmed and verified data (Sutton 1994; Fay,
Morissey, and Smyth 1999; Fay, Morrisey, and Smyth 1997; Ní Aoláin 2000). McKit-
trick et al. provide the definitive treatment of deaths (McKittrick et al. 1999; 2001;
reviewed by O’Leary 2000a).
The deaths-to-citizens ratio was very high given the region’s small population,
parliamentary conditions, and the absence of inter-state war (O’Leary and Mc-
Garry 1996: Ch 1). The most extensive phase of violence ended in 1994 when the
IRA called a ceasefire, later reciprocated by the major loyalist organizations. The
IRA broke its ceasefire in 1996 after a Conservative government failed to fulfil its
pledge to deliver speedy negotiations, but it was resumed when a new Labour
government took power in 1997. Despite intermittent break-downs and acts sug-
gesting poor discipline, the paramilitaries’ ceasefires are still formally in place at
the time of writing, and facilitated the construction of a comprehensive negoti-
ated inter-governmental and multi-party political settlement in 1998.
C Previous Attempts at Conflict Management

Three conflict-management regimes in and over Northern Ireland have existed
since its construction – each of which has had multiple contradictions. The first,
‘divide and partial quit’, was embedded in the Government of Ireland Act and
the 1921 Treaty and lasted until 1969–1972. The UK gave control to two new ma-
jorities – to Irish nationalists in independent Ireland and to Ulster Unionists in
Northern Ireland – and divested itself of direct responsibility for most domes-
tic governance in the island. It was happy to see cooperation between the two
new political units – both the Act and the Treaty envisaged the possibility of
the Council of Ireland linking the two units. And it would have been content
with Irish unification within the Empire. The settlement had right-protections for
both of the new minorities, but not as national minorities. Overall, it was biased
in favour of Britain’s geo-political interests, fettered the sovereignty of indepen-
dent Ireland, and enabled the Ulster unionists to abuse their local power. The
systems of rights-protection worked better in independent Ireland, with its small
minority, than in Northern Ireland, with its much larger minority. The settlement
was, eventually, more successful in independent Ireland, because its minority was
small, integrated, and eventually extensively assimilated, and because indepen-
dent Ireland established its full constitutional and international sovereignty and
personality in 1937, becoming a Republic in 1949. The settlement in Northern Ire-
land was much less successful because it facilitated majority dominance through
its local version of the Westminster model of parliamentary government, because
there was a much larger minority, and because that minority generally sought
neither integration nor assimilation with the UK.
The second system of conflict-management, ‘arbitration with intermittent ef-
forts to promote power-sharing’, was in effect throughout most of the period be-
tween 1972 and 1985. The UK government, through its Northern Ireland Office,
presented itself as a neutral authority above the rival sub-cultures of Protestants
and Catholics. The intervention and presence of the British Army was justified
in this manner. There would be, despite some occasional reconsiderations, no re-
More nuanced and detailed accounts of British policy-making after 1969 are avail-
able (Cunningham 2000; Mitchell and Wilford 1999; O’Leary 1997a; 2002b).
turn to simple majority dominance of the type exercised by the stormont Parlia-
ment – a vista unacceptable to Northern nationalists, the Irish government, and
international opinion, especially Irish Americans. Northern Ireland would be re-
formed: discrimination against Catholics would be effectively remedied. Security
policies, it was said, would be impartial, directed against both anti-state republi-
can and pro-state loyalist paramilitaries, whose actions would, it was eventually
decided, simply be criminalized.
Arbitration was a partial success. Though direct rule denied self-govern-
ment to Northern Ireland, the region was partly reformed. The agenda of the
civil rights movement was implemented. Cultural Catholics began to vote and
to stay in greater numbers. Funding of Catholic education gradually improved
(McGrath 2000). Security policy was no longer run at the direct behest of UUP
politicians. But, in two respects, arbitration was a failure. First, UK security poli-
cies – internment without trial followed by criminalization regimes, and the use
of shoot-to-kill policies – were initially incompetent, subsequently counter-pro-
ductive, and always involved human rights abuses (and sustained – and convinc-
ing – allegations of collusion between state agencies and loyalist paramilitaries).
UK security policies helped build a significant core of support for the IRA and
the political party that spoke for it, Sinn Féin, which started to grow significantly
in the early 1980s after Mrs Thatcher’s government mismanaged its response to
republican hunger-strikers. Second, efforts to promote power-sharing or conso-
ciational initiatives were persistently unsuccessful (McGarry 1990; O’Leary 1989).
A brief experiment in voluntary power-sharing within Northern Ireland, linked
to an all-island Council of Ireland, known as the Sunningdale agreement, col-
lapsed in 1974 after a general strike organized by loyalist paramilitaries and the
withdrawal of unionist support for the initiative (for details of its content and
collapse, see Wolff 2000; 2001).
A voluntary consociational settlement amongst moderates was inhibited by
several factors (O’Leary 1989). Hardliners opposed such a settlement: republicans
wanted Irish national self-determination and a withdrawal of the UK state; hard-
line unionists either wanted a return to local majority-rule or full-scale integra-
tion into the UK, i.e., no local power-sharing. Hardliners constrained moderates
who feared losing electoral support within their respective ethno-national blocs
(Mitchell 1991; 1995; 1999). Second, moderate nationalists wanted any settlement
to be internationalized, i.e. to have a linkage to Ireland, and a role for the Irish
government, a position that moderate unionists opposed. Moderate unionists
preferred voluntary power-sharing devolution, exclusively within the UK, i.e.,
with no all-Ireland dimension; others preferred Northern Ireland’s full ‘integra-
tion into the UK’, though they differed on what this meant. Third, the precondi-
tions of a voluntary consociational settlement were absent: namely, appropriately
motivated political elites with sufficient predominance within their blocs to make
a settlement, and with sufficient stability within their blocs to maintain such a
settlement (O’Leary 1989). Underlying the stalemate was the unresolved national
self-determination dispute.
The third type of conflict management, ‘inter-governmentalism and coercive

consociation’, emerged in 1985 (O’Leary 1987). It established the institutional and
policy environment that made the complex-power sharing agreement of 1998
possible. In the Anglo-Irish Agreement, the two sovereign states created an inter-
governmental conference which gave the Irish government the right of regular
consultation and deliberation over all public policy affecting Northern Ireland
(McGarry and O’Leary 1990b Appendix A; Hadden and Boyle 1989). They did so
for several reasons. They wanted to stop the growth in support for Sinn Féin, and
to create appropriate incentives for moderate nationalists and unionists to share
power. The Agreement made it easier for the SDLP to negotiate an internal agree-
ment because an Irish dimension was in place. The Agreement was, however,
coercively consociational toward unionists. They were told (Article 4) that if a
power-sharing devolved government was established then in the devolved func-
tions, there would be no role for the Intergovernmental Conference. Unionists
responded with widespread hostility to the Agreement, but failed to bring down
its key institution. They sought the termination and then the suspension of the
Intergovernmental Conference and its Secretariat before they would consider en-
tering negotiations on an agreed settlement. Eventually, such negotiations began
in 1991-1992, though without immediate success. Nevertheless, the processes at-
tached to the negotiations, and the substantive content of the future agreement,
were foreshadowed in these discussions (O’Leary 1995b; 1995c; 1995a).
II The Making of the Belfast and the British-Irish Agreements

The two sovereign governments had gradually improved their cooperation, de-
spite regular difficulties (especially during Margaret Thatcher’s long premiership
(1979–1990). Their shared interests in stability and a political settlement provid-
ed incentives to consider a comprehensive settlement. But, until 1991–1992 in-
ter‑governmental initiatives were premised upon building settlements with mod-
erates and marginalizing hardliners. In the 1990s, this environment was changed,
largely at the initiative of nationalists. Isolated by the Anglo-Irish Agreement,
Professor Paul Bew claims to have refuted this argument through ‘earlier research’
that he published with others (Bew 2000: 40; citing Bew, Gibbon, and Patterson
1996: Ch 6, 213-217). The cited sections provide no obvious refutation. Bew thinks
the 1985 Agreement had as its logical agenda joint authority, in which lip-service was
paid to devolution (for which he cites one British official), whereas in fact the 1985
Agreement was the result of; (a) the failure of the Irish government to achieve joint
authority; and (b) a joint determination of both British and Irish officials to force
unionists to choose between power-sharing devolution (consociation) and increas-
ing Irish governmental influence in the North (de facto joint authority). The officials
I interviewed in 1986 (British and Irish), confirmed my interpretation, though of
course they did not use the terminology of consociation. Bew and I appear to share a
common understanding of Northern Ireland’s recent institutional trajectory, from a
system of ‘control’ (which he calls ‘ethnic democracy’) to one of (attempted) conso-
ciation (Bew 2000: 48).
Sinn Féin sought to build a pan-nationalist alliance, and put out feelers to the
Irish government, the SDLP and the Catholic Church. Determined to encourage
the politicization of republicans, and to have negotiations in which all national-
ists could be present, the SDLP leader John Hume argued that Irish national self-
determination could be exercised through two referendums, North and South, to
endorse a comprehensive agreement that addressed all the relevant relationships
within the North; between the North and the South; and between the East and
the West, i.e., within Northern Ireland, across Ireland, and between Ireland and
the UK. This agenda would become Strand One, Strand Two, and Strand Three,
respectively, of the 1991–1992 negotiations, with separate chairs for each strand
of negotiations, including an international chair for Strand Two, and an agree-
ment that “nothing would be agreed until everything would be agreed.”
Republicans and nationalists built support for peacemaking and a negotiated
settlement amongst influential Irish Americans, including key figures in the Clin-
ton administration (O’Clery 1997 (1996)). Secret negotiations, denied at the time,
and misrepresented later, took place between the IRA and the UK government.
Republicans sought to ensure that an IRA ceasefire would quickly lead to Sinn
Féin’s inclusion in free-ranging negotiations. A joint prime ministerial statement,
the Downing Street Declaration of December 1993, issued by Taoiseach Albert
Reynolds and Prime Minister John Major, outlined the agreed principles to gov-
ern a negotiated settlement. They contained Hume’s agenda, enabled each party
with a democratic mandate and a peaceful agenda to negotiate for its prefer-
ences, but assured unionists that agreement would require the support of a ma-
jority within Northern Ireland. In August 1994, after a delayed response, the IRA
ceased all its ‘military operations’. Six weeks later, the major loyalist paramilitaries
reciprocated, promising a ‘no first strike policy’.
In February 1995, the two governments published two sets of documents. The
first, based on Strand One of the failed 1991–1992 negotiations, set out an agenda
for a devolved government in Northern Ireland, splitting the differences between
the known preferences of the SDLP and the UUP. The second, written by the two
governments, and covering Strands Two and Three of the previous failed negotia-
tions, set out an agenda for the creation of a strong multi-functional North-South
Ministerial Council across Ireland. They also indicated how the UK and Ireland
might resolve their dispute over sovereignty in the region. The ideas in the docu-
ments were consociational and confederal in character and had an implicit mod-
el for the double protection of the relevant majorities and minorities (O’Leary
1995b; 1995a). They anticipated much of what would be agreed in 1998.
In February 1996, the IRA ended its ceasefire. It did so, it said, in response to
the UK government’s refusal to convene promised comprehensive all-party nego-
tiations. The UK government, hamstrung by a diminishing parliamentary major-
ity, had insisted that decommissioning of weapons by paramilitaries precede the
entry of their parties into negotiations – an insistence known as ‘Washington 3’,
after its elaboration among other preconditions by the then Secretary of State Sir
Patrick Mayhew at a conference in the USA. The insistence on prior decommis-
sioning was motivated by a desire to get unionists to engage in negotiations, and
to appease Tory backbenchers, hostile to Irish republicanism. To resolve the im-

passe, the two governments requested US Senator George Mitchell to head an in-
ternational commission. It reported in January 1996 (Mitchell, de Chastelain, and
Holkeri 1996). It recommended that decommissioning take place during and par-
allel to negotiations, but that parties be required to sign up to exclusively peaceful
and democratic principles in and outside negotiations. In response to the Mitch-
ell Commission, Prime Minister Major proposed either prior decommissioning,
as before, or calling an election for a Peace Forum to mandate negotiators. This
was the immediate context in which the IRA broke its ceasefire. Major’s policy
seemed to play fast and loose with Mitchell’s recommendations, and to legitimate
a Northern Assembly before negotiations. Nevertheless, Sinn Féin participated
in the elections to the Forum. Elections to the Forum took place, and negotiations
about negotiations began, without Sinn Féin, but with the Mitchell Commission-
ers chairing them. Progress was tortuous. One innovation emerged: a unionist
proposal for a Council of the Isles to which the North-South Ministerial Council
would be subordinate.
The glacially paced negotiations were interrupted by the UK general election
of 1997, which brought a new Labour government to power with an overwhelm-
ing parliamentary majority. The Blair government sought and obtained a renewal
of the IRA’s ceasefire, and parked the issue of decommissioning into a separate
forum for negotiations. As Sinn Féin entered the negotiations, hardline union-
ists in the Democratic Unionist Party (DUP) and other parties left, accusing the
Labour government of appeasing terrorists. Negotiations nevertheless proved
agonisingly slow until the spring of 1998. There were negotiations about negotia-
tions – where the negotiations would take place, who would chair them, and what
procedures would be followed. Negotiations were interrupted by walk-outs and
intermittent suspensions of parties because of breaches in ceasefires. It had been
agreed, however, to proceed by “sufficient consensus”, following the model of the
South African peace process, the negotiating rule was that whatever was sup-
ported by a majority of nationalist and unionist negotiators would become part
of the Agreement. In the end, the two governments and Senator Mitchell created
a pressure-cooker atmosphere, with a specified deadline, to finalize a settlement
(Mitchell 2000).
Full multilateral negotiation on everything did not occur. The UUP and the
SDLP largely negotiated the details of internal power-sharing. The two govern-
ments and the SDLP and the UUP made the running in designing the North-
South Ministerial Council and the British-Irish Council. The two governments
agreed how they would handle their respective commitments to constitutional
change. The UK and Irish governments negotiated with Sinn Féin and the loyal-
ist parties, the PUP and the UDP, over arrangements for prisoner release pro-
grammes, decommissioning, and demilitarization. NGOs and others were con-
sulted over proposals made to address questions of human rights protection, and
a range of equality provisions. The administration of justice and policing arrange-
ments were left to special commissions to be appointed by the UK government,
in consultation with the Irish government. After several sleepless nights, on 10
April 1998 a text was not opposed by the negotiators from eight of the ten par-
ties elected to the Forum. Later, the text won the official endorsement of the two
nationalist parties, the moderate ‘others’, and a majority of the unionist represen-
tatives. In a referendum a month later, the Agreement was endorsed in simulta-
neous referendums by 71% of those voting in Northern Ireland, and 95% of those
voting in the Republic of Ireland.
III Complex Power-sharing: The Belfast or Good Friday Agreement and

The British-Irish Agreement
A The Name(s) of the Agreement(s)

The Agreement of 10 April 1998 is called the ‘Belfast Agreement’ by the UK gov-
ernment. Its most popular name, especially amongst nationalists, is the ‘Good
Friday Agreement’ – because it was finalized on the alleged anniversary of
Christ’s crucifixion. This name gives too much credit to Christianity, both as the
key source of conflict and as a source of resolution. It might best be called the
‘British-Irish Agreement’, because it fulfils and supersedes its predecessor, the
Anglo-Irish Agreement of 1985, but that name is now given to the 1999 intergov-
ernmental treaty that incorporates the 1998 Agreement. From, here reference is
made simply to ‘the Agreement’, distinguishing where necessary, the negotiated
text, the UK’s legislative enactment, the Northern Ireland Act 1998, and the trea-
ty, even though the three documents might be considered separate agreements.
B An Internal Consociation
The Agreement is consociational, i.e., it meets all the criteria: (1) cross-commu-
nity executive power-sharing; (2) proportionality rules throughout the govern-
mental and public sectors; (3) community self-government – or autonomy – and
equality in cultural life; and (4) veto rights for minorities (Lijphart 1977). A conso-
ciation may be built without any explicit theory to guide it. Most often consocia-
The conflict is primarily ethno-national rather than religious, contrary to the percep-
tions of many outsiders (McGarry and O’Leary 1995a: Ch 5 & 6; 1995b).
References to the negotiated text are to The Agreement: Agreement reached in the
multi-party negotiations (No place of publication, No date, UK Government), 30,
distributed to all households in Northern Ireland before the May 1998 referendum.
Cross-community executive power-sharing (rather than the rarer case of a compre-
hensive grand coalition) is the necessary hallmark of a full consociation (O’Leary
2005).
Lijphart claims that consociational rules were invented by Dutch politicians in 1917,
and by their Lebanese (1943), Austrian (1945), Malaysian (1955), Colombian (1958),
and Indian – in the 1960s – and South African (1993–1994) counterparts later in the
century. One does not have to agree with the citation of any of these cases to accept
tions are the outcomes of pacts between politicians; this Agreement was also the
product of tacit and explicit consociational thought.
But the Agreement was not just consociational. It has important external di-
mensions that reflect its status as a national self-determination settlement. It was
made with the leaders of national, and not just ethnic or religious, communities;
and majorities in two simultaneous and separate referendums endorsed it. It is the
first consociation endorsed in referendums that required concurrent consent in
jurisdictions in different states. The consociation was designed within overarch-
ing confederal and federalizing institutions, has elements of co-sovereignty in the
arrangements and oversight agreed between its patron states, and it promises a
novel model of ‘double protection’. These propositions are elaborated below.
1 Competencies and the Division of Functions

The Agreement proposed, and the 1998 Northern Ireland Act established, a sin-
gle-chamber Northern Ireland Assembly and an Executive. The Assembly and
Executive have full legislative and executive competence for economic develop-
ment, education, health and social services, agriculture, environment, and fi-
nance, including the local civil service. The Assembly may not legislate in contra-
vention of the European Convention on Human Rights or European Union law,
modify a specific entrenched enactment, discriminate on grounds of religious
belief or political opinion, or ‘deal with’ an excepted power except in an ‘ancillary
way’ – which roughly means that it may not enact laws that modify UK statutes
on excepted matters, such as the Crown. The internal security functions of the
state – policing and the courts – are retained by Westminster for now, but may be
devolved in principle (the meaning of ‘devolution’ is addressed below), and there
are proposals now published for the transfer of policing competence.
Through ‘cross-community agreement’ – defined below – the Assembly may
expand these competencies; and, again through such agreement and with the
consent of the UK Secretary of State for Northern Ireland and the Westminster
parliament, it may legislate for any currently non-devolved reserved function.
Within a traditional UK constitutional perspective, maximum feasible devolved
self-government is therefore within the scope of the local decision-makers. Ac-
cording to the Northern Ireland Act 1998, the Assembly can expand its autono-
that politicians are very capable of doing theory without the aid of theorists (Lijphart
1990b: viii; 1996).
Consociational thinking was reflected in the drafting of the Framework Documents
of 1995 (O’Leary 1995a), and the ‘novel’ executive design in the Agreement, based on
the d’Hondt rule, and reflected coalition principles used elsewhere in Europe and in
the European Parliament (O’Leary, Grofman, and Elklit 2001). Consociational think-
ing had had local resonance since the Sunningdale Agreement. The SDLP was espe-
cially interested in power-sharing devices, and was the prime initiator of proposals
in the internal negotiations (O’Leary 1999e; 1999d; 1999a; Horowitz 2001). Dr Mow-
lam, the UK Secretary of State for Northern Ireland in 1997–1999, had an academic
consociational heritage, and at least one of her former academic advisers has had an
abiding interest in the subject.
my only with regard to reserved, not excepted, matters. Reserved matters, most
importantly, include the criminal law, criminal justice, and policing. Excepted
matters include the Crown and the currency. A convention may arise in future
in which the Secretary of State and Westminster ‘rubber stamp’ the legislative
measures of the Assembly. Indeed, if the Agreement is fully implemented, most
public policy in Ireland, North and South, may eventually be made without direct
British ministerial involvement, though the British budgetary allocation will be
crucial as long as Northern Ireland remains in the UK.
2 Assembly Rules and Procedures

The Agreement required the 108 elected Assembly members (MLAs) to des-
ignate themselves as ‘nationalist’, ‘unionist’, or ‘other’. In this respect, Lijphart’s
injunctions in favour of “self-determination rather than pre-determination” are
violated (Lijphart 1985: 108ff; 1995). After the first Assembly was elected in June
1998, this requirement posed difficult questions for the Alliance Party (APNI)
and other ‘cross-community’ parties, such as the Women’s Coalition (NIWC),
who have both cultural Catholic and cultural Protestant leaders and voters. They
determined that they were ‘others’, though they were free to change their classifi-
cations once in each Assembly, and, of course, in future Assemblies.
Through standard legislative majority rule, the Assembly may pass ‘normal
laws’ within its devolved powers, though there is provision – the petition proce-
dure – for 30 of the 108 MLAs to trigger special procedures that require special
majorities. But ‘key decisions’ – that is, the passage of controversial legislation,
including the budget – automatically have these special procedures that require
‘cross-community’ support. Two rules have been designed for this support and
flowed from the negotiating procedures used to make the Agreement. The first
is ‘parallel consent’, a majority that encompasses a strict concurrent majority of
registered nationalists and unionists. It requires that a law be endorsed, among
those present and voting, both by an overall majority of MLAs and by majorities
of both its unionist and its nationalist members respectively. The second is the
‘weighted majority’ rule, described below. The UK legislation implied that the
parallel consent procedure must be attempted first, followed by the weighted ma-
jority procedure, though in the original Agreement the election of the premiers
may only be effected by the parallel consent rule (see below).
Table 1 records the numbers in each bloc returned in the June 1998 and Novem-
ber 2003 Assembly elections. Parallel consent with all members present requires
in both Assemblies the support of twenty-two nationalists and thirty unionists,
as well as an overall majority in the Assembly. Under parallel consent laws may
pass dependent upon the support of the ‘others’ – twenty-two nationalists, thirty
unionists, and three others enable the passage of a key decision. In short, it does
not automatically render the ‘others’ unimportant, though plainly their position
is not as pivotal as that of nationalists or unionists. The ‘weighted majority’ rule
requires, among those present and voting, that to become law a measure must
have the support of 60% of members. But it also requires the support of 40%
of registered nationalist members and 40% of unionist members, i.e., in both
Table 1 Bloc Performances in the June 1998 and November 2003 Northern
Ireland Assembly Elections
Bloc V Seats S V–S
(1st preference vote) (seats
% %)
Nationalists 1998 38.8 42 38.9 -0.1
2003 40.8* 42 38.9 -1.9
Unionists 1998 50.5 58 53.7 +3.2
2003 51.3+ 59 54.6 +3.3
‘Yes’ Unionists 1998 25.0 30 27.8 +2.8
2003 23.9 28 25.9 +2.0
‘No’ Unionists 1998 25.5 28 25.9 +0.4
2003 27.4 31 28.7 +1.3
Others 1998 8.4 8 7.4 -1.0
2003 7.9 7 6.5 +1.4
Source: O’Leary 1999d updated

Notes: * includes Eamon McCann as a Socialist candidate in Foyle; + includes the Con-
servatives.
The whole of the UUP in 1998 and 2003 is counted as ‘Yes Unionist’ because the party’s of-
ficial platform was pro-Agreement.
Assemblies at least seventeen nationalists and at least twenty-four unionists had

to consent. All nationalists (forty-two) and the minimum necessary number of
unionists (twenty-four) had the combined support for any measure to pass in
this way – without support from the ‘others’. By contrast, combinations of all the
others (eight or seven) and the minimum number of nationalists (seventeen) and
the minimum number of unionists (twenty-four) could not deliver a majority, let
alone a weighted majority. The operation of the rules obviously depend not just
on how parties register, but also on their internal discipline within the Assem-
bly. The outcome of the 1998 elections suggested that pro-Agreement unionist
MLAs (thirty) would be vulnerable to pressure from anti-Agreement unionists
(twenty-eight). Within the ranks of the pro-Agreement UUP were MLAs who
were de facto anti-Agreement. Indeed, one UUP MLA subsequently resigned
his party’s whip and later became a member of the DUP. But his rebellion left
room for the Agreement to function. The UUP could deliver a workable por-
tion of a cross-community majority under the weighted majority rule, even with
six dissidents, provided that party leader David Trimble could rely on the two
pro-Agreement Progressive Unionist Party (PUP) MLAs. The relative weakness
of the pro-Agreement unionists would prove very damaging to the stabilization
of the Agreement. In the 2003 elections, the numbers of formally pro- (twenty-
There is one important exception to this possibility: the death or the resignation of ei-
ther premier required that both be replaced under the parallel consent rule, see below.
eight) and anti-Agreement (thirty-one) unionists were almost exactly reversed

– though the UUP once again contained anti-Agreement MLAs within its ranks.
At the time of writing, it was unclear whether the strength of the DUP would
terminate the Agreement, lead to its re-negotiation (what it had campaigned for),
or generate a fresh intra-Agreement compromise, led by the more robust repre-
sentatives of each national tradition.
There is one additional ‘supermajority’ rule. The Assembly may, by a two-
thirds resolution of its membership, call an extraordinary general election before
its statutory four-year term expires. This rule was agreed by the parties, after the
Agreement, in preference to a proposal that the UK Secretary of State should
have the power to dissolve or suspend the Assembly – a sign of the local parties’
commitment to their self-government. Subsequently, to suspend the Assembly
in February 2000, the Secretary of State for Northern Ireland, Peter Mandelson,
had to pass new primary UK legislation, the Northern Ireland Act 2000, through
the Westminster parliament, and outside the remit of the Agreement – which
is why Irish nationalists and others regarded the suspension as a breach of the
Agreement, and indeed of the 1999 intergovernmental treaty (O’Leary 2000d;
and see below).
3 Executive Power-sharing: A Dual Premiership and D’Hondt Executive

The Agreement established a novel Executive Committee, and, at its head, two
quasi-presidential figures, a diarchy: a First Minister (FM) and a Deputy First
Minister (DFM), who are equal in powers. Once elected, this diarchy have presi-
dential characteristics, because it is almost impossible to depose them, provid-
ed they remain united as a team, until the next general election. The essence of
‘presidentialism’ is an executive that cannot be destroyed by an assembly except
through impeachment. The FM and DFM are elected together by the parallel
consent procedure. This rule gives very strong incentives to unionists and na-
tionalists to nominate a candidate for one of these positions that is acceptable to
a majority of the other bloc’s Assembly members. In the first elections for these
posts, in designate form, pro-Agreement unionists in the UUP and the PUP voted
solidly for the combination of David Trimble of the UUP and Seamus Mallon of
the SDLP. Naturally, so did the SDLP, which enjoyed a majority among registered
nationalists. The ‘No’ unionists voted against this combination, while Sinn Féin
abstained. The rule ensures, though it does not officially require, that a unionist
and a nationalist share the top two posts: it does not specify which must be First
Minister. The Agreement and the Northern Ireland Act 1998 make clear that the
two posts have identical symbolic and external representation functions. In the
negotiations, the SDLP conceded the difference in dignity in title between the
positions but successfully insisted on no differences in their respective powers
(confidential sources). The sole difference between the premiers is their titles:
both preside over the Executive Committee of Ministers and have a role in coor-
dinating its work. Clause 15 (10) of the Northern Ireland Act 1998 also enables the
top two ministers to hold functional portfolios. This dual premiership critically
depends upon the cooperation of the two office-holders and upon the coopera-
tion of their respective majorities – or pluralities under the weighted majority

rule. Article 16(7) of the Northern Ireland Act 1998 reinforced their interdepen-
dence by requiring that “if either the First Minister or the deputy First Minister
ceases to hold office, whether by resignation or otherwise, the other shall also
cease to hold office.” The Act also specified in Article 32(3) that if the Assembly
fails to elect two premiers within six weeks that “the Secretary of State shall pro-
pose a date for the poll for the election of the next Assembly.” The outcome of the
November 2003 Assembly elections produced a scenario that many had foreseen,
namely majorities for Sinn Féin and the DUP within the nationalist and unionist
blocs respectively, and with the DUP formally determined not to share the pre-
mierships with Sinn Féin (though Sinn Féin is fully prepared to share power with
the DUP). At the time of writing, it was unclear how this crisis would be resolved
– though technically Article 32(3) requires fresh elections within six weeks if the
DUP refuses to share the premiership with Sinn Féin (or a Sinn Féin-endorsed
SDLP premier).
The formation of the rest of the Executive Committee, according to the proce-
dure described below, did not go smoothly in the first Assembly. In the summer
of 1999, Seamus Mallon resigned as Deputy First Minister (designate), complain-
ing that the UUP was “dishonouring” the Agreement and “insulting its princi-
ples” by insisting upon the decommissioning of paramilitaries’ weapons before
executive formation.10 He did so to speed an intergovernmental review of the
implementation of the Agreement. The question arose, did Mallon’s resignation
automatically trigger Trimble’s departure from office and require fresh elections
to these positions within six weeks (a question rendered critical by the defection
of the UUP MLA mentioned above)? The (Initial) Presiding Officer’s answer to
this question was that it did not, because the Assembly was not yet functioning
under the Northern Ireland Act.11 This answer was accepted. It implied that if the
review of the Agreement succeeded and the Agreement’s institutions came into
force, either that there would have to be fresh elections of the FM and DFM under
the parallel consent rule or that Mallon’s resignation would have to be rescinded.
When the review succeeded and the Agreement’s institutions came on line, the
Assembly adopted the expedient of voting to nullify Mallon’s resignation, there-
10 See statement by the Deputy First Minister (Designate), Northern Ireland Assembly,
15 July (1999: 325).
11 “Members will recall that the First Minister (Designate) and the Deputy First Min-
ister (Designate) were elected, and I use the common parlance, ‘on a slate’, when we
were in a post-devolution situation. That means that under the Northern Ireland Act,
both positions would fall when one resigned, but the remaining individual would
remain in a caretaker capacity for up to six weeks. Before the end of that period the
Presiding Officer would call for a further election. However, we are still functioning
under the Northern Ireland (Elections) Act for these purposes and, therefore, the
position of the First Minister (designate), as I understand it – and you have simply
asked me for an immediate view – is unchanged. It is possible that some Standing
Order, or other arrangement, may already be on the way, but I have no knowledge of
it.” (Northern Ireland Assembly 1999: 326–327).
by preventing a vote under the parallel consent rule that might have prevented
Trimble’s and Mallon’s (re)installation in office. Later, it would not prove so easy
to elect the premiers when Mark Durkan replaced Seamus Mallon as the nomi-
nee of the SDLP.
The premiers have implicit and explicit coordinating executive functions, as
approved by the Shadow Assembly in February 1999 (Wilford 2001). The Depart-
ment of the First and Deputy First Ministers was created to fulfill them. It has an
Economic Policy Unit and an Equality Unit, and is tasked with liaising with the
other institutions of the Agreement, namely, the North-South Ministerial Coun-
cil, the British-Irish Council, the Secretary of State on reserved and excepted UK
powers, and EU/international matters, and, of course, with cross-departmental
coordination. Unlike executive presidencies, and unlike most prime ministers,
neither the FM nor the DFM formally appoints the other ministers to the Ex-
ecutive Committee. Instead, posts in the Executive Committee, or cabinet, are
allocated to parties in proportion to their strength in the Assembly, according to
the d’Hondt rule (O’Leary, Grofman and Elklit 2001). Under this rule, posts in
the rest of the Executive Committee are allocated to parties in proportion to their
strength in the Assembly. The rule’s consequences are clear: any party that wins a
significant share of seats and is willing to abide by the new institutional rules has
a reasonable chance of access to the executive, a subtly inclusive form of executive
formation, one that permits but does not mandate a ‘grand coalition government’.
It is a voluntary arrangement because parties are free to exclude themselves from
the Executive Committee. No programme of government has to be negotiated
in advance between the parties entitled to portfolios. The design, in principle,
creates strong incentives for parties to take up their entitlements to ministries,
because, if they do not, then the portfolios go either to their ethno-national rivals,
or to their rivals in their own bloc.12 The d’Hondt allocation procedure means
12 The rules of executive formation do not require any specific proportion of national-
ists and unionists. In the course of the crisis over executive formation in 1999, Dr
Mowlam introduced a new rule requiring that a well-formed executive consist of at
least three designated nationalists and three designated unionists. On 15 July 1999, in
a hand-written note to the Initial Presiding Officer, she added an additional Stand-
ing Order to the running of d’Hondt: “On the completion of the procedure for the
appointment of Ministers (designate) under this Standing Order, the persons ap-
pointed shall only continue to hold Ministerial office (designate) if they include at
least 3 designated Nationalists and 3 designated Unionists.” This order, authorized
under the Northern Ireland (Elections) Act 1998, in my view, was the first technical
breach of the letter of the Agreement. Given that the parties had previously agreed
that the executive should consist of ten ministers, in addition to the First and Deputy
First Ministers, the standing order gave a veto power to the UUP (and the SDLP)
over executive formation, because each party was entitled to three seats on the basis
of its strength in seats won. The standing order was introduced in a hurry to stop a
running of the procedure for executive formation leading either to an all-nationalist
executive, as actually transpired – given the decision of the UUP to fail to turn up to
the Assembly when the process was triggered and the decision of the ‘No’ unionists
that democratic parties get the absolute right to nominate ministers according
to their respective strength in seats; that is, no vote of confidence is required
by the Assembly either for individual ministers or for the Executive Committee
as a whole. A party as a whole may be excluded from a right to nominate if it is
deemed by the Assembly, through cross-community consent procedures, to be in
breach of the requirements of the Pledge of Office (see below). Efforts by unionist
MLAs to have Sinn Féin so deemed foundered in the first Assembly because the
moderate nationalists in the SDLP did not support them. Parties choose, in order
of their strength, their preferred ministries – which leads to fascinating strategic
decision-making (O’Leary, Grofman, and Elklit 2001). The Assembly under the
cross-community rules may depose an individual minister from office, but the
party that held the relevant ministry is entitled to appoint his or her successor
from amongst its ranks. In the course of 2000 the anti-Agreement DUP decided
to take advantage of this provision to rotate some of its MLAs through its two
ministerial portfolios. Its critics observed that they did not, however, resign their
entitlements to the two ministries.
Crises over executive formation and maintenance have been the major sig-
nals that the Agreement might fail. The first crisis arose for political and con-
stitutional reasons. Politically, because Trimble insisted that the IRA organize
some decommissioning of its weapons before Sinn Féin members could take their
ministries in the Executive Committee. ‘No government before guns’ became his
catchphrase. Otherwise, he threatened he would refuse to cooperate in the run-
ning of the d’Hondt procedure. Constitutionally, that is, under the text of the
Agreement, Trimble had no warrant to exercise this veto. No party is entitled
to veto another party’s membership of the Executive, though the Assembly as a
whole, through cross-community consent, may deem a party unfit for office; the
Agreement did not specify a starting date for decommissioning, though it did
require parties to use their best endeavours to achieve its completion within two
years of the referendum, that is, by 22 May 2000; and any ‘natural’ reading of the
text of the Agreement mandated executive formation as a necessary step to bring
all the Agreement’s institutions ‘on line’. Trimble rested his case on a communica-
tion he had received from the UK prime minister on the morning the Agreement
was made, indicating that it was Blair’s view that decommissioning “should begin
straight away.” Communications of opinions from UK premiers do not, of course,
have the force of law and the ‘should’ in Blair’s text was in the subtle subjunc-
not to take their ministerial entitlements – or to an executive in which there would

have been no pro-Agreement unionists (Northern Ireland Assembly 1999: 317). This
panic measure subtly changed the executive’s incentive structures as previously
agreed in the negotiation of the Agreement. It was consociational in spirit, but it was
not negotiated by the parties, was not endorsed in the referendums, and encouraged
moderates to over-bargain, knowing that they could veto executive formation. The
Standing Order no longer has force, so by implication, in future any boycott of the
executive by a party entitled to representation should lead to the re-allocation of its
ministries.
tive mood rather than a mandatory reading of the Agreement. Trimble’s concern
was to appease critics of the Agreement within his own party. His negotiating
team had split, with one of his Westminster MPs, Jeffrey Donaldson, walking
out on the morning of 10 April – he would later join the DUP; a majority of his
party’s Westminster MPs opposed the Agreement; and his new Assembly party
contained critics of aspects of the Agreement. The UK and Irish governments,
sympathetic to his exposed position, initially facilitated Trimble in exercising his
veto; and he took advantage of the fact that the SDLP did not make the forma-
tion of the rest of the executive a precondition of its support for the Trimble-
Mallon ticket for FM and DFM. One provision in the Agreement gave Trimble
further room for manoeuvre. The Agreement implied that there would be at least
six other Ministers apart from the premiers, but that there could be “up to” ten
(Government of the UK 1998: Strand One, paras. 14 (explicitly) and 3 (implicitly)).
The number of ministries was to be decided by MLAs through cross-community
consent, and that gave Trimble the opportunity to delay executive formation. It
would be December 1998 before the parties reached agreement on ten ministries,
when the UUP finally abandoned its demand for seven rather than ten depart-
mental ministries: with seven, unionists would have had an overall majority in
the Executive Committee (for details of the ministries, see Table 3).
Most of 1999 saw protracted bargaining, including a failed running of the
d’Hondt procedure to fill the executive in July, but no consensus on proceeding to
formation. Mallon’s resignation triggered a review of the Agreement, as permit-
ted by its terms, under US Senator George Mitchell. In mid-November, the crisis
looked as if, in principle, it would be resolved. The UUP accepted that executive
formation would occur – with the IRA appointing an interlocutor to negotiate
with the International Commission on Decommissioning – while actual arms de-
commissioning, consistent with the text of the Agreement, would not be required
until after executive formation. In concluding his review, and with the consent of
the pro-Agreement parties, Senator Mitchell stated that: “Devolution should take
effect, then the executive should meet, and then the paramilitary groups should
appoint their authorized representatives, all on the same day, in that order.” This
appeared an honourable resolution to what appeared a fundamental impasse.
The d’Hondt procedure was followed, and Northern Ireland had its novel power-
sharing Executive Committee – though the Ulster Unionist Council of the UUP
would later render problematic this settlement within the settlement.
Viktor d’Hondt devised a method of proportional representation used for
many purposes, including allocating political offices in the European Parliament.
The method works by iteration, using a simple series of divisors, 1, 2, 3, … n, that
are divided into a party’s share of votes or seats. Tables 2 and 3 show how the
allocation worked for the Northern Ireland Executive Committee in 1999. The
seats won by the largest political parties and the order in which ministries were
allocated are displayed in Table 2. (The principal change from Table 1 is that the
UUP had lost one member to the ‘No’ unionists, reducing the party’s member-
ship in the Assembly to twenty-seven).
Table 2 The Allocation of Ministerial Portfolios by d’Hondt, 1999

UUP DUP SDLP SF
Divisor S M S M S M S M
1 27 (1) 20 (3) 24 (2) 18 (4)
2 14 (5) 10 (7) 12 (6) 9 (9)
3 9 (8) 6.6 8 (10) 6
All M 3 2 3 2
The numbers in brackets in the M columns indicate the order in which parties
won ministries of their choice; S is the number of seats each party has during
each stage of the allocation.
In 1999, all parties entitled to ministries were willing to take them. The party
with the largest number of seats, the UUP, obtained the first ministry, and then
its seat share was divided by two, leaving it with 13.5. The next largest remaining
number of seats was held by the SDLP; it chose the second ministry, and its seat
share was divided by two, leaving it with 12. The next largest remaining number
of seats was held by the DUP; it chose the third ministry, and its seat share was
divided by two, leaving it with 10. The next largest remaining number of seats was
held by Sinn Féin; it chose the fourth ministry, and its seat share was divided by
two, leaving it with 9. The next largest remaining number of seats was the UUP,
with 13.5; it chose the fifth ministry, and its total seat share was divided by three,
leaving it with 9. And so on. Great foresight was shown in the legislative enact-
ment of the Agreement: where there is a tie in the number of seats held by parties
during any stage of the allocation, precedence is given to the party with the high-
er share of the first-preference vote. The tie-breaker was required at stage 8, when
both the UUP – 27/3 seats – and Sinn Féin – 18/2 seats – had a remaining seat
total of 9. In accordance with the rule, the UUP was given precedence in portfolio
choice. Unionists therefore obtained five ministries (three UUP and two DUP )
and nationalists obtained five (three SDLP and two SF), a mild disproportionality
by bloc, but not by party. After the 2003 elections, by contrast, unionists were
entitled to six ministries, and nationalists four, a result that is more proportional
by bloc as well as party.
In 1999, unionists did not fare as well as nationalists in strategic decision-mak-
ing over portfolio allocation. Nationalists obtained almost the entire welfare state
portfolio – education at all levels, health and social services, and agriculture – as
well as finance and personnel. What happened? Table 3 shows the actual portfo-
lios chosen by parties at each stage in the allocation.
Table 3 Party Choices of Ministerial Portfolios, 1999

Portfolio Nominee Party
1. Enterprise, Trade and Investment Empey UUP
2. Finance and Personnel Durkan SDLP
3. Regional Development Robinson DUP
4. Education McGuinness SF
5. Environment Foster UUP
6. Higher and Further Education, Training Farren SDLP
and Development
7. Social Development Dodds DUP
8. Culture, Art and Leisure McGimpsey UUP
9. Health, Social Services and Public Safety de Brun SF
10. Agriculture Rogers SDLP
There was no coordination between the UUP and the DUP, or between the SDLP
and Sinn Féin, but, since the negotiation of the Agreement, relations between the
SDLP and Sinn Féin have been more amicable than those between the UUP and
the DUP. The UUP did not, as expected, take the Finance and Personnel portfolio
with its first choice. This decision may have been affected by Empey’s own pref-
erences – as Trimble’s right-hand man during the negotiations he may have had
a free hand – or by the fact that Enterprise, Trade and Investment would mesh
well with the Economic Policy Unit in the First and Deputy First Ministers’ Of-
fice. More likely, the UUP may have calculated that it would be best to give the
SDLP the lead negotiation and arbitration role over the budget, knowing that
the SDLP would be better able to face down Sinn Féin’s suggestions. The SDLP’s
choice of Finance and Personnel was no surprise given that the post was avail-
able. The DUP then had the choice of the third ministry. The party had decided
to take office while refusing to interact with Sinn Féin ministers. It appeared to be
seeking to wreck the Agreement from within, while obtaining some of the perks
of office to which it was entitled. Its leader, Ian Paisley, nominated his deputy
leader, Robinson, to the Regional Development portfolio, consisting largely of
transport. Whether this was to prevent vigourous cross-border initiatives in this
area, or because the ministry could offer patronage and pork barrel opportunities
is not known. This choice did, however, leave Sinn Féin free to pick the Educa-
tion portfolio, which plainly shocked many unionist Assembly members. Sinn
Féin’s choice made strategic sense for a radical nationalist party. The education
ministry gave it access to a high-profile, big-spending, potentially redistributive
and socializing ministry. The UUP then chose Environment, and the party leader
nominated a relative unknown, Foster, to the portfolio. The ministry contains Lo-
cal Government within its remit and Foster is a local government councillor – he
was being rewarded by his party leader for delivering him crucial support. The
choice may partly have been motivated by a desire to block unwelcome changes
that might be proposed for local government. Whatever its rationale, it left the
SDLP free to pick the Higher and Further Education, Training and Development
portfolio. The DUP then claimed that the UUP’s decisions had left nationalists
in full control of education throughout Northern Ireland – this was not true be-
cause of the checks and balances in the Assembly, and in any case the DUP could
have picked either education portfolio for itself, but did not do so. It is not known
whether it wanted to benefit from unionists’ anxieties over nationalists’ grip on
the education portfolios – the Machiavellian view – or whether it prioritized oth-
er matters. The DUP then chose the Social Development portfolio, a choice that
left the UUP with a major headache. If it took either of Health, Social Services
and Public Safety or Agriculture then it would leave one nationalist in charge of
the Ministry of Culture, Arts and Leisure, with its potential agenda-setting con-
trol over items such as parades and binational and bilingual matters. The UUP
chose to sacrifice access to a big-spending ministry for this reason. Sinn Féin and
the SDLP then took the remaining portfolios, appointing women to the last two
ministries in a display of progressive politics.
This story suggests some major ideas for power-sharing. The d’Hondt rule –
and variations on it, such as a Saint-Laguë rule – is an efficient way of solving co-
alition-making problems, one that saves on the transactions costs of bargaining
over coalition government. The Northern Ireland Act stops parties from forming
post-election coalition pacts for the purpose of improving on their total number
of portfolios and the pecking order in which they receive ministries, but plainly
in other systems such pacting might be permitted.
The consociational criterion of cross-community executive power-sharing is
clearly met in the negotiated Agreement and its UK legislative enactment, but
there are special features of the arrangements that differ from other consocia-
tional experiments. Ministers take a ‘Pledge of Office’, not an ‘Oath of Allegiance’.
This cements what nationalists see as the binationalism at the heart of the Agree-
ment: nationalist ministers do not have to swear an Oath of Allegiance to the
Crown or the Union. The Pledge requires ministers to:
• discharge their duties in good faith;
• follow exclusively peaceful and democratic politics;
• participate in preparing a programme of government; and
• support and follow the decisions of the Executive Committee and the As-
sembly.
The duties of office include a requirement to serve all the people equally, to pro-
mote equality, and to prevent discrimination – which means, according to the
UK’s doctrine of ministerial responsibility, that civil servants will be bound to
run their departments consistent with these obligations (McCrudden 1999a;
1999b; 2001). They include a requirement that the ‘relevant Ministers’ serve in the
North-South Ministerial Council, a duty that, in conjunction with other clauses,
intended to prevent parties opposed to this aspect of the Agreement, such as the
DUP, from abusing their offices or taking offices in bad faith.
The UUP and the SDLP, in the negotiations over the Northern Ireland Act
1998, agreed that junior ministers could be created. They were put in place only in
the Office of the FM and DFM, one from the UUP and one from the SDLP; more
could be allocated places under the d’Hondt process, though they are not obliged
to be appointed in this way.13 Most of the leading members of the major parties
willing to participate in the executive, in consequence, ‘win prizes’ of one sort
or another – something intended to provide incentives for a shift of posture on
the part of ambitious anti-Agreement Assembly members. These incentives have
worked, at least in part. The anti-Agreement DUP took its seats in the Executive
and in the Assembly’s Committees, and fought the 2001 Westminster general
election not on a pledge to scrap the Agreement but to renegotiate it (Mitchell,
O’Leary, and Evans 2001; 2002); it fought the November 2003 Assembly elections
on the same platform. At the time of composition it remained to be seen whether
these incentives would operate further – the DUP now has an entitlement to the
First Minister position, but if, and only if, it accepts a Sinn Féin-approved Deputy
First Minister.
This inclusive executive design, of course, means that the new Assembly has a
rather small part of its membership free to function as an opposition for standard
adversarial parliamentary debating in the classic Westminster mould, though the
inter-party rhetorical engagement in the Assembly is sometimes difficult to rec-
oncile with the fact that the four largest parties – the UUP, the SDLP, the DUP,
and Sinn Féin – share the cabinet positions. The standard complaint of critics
of consociation – that it weakens the effectiveness of parliamentary opposition
– must, however, surely be tempered in this case by the fact that the backbench-
ers from other parties in the government are likely to hold the relevant minister
vigourously to account. Moreover, mechanisms for rigourous accountability ex-
ist because ministers face an Assembly Committee in their jurisdiction that is
headed by a representative of another party.
How should we appraise the executive design that is at the heart of the Agree-
ment? The special skill of the negotiators, building on their past experience and
comparative knowledge, was to create strong incentives for executive power-shar-
ing and power-division, but without requiring parties to have any prior formal
coalition agreement – other than the institutional agreement – and without re-
quiring any party to renounce its long-run aspirations. The dual premiership was
designed to tie moderate representatives of each bloc together and to give some
drive towards overall policy coherence. It was intended to strengthen moderates
and to give them significant steering powers over the rest of the executive. The
d’Hondt mechanism, by contrast, ensures inclusivity and was carefully explained
to the public as achieving precisely that: “The purpose is to ensure confidence
across the community … so that people know that their parties will, if they re-
ceive a sufficient mandate in the election, have the opportunity for their Members
to become Ministers and play their part in the Executive Committee.” (House of
Commons 1998: col. 1023). It also saves on the transaction costs of bargaining over
13 Section 19 of the Northern Ireland Act 1998 permits the First and Deputy First Min-
isters to determine, subject to Assembly approval, the number of junior ministers
and procedures for their appointment.
portfolios. Distinctive coalitions can form around different issues within the Exec-
utive, permitting flexibility but inhibiting chaos – given the requirement that the
budget be agreed by cross-community consent. The Executive successfully agreed
a budget and a programme of government through inter-ministerial bargaining
during 2000–2001. The DUP ministers agreed it, though they then supported
their colleagues in voting against it in the Assembly (secure in the knowledge that
the other parties would pass it)! These creative incentives to keep parties in the
executive despite strong disagreements means the Agreement differs positively
from the Sunningdale power-sharing experiment of 1973 that sought to maintain
traditional UK notions of collective cabinet responsibility.
What was not foreseen was that failure to timetable the formation of the rest of
the Executive immediately after the election of the FM and DFM could precipitate
a protracted crisis. Trimble availed himself of this loophole to prevent executive
formation until November 1999. If the Agreement survives, a future review might
agree amendments to the Northern Ireland Act 1998 that could be adopted by the
UK parliament or by the Assembly and that would be consistent with the Agree-
ment to prevent any recurrence of this type of crisis. In future, candidates for FM
and DFM could be obliged to state the number of executive portfolios that will
be available, and the formation of the executive should be required immediately
after their election. That would plug this particular constitutional hole. It may,
however, be unnecessary. It is not likely that future candidates for FM and DFM
will agree to be nominated without a firm agreement on the number of portfolios
and the date of cabinet formation. In the author’s view, it would be best that the
rest of the Executive were formed before the election of the premiers (after all,
the text of the Agreement created no necessary link between the election of the
premiers and the allocation of all other ministerial portfolios).
What was also not foreseen was that the dual premiership might prove the
most brittle of all the new institutions. Recall that the two large moderate parties
separately negotiated its creation, in which they had very direct stakes. Other
possibilities were excluded, such as filling all the top positions, including the pre-
miers, by the d’Hondt rule, or by another allocation rule. Also, the posts were lat-
er made tightly interdependent: the resignation or death of one triggers the oth-
er’s formal departure from office, and requires fresh elections within six weeks.
One consequence has been that all inter-communal tension has been transmitted
through these posts: Mallon deployed his resignation power before the executive
was fully formed; Trimble later deployed the resignation threat to precipitate a
unilateral suspension of the Agreement’s institutions by the UK; and was to do
so again in 2001, just before the Westminster general elections: a resignation that
became operative on 1 July 2001.14 The November 2003 elections have created a
14 The relationship between the first two premiers progressively worsened after a
promising beginning, and culminated in Trimble giving Mallon two minutes notice
of his intention to repeat his use of a post-dated resignation letter. One sage reporter
describes the Mallon-Trimble relationship as “poisonous,” compounded by Trimble’s
character traits – “unpredictable and mercurial, often bewildering, sometimes im-
fresh crisis, showing that the dual premiership has been a lightening rod for deep
tensions at least as much as it has been a mechanism for joint coordination and
creation of calm.
4 Proportionality Rules
Consociational arrangements are built on principles of proportionality. The
Agreement meets this test in four ways:
i) in the d’Hondt procedure for executive formation discussed above;
ii) in the Assembly’s committees;
iii) in the electoral system for the Assembly; and
iv) and in the recruitment and promotion policies within the public sector.
a The Assembly’s Committees

The Assembly has committees scrutinizing each of the departments headed by
ministers. Committee Chairs and Deputy Chairs are allocated according to the
d’Hondt rule. Committee composition is in proportion to the composition of
the Assembly. Each committee must approve any proposed new law within its
jurisdiction tabled by ministers, and the committee can itself initiate legislative
proposals. Consequently, a committee dominated by other parties may block the
legislative initiatives of a dynamic minister, and it may initiate legislation not to
that minister’s liking – though the success of such proposals is subject to cross-
community special procedures. So the committee system combines the two con-
sociational principles of proportionality and veto rights. In the passage of the
Northern Ireland Act 1998, the committees were explicitly prevented, by law,
from being chaired or deputy-chaired by ministers or junior ministers, and are
required, where feasible, to be organized in such a way that the Chair and Deputy
Chair be from parties other than that of the relevant minister. This ensures the
accountability of ministers at least to MLAs from other parties and inhibits full-
scale party fiefdoms in any functional sector. It also refutes the suggestion of con-
sociation’s critics that a power-sharing system precludes effective accountability.
b The Electoral System for the Assembly

Elections to the 108-member Assembly must be conducted under a proportional
representation (PR) system, using the single transferable vote (STV) method, in
six-member constituencies – though the Assembly may choose, by cross-com-
munity consent procedures, to advocate change from this system (subject to rati-
fication by Westminster). The ‘Droop’ quota used in STV is V/(N+1)/1, where V =
total valid votes, and N = number of Assembly members to be elected. The Droop
quota in each constituency is therefore one seventh or 14.3%, which squeezes the
very small parties, or, alternatively, encourages them to form electoral alliances.
Thus the smaller of the two loyalist parties, the UDP, won no seats in the first As-
possible” – and by the nature of his career – “a mixture of dashes and longeurs,
alternatively crisis-ridden and becalmed, of tacks towards moderation interspersed
with lurches to the confrontational” (McKittrick 2001: 27).
sembly election. Very small parties that can gather lower-order preferences from
across the unionist and nationalist blocs, such as the Women’s Coalition, have
shown that the system need not preclude representation for small parties (they
had two MLAs elected in 1998, though none in 2003).
This system, STV-PR, is not what Lijphart recommends for consociational
agreements (Lijphart 1990a). He is an advocate of party-list PR systems, princi-
pally because he believes they help make party leaders more powerful and better
able to sustain inter-ethnic consociational deals.15 Those who would like to have
seen Trimble in greater control of the UUP might hanker after Lijphart’s preferred
form of PR. The 1998 Northern Ireland Assembly suggests that a modification of
the consociational prescriptive canon is in order. Had a region-wide party-list
system been in operation in June 1998, the UUP would have ended up with fewer
seats, and with fewer seats than the SDLP. As a consequence, the implementation
of the Agreement would have been even more problematic. There is a further,
and less contingent, argument against party-list systems in consociations that is
especially important where the relevant ethnic communities are internally demo-
cratic rather than sociologically and politically monolithic. A region-wide party-
list election gives incentives for the formation of a wide variety of micro-parties.
It would have fragmented the votes of the major parties that made the Agree-
ment. Hardliners under party-list systems have every reason to form fresh parties
knowing that their disloyalty will penalize more moderate parties, but without
necessarily reducing the total vote and seat share of the relevant ethno-national
bloc. This objection to Lijphart’s favoured prescription is not merely specula-
tive. The 1996 elections to the Northern Ireland Peace Forum used a mixture of
a party-list system and ‘reserved seats’. Party proliferation and the erosion of the
UUP first-preference vote were among the more obvious consequences (Evans
and O’Leary 1997).16 STV, of course, does not guarantee party discipline, as mul-
15 Lijphart argues for this system rather than STV because it: (1) allows for a high dis-
trict magnitude, making possible greater proportionality; (2) is less vulnerable to
gerrymandering; and (3) is simpler for voters and organizers (Lijphart 1990a). By
contrast, I think high thresholds help reduce fragmentation, as a trade-off against
‘better’ proportionality. Contra Lijphart, I think that STV, legislatively enacted with
uniform district magnitudes and supervised by independent electoral commissions
charged with creating uniform electorates, is not more vulnerable to gerrymandering
than regional party-list PR. I concede that STV is suitable only for numerate elector-
ates, but otherwise its complexities are not especially mysterious – no more so than
the formulas used for achieving proportionality in party-list systems.
16 The nature of executive formation in the Agreement should act as one possible check
on the possibilities of fragmentation under party-list PR, because only large parties
are likely to win ministries under d’Hondt, but that is true of any electoral system
combined with this executive. In the November 2003 elections, the unification of
anti-Agreement unionist voters behind the DUP not only reduced fragmentation but
increased all unionists entitlements to ministries from five to six (what mattered
was the reduction in small unionist parties not the net increase (one) in unionist
MLAs).
tiple candidates for the same party in a given constituency may present, tacitly
or otherwise, slightly different emphases on party commitments, as indeed hap-
pened in Northern Ireland in 1998 and 2003. But I suggest that STV, combined
with higher effective thresholds than under most forms of party-list PR, makes it
more likely that parties will remain formally unified and therefore able to make
and maintain consociational deals if they want to do so. At the very least, the pre-
scriptive superiority of the party-list system for these purposes is unproven, and
Lijphart’s consistent counsel in this respect should be modified.17
As well as achieving proportionality, STV has the merit of facilitating inter-eth-
nic ‘vote-pooling’ (Horowitz 1985: 628 ff ). In principle, voters can use their lower-
order preferences – ‘transfer papers’ – to reward pro-Agreement candidates at the
expense of anti-Agreement candidates. In this respect, STV looks tailor-made to
achieve the ‘interethnic’ and ‘crossethnic’ voting favoured by Donald Horowitz,
a vigorous critic of consociational thinking, but an advocate of institutional and
policy devices to facilitate conflict-reduction (Horowitz 1985; 1989a; 1989b; 1991).
However, Horowitz believes that STV damages the prospects for interethnic co-
operation because the relatively low quota required to win a seat in six-member
constituencies makes it too easy for hardline parties and their candidates to be
successful (Horowitz 2001). He also thinks that the Agreement’s institutions, bi-
ased towards the key consociational partners, nationalists and unionists, com-
pound this effect by weakening the prospects of cross-ethnic parties, such as the
Alliance, which he believes is likely to impair conflict-reduction. The Northern
Ireland case challenges Horowitz’s reasoning (Horowitz 1991; 2001). Horowitz
would prefer the use of the Alternative Vote (AV) in single-member constituen-
cies, because its quota – 50% plus 1 – would deliver strong support to moderate
ethno-national and cross-ethnic candidates. But, this prescription is incoherent.
First, the outcomes it would deliver would be majoritarian, disproportional, and
unpredictably so. They would be disproportional both within blocs and across
blocs. They would, additionally, have much more indirectly ‘inclusive’ effects than
STV. In some constituencies, there would be unambiguous unionist and nation-
alist majorities (Mitchell, O’Leary, and Evans 2001) – and thus AV would lead
to the under-representation of minority voters within these constituencies, and
to local fiefdoms. Second, while candidates would often have to seek support
for lower-order preferences under AV, it would not be at all obvious that their
best strategy would be to seek lower-order preferences across the ethno-national
divide. Instead, the imperative of staying in the count would dictate building as
big an initial first and second preference vote tally as possible.18 Third, AV would
never be agreed to by hardline parties entering a constitutional settlement if they
17 John McGarry and I used to assume the superiority of the party-list system (Mc-
Garry and O’Leary 1990a: 297). Facts and reflection have made me reconsider the
merits of STV (O’Duffy and O’Leary 1995; O’Leary 1999c).
18 It may be that AV’s presumptively moderating effects materialize better in multi-eth-
nic political systems with no actual or potentially dominant group in given districts
– a situation that does not obtain in Northern Ireland.
believed it would be likely to undermine their electoral support. Since the Agree-
ment was made possible by encouraging ‘inclusivity’, by facilitating negotiations
which included Sinn Féin (the party that had supported the IRA) and the PUP and
the UDP (the parties that had supported the loyalist Ulster Defence Association
(UDA) and Ulster Volunteer Force (UVF)), it would have been perverse for their
leaders to agree to an electoral system that minimized their future prospects.
Indeed, STV arguably worked productively both before and after the Agree-
ment. It helped to moderate the policy stance of Sinn Féin. After its first phase
of electoral participation in elections in Northern Ireland in the 1980s and in the
Irish Republic in the latter half of the 1980s, the party discovered that it was in a
ghetto. Its candidates in some local government constituencies would pile up large
numbers of first-preference ballot papers but then sit unelected as a range of oth-
er parties’ candidates passed them to achieve quotas on the basis of lower-order
preferences.19 They received very few lower-order preferences from SDLP voters.
However, once the party moderated its stance, promoted the IRA’s ceasefire(s),
and became the champion of a peace process and a negotiated settlement, it found
that its first-preference vote, its transfer vote, and its seats won all increased.
The relevant constitutional design argument advanced here is this: where there
has been party fragmentation within ethno-national blocs, then STV can assist ac-
commodating postures and initiatives by parties and candidates, both intra-bloc
and inter-bloc.20 The primary normative objection to Horowitz’s position is that
proportionality norms better match both parties’ respective bargaining strengths
and their conceptions of justice. Once party pluralism has already emerged, some
form of proportionality is more likely to be legitimate than a shift to strongly ma-
joritarian systems, such as AV, or to systems with ad hoc distributive requirements
that will always be – correctly – represented as gerrymanders. Horowitz’s elector-
al integrationist prescriptions are most likely to be pertinent only at the formation
of a competitive party system. Once party formation and party pluralism within
blocs have occurred, there will be few agents with the incentives to implement
Horowitz’s preferences; and if a third party or outside power does so it would be
a provocation to the less moderate parties, and might re-ignite ethno-national
tensions. These arguments in defence of STV are, of course, qualified. STV alone
19 STV has been used in local government elections and European parliamentary elec-
tions in Northern Ireland since 1973 and 1979, respectively. Interestingly, the hard-
line unionist Ian Paisley has been most successful in the three-member district used
to elect Northern Ireland’s MEPs; in the more proportional five- or six-member local
government or Assembly constituencies the DUP did not fare as well before 2003.
20 The corollary is that STV’s positive effects apply to already polarized and pluralized
party systems in ethno-nationally divided societies. If there has been no prior history
of ethnicized party polarization within a state, or of pluralization of parties within
ethno-national blocs, the merits of its implementation may be reasonably doubted.
This consideration raises the key problem with Horowitz’s electoral integrationist
prescriptions: they apply best to forestalling or inhibiting ethnic conflict and are less
effective remedies for cases of developed, protracted, and intense ethnic and ethno-
national conflict.
is not enough, and it may not be appropriate everywhere. But it can help promote
accommodative moves and consolidate consociational deals in ways that the re-
gion-wide party-list systems and the AV in single-member district cannot.21
There has been some empirical confirmation of the merits of STV since the
Agreement was made. Some ‘vote pooling’ occurred within the first Assem-
bly elections, as we can surmise, from actual counts and from an opinion poll
(Mitchell 2001; Evans and O’Leary 2000). Some of the SDLP’s and Sinn Féin’s
voters found it rational to reward the UUP for making the Agreement by giving
its candidates their lower-order preferences. Likewise, in 1998, some of the UUP’s
and the PUP’s voters transferred their lower-order preferences to pro-Agreement
candidates within their own bloc, among the others and among nationalists. Of
course, large-scale transfers took place among the ‘No’ unionists and between
‘Yes’ unionists and ‘No’ unionists. Within-bloc rewards for moderation also oc-
curred: Sinn Féin won lower-order preferences from SDLP voters, and the PUP
had candidates elected on the basis of transfers from other candidates. Table 4
reports the outcome of the June 1998 elections to the first Assembly, and the
November 2003 elections – which may eventually produce a functioning second
Assembly. The proportionality of the results is evident with respect both to blocs
and to parties. In 1998, the deviations in seats won compared with the first prefer-
ence vote primarily benefited the pro-Agreement parties; in 2003, the deviations
primarily benefited the UUP and the Alliance, but not the nationalist pro-Agree-
ment parties. In both 1998 and 2003, the UUP was the principal beneficiary of the
transfer of lower-order preferences among the large parties, which in 1998 took
its seat share (25.9%) significantly above its first-preference vote-share (21.3%).
Its lower-order transfers came from voters who voted ‘No’ as well as those who
voted ‘Yes’ to the Agreement, as was evident in ballot papers and a survey (Mitch-
ell 2001; Evans and O’Leary 2000). In 1998, the Women’s Coalition was the great-
est beneficiary of lower-order preferences, winning two seats despite a very low
first-preference vote. In 2003, the Alliance party kept its six seats with almost
half the first-preference vote it had won in 1998. The net transfers by voters to the
pro-Agreement candidates, though not dramatic, helped perform one very im-
21 The Northern Ireland Act 1998 and the Northern Ireland (Elections) Act 1998 opened
one novelty in the practice of STV in Ireland. Both acts left it open to the Secretary of
State to determine the method of filling vacancies: this may be done through by-elec-
tions, substitutes, or whichever method the Secretary of State deems fit. By-elections
are anomalous in a PR system (Gallagher 1987). A candidate who wins the last seat in
a six-member constituency and who subsequently resigns or dies is unlikely to be re-
placed by a candidate of the same party or persuasion in a by-election, which becomes
the equivalent of the alternative vote in a single-member constituency. The Northern
Ireland Assembly (Elections) Order of 1998 has provided for a system of alternates or
of personally nominated substitutes with a provision for by-elections if the alternates
system fails to provide a substitute. The disproportionality possibly induced by by-elec-
tions, with its consequent ramifications for the numbers of registered nationalists and
unionists and the cross-community rules, needed to be engineered out of the settle-
ment, and it was a good sign that the parties cooperated with this concern in mind.
portant task in 1998. They converted a bare ‘anti-Agreement’ majority of the first
preference vote (25.5%) within the unionist bloc of voters into a bare ‘pro-Agree-
ment’ majority (27.7%) among seats won by unionists, a result that may have been
essential for the Agreement’s (partial) stabilization. In 2003, transfers may have
dampened the swing to anti-Agreement unionists, but confirmation of that pos-
sibility awaits a full count by researchers. Co-ordination failures are evident also:
in 2003, the failure of Sinn Fein voters to transfer to the SDLP in West Belfast was
probably decisive in letting the DUP win a seat.
Table 4 Party Performances in the 1998 and 2003 Elections to the Northern
Ireland Assembly
Party First Preference Vote % Seats % (Number)
1998 2003 1998 2003
SDLP 22.0 17.0 22.2 (24) 16.7 (18)
Social Democratic & Labour Party
SF 17.7 23.5 16.7 (18) 22.2 (24)
Sinn Féin
Other nationalists 0.1 0.3* – (–) – (–)
UUP 21.0 22.7 25.9 (28) 25.0 (27)
Ulster Unionist Party
PUP 2.5 1.2 1.9 (2) 0.9 (1)
Progressive Unionist Party
UDP 1.2 – – (–) – (–)
Ulster Democratic Party
Other ‘Yes’ unionists 0.3 – – (–) – (–)
DUP 18.0 25.7 18.5 (20) 27.7 (30)
Democratic Unionist Party
UKUP 4.5 0.8 4.6 (5) 0.9 (1)
UK Unionist Party
Other ‘No’ unionists 3.0 0.5+ 2.8 (3) – (–)
APNI 6.4 3.7 5.5 (6) 5.5 (6)
Alliance Party
NIWC 1.7 0.8 1.9 (2) – (–)
Women’s Coalition
Other others 1.3 3.5 – (–) 0.9 (1)
Source: O’Leary 1999e updated).

* Socialist Environmental Alliance candidate McCann counted as an independent na-
tionalist.
+ Conservatives counted as ‘No Unionists’.
c Proportional Recruitment and Representativeness in the Public Sector

The Agreement accepted past and future measures to promote fair employment
and affirmative action in the public sector that are likely to eventually ensure a
fully representative and non-discriminatory civil service. The civil service and the
rest of the public sector have already been subjected to fair employment legisla-
tion, but in the entirety of posts in the public sector the principles of representa-
tiveness or proportionality are to be applied, either in the form of party represen-
tatives holding others to account or of ascriptively representative bureaucracies
and public services. There was one initial exception: the judiciary. That significant
omission was eventually rectified by the report of the Criminal Justice Review
that commended a judicial appointments commission to ensure a more represen-
tative judiciary, a proposal that is being given legislative effect in 2003–2004.22
Most significantly, the Agreement envisaged a representative police force.
Democratic consociation cannot exist where those of military age in one com-
munity are almost the sole recruitment pool for policing all of those in another
community – a trait characteristic of control systems (Lustick 1979). Over 300
police officers were killed after 1969 – for whom there is significant sympathy,
especially amongst unionists. But nationalists remember that the outbreak of
armed conflict in 1969 was partly caused by an unreformed police responsible
for seven of the first eight deaths. Policing was so controversial that the parties
to the Agreement could not concur on future arrangements, and it was not made
a devolved function (McGarry and O’Leary 1999). They did agree the terms of
reference of a Commission, eventually chaired by Christopher Patten. The Re-
port of the Independent Commission – the ‘Patten Report’ – published in Sep-
tember 1999 fulfilled the Commission’s mandate under the Agreement (Patten
1999; O’Leary 1999b). Eight criteria were explicitly and implicitly mandated in
the Commission’s terms of reference. Policing arrangements were to be impar-
tial; representative; free from partisan political control; efficient and effective;
infused with a human rights culture; decentralized; democratically accountable
“at all levels”; and consistent with the Agreement. The Commission’s Report was
a thorough, careful, and imaginative compromise between unionists who main-
tained that the existing RUC met the terms of reference of the Agreement and
those nationalists, especially republicans, who maintained that its human rights
record required its disbanding.
The Report declared that responsibility for policing should be shared by citi-
zens and their representatives. This logic was apparent in the title and composi-
tion of the recommended Policing Board – bringing together ten elected politi-
22 The Agreement’s terms of reference for the Criminal Justice Review incorporated the
provision that “the arrangements for making appointments to the judiciary and mag-
istracy, and safeguards for protecting their independence.” The commissioners were
able to use this provision to commend an appointments commission and a judiciary
“reflective of Northern Ireland society.” Its key proposals are dependent on the de-
volution of the administration of justice to the Assembly and Executive (Livingstone
2001).
cians, drawn from the parties that comprise the new Executive according to the
d’Hondt rule, with nine appointed members representative of civil society, “busi-
ness, trade unions, voluntary organizations, community groups and the legal
profession” to be nominated by the FM and DFM. The elected members were not
to be ministerial office-holders. The Board was to be representative, but at one
remove from direct executive power. The Report intended to let police managers
manage, but to hold them ex post facto accountable for their implementation of
the Board’s general policing policy, and to enhance the audit and investigative ca-
pacities of the Board. It recommended giving directly elected local governments
opportunities to influence the Board though their own District Policing Part-
nership Boards. The Report displayed coherent communitarian, democratic and
pluralist ideas, and was informed by rigourous management practices. Segmental
policing, in which each community would be policed by ‘its own’, was not con-
sidered, nor seriously proposed. Instead, a representative but integrated service
was advocated, appropriate for a region with a high combination of both territo-
rial segregation and mixing. The Commission proposed recruiting Catholics and
non-Catholics in a 50:50 ratio from the pool of qualified candidates for the next
decade. This matches the population ratios in the younger age cohorts. Given
early and scheduled retirements of serving officers, this policy would ensure that
30% of the service would be of Catholic origin after ten years and between 17%
and 19% within four years – above the critical mass claimed essential to change
the police’s character. This is a slower pace of change than some advocated (Mc-
Garry and O’Leary 1999). The Commission proposed feasible policing arrange-
ments consistent with the spirit of the Agreement, including on recommenda-
tions for better-structured cross-border cooperation with the Garda Síochana in
the Republic, including secondments. Significantly, the Report’s recommenda-
tions mostly did not depend upon the Agreement’s institutions for their imple-
mentation. The commissioners explicitly recommended most of their changes,
come what may.
However, the Police Bill presented as the implementation of the Report to the
Westminster parliament in the spring of 2000 by Secretary of State Mandelson
was an evisceration of the Commission’s work. It was condemned as such by the
SDLP, Sinn Féin, the Women’s Coalition, the Catholic Church, human rights or-
ganizations (such as the Committee on the Administration of Justice), the Irish
government, the US House of Representatives (H. Res 447, 106th Congress), and
Irish-Americans, including President Clinton (O’Leary 2000e; 2000c; 2000b).
The veracity of the critics’ complaints can be demonstrated by comparing some
of Patten’s recommendations with the original bill
1. Names and Symbols. Patten recommended a neutral name, the ‘Northern
Ireland Police Service’. The Royal Ulster Constabulary’s non-neutral name
was to go. Patten recommended that the display of the Union flag and the
portrait of the Queen at police stations should also go. Symbols should be
‘free from association with the British or Irish states’. The original police bill,
by contrast, proposed that the Secretary of State have the power to decide
on the issues of names and emblems.
2. Affirmative action. Even critics of affirmative action recognized the need to

correct the existing imbalance in which over 90% of the police are local cul-
tural Protestants. But the original bill reduced the period in which the police
would be recruited on a 50:50 ratio of cultural Catholics and cultural Prot-
estants from ten years to three, requiring the Secretary of State to make any
extension, and was silent on ‘aggregation’, the proposed policy for shortfalls
in recruitment of suitably qualified cultural Catholics.
3. Policing Board and Political Control. Patten proposed a Policing Board con-
sisting of ten MLAs from political parties in the Executive allocated accord-
ing to the d’Hondt process, and nine members nominated by the FM and
DFM. These recommendations guaranteed a politically representative board
in which no bloc would have partisan control. The original bill required that
the Board should operate according to a weighted majority when recom-
mending an inquiry, tantamount to giving unionist or unionist-nominated
members partisan political control.
4. Efficiency and downsizing. Patten recommended downsizing the service, ad-
vocated a strong Board empowered to set performance targets, and enabling
local District Policing Partnership Boards to market-test police effectiveness.
The original bill empowered the Secretary of State, not the Board, to set per-
formance targets, made no provision for disbanding the police reserve, and
deflated the proposed District Policing Partnership Boards.
5. Rights Protection. Patten proposed that new and serving officers should have
human rights training and re-training, and observe new codes of practice.
In addition to the European Convention on Human Rights, due to become
part of UK domestic law, the Commission held out international norms as
benchmarks (Patten 1999: para 5.17). Patten’s proposals for normalization
– through merging the Special Branch into Criminal Investigations – and
demilitarization met the Agreement’s human rights objectives. The origi-
nal bill, by contrast, confined the new oath to new officers. No standards
of rights higher than those in the European Convention were to be incor-
porated into training, and Responsibility for a Code of Ethics was left with
the Chief Constable. Patten’s proposed requirement that the oath of service
“respect the traditions and beliefs of people” was excluded. Normalization
and demilitarization procedures were left unclear in the bill and the imple-
mentation plan.
6. Decentralization. Patten envisaged enabling local governments to influence
the Policing Board through their own District Policing Partnership Boards
and giving the latter powers “to purchase additional services from the police
or statutory agencies, or from the private sector,” and matching police in-
ternal management units to local government districts. The original bill, by
contrast, maintained or strengthened centralization: the Secretary of State
obtained powers that Patten proposed for the FM and DFM and the Board,
and powers to issue instructions to District Policing Partnership Boards; and
neither the bill nor the implementation plan implemented Patten’s proposed
experiment in community policing (Hillyard 2000).
7. Accountability. Patten envisaged a strong, independent and powerful Board

to replace the discredited Police Authority (Patten 1999: para 6.23). The po-
lice would have “operational responsibility” but be held to account and re-
quired to interact with the Human Rights Commission, the Ombudsman,
and the Equality Commission. The Bill watered down Patten’s proposals,
empowering the Secretary of State to oversee and veto the Board, and allow-
ing the Chief Constable to refuse to respond to reasonable requests from the
Board, and preventing the Board from making inquiries into past miscon-
duct.
The radical discrepancy between the Report and the original Police bill stemmed
from Northern Ireland Office’s officials drafting the bill under Mandelson’s su-
pervision. They took the views of the RUC and British security specialists more
seriously than those of the Patten Commissioners, treating theirs as a nationalist
report that they had to modify as benign mediators. They simply left aside what
they found unacceptable, premature, or likely to cause difficulties for pro-Agree-
ment unionists or the RUC.
The original bill suggested that the UK government was determined to avoid the
police being subject to rigourous democratic accountability, was deeply distrust-
ful of the capacity of the local parties to manage policing at any level, and, lastly,
concerned to minimize the difficulties that the implementation of Patten would
occasion for Trimble’s party. Under enraged nationalist pressure, Mandelson
beat a partial retreat. The UK government accepted more than sixty SDLP-driven
amendments to bring the bill more into line with Patten’s Report. The quota for
the recruitment of cultural Catholics became better protected. The Board was
given power over the setting of short-run objectives, and final responsibility for
the code of ethics. Consultation procedures involving the Ombudsman and the
Equality Commission were strengthened, and the FM and DFM were to be con-
sulted over the appointment of non-party members to the Board. The weighted
majority provisions for an inquiry by the Board went. While the Police (Northern
Ireland) Act 2000 fell significantly short of ‘the whole Patten’, it rectified some of
the original bill’s more overt deviations. But, on the two crucial issues of symbolic
neutrality and police accountability, vital for a new beginning, it remained at odds
with Patten’s explicit recommendations. The Police Act, because of a government
decision to accept an amendment tabled by the UUP, styles the service ‘The Police
Service of Northern Ireland (incorporating the Royal Ulster Constabulary)’, sure-
ly one of the longest names of a police service in the English-speaking world. The
Secretary of State promised an amendment to define the name ‘for operational
purposes’, and to ensure that the full title would rarely be used, but broke this
commitment. Mandelson declared he was following Patten’s wishes that the new
service be connected to the old and avoid suggestions of disbanding, but Patten
had proposed an entirely new and fresh name, and proposed linkages between
the old and new services through police memorials, and not by the renaming ad-
opted by the government. Patten unambiguously recommended that the police’s
new badge and emblems be free of association with the British or Irish states,
and that the Union flag should not fly from police buildings. The Act postponed
these matters. The UK government ‘passed the parcel’ to the local parties to reach
agreement while providing reassuring but vague words in Hansard. Since Man-
delson had already ruled that only the Union Jack, albeit only on specified days,
should fly over the buildings of the devolved administration, nationalists lacked
faith that he would deliver on cultural neutrality and impartiality.
Why have these symbolic issues mattered? Because they do in ethno-national
conflicts, and because the best strategy to win widespread acceptance for police
reform would have been to confirm Patten. Full renaming and symbolic neutrality
spell out a double message: that the new police are to be everyone’s, and the new
police are no longer to be, as they were, primarily the unionists’ police. To achieve
effective accountability, Patten recommended an Oversight Commissioner to
“supervise the implementation of our recommendations.” The UK government –
under pressure – put the commissioner’s office on a statutory basis, which it did
not intend to do originally, but confined his role to overseeing changes “decided
by the Government.” Patten had recommended a Board that could initiate inqui-
ries into police conduct and practices. The Police Act 2000 prevented the Board
from inquiring into any act or omission arising before the eventual act applies,
tantamount to an undeclared amnesty for past police misconduct. Many have no
objections to an open amnesty, especially as paramilitaries have received de facto
amnesties (see below), but the government’s chosen method appeared driven by
concerns to avoid state officials being held to account for their responsibilities for
the last thirty years of conflict (Ní Aoláin 2000). Additionally, the Secretary of
State has the authority to approve or veto the person appointed to conduct any
present or future inquiry (clause 58(9)). Whereas Patten recommended that the
Ombudsman should have significant powers (Patten 1999: para 6.42) and “exer-
cise the right to investigate and comment on police policies and practices,” in the
Act, the Ombudsman may make reports but not investigate – so it is not a crime
to obstruct her work. The Ombudsman was additionally restricted in her retro-
spective powers (clause 62), again circumscribing the police’s accountability for
past misconduct. The proposed arrangements sealed off avenues through which
the police might be held to account for misconduct – for example, in colluding
with loyalist paramilitaries or covering up assassinations – and appeared recipes
for leaving the police outside the effective ambit of the law.
Failure to deliver fully on police reform in the promised manner heralded pos-
sible disaster for the Agreement. The SDLP, Sinn Féin, and the Catholic Church
were unlikely to recommend that their constituents consider joining the police,
and considered boycotting the Policing Board and District Policing Partnership
Boards. The mismanagement of Patten Report’s meant that in 2000 the pressure
eased on Sinn Féin to require the IRA to go further in decommissioning than ar-
rangements for international inspections of its arms dumps. The argument was
made that the UK government had reneged on a fundamental part of the Agree-
ment, so the IRA was under no obligation to start disarming. In turn, this led to
a renewal of unionist calls for the exclusion of Sinn Féin from ministerial office,
leading to Trimble’s second resignation threat in the spring of 2001.
Mandelson’s conduct was partly motivated by the need to help Trimble and
the UUP, in a precarious and divided condition, and fearful of being outflanked
by the DUP on a critical issue. But while ‘Saving Trimble’ may account for the
tampering with Patten’s proposals on symbolic matters, it hardly accounts for the
blocking of the efforts to have a more accountable service. The decision to dilute
the content and the pace of Patten’s recommendations meant that policing re-
form, a core dimension of the Agreement, become a serious source of continuing
antagonism, and an excuse for delays on decommissioning. It was only later that
the UK would pedal back and come very close to the full implementation of the
Patten Report, a task rendered easier by Mandelson’s departure from office.
5 Communal Autonomy and Equality

Consociations seek, through bargaining, to manage differences equally, justly,
civilly. Liberal consociations do not prevent voluntary integration or assimila-
tion, and protect those who wish to have their identities counted differently or
not as collective identities. The Agreement is such a liberal consociation.
The Agreement left in place the arrangements for primary and secondary
schooling in Northern Ireland in which Catholic, Protestant, and integrated
schools are to be equally funded. In the past, Catholic schools received less capi-
tal funding, and before that had to raise a significant proportion of their own
staffing resources (McGrath 2000). In this respect, Northern Ireland is now
consociational but liberal: one can avoid Catholic and Protestant schools. Only
the very small minorities of non-Christian religious believers, amounting to less
than 1% of the population, lack full and equal funding for culturally distinctive
schools. The Agreement made new provisions for the educational use, protec-
tion, and public use of the Irish language, along the lines used for Welsh within
Wales, thereby adding linguistic to the educational protections of Irish national-
ist culture. It made analogous provisions for ‘Ulster Scots’. Most significantly, the
Agreement completes the equalization of both major communities as national
communities, that is, as British and Irish communities, and not just, as is so mis-
leadingly emphasized, as Protestants and Catholics.
The opening section of the Agreement’s ‘Constitutional Issues’ specifies that:
Whatever choice is freely exercised by a majority of the people of Northern Ireland,

the power of the sovereign government with jurisdiction there shall be exercised with
rigorous impartiality on behalf of all the people in the diversity of their identities and
traditions, and shall be founded on the principles of full respect for, and equality of,
civil, political, social and cultural rights, of freedom from discrimination for all citizens,
and of parity of esteem and of just and equal treatment for the identity, ethos and aspi-
rations of both communities (Government of the United Kingdom n.d.: ‘Constitutional
Issues’, 1. (v), p. 2).
Two words merit isolation here, viz, ‘ethos’ –“the characteristic spirit or attitudes
of a community, people, or system,” and ‘aspirations,’ “strong desires to achieve an
end; ambitions” (definitions from the Concise Oxford English Dictionary, 8th Edi-
tion). Judges and others must, in time, treat these words as requiring them to in-
terpret the Agreement as respecting the respective communities’ national identi-
ties, core parts of their respective ethos and aspirations, provisions that go beyond
those found in the European Framework Convention on National Minorities. The
European Convention on Human Rights, now incorporated in the UK’s domestic
law, is weak on the protection of collective rights, national minority rights, ethnic
rights, and equality rights. The new Northern Ireland Human Rights Commis-
sion (NIHRC) was tasked with advising on the defining of Westminster legisla-
tion supplementary to the Convention “to reflect the particular circumstances of
Northern Ireland” (Government of the United Kingdom, n.d.: ‘Rights Safeguards
and Equality of Opportunity’, 4: 16-17). Such provisions must protect both national
groupings and individuals, and ensure that none of the institutions of the Agree-
ment or the proposals of its commissions acting directly under its terms of refer-
ence are rendered invalid by such provisions (O’Leary 2001g).
The Agreement, sensibly interpreted, effectively makes Northern Ireland bina-
tional (even though it did not make it bi-statal, i.e., under the joint sovereignty of
both states). It has opened up the prospect of a fascinating and difficult jurispru-
dence, not least in the regulation of the use of flags and emblems, and of public
parades and marches. So far, the courts, with no nationalists on the most senior
Northern Ireland bench, have chosen to be conservative on this matter, but this
stance will not be sustainable over time – especially if the courts become more
widely representative.
The Agreement did not neglect the non-national dimensions of local politics,
nor does it exclude the ‘others’ from effective political participation. All aspects
of unjustified social equalities, as well as inequalities between the national com-
munities, are recognized in the text of the Agreement and given some means of
institutional redress and monitoring. The Agreement addresses national equality,
the allegiances to the Irish and British nations, and social equality, i.e., dimen-
sions that differentiate groups and individuals in Northern Ireland; including re-
ligion, race, ethnic affiliation, sex, and sexuality. Equality issues, be they national
or social, are not left exclusively to the local parties to manage and negotiate,
which might be a recipe for stalemate. Instead, under the Agreement and Sec-
tion 75 of the Northern Ireland Act 1998, the UK government has created a new
statutory obligation on public authorities. They must carry out all their functions
to promote equality of opportunity in relation to people’s religious background
and political opinions, and with respect to their gender, race, disabilities, age,
marital status, and sexual orientation. This commitment “mainstreams equality”
(McCrudden 1999a; 1999b). The new Human Rights Commission is charged with
a role that is extended and enhanced, compared with its predecessor, though it
is still arguably deficient in resources. Its role includes monitoring, the power to
instigate litigation, as well as drafting the elements of a tailor-made local bill of
rights. In its consultations and deliberations, the Commission has “‘established
working groups in … social and economic rights; children and young people;
criminal justice; cultural rights and community identity; language; victims’ rights;
equality; education; and implementation” (Harvey 2001: 129).
6 Minority Veto Rights

The final dimension of a full consociation is the protection of minorities through
explicit or tacit veto rights. The Agreement achieved this through the Assembly’s
design, through the prospect of a new human rights regime, through a Civic Fo-
rum, and through enabling political appeals to both the UK and the Irish govern-
ments.
The Assembly has procedures (parallel consent, weighted majority, and the
petition) that protect nationalists from unionist dominance. Indeed, they do
so in such a comprehensive manner that the rules designed to protect the na-
tionalist minority have been and may be used by hardline unionist opponents of
the Agreement to attempt to wreck it. The ‘others’ are less well protected in the
Assembly, as they can be out-voted by a simple majority or by any nationalist-
unionist super-majority, and their numbers leave them well short of being able to
trigger a petition. Since the ‘others’ have not been at the heart of the conflict, it is
not surprising if they are not at the heart of its pacts – though it is not accurate to
claim that they are excluded. If they win sufficient electoral mandates their power
will be enhanced accordingly.23
In the courts, the others, as well as disaffected nationalists and unionists, will
have means to redress breaches of their human and collective rights. The package
of collective rights that the NIHRC has recommended for legislative enactment
by Westminster has followed recommendations to include proportional electoral
rights (O’Leary 2001g), but the rest of their proposals in controversial areas are
largely modelled on the European Framework Convention – less than that en-
tailed by the plain words of the Agreement on ‘ethos’ and ‘aspirations’. Whether
the Westminster parliament will enact legislation according to its counsel, im-
prove upon it, or worsen it, remains to be seen. The implications for the rest of
the UK may weigh heavily on the Westminster government. It is still likely that
the new policing arrangements will be infused with a much better rights culture,
and that the absence of legal personnel within the RUC with expertise in human
rights will be remedied. The composition of the local judiciary who will supervise
the new systems of rights-protection, however they are fleshed out, is obviously a
key issue. The proposal for a new judicial appointments commission seems likely
to achieve this goal.
Non-national minorities were not forgotten or excluded. Mechanisms have
been established to ensure that ‘others’, outside the local ethno-national blocs,
will be able to express their voices and ensure that the new ‘rights culture’ does
not exclude them: in the Civic Forum created in the North and inaugurated on
9 October 2000, with a prospective southern counterpart; and through the In-
ter‑governmental Conference of the British and Irish governments.
23 John McGarry and I argue for changes to the voting rules to make ‘designation’ less
obviously favourable to unionists and nationalists, and fairer to the others (McGarry
and O’Leary 2004b).
C A National Self-determination Settlement and an Implicitly

Autonomous Federacy
Complex power-sharing is not exhausted by the preceding description and analy-
sis of the elaborate consociational provisions in the Agreement. It also directly
and crucially addressed the core national self-determination dispute.
1 Recognition
The Agreement was based on multiple forms of recognition. It was an act of recog-
nition between states and national communities. The people of Ireland, through a
referendum that changed their constitution, recognized Northern Ireland’s status
as part of the United Kingdom, subject to the implementation of the Agreement,
while continuing to express their firm will to see Irish unity achieved through the
future majority consent of both jurisdictions. The sovereign governments of each
state recognized each other’s official names for the first time, ‘Ireland’ and the
‘United Kingdom of Great Britain and Northern Ireland’, respectively. The United
Kingdom has recognized, in a treaty, the right of the people of Ireland, meaning
the whole island, to exercise their self-determination, albeit conjointly and sever-
ally as ‘North’ and ‘South’, to bring about a united Ireland if that is their wish. It
has confirmed that Northern Ireland has the right to secede, by majority consent,
to unify with Ireland. Ireland has recognized unionists’ British political identity.
The UK has recognized Irish northern nationalists as a national minority, not
simply as a cultural or religious minority, and as part of a possible future Irish
national majority. The two states have, in effect, recognized the paramilitaries
that have organized ceasefires as political agencies expressing the violent antago-
nism of the previously unresolved national self-determination dispute. They have
not required them to surrender themselves or their weapons to their respective
authorities, and have organized the release of their respective prisoners on the
assurances of their organizations’ ceasefires. Unionists who made the Agreement
recognized nationalists as nationalists, not simply as Catholics or as the minor-
ity. Nationalists recognized unionists as unionists, and not just as Protestants.
Nationalists and unionists have recognized ‘others’, who are neither nationalists
nor unionists. If ethno-nationalist conflicts are rooted in identity politics then
this one has at last moved to the multilateral and institutional recognition of the
identities at stake.
2 Constitutional Change: The Foundations of a Federacy

In the Agreement, both governments agreed to balanced constitutional change.
That required Ireland to modify Articles 2 and 3 of its Constitution, which made
a claim of sovereign right over all of Ireland, irrespective of the wishes of the local
majority in Northern Ireland. It reciprocally required that the UK government
repeal the Government of Ireland Act 1920 – which had violated the right of
the people of Ireland to self-determination – while affirming Northern Ireland’s
place as part of the UK as long as a majority so wished.
UK commentators and unionists generally interpret the Agreement merely as

a form of ‘devolution within a decentralized unitary state’. Arguably, this is quite
wrong. Arguably, the Agreement means that two distinct unions make up the
UK: the Union of Great Britain and the Union of Great Britain and Northern Ire-
land. The Northern Ireland Act 1998 created an open-ended mechanism for the
Northern Ireland Assembly, by its consent, to expand its law-making autonomy
from the rest of the UK, albeit with the consent of the Secretary of State and the
approval of Westminster. No such open-ended provision was granted to the Scot-
tish Parliament or the Welsh Assembly. More importantly, the Agreement, unlike
Scottish and Welsh devolution, was incorporated in a treaty between two states,
based on the UK’s and the makers of the Agreement’s formal recognition of Irish
self-determination – the right of:
[T]he people of Ireland alone, by agreement between the two parts respectively and
without external impediment, to exercise their right of self-determination on the basis
of consent, freely and concurrently given, North and South, to bring about a united
Ireland, if that is their wish, accepting that this right must be achieved and exercised
with and subject to the agreement and consent of a majority of the people of Northern
Ireland (Government of the United Kingdom, n.d.: ‘Constitutional Issues’, 1 (ii), p. 2).
The UK officially acknowledged that Northern Ireland has the right to join the
Republic, on the basis of a local referendum, and recognized, in a subsequent
treaty, the authority of Irish national self-determination throughout the island
of Ireland. The critical constitutional question is this: in the meantime, must the
United Kingdom of Great Britain and Northern Ireland be understood as a state
in which Westminster’s sovereignty is absolute? Arguably not. The UK’s treaty
signature cannot simply be construed as meaning that Westminster’s sovereignty
overrides the Agreement itself. The Agreement’s institutions were brought into
being by the will of the people of Ireland, North and South, in concurrent ref-
erendums, and not just by the people of Northern Ireland. The functioning of
the Assembly was made interdependent with the functioning of the North-South
Ministerial Council, a cross-border and all-Ireland body, which is in the joint and
co-equal jurisdiction of the Irish government, by treaty. So, the UK’s relation-
ship to Northern Ireland, as long as it remains within the UK, and as long as the
Agreement is the basis of that membership, should be read as having an explicitly
federal character. More precisely, the makers of the Agreement were implicitly
making Northern Ireland into what Elazar (1987) terms a federacy. A federacy
is an autonomous unit of government whose relationship with its host state is
federal, even if the rest of the state is organized in a unitary fashion. A federal
relationship exists when there are at least two units of government over the same
territory and when neither can unilaterally alter the constitutional capacities of
the other. This definition is a necessary element of a federal system. A federation
normally has multiple federated units that are co-sovereign with the federal gov-
ernment throughout most of the territory of the state. The notion of a federacy, by
contrast, emphasizes that any system of constitutionally entrenched autonomy,
entrenched in a manner that constrains the central government, makes the rela-
tionship between that region and the centre functionally equivalent to a federal
relationship. It is my contention that the Agreement makes Northern Ireland such
a federacy, though not officially by that name, as long as it remains within the UK.
The Westminster parliament and executive cannot, except through breaking its
treaty obligations, and except through denying Irish self-determination, and the
referendum process, exercise power in Northern Ireland inconsistently with the
Agreement. Northern Ireland’s membership of the Union of Great Britain and
Northern Ireland, in plain language, is subject to the Agreement, and not to the
uninhibited sovereignty of Westminster. Westminster’s unconstrained authority
applies solely to explicitly excepted and reserved functions. Any changes in the
exercise or division of competencies envisaged by the Agreement requires due
Review, by the two governments and the Assembly parties, as provided under the
Agreement, and appropriate legal processes in both the Assembly and Westmin-
ster.24 This discourse is not merely academic. Its substance became critical when
the UK parliament decided to arrogate to itself the power to suspend the institu-
tions of the Agreement in 2000.
3 The First Moment of Suspension

In late 1999 and early 2000, to obtain the support of his party’s Ulster Unionist
Council, Trimble offered his party chairman a post-dated resignation letter, leav-
ing his position as First Minister, to become operative if the IRA did not start
decommissioning within his own specified timetable. The IRA did not deliver
on this unilateral deadline, at least not in the way that Secretary of State Man-
delson believed was required to stop Trimble making effective his resignation
threat – though the IRA did appear to clarify that decommissioning would occur.
In February 2000, Mandelson obtained emergency statutory powers to suspend
the Assembly and Executive and did so at 5.00 p.m. on 11 February 2000. In do-
ing so, he acted in classic Diceyan fashion, using the doctrine of parliamentary
sovereignty to arrogate to his ministry a power that had not been negotiated in
the making of the Agreement or granted in its legislative enactment in the UK.
The UK government’s officials knew that the power of suspension would breach
the formal treaty incorporating the Agreement, because in the summer of 1999,
when both governments contemplated a suspension mechanism, they proposed
that the treaty that was about to be signed by the two governments, which in-
corporated the Belfast Agreement, should be amended to make it compatible
with suspension. No such amendment was made. The Secretary of State’s jus-
24 Intra-UK federalizing processes would have been enhanced if the UK and Northern
Irish courts were to have treated Northern Ireland’s relationships to Westminster as
akin to those of the former dominions, which had a federal character, as they did in
the period of the Stormont Parliament (1921–1972). But legal colleagues advise me
that the UK’s legislative enactment of the Agreement may have modified the perti-
nent precedents in this previous jurisprudence by changing the nature of the ‘vires’
test that the courts will use to deal with jurisdictional disputes.
tification for suspension was that it was necessary to save Trimble: his threat to
resign would have become operative in an environment in which ‘Yes’ unionists
no longer commanded an absolute majority of the registered unionists in the As-
sembly. Therefore, it was feared, Trimble could not have been resurrected as First
Minister if he did resign. This reasoning was partial. The Assembly, by weighted
majority, was entitled to pass any measure to amend its current rules for electing
the dual premiers and to send this measure, after an agreed review of the Agree-
ment, to Westminster for statutory ratification. It could, for instance, propose
that when deadlocked under the parallel consent procedure the Assembly adopt
the weighted majority procedure for electing the premiers. So there was a mecha-
nism, through review and legal change within the terms and institutions of the
Agreement, under which Trimble could have regained the position of First Min-
ister. And if not, there could have been fresh Assembly elections (the most obvi-
ous intra-Agreement mechanism for resolving crises in agreeing the premiers).
Mandelson’s justification was widely criticized,25 and in Irish eyes the suspen-
sion was regarded as an unconstitutional and partisan act. The suspensory power
had not been endorsed with cross-community consent through the negotiation
of the Agreement, or in the referendums, or in the UK’s legislative enactment of
the Agreement. Neither the Agreement, nor the Mitchell Review of the Agree-
ment that took place in late 1999, had required Sinn Féin to deliver decommis-
sioning by the IRA by a deadline set by the leader of the UUP. The then formally
agreed deadline for decommissioning required all political parties to use their
best endeavours to achieve full decommissioning by 22 May 2000. One passage
of the Agreement referred to procedures for review if difficulties arose across the
range of institutions established on the entering into force of the international
treaty, the British-Irish Agreement:
If difficulties arise which require remedial action across the range of institutions, or
otherwise require amendment of the British-Irish Agreement or relevant legislation,
the process of review will fall to the two Governments in consultation with the parties in
the Assembly. Each Government will be responsible for action in its own jurisdiction
(Government of the United Kingdom n.d. 1998; ‘Review Procedures Following Implementa-
tion, 7, p. 26’, emphasis added).
The italicized passages, read in conjunction with the whole Agreement, suggest
that the UK government was obliged to consult the parties in the Assembly and
the Irish government over obtaining any power of suspension, and that any re-
medial action required the joint support of the two governments, especially as
regards their treaty. That did not happen. That each government would be “re-
sponsible for action in its own jurisdiction” did not mean that the Westminster
parliament had unilateral discretion to alter, amend, suspend, or abolish the insti-
tutions of the Agreement; it merely meant that for agreed remedial action there
25 For a critical dissection see ‘The Blame Game’, Spotlight BBC Northern Ireland, pro-
duced by Justin O’Brien, reporter Andy Davies, 22 February 2000.
would be parallel legislative procedures. For these reasons, the Irish government
did not officially recognize the suspension.
The suspension had at least three messages. First, it meant that every aspect
of the Agreement was vulnerable to Westminster’s sovereignty. Its institutions,
its confidence-building measures, its commissions, the promise that Irish unifi-
cation will take place if there is majority consent for it in both parts of Ireland,
are all revisable by the current UK parliament, and any future parliament, and
parliament’s Secretaries of State, irrespective of international law or the solemn
promises made by UK negotiators. Even had the Secretary of State’s motives been
entirely benign – and that has been questioned – his decision to obtain the power
of suspension destroyed the assumptions of nearly a decade of negotiation. Sec-
ond, the suspension suggested to some official Irish negotiators, and northern
nationalists, the necessity in future negotiations of both achieving the repeal of
the Suspension Act and entrenching Northern Ireland’s status as a ‘federacy’, per-
haps in the same manner as the UK’s courts are instructed to make European
law supreme over law(s) made by the Westminster parliament, through full do-
mestic incorporation and entrenchment of the relevant treaty. Third, unionists
may one day rue the constitutional consequences of the Northern Ireland Act
2000, the Suspension Act. What Westminster did on unionists’ behalf it may take
from them tomorrow – including membership of the Union. The suspension act
means that in UK public law the Union does not rest on the consent of its com-
ponent parts. Westminster, despite the referendums, is free, according to its con-
stitutional norms, to modify the Union in any way it likes: for example, through
full-scale joint sovereignty over Northern Ireland with the Irish government or
through expelling Northern Ireland from its jurisdiction. Therefore, both in logic
and for political reasons, the implementation of the Agreement requires the ter-
mination of the suspension power, i.e., the full repeal of the Suspension Act 2000
(O’Leary 2001a). Without that measure, the Agreement cannot be constitutional-
ized consistently with Irish national self-determination. The Joint Declaration of
May 2003 by the UK and Irish governments in which the repeal of the Suspension
Act is conditionally promised would seem to suggest that, if successfully imple-
mented, Northern Ireland’s federacy status may be reaffirmed.
The first suspension did not completely save Trimble from conflict with some
of his party activists, 43% of whom voted for a stalking horse to replace him,
the Reverend Martin Smyth MP, a hardliner and former Grand Master of the
Orange Lodge. Trimble remained leader of his party, but bound by a mandate
for reformation of the Executive that neither the UK government nor republi-
cans seemed likely to deliver. Yet, in May 2000, negotiations between the pro-
Agreement parties and the two governments produced a formula that appeared
to break the deadlock. Republicans promised to deliver a ‘confidence-building
measure’, which involved the IRA supervising international inspections of the
IRA’s arms dumps, the UK government promised to deliver fully on police reform
and demilitarization, and the UUP and the UK government agreed respectively
to withdraw Trimble’s resignation and to end the suspension. But, as we shall see,
the salience of the suspensory power would recur.
D Confederalizing and Federalizing Elements of the Agreement

The Agreement was externally confederalizing and federalizing. These assertions
rest on these definitions: confederal relations exist when political units volun-
tarily delegate powers and functions to bodies that can exercise power across
their jurisdictions; and federal relationships exist when: (1) there are at least two
separate tiers of government over the same territory; and (2) neither tier can uni-
laterally alter the constitutional capacities of the other.
1 The All-Ireland Confederal Relationship

The first confederal relationship is all-Ireland in nature: the North-South Min-
isterial Council (NSMC). Finally brought on line on the same day as power was
devolved to the Northern Ireland Assembly and Executive, 2 December 1999, it
brings together those with executive responsibilities in Northern Ireland and in
the Republic. Its first plenary meeting was held in Armagh on 12 December 1999;
the DUP Ministers did not attend.
The NSMC was the product of a bargain. Nationalists were concerned that if
the Assembly could outlast the NSMC, it would provide incentives for unionists
to undermine the latter. Unionists, by contrast, were worried that, if the NSMC
could survive the destruction of the Assembly, nationalists would seek to bring
this scenario about. The Agreement was therefore a tightly written contract with
penalty clauses. Internal consociation and external confederalism were welded
together: the Assembly and the NSMC were made ‘mutually interdependent’; one
cannot function without the other. Unionists were unable to destroy the NSMC
while retaining the Assembly, and nationalists were not able to destroy the As-
sembly while keeping the NSMC. The NSMC linked northern nationalists to
their preferred nation-state, and is one means through which nationalists hope
to persuade unionists of the attractions of Irish unification. The Irish government
successfully recommended a change to its constitution to ensure that the NSMC
and its delegated implementation bodies would be able to exercise island-wide
jurisdiction in those functional activities where unionists were willing to cooper-
ate. The NSMC functions much like the Council of Ministers in the European
Union, with ministers having considerable discretion to reach decisions but re-
maining ultimately accountable to their respective legislatures. The NSMC meets
in plenary format twice a year and in smaller groups to discuss specific sectors on
a ‘regular and frequent basis’. Provision exists for the Council to meet to discuss
matters that cut across sectors and to resolve disagreements. In addition, the
Agreement provided for cross-border or all-island ‘implementation’ bodies. The
Agreement provided an annex that listed twelve possible areas for implementa-
tion26 but left it open for others to be considered.
26 These were: agriculture (animal and plant health); education (teacher qualifications
and exchanges); transport (strategic planning); environment (protection, pollution,
water quality, waste management); waterways; social security/social welfare (entitle-
ments of cross-border workers and fraud control); tourism (promotion, marketing,
The scope and powers of these institutions was somewhat open-ended. The
Agreement, however, required a meaningful Council. It stated that the Council
‘will’ – not ‘may’ – identify at least six matters where ‘existing bodies’ will be
the appropriate mechanisms for cooperation within each separate jurisdiction,
and at least six matters where cooperation will take place through cross-border
or all-island implementation bodies. The latter were subsequently agreed to be:
inland waterways, food safety, trade and business development, special EU pro-
grammes, the Irish and Ulster Scots languages, and aquaculture and marine mat-
ters. The parties further agreed on six functional areas of cooperation, including
some aspects of transport, agriculture, education, health, the environment, and
tourism, where a joint North-South public company was established. These zones
and modes of cooperation were to be decided during a transitional period be-
tween the Assembly elections and 31 October 1998, but were not in fact resolved
until 18 December.
The NSMC differed from the previous attempt to establish a cross-border body
of a confederal kind, namely the Council of Ireland of 1974, which enraged many
Ulster Unionists and contributed to the collapse of the Sunningdale settlement.
The name change was significant, being a concession to unionist sensibilities:
Ireland is not in the title, the equality of North and South is implied. The NSMC
is a ministerial rather than a parliamentary council. There was no provision in the
Agreement to establish a North-South joint parliamentary forum as there was in
the Sunningdale Agreement of 1973, but the Northern Ireland Assembly and the
Irish Oireachtas27 are asked ‘to consider’ one.
Nationalists wanted the NSMC to be established by legislation from Westmin-
ster and the Oireachtas to emphasize its autonomy from the Northern Ireland
Assembly. Unionists preferred that the NSMC be established by the Northern
Ireland Assembly and its counterpart in Dublin. The Agreement split these dif-
ferences. The NSMC and the implementation bodies were brought into existence
by British and Irish legislation, but in the transitional period it was for the North-
ern executive and the Republic’s government to decide, by agreement, how coop-
eration should take place and in what areas the North-South institutions should
cooperate. Once these were agreed, the Assembly was unable to change these
agreements except by cross-community consent.
The Agreement explicitly linked Ireland, North and South, to another confed-
eration, the European Union (EU). It required the NSMC to consider the imple-
mentation of EU policies and programmes as well as policy proposals being con-
sidered by the EU, and makes provisions for the Council’s views to be ‘taken into
account’ at relevant EU meetings.
research, and product development); European Union programmes (such as SPPR,

INTERREG, Leader II, and their successors); inland fisheries; aqua-culture and ma-
rine matters; health (accident and emergency measures, and related cross-border
issues); and urban and rural development.
27 This is the collective name in Gaelic for the two chambers of the Irish Parliament,
Dáil Éireann and Seanad Éireann.
The signatories to the Agreement promised to work “in good faith” to bring the
NSMC into being. There was not, however, sufficient good faith to prevent the
first material break in the timetable scheduled in the Agreement occurring over
the NSMC, though this was a by-product of the crisis over executive formation
and decommissioning. The signatories are required to use “best endeavours” to
reach agreement and to make “determined efforts” to overcome disagreements
over functions where there is a “mutual cross-border and all-island benefit”.
Participation in the NSMC was made an “essential” responsibility for “relevant”
posts in the two Administrations. Here “relevant” meant any portfolio which has
any part subject to North-South cooperation. This left open the possibility that a
politician opposed to the NSMC might take a seat on it with a view to wrecking
it, but ministers were required by the Agreement to establish the North-South
institutions in “good faith” and to use “best endeavours” to reach agreement.
Since these requirements were subject to judicial review it made it unlikely that
potential wreckers would be able to take part in the NSMC for long. One of the
requirements for membership of the Executive is that ministers must “support …
all decisions of the Executive Committee” and they can be removed if they do not
– though that pre-supposes decisions being made by the Executive Committee,
and votes on exclusion by cross-community consent by the Assembly.
In fact, although the DUP boycotted plenaries of the NSMC, it did not perform
its ministerial duties illegally. The UUP First Minister David Trimble, by con-
trast, has twice been found by the Northern Ireland courts and by the House of
Lords to have unlawfully exercised his appointment powers (held jointly with the
Deputy First Minister) to the NSMC. In late 2000, to compel Sinn Féin to deliver
IRA decommissioning, he refused to nominate the two Sinn Féin ministers to
carry out their relevant duties on the NSMC. Since the relevant ministers were
fulfilling their obligations under the Agreement and the law, the courts found
in January 2001, and again in October 2001, that Trimble’s use of his discretion,
while relevant for ministers not prepared to engage in “normal participation” in
their duties, could not be used for the purposes in which he was engaged.
If the implementation of the Agreement succeeds, currently a moot point, eco-
nomic and sociological developments may underpin this new constitutional con-
federalism, and the salience of the NSMC. The Republic’s ‘Celtic Tiger’ economy
means that Northern Ireland’s ministers and citizens, of whatever background,
should see increasing benefits from North-South cooperation. And, if the EU
continues to integrate, there will be pressure for both parts of Ireland to enhance
their cooperation, given their shared peripheral geographical position and simi-
lar interests in functional activities such as agriculture and tourism, and in hav-
ing regions defined in ways that attract funds (Tannam 1999). Northern Ireland
may even come to think that it would benefit from membership of the Eurozone,
though the Northern Ireland 1998 Act, unlike the Agreement, made currency
matters non-devolved.
2 The British-Irish Confederal Relationship

There is a second, currently weaker, confederal relationship established by the
Agreement, affecting all the islands of Britain and Ireland. In the new British-Irish
Council (BIC), the two governments of the sovereign states and all the devolved
governments of the UK, and the neighbouring insular dependent territories of
the UK, can meet and agree to delegate functions, and may agree common poli-
cies. This proposal met unionists’ concerns for reciprocity in linkages, and pro-
vides a mechanism through which they might in future be linked to the UK even
if Northern Ireland becomes part of the Republic. Unionists originally wanted
the NSMC subordinated to a British-Irish, or East-West, Council. This was not
agreed. There is no hierarchical relationship between the two Councils. Indeed,
there are two textual warrants for the thesis that the NSMC is more important
and far-reaching than the BIC. The Agreement required the establishment of
North-South implementation bodies, while leaving the formation of East-West
bodies a voluntary matter; and it stated explicitly that the Assembly and NSMC
were interdependent, but made no equivalent provision for the BIC.
The development of this confederal relationship may be stunted by an Irish
governmental reluctance to engage in a forum where it may be outnumbered
by seven other British-based governments – of Westminster, Scotland, Wales,
Northern Ireland, Jersey, Guernsey, and the Isle of Man – though rules may de-
velop to ensure the joint dominance of the sovereign governments. The BIC may,
however, flourish as a policy formulation forum if the devolved governments of
the UK choose to exploit it as an opportunity for intergovernmental bargaining
within the UK, or to build alliances with the Irish government on European pub-
lic policy – in which case it will give added impetus to other federalist or quasi-
federalist processes. But perhaps the real significance of the BIC, as we shall see,
lies in its role as part of a model of double protection.
3 A UK-Northern Irish Federalizing Process

It has already been argued that the Agreement was at least tacitly federalizing
in as far as it was understood by at least some of its makers to constrain West-
minster’s sovereignty by the Agreement, making Northern Ireland a federacy.
The Agreement was a further blow to centralized unitary unionism in the UK
– already dented by the 1997–1998 referendums and legislative acts establishing
a Scottish Parliament and a Welsh National Assembly (Hazell and O’Leary 1999).
If the Agreement ‘beds down’, then the political development of a quasi-federal
relationship between the UK and Northern Ireland, with a Great Britain itself en-
gaged in quasi-federalizing change within a quasi-federalizing European Union
(King 2001), is assured whatever may be said in the drier Diceyian recesses of the
UK Constitution’s ancien regime.
4 Irish Federalizing Processes

The Agreement also opened federalist avenues in the Republic of Ireland, one
of the most centralized states in Europe. Nationalists see the NSMC, North and
South, as the embryonic institution of a federal Ireland: first confederation, then
federation, after trust has been built. This stepping-stone theory is most loudly
articulated and feared by ‘No’ unionists’, but they are not wrong in their calcula-
tion that many nationalists see the NSMC as ‘transitional’. Sinn Féin says so. Fi-
anna Fáil says so. The Irish people did not abandon their aspiration for unification
when they endorsed the Agreement, and modified their Constitution. Instead, it
became “the firm will of the Irish nation, in harmony and friendship, to unite all
the people who share the territory of the island of Ireland, in all the diversity of
their identities and traditions, recognising that a united Ireland shall be brought
about only by peaceful means with the consent of a majority of the people ex-
pressed in both jurisdictions in the island” (from the new Article 3). The amended
Irish Constitution therefore officially recognizes two jurisdictions that jointly en-
joy the right to participate in the Irish nation’s exercise of self-determination.
Unification is no longer linked to ‘unitarism’, and therefore is entirely compat-
ible with either full confederation or federation. Irish unification cannot be pre-
cluded because of present demographic and electoral trends, which have led to
a steady rise in the nationalist share of the vote across different electoral systems
(O’Leary 1990; McGarry and O’Leary 1995a: Ch 10; O’Leary 1997b; O’Leary and
Evans 1997; Mitchell, O’Leary, and Evans 2001; 2002). The unification envisaged
in the re-drafted Irish Constitution no longer has anything resembling a pro-
gramme of assimilation. Respect for “the diversity of … identities and traditions”
in the new Article 3 connects with both consociational and (con)federal logic.
The Republic is bound by the Agreement to structure its laws and its protection
of rights so as to prepare for the possibility of a (con)federal as well as a unitary
Ireland. Northern Ireland is recognized as a legal entity within the Irish Consti-
tution. So its eventual absorption or elimination as a political unit is no longer a
programmatic feature of Bunreacht na hÉireann (Constitution of Ireland 1937).
The Agreement envisages the subjection of both jurisdictions in Ireland to the
same regime for the protection of individual and group rights – a situation entire-
ly compatible with a subsequent formal confederation or federation. And, now
in accordance with the Agreement, there is an Irish Human Rights Commission
tasked with cooperating with its Northern counterpart, and possibly developing
a common Charter of Rights for the whole island.
What might happen if a majority emerged for Irish unification within Northern
Ireland, matched by one in the South ( a possibility that is not, of course, guar-
anteed)? If nationalists acquired local majority support it would not necessarily
be in their considered interests to promote the region’s immediate administrative
and legal assimilation into the Republic. They would then have a new interest in
preserving Northern Ireland as a political entity within a federated Ireland: after
all, they would be a local majority. So too would the governing coalition in the
Republic, whose calculations might be disturbed by the entry of northern partici-
pants. Conversely, some unionists faced with this prospect might prefer a unitary
Ireland as the lesser evil, calculating that their chances of being key participants
in government formation in a bigger arena might protect them better than be-
ing a minority in Northern Ireland. But that is simply one possible future, one of
many.
Meanwhile, the federalizing dimensions of the Agreement are not merely pan-
Irish or pan-British. They will evolve within a European Union that has its own
strong confederal relationships and many ambitious federalists. There will be no
obvious organizational or policy-making contradictions – though multiple net-
working clashes will arise from this extra layer of (con)federalizing – and they
might help to transfer some of the heat from binary considerations of whether a
given issue is predominantly controlled by London or Dublin.
E Double Protection and Inter-state Institutions

The subtlest part of the Agreement went beyond standard consociational, confed-
eral or federal thinking. This is its tacit ‘double protection model’, laced with ele-
ments of co-sovereignty and intergovernmentalism. It is an agreement designed
to withstand major demographic and electoral change. The UK and Irish govern-
ments promised to develop functionally equivalent protections of rights, collec-
tive and individual, on both sides of the present border. In effect, Northern Irish
nationalists are promised protection now on the same terms that will be given
to British unionists if they ever become a minority in a unified Ireland. National
communities are protected whether they are majorities or minorities, irrespective
of the sovereign stateholder – hence the expression ‘double protection’. The two
governments not only promised reciprocity for the local protection of present
and future minorities, possibly through establishing the functionally equivalent
protection of rights on both sides of the border, but they have also created two
intergovernmental devices to protect those communities. One is the successor
to the Anglo-Irish Agreement, the British-Irish Intergovernmental Conference
(B-IGC) that guarantees the Republic’s government access to policy formulation
on all matters not – or not yet – devolved to the Northern Ireland Assembly or
the NSMC. The B-IGC, in the event of the collapse of the Agreement, is likely to
resume the all-encompassing role it had under the prior Anglo-Irish Agreement.
The other device is the British-Irish Council (BIC). If Irish unification ever occurs
then the Republic’s government would find it politically impossible not to offer
the British government reciprocal access in the same forums, thereby converting
the BIC into a version of what the NSMC is now. This reciprocal set of arrange-
ments to provide for both present and possible future sovereignty arrangements
is perhaps the truest novelty in the Agreement’s design.
It is important to note what has not happened between the two sovereign gov-
ernments. Formal co-sovereignty has not been established. Unionists claim that
they have gotten rid of the 1985 Anglo-Irish Agreement in return for conceding
the formation of the NSMC. This claim is, at best, exaggerated. Under the Agree-
ment, the Irish government retains a say in those Northern Irish matters not
devolved to the Northern Ireland Assembly, as was the case under Article 4 of
the Anglo-Irish Agreement, and, as with that latter agreement, there continues
to be an intergovernmental conference, the B-IGC, chaired by the Irish Minister
for Foreign Affairs and the Northern Ireland Secretary of State, to deal with non-
devolved matters affecting Northern Ireland, and this conference continues to be
serviced by a standing secretariat – though the secretariat is no longer located

in Belfast.
The Agreement promised to “intensify cooperation” between the two govern-
ments on all-island or cross-border aspects of rights, justice, prison, and policing,
unless and until these matters are devolved to the Northern Ireland executive.
There is provision for representatives of the Northern Ireland Assembly to be in-
volved in the intergovernmental conference, but they do not have the same status
as the representatives of the sovereign governments. The Anglo-Irish Agreement
fully anticipated these arrangements (O’Leary and McGarry 1996: Chs. 6–7), so
it is more accurate to claim that the Anglo-Irish Agreement has been fulfilled
rather than simply removed.
IV Confidence-building and Responses to the Agreements

The Agreement was initially greeted with much euphoria. Nationalists over-
whelmingly endorsed it. By contrast, unionists were divided both during the
negotiations and afterwards. The absence of decommissioning of weapons by
paramilitaries before, during, and after the negotiations was the major issue for
unionist politicians, especially because they had made it one.
Describing constitutional architecture is one thing; informal political reality
is often different. What lay behind this Agreement? And can it hold together?
Everyone asks, ‘is it a house of cards, vulnerable to the slightest pressures?’ ‘Is
it vulnerable to the play of either Orange or Green cards by hardline loyalists or
republicans, or to miscalculations by less hard-line politicians?’ These are not
foolish concerns, far from it. The annual fracas at Drumcree, when the Orange
Order demands to march down the Garvaghy Road against the will of its pre-
dominantly nationalist residents; the massacre at Omagh in August 1998 car-
ried out by the Real IRA; intermittent breakdowns in the loyalist ceasefires; con-
tinuing punishment beatings by all paramilitaries; and the continuing crisis over
weapons decommissioning all revealed continuing high levels of ethno-national
antagonism.
A The Agreement on Ending the Armed Conflict

The Agreement was a political settlement that promised a path to unwind armed
conflict and thereby create a peace settlement, although, formally speaking, no
military or paramilitary organizations negotiated the Agreement. The Agree-
ment encompassed decommissioning, demilitarization, police reform, and pris-
oner releases. It addressed these issues in this order, and it is plain that although
all these issues are inter-linked they were not explicitly tied to the construction or
timing of the new political institutions – with one exception.
B Weapons Decommissioning
The Agreement was clear on decommissioning, despite the difficulties it occa-
sioned. No paramilitaries that abide by the Agreement have had to engage in
formal surrender to those they opposed in war. The Independent International
Commission on Decommissioning (IICD), chaired by Canadian General John de
Chastelain, assists the participants in achieving “the total disarmament of all para-
military organizations.” All parties, but particularly those parties that (informally)
represented paramilitary organizations in the negotiations, are required to “use
any influence they may have to achieve the decommissioning of all paramilitary
arms within two years following endorsement in referendums North and South of
the agreement and in the context of the implementation of the over-all settlement”
(Government of the United Kingdom n.d.: 20, para 3, emphasis added).
The italicized passages above clarified the termination point for decommis-
sioning, but not the moment of commencement. They also made it clear that
decommissioning was linked to the implementation of the overall settlement,
including the establishment of the governance structures (North, North-South,
and East-West) and to police reform. That is why Trimble’s demand that Sinn
Féin achieve a start to decommissioning by the IRA before executive formation
in the North was regarded as a breach of any reasonable interpretation of the text
of the Agreement. Without executive formation in the North, none of the formal
institutions of the Agreement that required the cooperation of the local parties
could get under way.
Sinn Féin nominated a representative to the IICD, issued a statement to the ef-
fect that the war was over; and for the first time issued an outright condemnation
of other republicans – of the ‘Real IRA’, whose members carried out the Omagh
bombing. It even assisted the Basque organization ETA in its organization of a
ceasefire and efforts to accomplish political negotiations in Spain. But, until No-
vember 1999, Trimble and some of his senior colleagues, harried by the DUP,
were not prepared to regard this activity as sufficient evidence of good intentions.
Each move on Sinn Féin’s part merely led the UUP to demand more, and we have
discussed the problems occasioned by the suspension precipitated by Trimble. In
response to the first suspension, the IRA withdrew its nominee to the IICD. But,
in May 2000, a package deal to restore the Agreement’s institutions and to avoid
the decommissioning deadline of 22 May was agreed. The deadline was shifted
for a year, the IRA agreed to organize confidence-building inspections of its arms
dumps and to put its weapons verifiably and completely beyond use, and the UK
government indicated it would honor the Patten Report in full. This scenario was
then run, and led to similar blockages and renewed suspensions, which is why
the term ‘Groundhog Day’ is now current amongst Northern Ireland journalists.
Trimble’s resignation threat led the IRA to withdraw its cooperation with the
IICD – and Sinn Féin made it plain that it did not think the UK had honoured its
pledges on police reform.
C Demilitarization, Police Reform, and Prisoner Release

The Agreement promised, and the UK government began, a series of phased de-
velopments to ‘demilitarize’ Northern Ireland. It did not, however, publish any
complete demilitarization plan until the Joint Declaration of May 2003. “Nor-
malization” is explicitly promised in the Agreement – reductions in army deploy-
ments and numbers, and the removal of security installations and emergency
powers, were promised “consistent with the level of overall threat”. They have
been delivered, albeit more slowly than some had expected. In the joint declara-
tion, the UK government committed itself to having no more troops in Northern
Ireland than was required by the normal defence of the UK, and that such troops
would be confined to ‘garrisons’, and most of them were to be located in areas not
known for their Irish nationalist sympathies. There was also a commitment to
address personal firearms regulation and control as an extraordinary proportion
of Northern Ireland’s citizens, mostly Protestants and unionists, have legally held
lethal weapons (Government of the United Kingdom n.d.: 21, paras. 1–4).
Police reform, as we have seen, was to be addressed through an Independent
Commission. It was to report, at the latest, some nine months before decom-
missioning was scheduled to finish. It is difficult to believe that the choice of this
timing on the part of the makers of the Agreement was an accident. The public
outline of police reform was to be available as a confidence-building measure for
republicans and nationalists before the major part of republican decommission-
ing could be expected.
The early release of paramilitary prisoners sentenced under ‘scheduled offenc-
es’, and of a small number of army personnel imprisoned for murders of civilians,
has, by contrast with decommissioning, police reform, and demilitarization, been
proceeding with less disruption than might have been anticipated.28 Measures to
assist the victims of violence have helped ease the pain occasioned in some quar-
ters by these early releases. The early-release scheme has even worked in creating
incentives for some loyalist rejectionist paramilitary organizations – such as the
Loyalist Volunteer Force (LVF) – to agree to establish a ceasefire in order to ben-
efit their prisoners (a ceasefire dishonoured more than all others combined).
D The Political Nature of the Agreement

The Agreement was an act of statecraft, and was accompanied by warm and gen-
erous language of reconciliation, but it was also based on hard-headed calcula-
tions, not pious sentiments. The Agreement rested on recognition of a changed
balance of power. The Anglo-Irish Agreement of 1985 had led to a new but ulti-
28 The definitive treatment of the management of paramilitary prisoners in Northern

Ireland has been written by a criminologist (McEvoy 2001). He observes that “the
prisoners question [i.e., their prospective release] remained the most commonly
identified reason for voting against the agreement,” especially by UUP voters (McE-
voy 2001: 335 and n. 64).
mately productive stalemate. Republicans were left with no immediate prospect

of significant electoral growth and their objective ‘to sicken the Brits’ into with-
drawal proved elusive. Loyalists reorganized in the late 1980s, and by the early
1990s were able to raise the costs of sustaining violence within the republican
constituency. Unionists had discovered the limits of just saying ‘No’ as British or
bi-governmental initiatives occurred over their heads. There was thus a military
stalemate as well as a political stalemate.
But underground structural changes, beneath the ‘frozen surface’, were noted
by the late John Whyte (Whyte 1993). These included greater equality of opportu-
nity and self-confidence among nationalists, and a shift in the demographic – and
therefore electoral – balance of power between the communities. Together these
changes underlined the fact that any political settlement could not return nation-
alists to a subordinate status. The initiatives of John Hume of the SDLP and Gerry
Adams of Sinn Féin in the late 1980s and early 1990s constructively responded to
this new stalemate. Much work had to be done before their initiative bore fruit
(Mallie and McKittrick 1996).
There was a bargain at the heart of the Agreement. Nationalists endorsed it
because it promised them political, legal, and economic equality now, plus in-
stitutions in which they have a strong stake, with the possibility of Irish unifica-
tion later, provided there is simple majority consent for that in both jurisdictions.
They were to co-govern Northern Ireland rather than being simply governed by
either unionists or the British government. They obtained this share in govern-
ment with promises of further reforms to redress past legacies of direct and indi-
rect discrimination. Republicans in Sinn Féin and the IRA therefore traded a long
war that they could not win, and could not lose, for a long march through insti-
tutions in which they can reasonably claim that only their means have changed,
not their end: the termination of partition. Sinn Féin has been extensively re-
warded for this decision as its vote has consistently increased with the peace pro-
cess, culminating in passing the SDLP as the largest nationalist party in the 2001
Westminster and local government elections (O’Leary and Evans 1997; Mitchell,
O’Leary, and Evans 2001) and in the November 2003 elections. Nationalist sup-
port for the Agreement was not difficult to comprehend. For them, it was a very
good each-way bet.
But why did the UUP and the loyalist parties make this consociational bargain?
It is probably a mistake to overemphasize personality matters, the transformation
of Trimble, the charms and latent threats exercised on him by Tony Blair, Bertie
Ahern, and Bill Clinton, or the impact of the diplomacy of American Senator
George Mitchell – though all had their registers. The mere process of multiparty
inclusive negotiations was not enough to account for Trimble’s decision to lead
his party colleagues where many of them were most reluctant to go, nor do these
factors allow for his intelligence. The unionists who supported the making of the
Agreement were concerned not so much to end the IRA’s long war but rather to
protect and safeguard the Union. As with their predecessors, they feared Brit-
ish betrayal of the Union and Irish ‘salami tactics’. Their calculus was this: they
believed that only by being generous now could they reconcile nationalists to the
Union and protect themselves against possibly seismic shifts in the balance of
demographic and electoral power. Their bet was that unionists would get a share
in self-government now, avoid the prospect of a British government making fur-
ther deals over their heads with the Irish government, and have some prospect
of persuading northern nationalists that a newly reconstructed Union offered a
secure home for them. They hoped they could work with the moderate SDLP and
inhibit the agenda of republicans. They made the Agreement, in short, to stave
off something worse.
It is not surprising therefore that there has been greater ‘rejectionism’ within
the unionist bloc: they were conceding more, and some amongst them main-
tained there was no need to concede anything, at least not yet (Evans and O’Leary
2000). Nevertheless, significant proportions of supporters of the ‘No’ unionist
parties, especially in the DUP, have told pollsters at key times they would like the
Agreement to work – which implies they are convertible to its merits, especially
if there is IRA decommissioning, and they are strongly in favour of the Assem-
bly rather than direct rule. The making of the Agreement has demonstrated that
major shifts in public opinion are possible in ethno-national conflicts (compare
O’Leary 1992; Evans and O’Leary 1997; and Evans and O’Leary 2000). Voters,
especially unionists, became more willing to accept certain compromises after
some of their leaders had endorsed them. Public enthusiasm and support for the
Agreement has, however, diminished over time, especially amongst unionists.
In one recent quota poll, only 36% of Protestants said they would still vote for
the Agreement if the referendum were held again (Irwin 2003: Table 3), and 37%
of Protestants said ‘No’ to the question ‘Do you want the Belfast Agreement to
work?’ Unionist ‘rejectionists’ believe they have seen nothing but concessions to
nationalists flow from the Agreement, but without wholesale decommissioning
by the IRA in return. They portray the entire process as appeasement. In return,
unionist opposition to the Agreement – despite the fact that the Agreement has
led to the formal recognition of Northern Ireland, the IRA’s ceasefire, and the
relative weakness of the NSMC – has reinforced hard-line republicans in their
conviction that unionists are unreformable supremacists who will not treat them
as equals.
The beauty of the Agreement appeared to be that both nationalists and union-
ists had sound reasons for their respective assessments of its merits, that is, for
believing that they were right about the long term. They could not be certain they
were right, and so they were willing to make this elaborate settlement. But will
the Agreement wither and die once it has become apparent who is right about the
long term? That possibility cannot be excluded, but that is why the Agreement’s
architecture repays careful inspection. It was not a consociation, like that of Leb-
anon, vulnerable to the slightest demographic transformation in the composi-
tion of its constituent communities. The Agreement contains incentives for each
bloc to accommodate the other precisely to make its vision of the future more
likely. That is, both have had reasons to act creatively on the basis of self-fulfilling
prophecies. So near and yet so far.
V Implementation
This chapter has deliberately treated implementation matters in the course of
outlining the institutional and confidence-building measures in the Agreement.
Their successes and difficulties by the autumn of 2003 can now be summarized,
differentiating between institutional implementation and the execution of confi-
dence-building measures.
We may say that the peace process has been, on balance, more successful than
the political settlement, to date. That is to say, the levels of political deaths are
remarkably lower than they were before the first IRA ceasefire of 1994 – though
some other indicators of violence have remained high (e.g., incidents of intimida-
tion and punishment beatings). That said, the remarkable formal political institu-
tions of the Agreement, albeit with delays in their scheduled timetables, were all
eventually established in the course of 1998–1999. Most of them have been shown
to be coherent and potentially robust and democratically accountable. But the
institutions have been plagued by instability.
Four key difficulties have been evident in the running of these institutions.
First, the UK government has obtained and used a unilateral power of suspen-
sion, not envisaged by the Agreement, a power that damages the Agreement’s
autonomy and federacy features. This power has been immensely disliked by na-
tionalists, but sought by unionists, especially in the UUP, as a bargaining chip to
compel the IRA to deliver decommissioning. The use of the suspension power
eventually bordered on the ludicrous – with the UK Secretary of State postpon-
ing scheduled elections under the power, through one-day suspensions (generat-
ing six week extensions), through permitting the executive to run despite legal
requirements for fresh elections, and finally, in May 2003, through suspending
the Assembly and scheduled elections at the same time as his Prime Minister
was promising to abolish the power (in the Joint Declaration of May 2003). If the
Agreement is to function, the suspension power has to be abolished. Second, the
dual premiership has been vulnerable to resignation threats from both the First
and Deputy First Ministers, creating reasons for either intra-Agreement Reviews,
or for extra-Agreement suspensions. Two features of the dual premiership have
made it a brittle institution. One is the interdependence of the premiers, which
gives each a ‘nuclear weapon’ resignation threat over the other (especially if the
UK’s suspension power weakens the threat posed by elections), and the other is
their election under the concurrent majority procedure, which may make filling
the posts unduly difficult, as is evident after the November 2003 Assembly elec-
tions. For these reasons, John McGarry and I have recommended that the resig-
nation power, which was not in the text of the Agreement, should be qualified,
and that a default rule (d’Hondt, modified by the requirement that the two pre-
miers should not come from the same bloc) should be applied if a concurrent ma-
jority does not exist. Third, both unionists who reject and unionists who support
the Agreement have tried to create difficulties within the Executive, by refusing
to attend plenary sessions of the NSMC and rotating their ministerial nominees
(in the case of the DUP); by refusing to nominate their MLAs to, or to oblige their
withdrawal from, their ministerial entitlements (in the case of both the UUP and
the DUP); and (on the part of the UUP) by refusing to establish or maintain the
Executive in the absence of IRA decommissioning. Pending IRA decommission-
ing and dissolution, these difficulties seem likely to persist. Lastly, First Minister
Trimble acted unlawfully in refusing to nominate Sinn Féin ministers to carry out
their duties on the NSMC – though otherwise the NSMC has been remarkably
free of anticipated frictions.
The upshot of these four institutional difficulties is an ironic historic rever-
sal. Whereas nationalists and republicans intermittently boycotted devolved ar-
rangements under the Stormont Parliament (1921–1972), it is now unionists who
regularly threaten to do so. Nationalists have become enthusiasts for autonomy/
federacy; whereas unionists are divided between the merits of devolution and the
restoration of direct rule.
All of these institutional difficulties have flowed directly from two sources. The
first has stemmed from unionist dissatisfaction, either with the failure of the con-
fidence-building measures attached to the Agreement, especially decommission-
ing, or with the inevitable repercussions of appropriate implementation of the
confidence-building measures, especially with regard to police reform, and pris-
oner releases. The second has flowed from internal political competition within
the unionist bloc. Competition within the UUP, and between the UUP and the
DUP, has rendered it extremely difficult for the UUP to be a confident and stable
coalition partner with the nationalists and the republicans who made the Agree-
ment, and has persuaded the DUP to pose as a champion of renegotiation of the
Agreement.
The confidence-building measures embedded in the making of the Agreement
have been of two kinds: the responsibilities of the two sovereign governments,
and the responsibility of agents within Northern Ireland. The Irish government
has fulfilled its immediate obligations, including the organization of the change
of its constitution through a referendum. But, it has been slow in building the
forms of human rights institutions and measures that would demonstrate its full
commitment to the form of double protection embedded in the Agreement. The
UK has had a much more mixed track-record, perhaps, if one is generous, be-
cause it has had the most to do. As the key state power, it has been the most
vulnerable to lobbying. Erratic conduct in policy has been occasioned in part
by a succession of different Secretaries of State (Mowlam, Mandelson, Murphy,
Reid, and Murphy). The UK has fulfilled its obligations with regard to prisoner
releases. It has organized better arrangements for the victims of violence. It has
made promising starts with respect to better human rights protections and laws
on equality. It has initiated some demilitarization, but, reasonably, awaits full de-
commissioning before full completion. On police reform, it has zigzagged dra-
matically. It radically diluted the Patten Commission’s proposals, then moved to
satisfy the SDLP’s complaints that it had done so, but lost Sinn Féin’s confidence
that it wanted it to be part of new policing arrangements. Its conduct, especially
over suspensions, made it seem partial to the interests and threats of the divided
moderate unionists. Its conduct, at times, has made it less likely that the IRA
would deliver on decommissioning.
The confidence-building measures that lie outside the two governments’ control
rest with paramilitaries who did not directly negotiate the Agreement. They have,
to varying degrees, maintained their ceasefires. Since 1998, the worst atrocities,
on a generally lower scale than before, have been carried out by smaller dissident
organizations. The IRA has played tit-for-tat with the UK on police reform and
decommissioning, and tit-for-tat with the UK and the UUP on the institutions of
the Agreement and decommissioning. Nothing in the Agreement warrants the
IRA’s current lateness in completing decommissioning. What inhibits the IRA is
the failure of the UK and the UUP fully to honour their obligations, reluctance to
abandon the most important republican bargaining chip, and the fear amongst
its constituents that they should not be left defenceless against unreformed police
and active loyalists. Loyalists have failed to prosper politically (Bruce 2001). By
contrast, Sinn Féin has prospered electorally (Mitchell, O’Leary, and Evans 2001),
because it is seen as the (increasingly constitutional) best nationalist champion,
and because it is the beneficiary of demographic transformations. Its successes,
despite no completion of decommissioning by the IRA, have magnified unionist
anxieties.
VI Current Status and Future Prospects

The author of this review is a critical admirer of the Agreement, a fact which
readers will have deduced.29 Five years after its initiation it is in serious difficul-
ties, but it would be very premature to conclude that it has failed. Realism obliges
consideration of how the Agreement may be best stabilized – and of appropriate
default options if it collapses. The Agreement will, of course, work best if all par-
ties and governments fulfill their remaining obligations on its implementation.
Under the Agreement, Sinn Féin and the loyalist political parties are obliged
to use their good offices to ensure the comprehensive decommissioning and dis-
armament of the paramilitary organizations associated with them. As and when
all major aspects of the Agreement for which it is responsible are fully imple-
mented by the UK government (including the repeal of the Suspension Act of
2000), it will be reasonable to have provisions enabling the exclusion from min-
isterial office of parties that maintain links with paramilitary organizations. This
should occur within the provisions of the Agreement if at all possible. A legal
path would enable the courts to determine whether parties have associations or
conduct activities in breach of the ministerial Oath of Office, and to suspend
such parties’ entitlements to ministerial office until such time as their conduct is
deemed fully democratic. This would probably require fresh primary legislation
at Westminster, passed outside the Agreement’s procedures, and would therefore
be open to the valid objection that it is ‘extra-Agreement’ (like the 2000 Suspen-
29 What follows draws from joint work with John McGarry (McGarry and O’Leary
2004a).
sion Act). ‘Juridification’ is, moreover, a difficult road. Once judges extensively
regulate political parties there may be undesirable repercussions. That might be
less pressing an issue if judges were widely regarded as impartial in Northern
Ireland, but they are not, in fact, widely representative, though that may change.
A second path is embedded in the Agreement. It provides for the Assembly to
determine whether a party entitled to ministerial nominations is in breach of
its Oath of Office – which incorporates commitments to exclusively democratic
means. But, complain unionists, this provision operates under the constraint that
it requires cross-community consent. And, now that Sinn Féin is the majority
party within the nationalist bloc, they can hardly be expected to provide such
consent. Fresh Assembly elections may create strong incentives for Sinn Féin to
deliver the IRA’s final dissolution or to disassociate itself from its twin (O’Leary
2001f; 2001e; 2001h). If that proves not to work, then other remedies should be
sought for republican failures to deliver on their obligations. The third path is
intergovernmental, and has commended itself to the two governments. It is to
empower a commission to determine, after due deliberation, whether a party is
in breach of democratic principles. This political mode of deciding on the merits
of finding a party in breach of the Mitchell principles is better than the juridical
route – though it too suffers from the fact that it has not been agreed inside the
procedures for review within the Agreement. One important proviso to the ex-
isting proposals might be suggested – the two governments in the British-Irish
intergovernmental conference should ratify any future suspension of a party’s
entitlement to office triggered by the determination of the international commis-
sion.
The loyalist parties that made the Agreement have proven electorally brittle.
The UDP has dissolved itself while the PUP has a tough future. Loyalists have
no immediate prospects of ministerial office. Consequently, the Agreement’s in-
centives do not affect their conduct in the same way as republicans. Their para-
military organizations are merely committed to decommissioning on receipt of
confirmation of the IRA’s dissolution. It is evident that loyalist – and republican
– organizations that reject the Agreement must be dealt with by the new police
service, fairly, impartially, and effectively. The ambiguous status of the loyalist
organizations that are on formal ceasefire should be reviewed by the new police
service in conjunction with the two governments. Firm and impartial handling
of current crimes by loyalists will considerably strengthen the IRA’s disposition
to dissolve.
The negotiation and the making of the Agreement divided the unionist com-
munity. It remained divided in the referendum over its adoption. Public support
for the Agreement has wavered significantly within the unionist community.
At the time of writing, it is low, and outnumbered by those disappointed by or
hostile towards the Agreement. But sufficient support to make the Agreement
work has been there when progress has been evident. Unionists are, and consider
themselves likely to be, supportive of the Agreement if it generates both peace
and prosperity (Evans and O’Leary 2000). It remains possible to vindicate this
belief. The unionist community’s political allegiances are mostly divided between
two parties. One of these, the formally pro-Agreement party, the UUP, has been
deeply divided internally. To manage the rejectionists within his party, its leader
has breached the Agreement at regular intervals. When due allowance is made
for Trimble’s tremendous difficulties in managing his party, one thought should
be uppermost in the minds of those who want to be clear-eyed about the Agree-
ment. We will only know the Agreement’s institutions are secure when their of-
fices are held and tested by those most initially opposed to or suspicious of them.
It would be strange if the Agreement’s stability required the permanent minority
status of the DUP and Sinn Féin within their respective blocs. The Agreement’s
institutions have provided incentives for both the DUP and Sinn Féin to maxi-
mize their vote and seat shares by moderating their platforms. The Agreement
cannot credibly endure if the Assembly is only brought out to play when West-
minster thinks it will be in the hands of utterly safe unionists and nationalists.
There is a reasonable prospect that the leadership of the DUP would think twice
about wrecking the Agreement if their party became the majority party – and if
other appropriate incentives are in place to clarify their leaders’ minds. There also
remains a serious prospect that ‘no unionism’ would be dealt a death-blow by the
IRA fulfilling decommissioning. It is a hypothesis that needs to be tested.
Elsewhere, I have jointly outlined how the two governments might encourage
the full liberalization of the consociational arrangements, as regards designation,
and electing the First and Deputy First Minster (McGarry and O’Leary 2004b).
The completion of policing reform will be crucial in consolidating the Agree-
ment. Without these steps, the IRA would be most unlikely to fulfill its necessary
‘acts of completion’ (of decommissioning), and there will not be political stability.
The UK is now committed, in the context of a peaceful settlement, to a robust
Policing Board, a representative police service, effective cooperation between the
new PSNI and Ireland’s Garda Síochana, the reform of the Special Branch, nor-
malized and community policing, and the devolution of policing and criminal
justice. The delegation of responsibility for policing to the Assembly and Execu-
tive will be the final proof that the settlement has taken root. It is to take place
provided it is ‘broadly supported’ by the local parties. There is no possibility of
such support unless the IRA decommissions fully. There is also little possibility
of such support if policing were to become the preserve of either nationalist or
unionist ministers. It would therefore be best if policing become a joint respon-
sibility of the two premiers, who could also take a justice portfolio, and organize
their joint office to have these two jurisdictions, justice and policing, separated
within their offices but reporting to both of them.30 The Joint Document, together
with previous proposals on the administration of justice, and developments in
the pipeline on human rights, prefigures a transformation of the administration
of justice along the necessary consociational lines. So both policing and justice
reforms now look primed to fulfill the promise of the Agreement. Public inqui-
30 This is a better idea than other options: a single Justice Department, headed by one
Minister, a Justice department rotated between different parties; and separate justice
and policing departments, each headed by a Minister from a different tradition.
ries, present and promised, may partially redress the grievances of the relatives
of the victims of unlawful state-sanctioned killings by the police and army or
through collusion between public officials and paramilitaries (for a comprehen-
sive analysis of these cases, see Ní Aoláin 2000).
Distrust characterizes inter-party relations over the final implementation of the
Agreement, but the governments of Ireland and the UK are the key guarantors
of the Agreement, and want it to work. If the Assembly and the NSMC – which
are mutually interdependent – cannot function, then the British-Irish Intergov-
ernmental Conference reverts to the functions and capacities its predecessor
enjoyed under the Anglo-Irish Agreement. Publicly highlighting this fact may
concentrate the minds of both the DUP’s and the UUP’s hardliners. The failure
of the Agreement’s Assembly would not preclude the UK and Irish governments
from deepening their cross-border and all-island cooperation, through or outside
the B-IGC. Any reasonable reading of the Downing Street Declaration (1993), the
Framework Documents (1995) and the Agreement (1998) places duties on both
governments to promote and extend such cooperation. Such cooperation would
be better than an immediate shift towards joint sovereignty arrangements – but
in this negative scenario demands for the Irish government to pursue joint sov-
ereignty will become vigourous. The incentives of this default scenario are clear.
The DUP needs to know that no working of the Agreement’s institutions by the
Northern Assembly means no devolution and the growth in the scope and influ-
ence of the B-IGC. Sinn Féin has proven, interestingly, to like devolution, albeit
as a ‘transitional arrangement’. It knows it will flourish best within the framework
of a working Agreement rather than one in default.
These incentives may prove powerful, but perhaps they are not powerful
enough. The sovereign governments might consider binding themselves. The
Agreement was intended to rectify the historic denial of Irish national self-de-
termination by the British state. According to the Agreement, and the correct
reading of Ireland’s laws and constitution, the partition of Ireland now rests on a
decision of the people of Ireland, North and South. The institutions of the Agree-
ment are a product of (constrained) Irish choices, North and South, and not the
choices of Great Britain’s parliament or people. The UK government has agreed
that, as part of the full implementation of the Agreement, it is willing to repeal
the Suspension Act of 2000. But, it is not enough that the Suspension Act be
repealed as and when all parties implement the rest of the Agreement. It would
be desirable to have the Agreement – without the UK’s Suspension Act – en-
trenched in a treaty attached as a joint and justiciable protocol to whatever new
European constitution may be proposed and agreed in the future. The European
Union’s new constitution has to be compatible with each member-state’s con-
stitution, and this would be the best way of ensuring no clash of laws between
the UK and Irish states. This proposal would constitutionalize the Agreement so
that a unilateral suspension of any of the Agreement’s institutions by the UK or
Ireland would be regarded as a breach of the EU constitution by the appropriate
court. It would also, arguably, be consistent with the volitions of two types of
nationalist, Irish nationalists and British unionists.
Postscript (Summer 2007)

It is a pleasure to report that the difficulties described in the preceding pages
have been largely resolved, and that the Agreement has been stabilized, largely
because the two sovereign governments and the major parties accomplished
what was said to be necessary. This is not the place to outline matters in any
detail, but three processes merit emphasis. First, from 2003 until 2007, the key
security questions were slowly and successfully resolved. More robust policing
reform was delivered through the conjunction of, amended Police Act (Northern
Ireland) 2000, and the Police Act (Northern Ireland) 2003, which strengthened
the powers of the Policing Board. In the Joint Declaration of 2003, the British
government promised to support the devolution of policing and justice powers,
providing it was agreed to by local parties. It detailed steps towards ‘normaliza-
tion of security arrangements’, i.e., demilitarization, over a defined time frame be-
tween then and May 2005, and promised to repeal the Northern Ireland Act 2000
(the ‘Suspension’ Act) – which it duly did. These ‘acts of completion’ by the UK
government were implemented amid matching acts by IRA, including compre-
hensive decommissioning and an end to violent and criminal activity, and Sinn
Féin’s acceptance of the police and taking of its positions on the Policing Board.
Full-scale IRA decommissioning was independently witnessed in July of 2005,
and Sinn Féin’s acceptance of the police followed in January of 2007. In October
of 2006, and again in March of 2007, the Independent Monitoring Commission
reported that the IRA had ‘abandoned’ terrorism and violence and was ‘firmly
committed to the political path’,31 paving the way for the Sinn Féin-DUP pact on
power-sharing.
Second, after 2003, the DUP and Sinn Féin decisively consolidated their elec-
toral dominance of their respective blocs, but on moderated platforms. In the Eu-
ropean parliamentary elections of 2004, the local government elections of 2005,
and the Westminster elections of 2005, Sinn Féin decisively worsted the SDLP,
and the DUP triumphed over the UUP. These contests left the two historically
hardline parties free of fear of making compromises that might cost them elec-
31 “We remain of the same view as we expressed six months ago about PIRA. We be-
lieve that it is firmly committed to the political path. It is not engaged in terrorist ac-
tivity; nor in our view does it contemplate any return to it. Its operational structures
have been disbanded and, in the absence of activity, the deterioration of terrorist
capability continues. The organisation does not engage in acts of violence and has
instructed its members not to do so. The leadership continues to encourage mem-
bers to undertake political or community activities. We have no reason to think that
it will be diverted from continuing along this path and we note the decision of the
Sinn Féin Ard Fheis on 28 January 2007 to support policing and the criminal justice
system and subsequent positive statements to that effect. We therefore conclude that
terrorism and violence have been abandoned and that PIRA does not pose a threat
relevant to security normalization”, 14th Report of the Independent Monitoring Com-
mission, 12 March 2007, http://www.independentmonitoringcommission.org/docu-
ments/uploads/14th_IMC_Report.pdf
torally, and both decided, to shift from being pure ‘tribunes’ of their respective
peoples, to being their respective consuls. In the 2007 Assembly elections, the
DUP took 30%, and Sinn Fein 26% of the first-preference vote, with the SDLP
and the UUP trailing with 15% each (the SDLP won less seats than the UUP on a
slightly higher vote share). These results preceded the formation of an Executive
with six unionist and four nationalist ministers, headed by Ian Paisley and Martin
McGuinness as the First and Deputy First Ministers, respectively. Their personal
relationship seemed remarkably more relaxed and warmer than that between Da-
vid Trimble and Seamus Mallon.
Lastly, to deliver this benign scenario, the two sovereign governments did what
was necessary. The UK government abandoned Mandelson’s strategy of shoring
up the UUP – and, under Secretaries of State Murphy, Reid and Hain, they went
back to the full agenda of the 1998 Agreement, delivered on police reform, and,
jointly with the Irish government, made it plain that any failure on the part of
unionists to work within the new institutions (after republicans had comprehen-
sively delivered the end of the IRA) would lead to a plan B – in which British
and Irish intergovernmental cooperation would resemble joint sovereignty. This
stance incentivized the DUP to share power, because it could claim its intran-
sigence had delivered the end to the IRA’s war-machine and tell its supporters
that it was necessary to share power with democratized republicans to prevent a
worse scenario. Republicans were, as suggested above, incentivized by the pros-
pects of an increased share in power and the leadership of northern nationalists.
The UK Treasury engaged in both arm-twisting and sugar-coating of funding to
ensure that the new Assembly would begin life with a manageable fiscal balance-
sheet. More importantly, however, in the Saint Andrews Agreement of 2006, the
two governments agreed and followed through on two of the ideas emphasized
here. With the agreement of the parties they made the election of the two pre-
miers the decision of the two largest parties, without requiring that they vote for
each other’s candidate, in effect applying d’Hondt to the premierships; and they
agreed the ending of the UK’s Suspension Act once the parties could agree on a
working Executive. They also adjusted the ministerial code and formally intro-
duced consociational decision-making within the Executive.
There are places and times where integration is the appropriate normative
strategy for resolving conflicts. Contemporary Northern Ireland is not one of
them. A British or an Irish integration strategy has been available and either im-
posed or demanded throughout the last century – orthodox Unionism and Irish
nationalism respectively. These rival programmatic visions have been at the heart
of conflict. If the current institutions endure, a common Northern Irish identity
may come to be shared by most unionists and nationalists, but that will be the
work of at least two decades, and it will be consociation that eases the path to this
shared identity. This complex consociational and territorially pluralist settlement
invalidates the proposition that consociations are never made in or never appro-
priate for deeply divided places. We shall see whether it will show that consocia-
tions can be stable over the longer run amid nationally divided peoples. Its chanc-
es are fair. Meanwhile critics of the Agreement are free to mobilize and transform
the identities that they hold culpable for the conflict. They are equally free to
transform the institutions that they believe – wrongly – freeze these collective
identities. If they succeed, I would say consociation enabled them to succeed. It is
feasible that, within this century, a majority will emerge within Northern Ireland
that favours an all-Ireland state. But, if that happens and if it is matched with
sufficient support in independent Ireland, the re-unification of Ireland should
take a confederal or federal form, preserve consociational arrangements within
Northern Ireland, and keep Northern Ireland linked to Great Britain through the
British-Irish Council – if that is what unionists want. The experience of stable
consociation within the sovereign framework of a UK with confederal links to
Ireland will ease any such transition.
References
Constitutional Texts, Agreements, Treaties and Reports, all of which can be found
on the Cambridge-Carnegie Project website.32
1) Ireland
Bunreacht na hÉireann 1937, Constitution of Ireland
2) United Kingdom
Government of Ireland Act 1920
Northern Ireland Act 1998
Northern Ireland Act 2000
Police (Northern Ireland) Act 2000
3) United Kingdom and Ireland

The Anglo-Irish Agreement 1985, known as the Agreement between the Government of
Ireland and the Government of the United Kingdom in Ireland, and known as the
Agreement between the United Kingdom of Great Britain and Northern Ireland and
the Republic of Ireland in the UK.
The British-Irish Agreement 1999
The Joint Declaration May 2003
The Saint Andrews Agreement 2006
4) Intergovernmental and inter-party Agreement

Government of the United Kingdom (1998) ‘The Agreement: Agreement Reached in the
Multi-Party Negotiations’, No place of publication listed: No publisher specified.
5) Reports
Patten, C. et al. (1999) “A New Beginning: The Report of the Independent Commission
on Policing for Northern Ireland”, Belfast: Independent Commission on Policing for
Northern Ireland, and published at http://www.belfast.org.uk/report.htm.
32 http://www.intstudies.cam.ac.uk/centre/cps/.
Chapter 5
Resolving the Bougainville Self-determination Dispute:
Autonomy or Complex Power-sharing?
Anthony J. Regan
I Introduction
The development of a distinct ethnic identity by the people of Bougainville, a
remote part of Papua New Guinea, is a relatively recent phenomenon, occurring
only during the twentieth century. It has, however, contributed to two separate
secession attempts. The first, in 1975, was resolved with almost no violence by a
1976 settlement. The second was part of a deeply divisive violent conflict from
1988 to 1997. While a political settlement was reached in August 2001, a signifi-
cant minority faction continues to espouse the secessionist cause.
The 1976 settlement involved territorial autonomy arrangements developed
without significant international involvement. Extended to the other eighteen
provinces of Papua New Guinea (“symmetrical” autonomy (Ghai 2000b)), the
arrangements operated reasonably well in Bougainville until the violent conflict
erupted in 1988. The Bougainville Peace Agreement signed on 30 August 2001
that formally ended the second conflict also gives a central place to territorial
autonomy. But not only is the autonomy far more extensive than under the previ-
ous arrangements, it is also both applicable only to Bougainville – what Ghai calls
“asymmetrical autonomy” (Ghai 2000b) – and is part of a much wider agreement
that provides a regime of special rights for a small minority of the Papua New
Guinea population.
Other major elements of the 2001 agreement include deferral of the self-deter-
mination question for Bougainville to a non-binding referendum on Bougainville’s
future political status (inclusive of independence), arrangements for amnesty and
pardon in relation to crimes committed during the conflict, a complex demilita-
rization process for Bougainville – the progress of which determines the imple-
mentation of the autonomy and referendum arrangements – and constitutional
entrenchment provisions that require Bougainville consent to any alteration of
the agreed arrangements. The international community has been extensively in-
volved in facilitating and monitoring the ongoing peace process established in
1997 from which the 2001 political agreement emerged, and in mediating aspects
126 Anthony J. Regan
of the negotiation of that agreement, and managing or monitoring key aspects

of it.
This chapter examines the origins, main features and implementation of the
power-sharing arrangements in the Bougainville case, with particular emphasis
on the main features and implementation of the 2001 Agreement. At the same
time, however, brief comparisons between the agreements of 1976 and 2001 and
the conflicts that produced them provide insights into both the reasons why the
two agreements were necessary, and the significant differences between these
two sets of power-sharing arrangements. The comparison also raises questions
about the sustainability of the 2001 agreement. The central issues examined in-
clude:
• the geographical, political and economic factors that contribute to the focus
on territorial autonomy (rather than mechanisms for integration into the
central government) as the central aspect of power-sharing arrangements in
1976, and a key aspect in 2001;
• whether the ending of the self-determination dispute in 2001 would have
been possible through other strategies – for example, some form of assimila-
tion, or arrangements for separation;
• whether the resolution of the conflict would have been possible without in-
ternational involvement in both the process that produced the 2001 agree-
ment and its implementation;
• the extent to which that agreement goes beyond territorial autonomy, and
whether it meets the criteria for complex power-sharing;
• whether the power-sharing agreements are more likely to entrench division
or resolve it;
• whether the particular form of power-sharing in the agreement is likely to be
permanent or transitional.
In relation to the latter two points, of central concern is the question of whether
the 2001 agreement offers the prospect of a lasting settlement of the secession-
ist dispute, particularly in light of the weakness of the state in, and the political
culture of, Papua New Guinea and Bougainville, and the inherent tension of a
power-sharing agreement that explicitly keeps open the secession issue by pro-
viding a constitutional guarantee of a referendum on independence within a de-
fined period.
II Background – The Self-determination Dispute
A Papua New Guinea and Bougainville – Origins of the Dispute

Bougainville, a group of islands 1,000 kilometres east of Papua New Guinea’s na-
tional capital, Port Moresby, is the most remote of nineteen provinces. Its 9,438
square kilometres is about 2% of Papua New Guinea’s land area. Geographically,
culturally, and linguistically Bougainville is part of the Solomon Islands chain
rather than the groups of islands that, together with the Eastern half of the island
5 Resolving the Bougainville Self-determination Dispute: Autonomy or Complex Power-sharing? 127
of New Guinea, make up Papua New Guinea. It was a consequence of some of the
last throes of European colonial competition in the late nineteenth century that
saw Bougainville became part of German New Guinea rather than the British
colony of Solomon Islands.
New Guinea’s approximately 5.1 million speak twenty-one distinct languages,
eight sub-languages and thirty-nine dialects (Allen and Hurd 1963; Regan 2002c:
21), with considerable cultural diversity both within and between language groups
(Ogan 1991). Cultural and linguistic diversity exists, however, throughout Papua
New Guinea (home to over 800 language groups), and the main distinctive feature
shared by most (but not all) Bougainvilleans is very dark skin colour. It was almost
certainly only during the last 100 years, originally in the context of colonial devel-
opment, that skin color became the marker of a Bougainvillean identity distinct
from the rest of Papua New Guinea. That identity is asserted mainly in dealings
with the rest of Papua New Guinea, while many quite different ethnic identities,
based on local language or culture groups, are asserted in dealings between Bou-
gainvilleans.
Bougainville’s close contact with the outside world and its integration into
Papua New Guinea are recent, with the first permanent colonial administrative
post (under German New Guinea) being established in 1905. Australia was the
main colonial power from 1914 until Papua New Guinea’s independence in 1975.
It was under Australian rule that one of the world’s largest copper and gold mines
was established at Panguna in central Bougainville, operating from 1972 to 1989
when it was closed by violent conflict. It was the only large mine in Papua New
Guinea until the mid-1980s, and was crucial to the economy of both Papua New
Guinea and Bougainville. The imposition of the mine by the colonial regime with-
out consultation with Bougainville and for the economic benefit of Papua New
Guinea rather than Bougainville was widely resented in Bougainville. Neverthe-
less, together with widespread plantation and smallholder cocoa production, the
mine made Bougainville Papua New Guineas’ wealthiest province prior to the
conflict.
The state in Papua New Guinea (including Bougainville) has always been weak,
at all levels, with limited impact on local communities and difficulties in impos-
ing policies on those determined to oppose them. Like the rest of Papua New
Guinea, there were no pan-Bougainville political structures under the centralized
colonial administrative structure. Indeed, elected local governments were only
established gradually in Bougainville from the late 1950s onward, and in some
areas were resented and resisted as a form of colonial control (Connell 1977; Grif-
fin 1977).
Pre-colonial social structures based on small land-holding clan lineages have
been resilient. A minority of societies have hereditary leadership, the rest largely
have performance-based leadership, though often with a hereditary element.
Most societies are matrilineal, with women regarded as the custodians of land
For more on the development of Bougainvillean identity and the process of its
politicization, see Griffin 1977; 1982; Nash and Ogan 1990; Ghai and Regan 2000.
and most other forms of wealth in customary society. While women tend to have
high stature within their societies, they tend to play limited public roles under
customary arrangements, with uncles and brothers usually speaking on their
behalf in public discussion, even on land matters. It is still often far from easy
for women to play public roles. Most Bougainvilleans still live in isolated rural
communities that rely heavily on subsistence agriculture – both contributing to
their high level of de facto autonomy from the state and continued cultural and
linguistic diversity. (Ghai and Regan 2000; Ogan 1991; Terrell 1986). At the same
time, the fact that Bougainville society comprises hundreds of small societies
with numerous identities around which much political mobilization occurs adds
to the complexity of ‘modern’ politics, government decision-making processes,
and institution building (political and bureaucratic).
A Bougainville movement for separation from Papua New Guinea developed
slowly, based on a growing sense of a distinct identity, continuing links with Solo-
mon Islands (reinforced by links of the two main churches to ‘parent’ houses
in Solomon Islands), and reinforced by grievances against the colonial regime,
which contributed to movements of resistance to colonialism (Kiki 1968: 104-
125; Ogan 1972; Oliver 1991; Rimoldi and Rimoldi 1992; Regan 2002c). Separatism
came to the fore in the late 1960s and intensified in the early 1970s, largely be-
cause of resentment about the Panguna mine. It was seen as having been imposed
without consultation, paying a minimal revenue share both to Bougainville as a
whole and to the owners of land used for mining purposes, and contributing to an
influx of people from other parts of Papua New Guinea attracted by employment
and other economic opportunities associated with the mine (as well as with plan-
tations). Outsiders were seen as undermining Bougainville culture, threatening
Bougainville identity, and challenging Bougainvilleans’ rights to land as well as to
employment and economic opportunities associated with the mine.
B The First Self-determination Dispute and the 1976 Agreement

While high expectations of secession as the solution to all Bougainville’s problems
tended to be generated among ordinary people from the late 1960s, for many of
the more educated Bougainvillean leaders secessionist claims largely involved an
ambit claim for a better share of the Panguna mine revenue and a reasonable de-
gree of autonomy for Bougainville. Strident secessionist demands made from late
1972 were gradually modified into a push for autonomy under pan-Bougainville
political structures. Disagreements with the central government over funding
and the constitutional entrenchment of arrangements for political decentraliza-
tion to elected provincial governments consolidated support for the secessionist
cause, and Bougainville’s unilateral declaration of independence (UDI) was made
There is an extensive literature on the impact of the mine, including Bedford and
Mamak 1977; Denoon 2000; Griffin 1982: Mamak and Bedford 1974; Ogan 1972; Con-
nell 1991; Thompson 1991; Griffin and Togolo 1997, as well as other works cited in the
bibliography in Wesley-Smith 1992.
on 1 September 1975, only days before Papua New Guinea’s own gaining of inde-
pendence from Australian rule on 15 September 1975.
The 1970s Bougainvillean leadership was essentially moderate, with little stom-
ach for protracted conflict, especially when it became clear that there was no
prospect of international support for secession. Fearing the divisiveness of inter-
nal conflict and the possibility that other areas might follow Bougainville’s exam-
ple, Papua New Guinea was conciliatory. Negotiations in 1976 saw Bougainville
accepting Papua New Guinea sovereignty and provincial government arrange-
ments constitutionally entrenched.
Although Bougainville wanted to be treated as a special case, the national gov-
ernment feared the divisiveness of such arrangements, and the new system of
provincial governments with potentially extensive legislative powers was gener-
alized to the other eighteen provinces. The provincial government system oper-
ated reasonably effectively in Bougainville (Griffin and Togolo 1997; Ghai and
Regan 2000). But, by the mid-1980s, Bougainville’s provincial government was
finding that it had limited power to respond to either the ongoing tensions as-
sociated with the unresolved grievances of ordinary people concerning the mine
or to the tensions associated with rapidly increasing economic inequality, due to
both the mine and related economic activity and escalating small-holder cash
crop activity (Regan 1998).
These tensions were evident in the closely related phenomena of growing re-
sentment of outsiders and concerns about the undermining of traditional author-
ity and culture, something probably due mainly to rapid social and economic
change, but frequently attributed to outsiders. Resentment against outsiders was
fuelled by their involvement in competition with Bougainvilleans for economic
resources in circumstances of growing inequality. The general direction of eco-
nomic change – especially the growing economic inequality that was so much
at odds with the relatively egalitarian nature of traditional society (Regan 1998)
– was itself undermining traditional society. Concerns people expressed in this
regard and the blame apportioned to outsiders were merely symptomatic of the
deeper tensions that the provincial government was not able to manage (Ghai
and Regan 2000). Nevertheless, while interest in secession continued, there was
little evidence of active political mobilization around secessionist goals for most
of the 1980s.
For more detailed evaluations of the origins and operation of the provincial
government arrangements generally, see Conyers 1976; Ballard 1981; Ghai and Regan
1992; May and Regan 1997; May 1999; and Ghai and Regan 2000. For discussion of the
operation of the arrangements in relation to Bougainville in particular, see Griffin and
Togolo 1997.
C The Second Self-determination Dispute 1988–2001

The catalyst for a rapid and generally unexpected renewal of widespread support
for secession was the police response to damage to Panguna mine property in
November 1988. In turn, this was precipitated by an intergenerational dispute
among mine lease landowners that had developed from the mid-1980s over the
distribution of mine rents and compensation. Leaders of the frustrated younger
landowners unsuccessfully seeking access to increased revenue shares developed
coalitions with other groups with mine-related grievances. Damage to mine
property was probably intended to apply pressure to the mine and to the central
government to increase mine revenue shares rather than to close the mine per-
manently. There were, however, groups in the emerging coalition of dissent that
did seek mine closure (Regan 2003).
The weak Papua New Guinea state had limited capacity to analyze and respond
to the problems facing it, and was under intense pressure due to fears of the severe
economic impacts of mine closure. Although there were always some in both the
political, bureaucratic, and military leadership that opposed the use of violence
against the Bougainville dissidents, many regarded the emerging Bougainville
conflict as simply a law-and-order issue. The initial responses involved sending
in the Police Mobile Squads and, when it became clear that was not working, the
Papua New Guinea Defence Force (PNGDF) was deployed. It was the almost ran-
dom police violence that welded the emerging coalition of Bougainville dissidents
together around an ethnic divide and put secession on the agenda again as the
simple answer to a myriad of problems.
The mine closed permanently in May 1989, and much government and planta-
tion infrastructure was destroyed. Some non-Bougainvilleans were harassed, and
most fled Bougainville, resulting in an effective ‘ethnic cleansing’ of 15,000 to
20,000 people, mainly from other parts of Papua New Guinea. They were accom-
panied, however, by a large part of the educated Bougainvillean elite, elements
of which were blamed by ‘levelling’ elements of the dissidents for the growing
economic inequality of 1980s Bougainville.
By early 1990, the situation was out of control. A ceasefire agreement reached
in March 1990 resulted in the withdrawal of all national government forces, but
failed to provide the basis for some form of negotiated settlement as the national
government had expected. Instead, the rebel fighters, by now known as the Bou-
gainville Revolutionary Army (BRA), sought to take control. The BRA was made
up of locally based and loosely affiliated groups developed to fight a guerrilla war.
BRA members held a wide range of ideological positions. Some had as much
For more on the issues concerning the origins and unfolding of the conflict, see May
and Spriggs 1990; Spriggs and Denoon 1992; Regan 1998; 2001a.
Notably mine workers disgruntled about what was seen as lack of opportunities for
advancement on the part of Bougainvillean employees, as well as other landowner
groups who felt under threat from squatter settlements and purchasers of customary
land from elsewhere in Papua New Guinea.
criminal motivation as anything else. The BRA leadership saw both the Papua
New Guinea and Bougainville governments as opponents, and, with the depar-
ture of the Papua New Guinea forces, virtually all forms of government adminis-
tration and services collapsed quickly. Without hierarchical and well-disciplined
structures, the BRA leadership experienced great difficulty in asserting control,
and a climate of semi-anarchy prevailed.
The national government responded to the BRA by imposing a sea and air
blockade in April 1990. The BRA proceeded, however, to set up its own civilian
government – the Bougainville Interim Government (BIG) – and in May 1990
made Bougainville’s second unilateral declaration of independence (UDI). Over
the next few years, BRA and BIG efforts to seek international recognition of Bou-
gainville’s secession were unsuccessful.
From mid-1990, the conflict gradually developed two main dimensions –be-
tween Papua New Guinea and Bougainville, and between Bougainville groups.
Localized armed opposition to the BRA developed. Leaders in various areas
sought the return of the Papua New Guinea forces to assist them. The first troops
landed in Buka (a large island just north of the main island of Bougainville) in
September 1990, and their presence gradually expanded to mainland Bougain-
ville areas from 1991. The situation soon developed into something of a low inten-
sity civil war between armed Bougainvillean groups – the BRA versus a gradu-
ally developing coalition of armed groups eventually known as the Bougainville
Resistance Forces (BRF). At the same time, the BRA continued their ever-more-
bitter secessionist guerrilla war. The BRF, and the leaders of the communities
they were based in, opposed the secessionist goals of the BRA, some because of
a recognition of economic and other benefits of belonging to Papua New Guinea
(something particularly true of the educated elite leadership from Buka, an area
where secessionist support had long been less pronounced than in other parts of
Bougainville), but many not so much for ideological reasons but simply because
they feared being part of an independent Bougainville dominated by their oppo-
nents in the BRA and BIG.
The years 1990 to 1997 witnessed complex and costly conflict. An unknown
number of people died – perhaps as many as several thousand. There was mas-
sive disruption of normal life, with little in the way of basic services available.
About one third of the Bougainville population was displaced, and was forced to
live in refugee camps. Bougainville factions opposed one another in part through
parallel government and military institutions – BIG, councils of chiefs (local gov-
ernments established by the BIG), and the BRA, versus the Bougainville Transi-
Probably well over 1,000 in combat, others through extra-judicial executions on

all sides, and an unknown number of deaths as a result of, or contributed to, by
the Papua New Guinea blockade. However, reports of 15,000 to 20,000 deaths are
almost certainly exaggerated.
tional Government (BTG), councils of elders (local governments established by

the BTG), and the BRF.
By early 1996, it appeared that the parties were so far apart and the conflict
and divisions escalating so rapidly that a peace process and a settlement seemed
most unlikely. Further, the BRA seemed to have a decided advantage in the con-
flict with the PNGDF. At the same time, however, there was increasing evidence
of war-weariness among the population of PNG generally, and Bougainville in
particular. However, although the conflict escalated during 1996, developments
from late 1996 and in the first part of 1997, documented elsewhere, resulted in a
peace process developing from mid-1997.
III Peace Process – Internal Dynamics and External Actors

One of the key aspects of complex power-sharing is the active engagement of
external actors in the making, monitoring, and management of the agreement
that ends a self-determination dispute. In the Bougainville case, such actors have
played major roles in the development of the key agreement, the Bougainville
Peace Agreement of August 2001.
It must be emphasized, however, that despite the significant roles of external
actors, it was in fact the parties to the conflict that provided the impetus for the
peace process. Despite the deep divisions generated by the conflict, there were
enough leaders on all sides convinced that the dangers of continued violent con-
flict outweighed any likely benefits (Regan 2001a). Further, it was also the parties
themselves that recognized that the depth of divisions between them was such
that external actors were required to assist in various aspects of facilitating the
peace process, as well as aspects of negotiation and implementation of the politi-
cal settlement. There was no need for coercion, or even significant pressure, from
the international community.
On the other hand, the divisions between the parties were so significant that
the first two years of the process – mid-1997 to mid-1999 – were devoted to con-
solidating the process rather than to efforts to negotiate a political outcome.
Both sets of divisions had to be addressed – those among Bougainvilleans and
those between Bougainville and Papua New Guinea (Regan 2002b). Before po-
litical negotiations between Bougainville and Papua New Guinea could be held,
a common Bougainville negotiating position had to be developed. The process
of unifying Bougainvillean positions was not without its difficulties. Splits in the
Bougainville groups supporting the peace process occurred during 1998–1999,
resulting in significant elements of those that supported integration into Papua
New Guinea refusing to join the coalition of ‘moderate’ groups (including most
The BTG was the provincial government for Bougainville, re-established under
national law with a name change in 1995, the prior Bougainville provincial government
having been suspended in 1990.
See Regan 1997; 2001a; Dinnen, May, and Regan 1997; Dorney 1998; O’Callaghan
1999; Adams 2001; Carl and Garasu 2002.
BRA and BIG leaders) that began negotiating with the national government in
June 1999 for a political settlement. Negotiations and reconciliations among the
leadership were required before the dissident elements joined the negotiations in
December 1999 (Regan 2001a; 2002b).
However, the most serious and, in mid-2003, ongoing unity problem has in-
volved Francis Ona, the original leader of the BRA and president of the BIG.
He has refused to support the peace process, claiming Bougainville is already
independent as a result of the UDI he made in May 1990, and declaring that he
leads a Republic of Me’ekamui. Elements of the BRA supporting Ona formed
the Me’ekamui Defence Force (MDF) in 1998, and do not support the weapons
disposal process (below). His group has some control over an area of the moun-
tainous interior of Bougainville around the mine site, and some pockets else-
where. While the refusal of Ona and his supporters to join the process probably
assisted the Bougainvilleans in their negotiations with the national government
(highlighting the continued depth of support for secession), they remain a desta-
bilizing element in the Bougainville political situation, and could contribute to
what can be expected to be considerable difficulties in the implementation of the
political agreement.
Turning to the roles of external actors, several aspects require brief comment.
First, the main external participants have been the governments of countries
from the Pacific region – especially Australia and New Zealand, but also Solo-
mon Islands, Fiji, and Vanuatu. The United Nations has provided a small observ-
er mission since 1998 – the United Nations Observer Mission on Bougainville
(UNOMB). Without extensive international input to facilitation, trust-building,
mediation, and some support in implementation, it is difficult to see how the
peace process could have continued for long. Division, distrust, and bitterness
generated by the conflict made all sides look to external actors to assist them in
the process of resolving the conflict, and to do so from as early as 1990. For their
part, the governments from the region have not felt any significant pressure to
impose a settlement. They have, in general, been cautious about their involve-
ment, generally acting only in response to invitation from the Papua New Guinea
government and the Bougainville groups.
Both Australia and New Zealand had had little prior experience of responding
to conflict in their region and so needed to develop new policies and capacities.
The actual roles played by the external actors have been discussed elsewhere (Re-
gan 2001a; 2001b; Adams 2001; Carl and Garasu 2002; Regan and May 2001).
Their main direct contribution has been the provision of unarmed monitors of
the truce and ceasefire arrangements agreed between the parties in October 1997
and April 1998, respectively (the Truce Monitoring Group (TMG) and the Peace
Monitoring Group (PMG)). 10 In this role, they have been supported by the gov-
Something akin to ‘sacred land’ in the Nasioi language of Central Bougainville.

10 On the origins and development of the TMG and the PMG, see Regan 2002b, for
discussion of the operation of the TMG, see Adams 2001, and for operation of the
PMG in its first two years, see Wehner and Denoon 2001.
ernments of Fiji and Vanuatu, who have provided small numbers of personnel,
thereby ensuring that support for the peace process had a strongly regional fla-
vour, and also ensuring that Australia and New Zealand could not be criticized
for excessive interference in the affairs of Pacific Island states (Regan and May
2001). Solomon Islands played an important role early in the process, especially
through the taking of chairing roles in early negotiations. The roles of New Zea-
land and Australia went well beyond the provision of monitors. They also facili-
tated the peace process, through hosting talks and funding many aspects of the
process within Papua New Guinea. They also mediated in negotiations at crucial
points – New Zealand in the critical period when the process was being estab-
lished (1997–1998), and Australia during the political negotiations.
Australian and New Zealand involvement was driven largely by concerns about
the Bougainville conflict contributing to the risks of instability in the region.
Since the 1986 coups in Fiji, there has been a growing perception that instability
in the southwest Pacific could threaten Australian and New Zealand regional in-
terests. Both countries also share concerns that the state in Papua New Guinea is
so weak that it could be at risk of collapse in the not-too-distant future, and that
the Bougainville conflict was placing additional stresses on an already weakened
state that needed to be ameliorated. An additional factor for New Zealand was
an interest in asserting a greater role in the Southwest Pacific, as opposed to its
‘traditional’ sphere of influence in the Polynesian Pacific (Regan and May 2001).
Since the peace process has begun, significant instability in neighbouring Solo-
mon Islands – that has in some ways been influenced by the example of the con-
flict in Bougainville – has added to such concerns, especially as there continues
to be some interest on both sides of the border in the possibility of union between
Bougainville and the western parts of Solomon Islands. In June 2003, concern
about the possible collapse of the state in Solomon Islands in the aftermath of
ethnic violence, a June 2000 coup and the subsequent criminalization of the state
by ethnic militias has encouraged the Australian government to announce a new
policy initiative of ‘cooperative intervention’, which from July 2003 has seen it
become far more directly involved in the running of Solomon Islands, including
initial police action (with military support) to disarm ethnic militias and restore
the integrity of the Solomon Islands police system, and the provision of personnel
and other forms of support to build state capacity.11 New Zealand is also actively
involved in these actions, and, following the model that worked so well in Bou-
gainville, in an effort to ensure the legitimacy of the initiative, other Pacific states
(including Papua New Guinea) are contributing either or both military and/or
police personnel.
Second, it was the BRA and the BIG, rather than the Papua New Guinea gov-
ernment, that initially sought the involvement of international actors. This was
in part because of distrust of the Papua New Guinea forces, but was also part of
11 For more on the problems in Solomon Islands, see Dinnen 2002; Wainwright 2003.
For more on the new Australian policy, see Wainwright 2003; O’Callaghan 2003;
Kelly 2003.
broader efforts made by the BRA and BIG from the early 1990s to ‘international-
ize’ the conflict, in the hope of fostering international support for secessionist
goals. Recognizing this strategy, Papua New Guinea, although open to regional
facilitation of the process, was also concerned that involvement of external actors
might undermine national sovereignty. In particular, there was opposition to-
wards direct involvement of the United Nations until early 1998, and very strong
pressure from the Bougainville side was needed to get Papua New Guinea agree-
ment for establishing the UNOMB in the January 1998 Lincoln Agreement. On
the other hand, once the involvement of the United Nations was agreed, the na-
tional government was scrupulous in its efforts to facilitate that involvement.
Third, mediation by external actors was of critical importance at two points
during the political negotiations in resolving the most divisive issue in the po-
litical negotiations – Bougainville’s demand for a referendum on independence.
First, in March 2000, when the Papua New Guinea government refusal to ac-
cept that the referendum issue should even be on the agenda for discussion was
threatening to lead to deadlock, the UNOMB mediated an agreement to include
that issue on the agenda. Then, in December 2000, there was deadlock again, this
time over the way in which the referendum issue should be dealt with. At this
point, the Australian Minister for Foreign Affairs and Trade, Alexander Downer,
played a crucial mediating role in proposing that the national government accept
a referendum deferred for at least ten years, provided that the Bougainvilleans
agreed that the outcome was not to be binding but would instead be subject to
the ultimate authority of the national parliament. The essence of his argument to
the national government was that if the outcome of the referendum was subject
to the authority of the parliament, then national sovereignty would not be com-
promised. The essence of his argument to the Bougainvillean groups was that
there was never any legal provision making the outcome of the East Timor refer-
endum of 1999 binding, and that rather it was the international community that
ensured that the outcome was honoured once it was clear that an overwhelming
majority of East Timorese wanted independence. Anxious to avoid the dangers
of deadlock, both sides accepted the proposed compromise, opening the way to
resolution of all other outstanding issues.12 In the process, however, high expecta-
tions were created in Bougainville about the international community playing an
ongoing role as guarantor of the Bougainville Peace Agreement, a role expected
to extend for the ten to fifteen years within which the referendum must be held.
IV The 2001 Agreement – Beyond Autonomy

The 2001 Bougainville Peace Agreement (the 2001 Agreement) aims to end the
self-determination dispute in two distinct ways. The first is by an agreement for
the holding of a non-binding referendum for Bougainville on the self-determina-
tion question. The second is by providing Bougainville with the ability to decide
12 For discussion of the negotiation of the political settlement, see Wolfers 2002; Regan
2002b.
its own future in another way, through the exercising of a high level of autonomy
in the period until the referendum is held. Papua New Guinea’s reluctant conces-
sion of the almost unique combination of constitutionally-guaranteed referen-
dum and complex autonomy arrangements is, in part, based on the hope that the
operation of those latter arrangements will not just reduce the likelihood of re-
newed support for opposed secession (including support for the dissident group
based around Francis Ona’s Republic of Me’ekamui), but will also encourage sup-
port for integration when the referendum is eventually held (Regan 2002a).
The 2001 Agreement also provides for immunity from prosecution for crimes
committed in connection with the conflict by members of all combatant groups,
and for the demilitarization of Bougainville. These, and a number of other as-
sociated arrangements, are provided for in constitutional laws implementing
the Agreement that were passed by the Papua New Guinea parliament in March
2002 – an amendment to the Papua New Guinea Constitution and the Organic
Law on Peace Building in Bougainville – Autonomous Bougainville Government
and Bougainville Referendum (the Organic Law).13 The former law offers unusual
protection to the arrangements in that it provides that they cannot be changed
unilaterally by the Papua New Guinea parliament, but also require a vote of the
Bougainville legislature – a form of protection for a ‘sub-national’ political unit
usually found only in highly federal systems.
There are a number of other unusual characteristics of the 2001 Agreement
that together take it far beyond mere autonomy of the kind provided for under
the 1976 Agreement. These include: complex dispersal of power in relation to
Bougainville among the Papua New Guinea national government and a proposed
autonomous Bougainville government; defined arrangements for management of
intergovernmental relations between the national government and Bougainville,
including dispute settlement mechanisms and guarantees of a high degree of in-
dependence for Bougainville from national government direction and control;
and provision for roles for external actors, especially, but not only, in the imple-
mentation of aspects of the Agreement.
A Responding to the Self-determination Dispute through a Deferred

Referendum on Independence
The extent to which the 2001 Agreement and the constitutional laws implement-
ing it provide for sharing of power between Papua New Guinea and Bougainville
is evident in the fact that even the conduct of the referendum is to be a respon-
sibility shared between the electoral authorities of both governments. The ref-
13 See Papua New Guinea National Gazette No.G143, 12 November 2001. Most
provisions of those laws came into operation only from August 2003, as provisions
linking the weapons disposal process and the coming into operation of the bulk of
the national constitutional laws provided that they came into operation only when
the second of the three stage weapons disposal process was verified by the UNOMB
as completed.
erendum is to be on the future political status of Bougainville,14 and the choices

offered must include independence for Bougainville. It is to be held among Bou-
gainvilleans (people entitled to vote in national elections in Bougainville as well as
non-resident Bougainvilleans) no earlier than ten years, but in any event no later
than fifteen years, after the first election of the autonomous Bougainville govern-
ment. The actual date for the referendum within that ten to fifteen year period is
to be agreed between the autonomous Bougainville government and the national
government, taking account of the achieving of standards of good governance in
Bougainville and progress on weapons disposal. International observers are to be
invited to observe the conduct of the referendum, and are guaranteed full and
free access to all aspects of the process. The outcome of the referendum will be
subject to the final decision-making power of the national parliament, but must
also be the subject of consultation between the autonomous Bougainville govern-
ment and the national government. To ensure that the holding of the referendum
was fully guaranteed as soon as the constitutional laws came into operation, the
entire machinery and procedural requirements for the holding of the referendum
have been included in the Organic Law.
It can be argued that the combination of a referendum on independence to-
gether with power-sharing may undermine the effectiveness of power-sharing as
a device for responding to a self-determination dispute. The main issue here is
that the promise of a referendum explicitly keeps the self-determination dispute
on the agenda, so that power-sharing is not accepted by the secessionists as the
complete answer to their self-determination claims. There are two main respons-
es to this view. The first is that in circumstances such as prevail in Bougainville,
where the secessionist fighters believed that they had the military advantage be-
fore the peace process and where a large (but never accurately measured) pro-
portion of the population continue to support secession, it can be difficult to
respond to the self-determination dispute by power-sharing arrangements alone
– there may also need to be some other provision enabling the self-determina-
tion issue to be addressed. Second, there is a logic to the pairing of referendum
and power-sharing through territorial autonomy as a solution to a bitter conflict.
This is because the provision for the referendum should encourage the national
government to do all that it can to ensure that the autonomy arrangements work
well, in that way maximizing the likelihood that Bougainvilleans will recognize
the advantages of Bougainville being integrated into Papua New Guinea, and in
this way encourage the maximum possible vote in favour of integration when the
referendum is held.
B Responding to the Self-Determination Dispute through Complex Power-

sharing
That the agreed autonomy arrangements are intended to provide one of two av-
enues to self-determination is explicitly recognized in the Preamble to the draft
14 Constitution ss338-343, and Organic Law, ss52-63 and Schedule 1.

Constitution for the Autonomous Region of Bougainville that is in the process of

being developed in mid-2003.15 The autonomy arrangements are radically differ-
ent from either the existing or previous Papua New Guinea provincial govern-
ment arrangements, and apply only in Bougainville, and not in any of the other
parts of Papua New Guinea. There are at present nineteen provinces in Papua
New Guinea, including Bougainville. Under the autonomy arrangements, Bou-
gainville will no longer be regarded as a province and its government will not be
a provincial government. Indeed, Bougainville will have the power to determine
the names for the autonomous area and of its government. Although the arrange-
ments are intended only to apply in Bougainville, there is, however, pressure for
the arrangements (or parts of them) to be extended to other parts of Papua New
Guinea, pressure that the national government can be expected to resist.
1 Territorial Autonomy
It is because the Bougainville case involves an ethnic group that comprises a
small minority of the Papua New Guinea population that is concentrated in a
defined geographical area where that group constitutes an overwhelming major-
ity, remote from other parts of Papua New Guinea, that both the 1976 and 2001
agreements have concentrated on territorial autonomy, and have taken little ac-
count of measures for integration into the central government. While the influx
of migrants from other parts of Papua New Guinea was a factor in the intensifica-
tion of ethnic tensions in Bougainville in the 1980s, the effective ‘ethnic cleansing’
that occurred in 1989–1990 meant that there was little need to consider ethnic
minorities in the 2001 settlement. Because of the focus of the Bougainville lead-
ership on territorial autonomy rather than integration, they paid little attention
to provisions on proportionality in the operation of national public sector insti-
tutions or areas of activity. However, the Agreement does demonstrate concern
for a degree of integration in relation to key institutions (public service, police,
judiciary) in respect of which the national government conceded to Bougainville
the right to establish separate institutions.
Further, the combination of the geographic concentration of the Bougainville
population with the considerable cultural and linguistic diversity within that pop-
ulation meant that there was little need for either of the agreements to deal with
many of the difficult issues about powers over language, education, culture, and
so on that so commonly need to be addressed in power-sharing agreements.
There are, of course, some people from other parts of Papua New Guinea still
resident in Bougainville, and some thousands of Bougainvilleans live elsewhere
in Papua New Guinea. Their rights and interests are not the subject of specific
protections in the 2001 Agreement. Rather, the human rights provisions of the
Agreement – which ensure the continued operation of the quite extensive pro-
tections of the Papua New Guinea human rights regime in Bougainville – were
assumed to provide sufficient protection to both groups.
15 ‘Second draft of the Constitution of the Autonomous Region of Bougainville. Bou-

gainville Constitutional Commission, 25 March 2003.
2 Institutional Arrangements – Bougainville Choices of Government

Structures
Bougainville has the power to devise its own constitution, which can make provi-
sion for structures of the proposed autonomous Bougainville government, inclu-
sive of the name of the government (Constitution ss279-287). It must do so within
the broad parameters provided by the national constitutional laws (for example,
requirements for a mainly elected legislature and adherence to basic standards
of good governance). The Agreement requires the proposals for the new Bou-
gainville Constitution to be developed by a broadly representative Constitutional
Commission, and then considered and adopted by a Constituent Assembly, before
being endorsed by the national government (which can only refuse to endorse it
if the proposed Bougainville Constitution fails to meet the requirements of the
national constitutional provisions). The Constitutional Commission was estab-
lished in September 2002, and at the time of writing (August 2003) has published
two drafts of the proposed Constitution, but is yet to finalize its proposals.
3 Institutional Arrangements – Bougainville Representation at the Centre

The original Bougainville negotiating position, adopted in June 1999, involved
proposals for essentially confederal arrangements under which Bougainville
would have had power over almost all matters other than defence and foreign af-
fairs, and would also have had very limited representation at the centre – just one
representative in the national parliament instead of the present four, and that one
would have been chosen by the Bougainville legislature rather than by general
election by Bougainville voters as at present. The Papua New Guinea government
opposed the proposal for reduced Bougainville parliamentary representation at
the centre, arguing for the need for continued Bougainville representation at the
same level as other parts of the country.
It was, however, the need to build unity among the Bougainville parties that
resulted in the dropping of the proposal for a single representative in the par-
liament. The divisions that developed in 1998–1999 among Bougainville groups
supporting the peace process saw three of the four elected Bougainville mem-
bers of the national parliament aligning with Bougainville groups that supported
integration into Papua New Guinea. In December 1999, the coalition of groups
that had developed the initial Bougainville negotiating position, and had begun
negotiations without some of the pro-integration elements, decided to drop the
demands for reduced representation at the centre. They did so as part of efforts
to make the negotiating position more attractive to the pro-integration groups,
and especially the three members of the parliament.
The issue remains controversial, however, and the first draft of the Bougainville
Constitution released by the Bougainville Constitutional Commission in January
2003 included provision for the autonomous Bougainville government to negoti-
ate with the national government with a view to securing change to the national
Constitution providing for a reduction in Bougainvillean representation in line
with the original 1999 proposals. The Commission was persuaded to drop this
proposal from the second Draft, on the basis that in the early years of the im-
plementation of the autonomy arrangements it was likely to be to Bougainville’s

advantage to have the strongest possible representation at the centre.
There are a number of additional provisions in the Agreement that emphasize
Bougainville’s right to participate at the national level. Existing national institu-
tions, such as the police, the defence force (PNGDF), and the public service, will
continue to be national institutions that recruit in and from all parts of Papua
New Guinea (including Bougainville) even if Bougainville chooses to establish
separate institutions. The inclusion of these provisions was at the initiative of
the national government, which was at pains to ensure that, as far as possible,
Bougainville should be integrated into the national institutional arrangements
in the same ways as all other parts of the country. It was the same concern that
underlies a number of complex provisions in the 2001 Agreement intended to
encourage cooperation by Bougainville with the national government in the ex-
ercise of powers to establish separate institutions such as the police, courts, and
public services.
4 Complex Layering of Authority – Powers and Functions

During negotiations for the 2001 Agreement, the national government opposed
Bougainvillean proposals that there should simply be one list of a few powers
vested in the national government, with all other powers subject to overriding
Bougainville laws. While the national government was prepared to concede that
most powers should be available to Bougainville, it wanted both a far more ex-
tensive list of national powers in relation to Bougainville, and two separate lists
of powers, one for each government, and provision for later allocation of any
unspecified powers (those not on either list, mainly involving powers not then
considered) on the basis of the principles underlying the two lists. The matter
remained unresolved until late in the negotiations, when Bougainville reluctantly
conceded the two lists arrangement with the later allocation of unlisted pow-
ers, but only once it had gained agreement to the inclusion on its list of powers
those it regarded as of central importance (including police, judiciary, and public
service).
The result is that the powers and functions relating to Bougainville will be
shared, although the bulk of the powers that might be dealt with by law will be
available to Bougainville. The national Constitution (s289(1)) provides a list of
seventeen national government powers, mainly concerning subjects related to
international relations or maintaining the integrity of the state and of a single
economic system. These include defence and foreign affairs, external migration,
currency, central banking, customs, international trade, international shipping,
and international civil aviation. Most other subjects conceivably capable of being
covered by law are included in a list of fifty-nine subjects available to Bougainville,
including land, mining, environment, local government, and so on (Constitution
s290). While each government can make laws on the subjects on the list reserved
to the other, such laws cannot be inconsistent with those of the government on
whose list the powers appear (Constitution s296(2)) – and so ultimate authority
will lie with the government on whose list the power in question appears. Re-
sponsibility for subjects not on either list (in practice, those not thought of at the
time the two lists were negotiated) will be divided between the two governments,
in line with the principles underlying the two lists (Constitution s292).
The powers on the Bougainville list are available to rather than directly vested
in the Bougainville government in that most will initially remain vested in the
national government, but will be transferred to Bougainville at its initiative under
a process that requires the giving of twelve months notice to the national govern-
ment (to allow for necessary arrangements to be made in relation to budgets,
staff, assets, etc.) and the drawing up of joint implementation plans providing
for the development of the capacity in Bougainville needed to exercise the new
powers and functions (Constitution ss295 and 297; Organic Law ss4-6). These
provisions, when combined with the limited guarantees made in relation to the
provision of fiscal resources to Bougainville, add to the complexity of the arrange-
ments. They mean that there will probably be some uncertainty about the proc-
ess of transfer of authority to Bougainville, and a need for ongoing consultation
and even negotiation, especially in relation to significant areas of power such as
police, courts, public service, land, mining, and so on. The Agreement envisages
that primary responsibility for dealing with such matters rests with the Joint Su-
pervisory Body.
While the focus in the Agreement and the constitutional laws is on the two lists
of powers, in fact complex provisions about how a number of powers that appear
on both lists are to be exercised mean that although there is no specific category
of concurrent or shared powers, in effect there is a third category of powers that
could be classified in that way. This is a result of the inclusion of a number of pro-
visions that aim for the protection of the interests of one or other government de-
spite the location of a particular power on the list of the other government. These
are provisions intended to place limits on, or encourage cooperation between the
governments in, the exercise of a power.
Thus, there are specific limits on the use of some powers vested in the na-
tional government, or provision for a limited Bougainville role in the exercise
of some aspects of particular national government powers in relation to Bou-
gainville. For example, while defence, foreign relations (including foreign aid),
and telecommunications are national government powers, the Agreement limits
the future deployment of the PNGDF to Bougainville without the consent of the
Bougainville government, and gives Bougainville both some rights in relation to
the development of telecommunications, and some foreign affairs-related pow-
ers. Conversely, there are a number of subjects where Bougainville will be able
to hold authority only subject to specific limits, or to detailed requirements on
cooperation with the national government, which effectively means that these
powers will be shared between the levels of government. These include criminal
law, human rights, regulation of foreign investment, judicial administration, po-
lice, the public service, and prisons.
The most complex cooperative arrangements were proposed by the national
government as a way of making more acceptable (to the national government)
some of the more controversial aspects of Bougainville’s autonomy demands
– regarding powers to establish separate police, public service, correctional serv-

ice, and judicial bodies. Such powers were only agreed subject to provisions for
cooperative links with equivalent national institutions, arrangements intended
to maintain similar standards in Bougainville to those that apply at the national
level, thereby ensuring a degree of integration between key national government
administrative and legal and judicial institutions and their equivalent institutions
in Bougainville (Constitution ss305-318; Organic Law ss7-37).
a Administration of Justice
Under the 2001 Agreement, Bougainville will have the power to establish its own
courts system, up to a level equivalent to the National Court – the highest court
that at present can hear matters in the first instance. The highest Papua New
Guinea appeals court – the Supreme Court – will remain the ultimate court of
appeal in relation to any new Bougainville courts, thereby ensuring a reason-
able degree of integration of any Bougainvillean judicial system into the national
system. Judges for the Bougainville courts must be appointed by an independent
judicial appointments body, and, in order to minimize the need for duplication,
may hold concurrent appointments in the national government judicial system.
The capacity of the Bougainville government to develop its own judicial system
will depend greatly on its financial resources.
b Police
Bougainville will have the power to establish its own police service, separate from
the Papua New Guinea police. It must be headed by a person appointed through
an independent appointment process, and members of the Bougainville police
are to be appointed on merit. Policing standards compatible with those applica-
ble at the national level will apply in Bougainville. A Bougainville police service is
to have access to specialist services available from the national police. The func-
tions of the Bougainville police will include enforcement of both Bougainville and
national laws. Under transitional arrangements, prior to establishing a separate
police force, Bougainville will have the power to direct and control the national
government’s provincial police commander for Bougainville. The ability of the
autonomous Bougainville government to develop a separate police service will
also depend greatly on financial capacity, as additional costs (above the existing
costs of policing adjusted upwards to normal policing levels in other parts of
the country) involved in establishing such an service will not be met by national
government grants.
c External Relations and Trans-border Cooperation

Among the more unusual aspects of the 2001 Agreement and the implementing
constitutional laws are those that vest the autonomous Bougainville government
with some powers in relation to foreign affairs, and impose some limits on the
exercize of national government foreign affairs powers in relation to Bougainville.
These aspects of the Agreement make the quality of Bougainville’s autonomy sig-
nificantly different from most autonomy agreements and blur the boundaries of
sovereignty.
First, the autonomous government will be able to enter into international agree-
ments under the authority of the national government (Constitution s293(7)). Sec-
ond, it will be able to obtain foreign aid, with the national government required
to approve such aid secured by Bougainville that does not reduce the value of aid
already available to Papua New Guinea, and “subject to overriding foreign policy
considerations” (Agreement clause 175; Constitution s327). Third, Bougainville
can send representatives or observers to regional meetings and organizations
with consent from Port Moresby (Agreement clause 70). Fourth, arrangements
are to be developed to allow Bougainville to “participate or engage directly in the
negotiation of international agreements of particular relevance to Bougainville”
(Agreement clause 73; Constitution s293(7)). Fifth, Bougainville will have a repre-
sentative in any national government delegation negotiating either border agree-
ments with Solomon Islands (Agreement clause 77) or access and other fisher-
ies agreements regarding Bougainville maritime territory or areas of the Papua
New Guinea territorial seas beyond the three nautical mile limit of Bougainville
maritime territory or in the area of the Exclusive Economic Zone (EEZ) and the
continental shelf associated with Bougainville territory (Agreement clause 85).
With regard to limits on the national government powers in relation to for-
eign affairs, while all Bougainville powers will be subject to Papua New Guinea’s
international obligations (Constitution s293(2)), the national government will be
required to consult Bougainville before entering into new obligations, and inter-
national agreements with a purpose of altering the autonomy arrangements will
take effect only with Bougainville consent (Constitution s293). The Agreement
also provides for development of equitable arrangements for sharing revenues
from fishing and other activities in the EEZ (clauses 8 and 86).
d Human and Minority Rights

It was with considerable reluctance that the Bougainville parties negotiating the
2001 Agreement dropped their original demand for the power to determine a hu-
man rights regime separate from that provided for under the Papua New Guinea
Constitution. Of particular interest here was the ability to limit rights of freedom
of movement into and within Bougainville for non-Bougainvilleans. Of course,
the Papua New Guinea government negotiators were opposed to such propos-
als, and argued for the need to protect the rights of all minorities – including
non-Bougainvilleans in Bougainville, and Bougainvilleans living in other parts of
Papua New Guinea.
Interestingly, it was not arguments about protection of minorities that eventu-
ally persuaded the Bougainville leaders to accept the existing Papua New Guinea
rights regime and abandon ideas of direct limits on freedom of movement. Rath-
er, it was concern about the need to maintain the support of the international
community if the 2001 Agreement is to be implemented in full. The leadership
accepted the arguments of their own advisers that such support could be endan-
gered if Bougainville were to develop policies that could be regarded as discrimi-
natory. Bougainville therefore agreed to accept the existing Papua New Guinea
human rights regime, inclusive of the right to freedom of movement.
Bougainville demanded, however, and the national government eventually ac-
cepted, that not only would the autonomous government be able to provide its
own regime for enforcement of rights (Constitution s304(2)), but also that the
Bougainville legislature would have the same extensive powers as the national
parliament to pass laws qualifying human rights (Constitution s303). This last
power gives Bougainville potentially extensive powers to limit rights, provided
this is not done in a discriminatory way, and is done within the limits laid down
in s38 of the Constitution. This power may enable Bougainville to pass laws that
involve indirect limits on most rights, including that to freedom of movement
– for example, laws on such subjects as land, town planning, and employment
could provide restrictions that act as effective limits on that right.
e Financial Arrangements
The funding arrangements (Constitution ss324-329; Organic Law ss38-51) are in
many ways the weakest part of the autonomy arrangements, mainly a result of
a combination of the poor economic base in post-conflict Bougainville without
the Panguna mine, and the fiscal crisis facing a Papua New Guinea government
to some degree crippled by ‘Dutch disease’ after many years of reliance on rev-
enue from large mining and petroleum projects. In summary, the national gov-
ernment has promised to fund Bougainville’s autonomous government through a
main grant calculated on the basis of the costs of existing functions as well as the
costs to the national government in the year before the transfer of new powers
or functions transferred to Bougainville and adjusted to take account of the need
for services in Bougainville to meet average national standards. In addition to
certain other grants, Bougainville will have far wider taxing powers than existing
provincial governments, extending to personal income tax, export taxes, excise,
and so on. These arrangements are unusual in comparison with most autonomy
arrangements, and reflect the intense pressure the Bougainville negotiators ap-
plied for the arrangements to be as close as possible to independence.
The arrangements are ‘weak’ in the sense that Bougainville will not be able to
readily develop its own revenue base sufficient to support a high degree of opera-
tional autonomy, nor is it likely that the national government will have available
significant fiscal resources to transfer to Bougainville, nor is there a strong consti-
tutional guarantee that the national government will make payments in accord-
ance with the grant calculation arrangements in the national constitutional laws.
The weakness of the financial arrangements is a concern, given the high expecta-
tions of autonomy among Bougainvilleans. There is a perception in Bougainville
that donors among the international community can be expected to provide the
funding needed to ensure that the autonomy arrangements work reasonably well,
and, under the Agreement, Bougainville has some rights to deal directly with
aid donors. However, the main donor at present, the Australian government, is
unlikely to be willing to allocate more funds to Bougainville, both because Papua
New Guinea’s fiscal crisis is putting pressure on all Australian aid funds to the
country, and because the Australian-led intervention in Solomon Islands will be

costly, and so will limit the availability of Australian aid funds in the Pacific re-
gion as a whole. There are unlikely to be any other major aid donors that will see
Bougainville as a priority.16
5 Management of Complexity – Dispute Settlement and Limited

National Control
The Agreement provides for a system for intergovernmental relations intended to
ensure that the problems that can be expected to arise as a result of the complex
layering of authority can be managed. It is a system comprising several main ele-
ments:
• A ‘joint supervisory body’ (Constitution s332) made up of representatives
of both the national government and Bougainville that is to be responsible
for:
– first, managing the implementation of autonomy;
– second, providing the first stage in the dispute settlement process; and
– third, for dealing with some aspects of the division of powers and func-
tions unable to be resolved during the negotiations for the 2001 Agree-
ment.
• Two separate dispute settlement mechanisms – one for the period when the
Agreement is being implemented, and another applying once the autonomy
arrangements are in place – which can be used to resolve both disputes re-
lating to autonomy and any that may arise later in relation to the implemen-
tation of the referendum arrangements.
• Strict limits on the power of the national government to direct and control
the Bougainville government.
• Provision for a number of controversial issues that could not be resolved
during the negotiations to be deferred to later processes, including:
– consultation both on the extension of Bougainville’s maritime territory
beyond the limit of three nautical miles reluctantly agreed to by Bou-
gainville (2001 Agreement clause 9), and on the development of a new
Criminal Code statute by Bougainville (2001 Agreement clause 297);
– reference to the Joint Supervisory Body of any concern by Bougain-
ville to either change national laws on telecommunications or assume
additional powers in relation to telecommunications (2001 Agreement
clauses 93-100);
– reference to a joint commission of issues about human rights and other
implications of Bougainville proposals to enable courts and local gov-
ernments to require clan-groups to meet customary obligations in rela-
tion to clan members convicted of criminal offences and certain other
16 Australia has been contributing about AUD 20 million per year in aid to Bougainville
since 1998. The next largest donor is New Zealand, which contributes about AUD 5
million per year.
issues relating to Bougainville powers over human rights (2001 Agree-

ment clauses 128-133); and
– provision for development of a mechanism to enable Bougainville to
participate in the negotiation of international agreements of relevance
to it or to negotiate such agreements on its own behalf (Constitution
s293(7)).
In relation to disputes over the implementation of the Agreement occurring be-

fore the autonomous Bougainville government is established, the parties are re-
quired to consult where differences occur, and if that fails to resolve the matter,
a “neutral third party” may be appointed by agreement, with involvement of the
courts available as a last resort (Agreement clause 332). Further, the Agreement
(clause 335) records the fact that the main body established under earlier agree-
ments to facilitate the cooperation of the parties in the peace process – the Peace
Consultative Committee (which is chaired by the Director of the UNOMB) – can
deal with matters within its mandate or that the parties refer to it during the
implementation period.
Once the autonomous Bougainville government is established, the joint su-
pervisory body is to be used as the first stage in a dispute settlement procedure
which either side can invoke, where consultation is required as a first step, failing
that mediation or arbitration may be attempted. As a final stage in the process,
the courts may become involved in some circumstances (Constitution ss333-6).
In relation to coercive powers, not only will the national government have no
power to unilaterally alter the constitutional arrangements but it will also have
no power to suspend or to withdraw powers and functions from the autonomous
Bougainville government (Constitution s331(c)). However, the national govern-
ment will have strictly limited powers to withhold funds as a last resort in cases of
serious financial mismanagement by the autonomous government of grant fund-
ing (Constitution s329; Organic Law s51). The severe limits on control powers
contrast sharply with the extensive powers the Papua New Guinea government
has to suspend and withdraw powers from existing provincial governments. In-
deed, it is not uncommon for national governments in federal systems to have
suspension powers over state governments, as in the cases of India and Malaysia,
and the absence of such powers in relation to Bougainville is an indication of the
unusually high degree of autonomy provided by the 2001 Agreement.
C Protecting the Constitutional Arrangements from Unilateral Change

The Bougainvillean negotiators sought assurances that the constitutional ar-
rangements implementing the 2001 Agreement (particularly those concerning
the referendum and autonomy) could not be changed unilaterally after enact-
ment. Hence, they argued for, and the national government eventually conceded,
what became known as ‘double entrenchment’. This involves provisions concern-
ing the amendment of the constitutional provisions relating to Bougainville that
are additional to those applying to other parts of the constitutional laws.17 The
Bougainville legislature would need to approve any proposed amendment before
it becomes law – a two-thirds majority vote in the Bougainville legislature is re-
quired for amendments to the provisions of the Constitution concerning the ref-
erendum, and a simple majority vote of that body in relation to other provisions
of the Constitution or to the Organic Law (Constitution ss345-346).
D Demilitarization of Bougainville
The goal of demilitarization of Bougainville is in large part related to self-de-
termination, in that demilitarization is seen as essential if there are to be fair
and democratic processes in establishing and operating autonomy and in holding
the referendum. The demilitarization arrangements also involve important and
unusual restrictions on the powers of the national government in relation to de-
fence. The goal is pursued through a number of separate arrangements, the key
features of which are:
• A multi-staged process of disposal of weapons by armed Bougainville fac-
tions (Agreement clause 329);
• In association with the weapons disposal process, complete withdrawal of
the PNGDF and the Police Mobile Squads from Bougainville;
• Limits on the future deployment of PNGDF (Agreement clauses 60-68) and
Police Mobile Squads (Agreement clause 222); and
• Limits on Bougainville’s ability to establish armed police mobile squads
(should it choose to exercise its power to establish a separate police force).
The weapons disposal process has three main stages, and responsibility for su-
pervision of its implementation is vested in external actors – the UNOMB (with
assistance from the PMG until its withdrawal in June 2003). There are complex
and inventive two-way linkages between the implementation of those various
stages, the implementation of the autonomy arrangements provided under the
2001 Agreement, and the beginning of the period within which the referendum
must be held. The linkages are intended to provide incentives to the parties to
implement core provisions of the Agreement – first, to the national government
to pass the constitutional laws implementing the Agreement, and, second, to the
Bougainville factions to implement the weapons disposal process.
The first stage of the process involved weapons being held in locked containers
with the keys under the control of the BRA or BRF unit commanders whose per-
sonnel are submitting the weapons, with the PMG registering the weapons, and
the containers being sealed by the UNOMB for verification purposes. The second
stage involved the weapons being moved to a smaller number of double-locked
17 Under the Papua New Guinea Constitution, these include two separate votes of the
national parliament at least two months apart and with special majorities required,
ranging from an absolute majority for provisions of less central importance to three
quarters absolute majorities for provisions seen as of fundamental importance.
containers, with one of the keys held by the UNOMB. It is here that the two-way
linkages first came into operation, in that the movement of weapons from stage
one to stage two containment only began when the constitutional laws imple-
menting the Agreement had been passed. On the other hand, most provisions of
those laws only came into operation when the UNOMB verified that stage two
was complete. Verification was announced by the Director of the UNOMB on 30
July 2003, and the bulk of the provisions of the national constitutional laws came
into operation at the beginning of August 2003.
The third stage involves consideration of the final fate of the weapons, with
the parties required to reach a decision within four and a half months of the
constitutional laws coming into operation (by December 2003). This stage has
been left extremely open-ended, even ambiguous, reflecting difficulties in reach-
ing agreement on the final fate of the weapons when the parties were negotiating
the weapons disposal agreement in April–May 2001. While the BRF and the na-
tional government wanted destruction of weapons before election of the autono-
mous government, the BRA still suspected the intentions of the PNGDF (who, at
that stage, had not withdrawn from Bougainville), and also envisaged that their
weapons might ultimately become those of the defence force of an independent
Bougainville. The national government and the BRF were able to live with the
ambiguity of what was eventually agreed because of agreement on the further
linkages between weapons disposal and other parts of the Agreement – includ-
ing a provision that elections for the autonomous government can be delayed if a
decision on the final fate of the weapons has not been reached, or if the UNOMB
(with assistance from the PMG) judges that there is not substantial compliance
with arrangements for the handing in of weapons or for the security of already
contained weapons.
The disposal process has operated reasonably well to August 2003, with over
1,900 weapons contained, over 300 of which are classified as high-powered, over
300 as ‘sporting’, and the balance either refurbished World War II or home-made
weapons. About 150 weapons have already been destroyed, ahead of agreement
being reached on the final fate of the weapons. On the other hand, the process
has taken far longer than was expected when the Agreement was made. There
have been a number of reasons for the slow speed of the process, including divi-
sions and tensions within the BRA, and the refusal of Francis Ona and the MDF
to support the weapons disposal process, a factor that probably puts pressure on
some BRA and BRF elements to make less than full commitment to the disposal
process. Several weapons containers were broken into late in 2002, with about
300 weapons removed. There was concern that MDF elements might be involved,
a matter never entirely resolved. However, there were no further incidents after
that, perhaps due to strong pressure from the BRA leadership on the MDF lead-
ership. Intensive efforts by the UNOMB, PMG, BRA, and BRF leaders resulted in
most of the stolen weapons being re-contained in the first half of 2003.
In announcing the verification of the completion of stage two of the weapons
disposal process in July 2003, the UNOMB had to take account of the fact that an
unknown number of weapons held by the MDF and some others held by reluctant
BRA and BRF members, criminals, and business people had not been contained.
The MDF weapons were the most serious matter, but might be regarded as a
manageable issue following statements by Francis Ona to the effect that he has
no intention of disrupting the peace process. Discussions within and between the
BRA and the BRF concerning the final fate of the weapons had begun even before
the announcement of verification of stage two, but at the time of writing (August
2003) it remains unclear how close the parties are to agreement. Nevertheless,
as there is such strong commitment among the Bougainville groups involved in
the peace process to the earliest possible establishing of the autonomous gov-
ernment, it would be expected that the linkages between weapons disposal and
the provisions on the establishing of the autonomous government will encourage
those groups to work to the containment of weapons not yet contained, maintain
the security of the weapons already contained, and reach early agreement on the
final fate of the weapons.
E External Participants’ Roles – Making, Monitoring and Managing

Agreement,
The roles of external participants in the ongoing peace process that generated the
2001 Agreement have been discussed elsewhere in this chapter. The Agreement
also provides for roles for external participants in a number of aspects of moni-
toring and managing the implementation of particular parts of the Agreement.
The roles of the UNOMB and the PMG in relation to the weapons disposal pro-
cess have already been discussed. Other matters provided for in the Agreement
include:
• A requirement for the tabling by the national government of the Agreement
in the United Nations General Assembly (clause 334), a provision directed to
underlining the interest of the parties in the process being monitored by the
international community (the tabling occurred late in 2001);
• Potential roles for the UNOMB and a ‘neutral third party’ in dispute settle-
ment during implementation of the Agreement, prior to the establishing of
the autonomous Bougainville government (Agreement clauses 333 and 335);
and
• Roles for international observers in observing of the conduct of the referen-
dum on the future political status of Bougainville.
Quite apart from formal roles in the peace process and the monitoring and man-
aging aspects of the Agreement, the key external participants (Australia and New
Zealand) have a number of other roles. First, they are significant donors to both
the Papua New Guinea government and to Bougainville. Australia, in particular,
spends about AUD 25 million per year in aid to projects for Bougainville, much of
it on infrastructure, but significant amounts in supporting the peace process, and
some in assistance to the Bougainville administration. New Zealand funds as-
pects of the peace process and also a number of Bougainville projects, including
one supporting development of community-based policing. Both countries have
‘invested’ considerable ministerial and bureaucratic effort in the peace process,

in addition to the money, towards the goals of reducing regional insecurity and
reducing the pressures on the weak state in Papua New Guinea. The extensive
roles played by the various external actors have given them considerable influ-
ence, especially upon the Bougainville groups supporting the peace process, who
are very conscious of the need to maintain international support, as evidenced
by the decision during the negotiations for the political agreement to accept the
Papua New Guinea human rights regime.
There is an expectation in Bougainville, in particular, that these countries, as
well as the United Nations, can be expected to be the principal monitors, and
even guarantors, of the peace process and the implementation of the Agreement
(including funding of the autonomy arrangements, to the extent that the national
government may have difficulties in that regard). Further, the East Timor anal-
ogy used by the Australian Minister for Foreign Affairs and Trade, Alexander
Downer, in mediating the compromise solution in relation to the referendum has
created expectations in Bougainville that Australia, in particular, and the wider
international community in general, can be expected to be the guarantors of the
referendum outcome should the vote be overwhelmingly in favour of independ-
ence. As a result, it can be expected that there will be some interest on the part
of the Bougainville parties to keep the attention of the international community
focused on Bougainville.
V Arrangements for Transition

The 2001 Agreement is still in the early stages of implementation, but as most
provisions of the constitutional laws giving effect to the 2001 Agreement came
into operation only in August 2003, the autonomy arrangements have only just
begun to operate. Completion of the Bougainville constitution is not expected
until the second half of 2003, and the holding of the first election for the autono-
mous government is unlikely until July 2004. Consequently, it is not yet possible
to discuss how the practice of most aspects of the 2001 Agreement is developing.
Rather, we can comment briefly on implementation to date, outline the amnesty
arrangements in relation to crimes committed by combatants, and outline as-
pects of the intended operation of provisions relating to designated sectors of
government activities.
A Implementation to Date
The first steps in implementation of the 2001 Agreement involved the drafting,
and then the passing by the national parliament, of the amendments to the na-
tional Constitution and of the new Organic Law, steps completed by March 2002.
The clear commitment to the Agreement thereby demonstrated by the national
government greatly enhanced the credibility of the Agreement in Bougainville.
First steps towards making the Bougainville Constitution have begun with the
Bougainville leaders establishing a twenty-four member Constitutional Commis-
sion in September 2002. The Commission has consulted widely with the people
of Bougainville in developing successive drafts of a Constitution, with a final draft
expected during the second half of 2003. Under the constitutional laws imple-
menting the 2001 Agreement, only the Constitutional Commission could be es-
tablished before the main parts of those laws had come into operation. As a re-
sult, the Constituent Assembly that is required to consider the Commission’s final
draft could not have been established before the verification of stage two of weap-
ons disposal brought the bulk of the provisions of the new national constitutional
laws into operation in August 2003.Transitional arrangements for the transfer of
powers and functions in relation to police, public service, and correctional serv-
ices can be made to the Bougainville Interim Provincial Government (BIPG), well
in advance of the autonomous government being established. Despite a formal
request with regard to police power having been made by Bougainville early in
2002, no action has yet been taken by the national government, although talks
between Bougainville and national government officials about how to achieve the
initial transfers took place in June 2003.
On the Bougainville side, there are some concerns that the slow progress on the
transfer of police powers indicates lack of commitment to the Agreement on the
part of the national government. It is true that there have been some problems for
Bougainville in getting responses relating to a number of implementation mat-
ters. In part, the explanation is that staff in key national institutions have changed
since the Agreement was negotiated. Further, capacity at the national govern-
ment level is limited, so much so that there is sometimes doubt whether there is
the necessary ability to fully understand and implement the rather complex pro-
visions of the Agreement. The serious financial crisis facing the national govern-
ment adds an additional level of uncertainty about implementation. On the other
hand, the Bougainville administration also has limited capacity. Perhaps more
important is the fact that progress towards establishment of the autonomous
government has been slower than expected, mainly because of the slow move-
ment in the weapons disposal process. Pressure for more rapid implementation
of the autonomy arrangements can be expected now that verification of stage two
of weapons disposal has occurred.
B Amnesty Arrangements
A major problem in the aftermath of most violent secessionist conflicts is how to
respond to the crimes and human rights abuses committed during the conflict,
usually on all sides. Alternatives include some form of amnesty arrangements or
various ‘transitional justice’ processes such as truth commissions or special judi-
cial processes. In the Bougainville case, the 2001 Agreement provides for immu-
nity from prosecution to be granted for all criminal offences committed in rela-
tion to the conflict and for pardons to be granted to persons already convicted of
such offences. Such arrangements were judged necessary to encourage the sup-
port of all the combatant groups for the peace process and for the political settle-
ment in general, and in particular for the weapons disposal process. However, the
amnesty arrangements in no way restrict people from making claims relating to

matters arising from the conflict either in the civil courts or through the judicial
human rights enforcement processes provided for under the Papua New Guinea
Constitution. Further, there is no restriction on the possibility of establishing
some form of truth and reconciliation commission, a possibility which remains
open, and has some support in both Port Moresby and Bougainville (Kabui 2002:
65). The amnesty provisions of the constitutional laws implementing the 2001
Agreement came into operation in 2002, and the necessary notice defining the
period of the amnesty and the offences to which it applies has been published in
the National Gazette.
C Aspects of Anticipated Implementation

In presenting such an outline, it needs to be remembered that until the consti-
tutional provisions implementing the 2001 Agreement come into operation and
the autonomous Bougainville is established under them, the existing political au-
thority in Bougainville, the Bougainville Interim Provincial Government (BIPG)
continues to be responsible for the execution of powers and functions under the
provincial government system operating throughout Papua New Guinea. The
BIPG has responsibility for aspects of some powers and functions in relation to
the sectors in question, and it would be expected that such arrangements would
be the starting point for any exercise of further relevant powers made available
under the new autonomy arrangements. Further, the limited financial resources
likely to be available to both the Papua New Guinea government and the pro-
posed autonomous Bougainville government over the next few years can be ex-
pected to be a factor limiting the extent to which the latter will be able to depart
from existing administrative, policy, and other arrangements.
Brief comments have already been made about division of powers and func-
tions regarding administration of justice, police, human and minority rights, and
external relations and trans-border cooperation, and no more needs to be said
about them here except to note that existing provincial governments, including
the BIPG, have almost no existing responsibilities in relation to these sectors, so
that implementation of the new arrangements would result in significant change
in Bougainville, and in relations between Bougainville and the national govern-
ment.
1 Communication and Media

The existing BIPG does not exercise any legislative powers relating to commu-
nications and the media. It does have a small Government Information Office,
but the BIPG funding is so limited that that office is essentially moribund. The
2001 Agreement and the implementing constitutional laws provide for “com-
munications and information services within Bougainville”, “information tech-
nology”, and “censorship” to be powers available to Bougainville (Constitution
s290(1)(i), (z) and (e)), while telecommunications will be a national government
power (s289(2)(p)). However, Bougainville will also have the power to develop
its own arrangements for establishing a telecommunications carrier within the

national government regulatory framework. Further, as already noted, one of the
functions of the Joint Supervisory Body will be to deal with any proposals that the
Bougainville government may develop for changing national government laws or
policies on, or for assuming additional powers and functions in respect of, tele-
communications (Agreement clauses 93-100).
2 Health and Social Services

The BIPG exercises administrative responsibility for the delivery of rural health
services in Bougainville, and for the minimal social services that are provided
through the public sector in Papua New Guinea. The major hospital in Bougain-
ville is a national government responsibility. All present laws relating to health
and social services are national government laws. Under the 2001 Agreement and
the constitutional laws implementing it, the subjects of “health” and “home af-
fairs, including youth and social welfare” will be open to Bougainville laws. There
is no indication at this stage as to what priority Bougainville will accord to health
and social services policy development or when they might choose to assume
responsibility for such matters.
3 Education
The BIPG exercises administrative responsibility for the delivery of primary, sec-
ondary, and technical education services, and under the new arrangements all
aspects of the broad subject matter of “education” will be open to Bougainville
laws. Successive provincial governments in Bougainville have regarded primary
education as a priority. Before the conflict Bougainville was unique in Papua New
Guinea in achieving almost 100% school attendance for lower primary school
age children. Further, in the 1980s, Bougainville developed new education poli-
cies giving priority to education in local languages for an additional two years of
schooling before normal primary school (which has a starting age of seven years).
These were meant to enable children to achieve literacy in their first language
before beginning primary school. In the re-establishing of education services
that began in ‘government-controlled areas’ from the 1990s, the local language
schools have largely ceased, and the main focus has been on implementing na-
tional government education policies rather than development of policies that
meet the special needs of post-conflict Bougainville (such as those of the age
groups that in many areas largely missed all opportunities for formal education
over the period 1989 to 1998). However, a number of NGOs are operating projects
providing adult literacy and technical education programmes in several parts of
Bougainville.
4 Economic Policies
The BIPG exercises little authority and allocates almost none of its limited re-
sources to the development or promotion of particular economic policies or to
economic management. It does have a small commerce office, which provides
limited technical support to small businesses. It also has a planning office, which,
with very limited resources, attempts to develop plans for the future develop-
ment of Bougainville. But, with limited provincial government funds to allocate,
and very little capacity to coordinate or even to influence the allocation of official
donor funds (about pgk 50 million18 per year, considerably more than the amount
currently available under the BIPG annual budget) and funds spent by NGOs, the
impact of the BIPG on economic development is extremely limited. All relevant
laws and most policies relating to economic management and development are
made by the national government, though in fact they have limited impact in
post-conflict Bougainville.
In negotiations for the 2001 Agreement, the Bougainville parties reluctantly ac-
cepted that Bougainville would continue to be part of a single Papua New Guinea
economic system, a system that is in any event largely subject to the vagaries
of global forces. As a result of these and related considerations, it was agreed
that the national government should retain control over “central banking”, “cur-
rency”, “customs (imposition, administration, and collection)”, “foreign aid”, and
“international trade” (Constitution s289(2)(a),(b), (c), (e) and (j)) and would share
responsibility with Bougainville in relation to control of “foreign investment”,
(s289(5) and s290(6)). On the other hand, Bougainville will be able to pass laws on
a range of subjects relevant to economic management and development, includ-
ing “agriculture”, “corporation law”, “energy”, “environment”, “fisheries”, “forestry”,
“land and natural resources”, “mining”, “oil and gas”, “trade, commerce and in-
dustry”, and development of infrastructure related to its functions (Constitution
s290(2)(a),(k),(n), (o),(r), (s),(zd),(zm),(zo),(zza), and (6)).
The autonomous Bougainville government can be expected to give consider-
able attention to economic policy, both because of its need to raise revenue, and
because of the general concern that, as far as possible, Bougainville should control
future patterns of economic development with a view to avoiding the inequalities
and other problems that were the consequences of the patterns of development
before the conflict, and which are widely perceived as having contributed to the
origins of the conflict. There is no consensus yet on what approach to future eco-
nomic development should be followed. The expectation of many is that it should
be based primarily on agriculture rather than mining, but there are some leaders,
especially in Buka and north Bougainville, who believe that renewed mining is
going to be necessary if Bougainville is to generate enough economic activity for
the autonomous government to derive the revenue needed to meet the expecta-
tions of ordinary people over the delivery of basic services.19
5 Humanitarian Cooperation
There is no specific provision in the 2001 Agreement concerning humanitarian
cooperation, nor any existing arrangements under which the BIPG is involved
directly in such matters.
18 In mid 2003, the exchange rate for the pgk was about usd 0.25.
19 For further discussion of the dilemmas involved in consideration of the future of
mining in Bougainville, see Regan 2002a and 2003.
VI Conclusions
Although development of political mobilization around a Bougainville identity
may be a relatively recent phenomenon, it has made Bougainvillean self-deter-
mination one of the most difficult issues that Papua New Guinea has had to deal
with in its almost thirty years of independence. While the extent of the commit-
ment of the Bougainvillean leadership to self-determination in 1975–1976 was
never entirely clear, there is little doubt that, even then, most Bougainvilleans saw
themselves – and their problems and needs – as so distinct from the rest of Papua
New Guinea that, at the very least, considerable autonomy was required. Further,
the conflict of 1975–1976 contributed to the development of expectations that
independence offered a simple solution to Bougainville’s complex problems.
Nevertheless, the autonomy arrangements under the 1976 Agreement worked
reasonably well to end the self-determination dispute at that time. While neither
the autonomy arrangements nor the way that they operated contributed directly
to the development of the renewed self-determination dispute in 1988, never-
theless they could not accommodate the forces unleashed by rapid social and
economic change that contributed to growing dissatisfaction among Bougainvil-
leans. On the other hand, although many Bougainvilleans probably felt that their
situation would have been better if the 1975 UDI had succeeded, there was no
evidence of a development of a movement for independence in the 1980s until
that was generated by the behaviour of first the Police Mobile Squads and later
the PNGDF from late 1988 to early 1990. In the course of the conflict to 1997,
political mobilization around identity intensified dramatically, and in the process
self-determination became the only acceptable solution for a large – but never
defined – proportion of Bougainvilleans. As a result, it was inevitable that resolv-
ing the 1988–1997 conflict would require much more extensive concessions from
the Papua New Guinea government than was the case with the 1976 Agreement.
The reasons why the Bougainvillean negotiators of both the 1976 and 2001
agreements preferred territorial autonomy rather than some form of integration
are clear. As a small minority of the Papua New Guinea population that shares a
sense of a distinct identity, and concentrated in an area not only where they con-
stitute the overwhelming majority but also remote from the rest of the country,
there was little reason for Bougainvilleans to be interested in integration into
central government arrangements. It is therefore not surprising that in the nego-
tiations for the 2001 Agreement it was the national government that pushed for
what integrationist elements that were included, such as normal arrangements
for Bougainvillean representation in the national parliament and the linkages be-
tween national government and Bougainvillean public service, police, and judi-
cial institutions. In fact, it can be argued that in negotiating and implementing
the 2001 Agreement, the Bougainville leadership has been unwise in its lack of
interest in such things as elected Bougainvillean representation in the national
parliament. After all, in a weak state, facing fiscal crisis, Bougainville can be ex-
pected to need a reasonably strong voice at the centre to attract attention and
resources.
It is hard to see how the 1988–1997 dispute could have been successfully end-
ed – even in the short to medium term – by anything other than some form
of complex power-sharing arrangements. Neither assimilation nor separation
could have been expected to have succeeded. Assimilation was not acceptable
even to the Bougainvillean elements who opposed the BRA and their secessionist
demands – for although they feared independence under the BRA, they never-
theless saw Bougainvilleans as separate and unique, and supported the highest
possible level of autonomy. On the other hand, separation was also not viable.
Divisions among Bougainvilleans were deepened by the nine years of war, and
have not been fully susceptible to resolution even with a six-year peace process.
A proportion of Bougainvilleans – again, one that has never been defined – has
always opposed independence, but that proportion has undoubtedly increased as
a result of the internal conflict in the period 1990–1997.
There is also the question of the extent to which Bougainvillean identity is a
phenomenon generated by opposition to outsiders, in the absence of which more
locally-based identities among Bougainvilleans not only tend to be more impor-
tant, but also become foci for tensions and conflict. It is yet to be seen whether
the Bougainvillean leadership can build a unified nation by mobilizing political
support around a single Bougainvillean identity. Without that occurring, there
may be dangers of internal conflict again becoming hard to manage. Develop-
ment of the unity needed to make independence viable may not be an easy task
when divisions remain so strong, with significant groups still refusing to support
the 2001 Agreement.
Indeed, Bougainville’s dilemma is illustrated by the opposing positions that
tend to be put in discussion of the advantages and disadvantages of its being part
of Papua New Guinea. On the one hand, it is argued that there may be advan-
tages for Bougainville to persist with special status within Papua New Guinea
rather than having independence, in that a unifying impetus may be provided by
the inevitable tensions involved in dealing with a remote national government.
On the other hand, it is said that the tensions involved in being part of Papua
New Guinea are more likely to result in a renewed self-determination conflict
than give rise to mild tensions sufficient only to both encourage unity and en-
able Bougainville to accept integration into Papua New Guinea. The assumption
of the 2001 Agreement is that, with enough autonomy, Bougainville will be able
to manage enough of its own affairs that the tensions involved in dealing with
the national government will be manageable. But the complexity of the arrange-
ments, the heavy dependence that Bougainville will have on national government
funding, the weakness of the state at all levels, and the ambiguous degree of sup-
port for the 2001 Agreement at the national level all contribute to doubts about
whether the Agreement will work as hoped.
Would it have been possible to base the 2001 Agreement on territorial au-
tonomy alone? With a significant proportion of the population strongly support-
ing secession, it was essential to include a mechanism that dealt with that issue
explicitly. It was fortunate, however, that there were enough moderate leaders
within the BRA and the BIG who recognized that the extent of the division in
Bougainville on the question of independence was such that the issue could not
be dealt with in the short term, and was best left to the democratic device of a
referendum held after enough time had elapsed to enable divisions to be healed,
for without such a compromise there was no way of resolving the self-determina-
tion dispute amongst Bougainvilleans.
It is clear that external actors have been of critical importance to both the
making and implementing of the 2001 Agreement. The divisions generated by
the two dimensions of conflict (between Bougainville and Papua New Guinea
and within Bougainville) were too deep to be dealt with otherwise. External ac-
tors have made major contributions to: establishing and monitoring the peace
process that enabled the 2001 Agreement to be reached; achieving agreement
on the most difficult issue in the negotiation of the Agreement (the referendum
on independence); and implementing the Agreement (especially, but not only, in
relation to the weapons disposal process). On the other hand, while there are now
high expectations in Bougainville that the international community will continue
to guarantee the Agreement, through funding of autonomy and by ensuring that
the referendum is held and its outcome honoured, it is far from certain that these
expectations will be met.
In the context of this book, a key issue concerns whether the 2001 Agreement
goes beyond mere territorial autonomy, and meets the criteria for complex pow-
er-sharing. Commentators on power-sharing as a response to self-determination
disputes tend to regard integrationist arrangements as being the most important.
Bougainville’s experience, however, indicates that, in limited circumstances (such
as the geographical concentration of a minority population), even bitter self-de-
termination disputes can be responded to by territorial autonomy. However, it
is a case that also indicates that, in such circumstances, the autonomy arrange-
ments themselves need to be of a degree of complexity similar to constitution-
ally-provided integration arrangements.
A key criterion for complex power-sharing is the distribution of public power
in such a way that there is a sharing, or blurring of the location, of sovereignty.
The 2001 Agreement divides authority between Bougainville and the national
government in such a way that this criterion is certainly met. The degree of power
available to Bougainville will be remarkable, extending even to powers relating
to foreign affairs and human rights. In fact, under the unusual arrangements for
division of powers and functions, most powers of government will be available
to Bougainville. The complexity, however, is not just in the division of powers in
relation to Bougainville into two formal categories (a small list of national gov-
ernment powers and a large list of powers available to Bougainville), but the ef-
fective existence of a third category of powers resulting from complex arrange-
ments for limits or conditions applicable to the exercising of those powers by the
government on whose list they appear. Not only are these and the other complex
power-sharing arrangements discussed in this chapter constitutionally provided,
but those arrangements cannot be changed without the consent of the Bougain-
ville legislature (‘double entrenchment’).
The Bougainville case does not, however, involve just territorial autonomy,
but it also meets several of the other criteria for complex power-sharing. First,
external actors have played critical roles in making, monitoring, and managing
the Agreement. Second, constitutional provisions are made for the machinery to
manage intergovernmental relations – provisions intended to deal with the com-
plexity of the arrangements for dividing public authority.
An additional element of complexity in the Bougainville case involves one
that in some ways may be seen as detracting from the use of power-sharing as
the main means of resolving self-determination disputes. The complex territo-
rial autonomy arrangements sit together with an explicit constitutional provision
for a referendum – something that might be seen as contradicting the idea of
power-sharing as the solution to a self-determination dispute. But, as discussed
in this chapter, an important part of what might be termed the ‘internal logic’ of
the Agreement is that the existence of the guarantee for the referendum should
encourage the national government to do all it can to make the autonomy ar-
rangements work well so as to maximize the likelihood of a high level of voting in
favour of integration into Papua New Guinea when the referendum is eventually
held. Seen in this light, and in the particular circumstances where strong support
for secession continues, there is a basis for the two approaches to be included
in the same agreement, and without detracting from the case for these arrange-
ments to be classified as a case of complex power-sharing.
A criterion for complex power-sharing that the Bougainville case does not meet
so clearly involves the inclusion of provisions for human and minority rights pro-
tection. The 2001 Agreement does provide a high degree of protection for the
rights of Bougainvilleans living in Bougainville. However, it gives little specific
attention to the protection of either minorities living in Bougainville or of the
Bougainvillean minority living in other parts of Papua New Guinea, mainly be-
cause the focus of the Agreement is on territorial autonomy. On the other hand,
the Agreement is also based on the assumption that existing constitutional pro-
tections should be adequate to meet the needs of both minorities.
Finally, it is necessary to touch briefly on two closely related questions that can-
not yet be answered with any degree of certainty, mainly because the Agreement
is still at such an early stage of implementation. The questions concern whether
the power-sharing arrangements are more likely to entrench or to resolve the
divisions on the self-determination issue, and whether the arrangements can be
expected to be permanent or transitional in nature. Some aspects of these ques-
tions have already been touched upon in the discussion of the opposing positions
on the advantages and disadvantages of Bougainville belonging to Papua New
Guinea, and the apparent contradiction of pairing territorial autonomy with a
referendum on independence. In relation to the latter point, there is as yet little
evidence that the ‘internal logic’ of the Agreement is widely understood in the
senior levels of the Papua New Guinea national political or bureaucratic leader-
ship. If that proves to be the situation in the long-term, there will be a danger that
popular support for the Agreement in Bougainville will dissolve, perhaps well
before the referendum is held.
On the other hand, of course, even if support for the Agreement is maintained,
and the referendum held, there will always be a risk that even if the national
government has put a great deal into supporting the autonomy arrangements,
that a strong vote in support of independence will lead to the self-determina-
tion conflict being renewed, or strong pressure being applied to the Papua New
Guinea government to agree to independence for Bougainville. These are just two
of the possible outcomes that demonstrate the difficulty of predicting with any
certainty whether the power-sharing arrangements in the 2001 Agreement can
be expected to be permanent.
Attempting to answer these last questions, however, involves speculation that
is not particularly fruitful. At present, the best guides to the future of the 2001
Agreement probably lie in its development and implementation to date. The
Agreement involves considerable attempts on the part of diverse and previously
opposed groups to resolve significant conflict. Since the Agreement was reached,
the groups involved have shown serious commitment to its implementation
(even if implementation has not been without difficulties), and despite concerns
about the complexity of the autonomy arrangements and the limited capacity
available to implement them, so far there are no indications that the degree of
commitment is likely to change to any substantial degree. The involvement of the
international community has been critical to the success in reaching the power-
sharing agreement, and will probably be equally important to its ongoing im-
plementation. While that involvement is unlikely to be able to provide the same
degree of support as it has managed in the first six years of the peace process,
some support can be expected to continue, adding to the prospects that progress
in implementation will continue, albeit probably with continuing problems. In
summary, then, the Bougainville case involves not just an instance of complex
power-sharing, but also one where there are reasonable prospects that the agreed
arrangements will provide the basis for successful resolution of the self-determi-
nation dispute.
Chapter 6
Resolving Self-determination Disputes through
Complex Power-sharing Arrangements: The Case of
Mindanao, Southern Philippines
Mark Turner
I The Origins of the Conflict
A Introduction
The Philippines is a colonial invention. There was no concept of the Philippines
when Spanish explorers arrived in the archipelago in the sixteenth century. The
scattered islands which make up the present-day country were populated by small
independent communities. Colonization of these lowland communities was eas-
ily accomplished, their inhabitants converted to Christianity and the island chain
named Las Islas Felipinas in honour of the crown prince of Spain, a distant figure
unknown to the indigenous inhabitants.
There was a notable exception to this experience. More than two centuries be-
fore Spanish galleons first anchored off Philippine shores, traders from the Mid-
dle East introduced Islamic beliefs and practices into the archipelago. The effect
was strongest in the south on the large island of Mindanao and in the Sulu archi-
pelago. Part of this “gradual Islamization” (Tan 1993: 7) was the consolidation of
neighbouring micropolities known as datuships into larger political units known
as sultanates. These strong trading states were not amenable to Spanish colonial
advances. A combination of military prowess and international diplomacy meant
that they were able to resist European control for several centuries (Gowing 1979;
Warren 1985; Laarhoven 1989).
It was these early colonial encounters that established the theme for more re-
cent struggles for secession or autonomy by Muslim (Moro) groups on Mind-
anao. One of many contemporary illustrations of this can be seen in the opening
statement of the 1989 Bangsamoro National Congress: “It is a historical fact that
Bangsamoro people have for four centuries remained steadfast in the defense
of their fundamental right to self-determination” (Tan 1993: 132). First, it was
the Spanish, then the Americans, and finally the independent Republic of the
Philippines. In all cases, Muslims in Mindanao have perceived these regimes as
162 Mark Turner
alien impositions. The contexts in which struggles have taken place have certainly
changed over the centuries but the overriding theme has remained intact.
B The Colonial Experience: Creating the Moro Nation

Pacification of the Muslim populations has been the policy of both colonial re-
gimes and the Republic of the Philippines. The Spanish made their intentions
clear early on. They lumped the thirteen different ethnic groups who were Mus-
lim under the collective name of ‘Moro’ after the Islamic Moors who had fought
Spain over centuries for control of the Iberian peninsula. The Spanish also trans-
ferred their hatred for the Moors to the Philippines. The Moros were thus as-
cribed a long list of negative characteristics and were to be made to acknowledge
Spanish sovereignty over their territory, to restrict their trade to the Philippine
islands, and to submit to “Hispanization and Christianization” (Gowing 1979: 29-
30). Over the next three centuries, Spain attempted to convert this policy into
reality through the Moro Wars, which “shaped the character of Muslim-Christian
relations down to the present day” (Gowing 1979: 31-32). But only in the sixth and
final stage of these sporadic wars in the mid-nineteenth century did the Spanish
succeed in penetrating deeply into Muslim territories and activities.
There was some respite for the Muslim polities as military forces were with-
drawn from Mindanao to deal with the outbreak of the Philippine Revolution in
Luzon in 1896 and the subsequent spillover of the Spanish-American War from
the Caribbean into the Philippines. But the outcome of these struggles was a new
colonial master for the Muslims in Mindanao – the Americans. From 1899 on-
wards, the Americans pursued a policy of pacification using superior military
technology, which eventually overwhelmed the Muslim forces. They then com-
menced the ‘civilizing’ process, involving the imposition of alien laws, new forms
of government, and formal education. A “policy of attraction” was launched in
1914 to ensure the full incorporation of Mindanao’s Muslims into mainstream
Filipino life. Occasional uprisings reminded the American colonists of the tenac-
ity of Moro opposition to external domination, but this was a time of relative
peace in Mindanao (Tan 2000).
The most significant legacy of the colonial era was the creation of a “tran-
scendent Muslim (Moro) identity” (McKenna 1998: 275). Maguindanao, Tausug,
Maranao, and other more local identities were subsumed under the one collec-
tive identity. Muslims also adopted the transcendent identity and utilized it for
their own purposes. Following pacification by America, accommodating Mus-
lim intellectuals and elites “sifted out the favorable attributes of ‘Moro’ culture
for administrative enhancement” (McKenna 1998: 275). The second generation
of Muslim intellectuals maintained the idea of a transcendent Muslim identity
and embraced the once pejorative term Moro. But they rejected the American
notion of Muslim self-improvement as “the principal cure for Muslim underde-
velopment” and focused on “a glorious history [of ] unified Muslim resistance to
Western imperialists” (McKenna 1998: 276).
6 Resolving Self-determination Disputes Through Complex Power-sharing Arrangements 163
C The Republic of the Philippines: Independence or Internal Colonialism?

When Philippine independence came onto the political agenda in 1934 the Mus-
lims were poorly placed to wield power in the new state. They had either to “ac-
cept subordination to Christian Filipino leadership or demand separation from
the envisioned independent Philippine republic” (Tan 1993: 11). Some Muslim
leaders even made overtures to the USA in efforts to make Mindanao another
American state. Remote from the centre of power in Manila, Muslim elites re-
mained peripheral to the political manoeuvring that was determining the struc-
ture and officeholders of the new state. They saw internal colonialists replacing
foreign ones.
There was a yawning gulf between the “New Nationalists” (George 1980: 82) in
Manila and the Muslims in Mindanao. These New Nationalists were ignorant of
Muslim socio-economic systems, their cultures, and their history of anti-colonial
struggle. Independence in 1946 brought no improvement. Instead, events pro-
pelled Muslims inexorably into renewed armed conflict.
The education system maintained its bias against Muslim Filipinos with “the
provocative slant of textbooks” working against building “the confidence of a
self-conscious minority” (George 1980: 94). Popular negative stereotypes of Mus-
lims propagated during colonial times were firmly embedded in the Christian
majority’s minds in the independence era (Bulatao 1975). Poor levels of educa-
tion among Muslims meant that northern Christian officials came to the south to
exercise their administrative authority. Some Muslims who did assume positions
alongside the northern bureaucrats used the opportunities for self-advancement
but few benefits trickled down to the community. A 1963 Senate Committee re-
port on the situation of Muslims made grim reading. It spoke of frustration and
alienation and pointed to “worsening problems” in education, health, livelihood,
transportation, and communication (Tan 1993: 33). The report also gave emphasis
to land-grabbing, which was depriving Moros of their ancestral rights.
The loss of land by Muslims was a direct result of a migration of Filipinos from
the northern islands. The colonial governments, and especially the independent
government, encouraged these settlers to come to Mindanao, “the land of prom-
ise” (Turner, May and Turner 1992), seeking economic opportunity and security.
A stream of Christian migrants caused a population explosion on Mindanao and
transformed the Muslim majority into a minority in its own land. In 1903, Mind-
anao’s population was 670,833 (Costello 1992). By 1939, it had reached 2,244,421.
In 1960, 3,200 land-hungry northern migrants were disembarking in Mindanao
each week (George 1980) swelling the population to 5,384,164, and by 1980 it had
reached 10,905,243. In 1903, Muslims comprised 76% of Mindanao’s population.
In 1980 the figure was 23%.
This northern invasion created friction and conflict, which erupted in the 1960s
in armed actions against Muslims by fanatical gangs of Ilagas (rats) sponsored by
local Christian politicians and seemingly blessed by the Philippine armed forces.
The counter-reaction came in the form of gangs called Barracudas and Black-
shirts, which recruited from Muslim populations to terrorize Christian commu-
164 Mark Turner
nities. The situation degenerated into what May (1992) has described as “the Wild
West in the South”, or as Glang puts it, “a free-for-all” with “Muslims fighting
Christians; government troops fighting Muslims; political private armies fighting
Muslim or Christian farmers; private armies or hired goons fighting army men”
(Glang 1972: 7, as cited in May 1992). Without guns, Muslims believed they would
be even more vulnerable to terror and dispossession.
D The Rise of the Moro National Liberation Front

It was in this context of disorder that the Moro National Liberation Front (MNLF)
emerged in the early 1970s to assume leadership of the Muslim cause. Several
events precipitated the slide into open war between the MNLF and government
forces. These were the Jabidah Massacre of Muslim military recruits in 1968, the
Manili Massacre of sixty-five Muslim men, women and children at a mosque in
North Cotabato in 1971, the intensified hostilities between Muslims and Chris-
tians accompanying the 1971 elections, and the declaration of Martial Law in 1972
with the associated attempt to disarm civilians (Che Man 1990).
The MNLF had been formed in 1969 by a group of young secular-educated
men who rejected the traditional aristocratic leaders of Mindanao’s Muslims.
According to MNLF leader, Nur Misuari, “our armed struggle today is a jihad
for national salvation from colonialism” and involved setting up an independent
“Bangsa Moro Republik” (Che Man 1990: 87). Full-scale war between the MNLF
and government forces lasted from 1972 until the mid-1970s, resulting in approxi-
mately 50,000 deaths and the displacement of at least 200,000 people with an ad-
ditional 140,000 refugees in Sabah (May 1992). Other estimates put these figures
considerably higher. For example, Vitug and Gloria (2000: 27) claim the death toll
was 120,000, while over 300,000 were displaced.
The death and destruction in Mindanao led to MNLF claims of genocide be-
ing perpetrated by the Philippine government. Thus, President Marcos hosted
delegations from Muslim states to prove there was no genocide in Mindanao,
while the MNLF sought foreign assistance to prosecute their armed struggle. Fi-
nance and weapons were initially secured from Libya, while logistical support
was acquired from the Malaysian state of Sabah, which lay adjacent to Mindanao.
Neighbouring Indonesia was not amenable to MNLF requests for assistance. The
leading international body of Islamic states, the Organization of the Islamic Con-
ference (OIC), was more sympathetic. In 1972, it expressed “serious concern for
the plight of Moslems living in the Philippines” (OIC 1972, quoted in Santos 2001:
57) and in 1974 urged the Philippine government to find a peaceful solution to the
conflict, particularly with the MNLF. This came in 1976 in the form of the Tripoli
Agreement, signed in Libya by the Philippine Government and the MNLF and
brokered by the OIC.
Under the terms of the agreement, there would be no independent Bangsa
Moro Republik. Instead, thirteen provinces in Mindanao, Palawan and the Sulu
archipelago would become “areas of autonomy for the Muslims” (GRP-MNLF
1976: 2). There was disagreement on the implementation of the Agreement be-
tween the MNLF and the government. The government created two autonomous
regions rather than one envisaged by the MNLF; officeholders were often viewed
as opportunists, and it is doubtful whether real power was actually devolved from
Manila. Hostilities between the government and MNLF continued on a much re-
duced scale. Some commanders returned to the government fold (for a fee), while
others transformed their units into ‘lost-commands’, which engaged in criminal
activities. The MNLF fragmented along both ethnic and ideological lines. The
most significant breakaway group was the Maguindanao-based Moro Islamic
Liberation Front (MILF), which, under Hashim Salamat, professed a more Is-
lamic orientation than the MNLF. The separatist movement had run out of steam
and “was considered, for all practical purposes, a minor local problem” (Mercado
1992: 163).
With the overthrow of Marcos in 1986 and the inauguration of Corazon Aquino
there were renewed efforts to negotiate a lasting peace. The ‘Mindanao problem’
was returned to central stage in the national political arena. A new Constitution
in 1986 mandated “the creation of an autonomous region in Muslim Mindanao”
(Republic of the Philippines 1986, Art X(1)). Both the MNLF and MILF lobbied
for delay in ratifying the sections dealing with Muslim Mindanao but the govern-
ment submitted the Constitution to the people for approval in February 1987.
There was overwhelming acceptance of the new Constitution even in predomi-
nantly Muslim areas. A Regional Consultative Committee (RCC) was appointed
to draft the “Organic Act on Muslim Mindanao”, but the MNLF and MILF refused
to participate. Despite this opposition, a draft act was produced by the Commit-
tee and subsequently passed by Congress (Republic Act 6734) in July 1988.
The Act was not only intended for Muslims but also contained a minority pro-
tection clause. To assuage Christian and Lumad (tribal minorities) fears of Mus-
lim domination, Article III, Section 5 stated that “no person in the autonomous
region shall, on the basis of creed, religion, ethnic origin, parentage or sex, be
subjected to any form of discrimination.” The Act provided for a regional assem-
bly and executive organizations. It also devolved responsibility for a wide range of
developmental and administrative functions to the region, including special Sha-
riah and tribal courts, but central government retained control of critical items
such as foreign affairs, defence, general auditing, civil service, and elections. Cen-
tral government also agreed to provide ‘seed’ money of php 2 billion per year for
five years (Mercado 1992).
While the Act was enthusiastically promoted by the government, it met with
considerable opposition in Mindanao. The MNLF and MILF rejected it and rec-
ommended a boycott of the plebiscite to determine which territories would be
incorporated into the autonomous region. Many Christians were opposed to
joining the region, believing that it would bring Muslim domination. When the
votes were cast in November 1989, of the thirteen provinces and nine cities which
were eligible for regional membership only four non-contiguous provinces opted
to become partners in the Autonomous Region of Muslim Mindanao (ARMM).
This has been characterized as an “unsatisfactory arrangement contributing little
to the resolution of the ‘Moro problem’ or to the promotion of socio-economic
166 Mark Turner
development” (Turner 1995a: 10). The MNLF and MILF remained steadfastly op-
posed to the ARMM and sporadic armed clashes continued to disrupt life on
Mindanao and hold back its developmental potential.
II The Origins of the Peace Agreement
A President Ramos: Peace for Prosperity

When President Ramos assumed office in 1992 he regarded the promotion of
economic development and of peace and order as primary policy objectives. In
Mindanao, the two objectives coalesced. Ramos saw Mindanao as having consid-
erable economic potential, which could be realized only if peace prevailed. His
project to make Mindanao an investment focus of the eastern areas of the As-
sociation of Southeast Asian Nations (ASEAN), comprising Mindanao, parts of
Indonesia and Malaysia, and Brunei. This ASEAN Growth Area (EAGA) required
the cessation of armed hostilities and the demobilization or elimination of vari-
ous armed groups (Turner 1995b). The largest of these groups and the greatest
threat to peace was seen to be the MNLF. Thus, Ramos made immediate efforts
to settle the ‘Mindanao problem’ once and for all by seeking a lasting peace with
the MNLF.
Exploratory talks between the government of the Philippines and the MNLF
were held in Tripoli in October 1992 and then in Indonesia in April 1993 with
OIC assistance. A Statement of Understanding emerged from these discussions
leading to formal negotiations aimed to bring about the full implementation of
the 1976 Tripoli Agreement. A three-tier structure was adopted to conduct the
formal negotiations, although informal face-to-meetings were reportedly impor-
tant in securing consensus (Santos 2001). At the top level were the formal panel
talks held in Jakarta with the leading figures from the Philippine government and
MNLF. Below were the mixed committees and support committees. These two
layers of lower level committees met in the Philippines. The mixed committees
consolidated the technical work of the support committees on nine substantive
issues. The mixed committee’s outputs of ‘consensus points’ were forwarded to
the top-level panels at the formal talks. At all levels, meetings were facilitated
by diplomats from Indonesia, which was chair of the OIC Ministerial Commit-
tee of the Six. Indonesia and Bangladesh had been added to the original four
countries of Libya, Saudi Arabia, Somalia, and Senegal, which had comprised the
Quadripartite Committee in previous negotiations. Making Indonesia chair was
described by its ambassador as “throwing the whole problem to Indonesia” but he
did claim that it allowed the committees to “work very fast” (Santos 2001). Others
saw the negotiations as “tortuous” (Muslim and Cagoco-Guiam 1999).
President Ramos’s strategy towards the MNLF was “accommodation and co-
optation”, with government negotiators empowered in advance to give political
and economic concessions to lure the MNLF into the Philippine mainstream (Vi-
tug and Gloria 2000: 43). This was not a new strategy. American colonial admin-
istrators and previous Philippines presidents had also adopted such an approach.
The strategy was an exclusively executive initiative. The legislature had little or
no part, preferring to distance itself from negotiations with the MNLF. The ad-
vantage of this strategy for Ramos was a unified voice in negotiation. The danger
was that, lacking ownership, the Congress would oppose, water down, or drag
out discussions on any required legislation. Given the fractiousness of Philippine
politics and strong opposition to accommodation among some constituencies
in Mindanao, it was not surprising that when the agreement was brought into
public view this scenario would be enacted.
Opposition in the House of Representatives was led by three Congresswomen
from Mindanao, the “Tres Marias.” Among their Christian supporters, the wom-
en railed against President Ramos for surrendering Mindanao to Muslim rebels.
They accused him of selling out the Christian majority on Mindanao. Other Min-
danao politicians joined the chorus and whipped up further opposition. There
were also problems for Ramos in the Senate. For the whole of July 1996, it sat as
committee to discuss the council proposed in the peace agreement (Vitug and
Gloria 2000). The military was also wary of integrating MNLF troops into the
police and Armed Forces of the Philippines (AFP). The initial numbers suggested
by Misuari – up to 30,000 – were seen as ludicrous by AFP leaders. Some MNLF
soldiers were wary of joining the very forces they had been fighting for twenty-
five years. Other MNLF supporters also wondered whether the agreement would
bring the developmental gains so urgently needed in this zone of poverty and
insecurity. Civil society organizations on Mindanao shared these concerns but
were not consulted for inputs into the peace process negotiations This allegedly
“stoked resentment among Mindanao’s Christian and Lumad populations” and
among the nongovernmental organizations (NGOs), which “could be credited
with preventing conflict spawned by poverty and displacement” (Cagoco-Guiam
1999).
Another bystander in the peace process was the MILF. At the start of his
presidency, Ramos had established the National Unification Commission (NUC)
to advise on peace policy and amnesty arrangements for Muslim rebels, com-
munist insurgents, and military dissidents. Naturally, the MILF was one of the
organizations that the NUC approached. However, the MILF realized that the
government’s attempts to resolve the ‘Mindanao problem’ focused on the MNLF.
Exploratory talks with the MILF came to a standstill and the MILF waited to see
what would happen in the Jakarta talks with the MNLF. Meanwhile, the MILF
continued to expand its armed forces and area of influence.
The MILF would have been a major actor in the peace talks if it had had formal
recognition by the OIC. But only the MNLF had that status and jealously guarded
it. As the OIC was the principal foreign actor in the Mindanao peace process, the
MILF was effectively sidelined. This suited both the government and the MNLF.
The government greatly appreciated the OIC position, established in the 1970s,
that the peace process was about autonomy and not independence. The nomina-
tion of Indonesia as chair also suited the Philippine government as it relegated
Mindanao to a regional issue while conflicts in Bosnia, Algeria, and other hot-
spots assumed greater importance among OIC members. Furthermore, it has
168 Mark Turner
been alleged that “Indonesian diplomacy is not so much Islamic as it is Asian or

more precisely, ASEAN [Association of Southeast Asian Nations]” (Santos 2001:
80). The Philippines and Indonesia are both members of ASEAN, where non-in-
terference in the domestic affairs of fellow-members is a rule. It should also be re-
membered that the OIC was not an economic organization like the Organization
of Petroleum Exporting Countries (OPEC). It was not in the business of imposing
sanctions and preached a “pacific” approach to conflict resolution (Wadi 1993),
although common membership of OIC and OPEC has led some to suggest that
economic pressure was exerted on the Philippine government by individual OIC
members such as Saudi Arabia (Rodil 2001).
The MNLF liked OIC backing as it gave international legitimacy to its cause
although it never managed to rise above observer status and become a full mem-
ber of the OIC. Member countries also provided convenient bases from which
the MNLF could operate. No other multilateral organization was ever involved in
Mindanao peace negotiations. ASEAN was once approached but its firm policy
of non-involvement in its members’ domestic affairs doomed this effort from the
start. The UN was never brought in to mediate the dispute. The OIC’s consulta-
tive relationship with the UN prevented the MNLF from approaching the UN di-
rectly. As the OIC regarded Mindanao as a matter of autonomy, the MNLF could
not claim status as a “non-self-governing territory” (Santos 2001: 76). The MNLF
was thus prevented from taking its case to the UN.
The involvement of both Libya and Indonesia in Mindanao as the brokers of
peace negotiations in the 1970s and 1990s was conducted under the auspices
of the OIC. In the case of Indonesia and the 1996 agreement, President Ramos
had enjoyed a meeting of minds with President Suharto. The latter was keen on
Ramos’s plans for EAGA, the new sub-regional economic zone in which Mind-
anao and parts of Indonesia were enmeshed. Suharto was also intent on retaining
central control over potentially subversive outer islands. Secessionist movements
in Mindanao might give ideas to disaffected groups in neighbouring parts of In-
donesia.
III Power-sharing Arrangements
A The 1996 Peace Agreement: Phase One

On 2 September 1996, with much fanfare, the Philippine government and the
MNLF signed the historic peace agreement, which would end “decades of war”
and usher in an era of peace and development, acknowledged by both leaders as
“a path more difficult than the path of war” (Mercado n.d.). The agreement de-
clared itself to be “the full implementation of the Tripoli Agreement” of 1976.
The 1996 agreement identified two phases to the peace process. Under the
first phase, there was the immediate establishment of the Special Zone of Peace
and Development (SZOPAD), comprised of the fourteen provinces and nine cit-
ies occupying the territory originally identified as the autonomous region in the
Tripoli Agreement 1976. Overseeing the SZOPAD was the Southern Philippines
Council for Peace and Development (SPCPD), comprised of a chair, a vice-chair,

and three deputies, each one representing the Muslims, the Christians, and the
cultural communities (Lumads). No indication was given about how the officials
would be chosen. An SPCPD advisory council of unspecified numbers was to be
created by the Chair of the SPCPD.
The function of the SPCPD to oversee peace and development was broadly
defined and included promoting, monitoring, and coordinating the improve-
ment of peace and order; monitoring, promoting, and implementing peace and
development projects; attracting foreign investment; supporting local govern-
ment units; and assisting in electoral organization. The ‘peace and development’
mission seemed to encompass every activity from crime prevention to electoral
administration and construction of infrastructure. Various official taskforces and
regional development entities such as the Southern Philippines Development
Authority (SPDA), the regional and field offices of the Office of Muslim Affairs
(OMA), and Task Force Basilan were placed under the control and/or supervi-
sion of the SPCPD as far as their activities in the fourteen provinces and nine
cities were concerned. In order to pursue its diffuse mission, the SPCPD was
empowered by the 1996 peace agreement to set up the necessary “offices and
instrumentalities for the effective and efficient administration of the affairs of
the area” but only after budgetary approval from the Office of the President. This
maintenance of central budgetary control over SPCPD activities is reminiscent of
President Marcos’s methods of keeping the previous autonomous region authori-
ties in line with his wishes.
In addition to the SPCPD, the peace agreement established a Consultative As-
sembly (CA) with eighty-one members. Forty-four places were reserved for the
MNLF, thus giving it an automatic majority. Other seats were occupied by the
political leaders of provinces and cities in the SZOPAD and the ARMM. Eleven
members were drawn from civil society. The Chair of the SPCPD was the CA’s
head. The CA was to be a consultative body where concerns could be ‘ventilated’
and from where advice could be drawn by the SPCPD and the president. It was
also awarded the authority to make rules and regulations “for the effective and
efficient administration of the affairs of the area” – just like the SPCPD. This du-
plication was rather confusing as there was no indication of what regulatory pow-
ers the CA possessed and where they might differ from the equally vague SPCPD
authority in this field.
Much more specific were the instructions for integrating elements of the
MNLF with the Philippine National Police (PNP) and the AFP. The PNP was to
receive 1,500 MNLF members plus another 250 for auxiliary services. The AFP
was to absorb 5,750 MNLF members. No timetable was prescribed except that
the integration should begin at once. Members of the MNLF not absorbed into
the AFP and PNP were to be catered for in a “special socioeconomic, cultural and
educational program.” The idea was to prepare these MNLF fighters and their
families for developmental endeavours. Priority hiring of MNLF members was to
be a feature of development projects in the SZOPAD. A Special Regional Security
170 Mark Turner
Force (SRSF) was also indicated as a further organization for employing out-of-
work MNLF guerrillas.
B The 1996 Peace Agreement: Phase Two

The first phase of the peace agreement was scheduled to last for three years, af-
ter which phase two would begin. This involved repealing Republic Act 6734 on
the ARMM. A new act of Congress would establish an extended autonomous
region. Presumably, government publicity and the achievements of both the old
ARMM and the SPCPD would persuade other provinces and cities to join the
new venture. Once again, there would be a plebiscite to decide which additional
territories would join the extended ARMM. This was a high risk venture given
the overwhelming popular rejection of the first version of the ARMM and its
lacklustre performance since.
The 1996 peace agreement contained recommendations of the government
to the Congress for incorporation into the new or revised act for the ARMM.
The legislation was supposed to be completed and subjected to popular approval
within two years (1998) of the establishment of the SPCPD. A “convoluted pro-
cess” in Congress ensured that this timetable was not kept (Ferrer 2000). The
lower house only succeeded in passing its bill at the third reading in July 1999
while the senate’s self-imposed deadline of completing the work by mid-2000
was not met. Final passage of Republic Act 9054 was still a year away and four
years behind schedule. It was eventually passed by the legislature and signed into
law by the president on 31 March 2001.
1 The Structure of Regional Government

Republic Act 9054 provides for an ‘expanded’ autonomous region, membership of
which would be determined in a plebiscite. The ARMM is clearly identified as “an
integral and inseparable” (Art 3(1)) part of the Republic of the Philippines. Cer-
tain powers are devolved to the regional government, particularly in education,
health, human resources, science and technology, and people empowerment. The
national government retains foreign affairs; national defence, and security; postal
service; coinage and fiscal and monetary policies; administration of justice; quar-
antine; customs and tariff; citizenship; naturalization, immigration and deporta-
tion; general auditing; national elections; maritime, land, and air transportation
and communications, except intra-regional operations; patents, trademarks, and
copyrights; and foreign trade.
The ARMM slots into the territorial hierarchy between central government
and province or city. This makes for a long supervisory chain. Formally, the presi-
dent supervizes the regional governor, who looks after the provincial governors
and city mayors. The provincial governors supervise the municipal mayors, who
like their city counterparts supervise the barangay (local community) heads. In
turbulent parts of Mindanao, it is sometimes difficult for these formal arrange-
ments to be realized on a permanent basis. Because of this, and to reflect the
ARMM’s particular characteristics, the regional government enacted its own lo-
cal government code in 1993. However, this code has never been implemented
and was not revised to accommodate RA 9054. Local government units (LGUs)
thus continue to operate according to the 1991 Local Government Code.
The regional governor is an ex officio member of the National Security Council.
A couple of other intergovernmental matters are qualified by the phrase ‘as far as
practicable’. These are the appointment of the regional governor’s recommendees
to the national cabinet and executive departments and constitutional bodies.
The regional assembly is the ARMM’s legislative body, comprised of popularly
elected members and elected sectoral representatives from among agriculture,
labor, urban poor, disabled, indigenous cultural communities, youth, and women.
The number of sectoral representatives cannot be more than 15% of the regularly
elected members. All members have three-year terms and cannot serve more
than three consecutive terms. Bills require three readings for approval. Chief
among these are the annual budget for the regional government and a regional
public works act . The regional governor can sign into law or return a bill to
the assembly for reconsideration. A two-thirds majority can then make the bill
into law whatever the governor thinks. The assembly can change local govern-
ment boundaries subject to the standards set in the Local Government Code 1991
and the approval of a plebiscite. The assembly can also enact a civil service law
concerning appointments to government positions. However, for up to six years,
national government rules on eligibility, conduct, and other civil service matters
will apply.
The chief executive (regional governor) and deputy (vice governor) are popu-
larly elected as a team from the same party or coalition of parties. Both officials
can serve a maximum of three terms of three years each, just like provincial gov-
ernors, mayors, and congressional representatives. The governor selects a cabinet
of not more than ten persons, six of whom must come from “indigenous cultural
communities,” which are comprised of Muslim groups and tribal communities
(Lumads). An executive council of three deputies representing the Christians,
indigenous cultural communities, and the Muslims is appointed by the governor.
The governor controls the regional executive commissions, agencies, boards, bu-
reaus, and offices and is responsible for senior appointments.
RA 9054 declares that the ARMM will “enjoy fiscal autonomy.” In practice,
national government transfers still account for 97% of the ARMM government’s
income, even though there is provision for retention of 70% of internal revenue
taxes generated in the region. ARMM government income is subject to semi-an-
nual and annual audits by the National Commission on Audit and the rules of
the Department of Budget and Management. If there is a failure to comply with
these rules, the president can suspend or dismiss offending officials. Rules of the
Department of Budget and Management also govern revenue generated by the
regional government and that received from donors.
The ARMM government has enacted its own local revenue code, but it is only
partially implemented. Most of the LGUs in the ARMM fall into the two lowest
income categories and obtain over 90% of their revenue from central government.
Real property taxes should be a major source of local government income, but a
172 Mark Turner
thorough property assessment has not been carried out since the mid-1990s. The
possibility of raising additional income is hampered by economic underdevelop-
ment, poor institutional capacity and security concerns. The situation is exacer-
bated by poor awareness of tax laws and low levels of compliance with them.

Health is mentioned only in passing in RA 9054. In one of the guiding principles
and policies it is stated that the regional government “shall provide maintain, and
ensure the delivery of … basic and responsive health programs” (RA 9054 Art
III(11)). Health services were devolved to local government units under the Lo-
cal Government Code 1991. Health programmes, projects, and monitoring and
supervisory activities are evident in the planning forums and documents that
are prescribed in the act for the regional government. Health service delivery
remains a provincial, city, and municipal responsibility.
Health gets a few further mentions in association with social services and wel-
fare provisions. Social welfare functions were also devolved in 1991 under the na-
tional Local Government Code. Nevertheless, in an abrupt sweeping statement,
the regional assembly is enjoined to “enact measures to provide and promote
social services” (RA 9054 Art XV(2)). It must also maintain an ‘effective’ food
and drug regulatory system. Further legislation is required on child health and
development, including support for the ‘physically challenged’ and other disad-
vantaged persons; and the protection and development of the rights of women
and indigenous populations. The regional government is to ‘evolve’ a housing
programme in cooperation with the private sector to provide adequate low-cost
housing and other basic services. The regional government must also not only
recognize the Filipino family as the foundation of the nation but seek ways to
“strengthen its solidarity and actively promote its total development”, however
that may be interpreted (RA 9054 Art XV(5)). Finally, the regional government
“may” create an office of youth affairs.
3 Education
Providing “quality education” is identified as a top priority of the ARMM in RA
9054. The regional government assumes responsibility for the schools, colleges,
and universities in the ARMM as a ‘subsystem’ of the national education system.
The act identifies a long list of policies and principles that “the integrated system
of quality education” should follow. Values assume an important position – Fili-
pino and Islamic values are to be perpetuated; peaceful settlement of disputes
should be inculcated in students; patriotism and nationalism are to be promoted;
and consciousness and appreciation of ethnic identity developed. Religious in-
struction is optional. English and Filipino are the official languages of instruction
with Arabic as an auxiliary. The latter is important in the Islamic madrasah edu-
cational system, where religion is of central concern. Other regional languages
may also be used as auxiliary languages for instruction and there is even provi-
sion for the highly unlikely evolution and development of a regional language
among the far-flung and linguistically diverse population.
The structure and administration of education retains a close connection with

the national system. The act confirms the creation of a regional Department of
Education, Culture and Sports to supervise and regulate both public and private
schools. State colleges and universities are parts of the regional education subsys-
tem but have academic and fiscal autonomy and remain governed by their own
charters. The organizational structure of the ARMM subsystem must follow the
basic structure of the national system with the same number of years required at
each level. The regional education department should develop school curricula
“that are relevant to the economic, social, political, cultural, moral, and spiritual
needs of the people in the autonomous region” (RA 9054 Art XIV(3)). Responsi-
bility for the provision of learning aids and instructional materials is also given to
the regional government but textbook preparation, writing, revision and printing
is a joint responsibility of central and regional governments.
Personnel matters are under regional government control but national mini-
mum standards must apply. Thus, regional government is responsible for selec-
tion, recruitment, appointment, and promotion of both teaching and non-teach-
ing personnel. Regional standards for personnel matters may be imposed by the
regional government but must not be below those of the national Department of
Education, Culture and Sports and the Technical Education and Skills Develop-
ment Authority. Existing personnel cannot be removed if they have already satis-
fied national civil service eligibility. Whether national standards actually apply
in the region’s various zones of conflict is doubtful. For example, it is difficult to
recruit teachers with the prescribed formal qualifications.
While “management, control and supervision” of education are placed in re-
gional hands, it is the national agencies for education that are given the job of
monitoring compliance with national education policies, standards, and regula-
tion. The regional government is invited to participate in the “policy and deci-
sion-making activities of their counterparts of the central government or national
government in matters that affect the regional educational subsystem” (RA 9054
Art XIV(4)). Curiously, the Koranic-based madrasah system is to be supervised
and periodically reviewed by the national Department of Education, Culture and
Sports. Policy for the madrasah is a regional responsibility, although there have
been recent national urgings to “mainstream the madiris” by standardizing the
curriculum, introducing new subjects, and upgrading the teachers (Bagayaua
2002).
Funds for education are provided by the national government to the region-
al government in the General Appropriations Act and are released directly to
the regional treasurer. Donations to public or private universities, colleges, and
schools are tax deductible from regional government charges. Donations can be
sought and received from overseas as well as from domestic sources. The regional
government is obliged to provide scholarships to qualified but poor students at
all levels of education, and at least 15% of the education budget must be devoted
to these scholarships. The scholarship programmes of the central government
Commission on Higher Education and Technical Educations and Skills Develop-
ment Authority also pass to regional control. The regional government is also
174 Mark Turner
called upon to provide financial assistance to “disadvantaged but deserving” stu-

dents in addition to those covered by the 15% of the educational budget suppos-
edly allocated to scholarships.
4 Economic Policies
RA 9054 has much to say on the regional government’s role in the economy and
gives it a wide-ranging mandate for promoting economic development in an eco-
nomically underdeveloped region. The ARMM is the poorest region in the coun-
try, with official figures showing 68% of the population living in poverty. Accord-
ing to the Filipino Report Card of the Asian Development Bank, the ARMM has
the country’s highest levels of dissatisfaction with government services and has
welfare indicators that are all below the national average. It is therefore a priority
of government to promote economic development and poverty alleviation. The
regional government can encourage and support the establishment of economic
zones, industrial centres, ports, and growth centres. Incentives can be offered
to investors, such as tax rebates and holidays. The construction of transport and
communications facilities to ‘expedite’ economic development is identified as a
priority of the regional government. Delivery of power services is another pri-
ority. To facilitate this, the regional government may also establish and operate
‘pioneering public utilities’, which it may later sell to cooperatives or other col-
lective organizations. This seems to be an extension of the regional government’s
obligation “to promote profit sharing and broaden the base of ownership of busi-
ness enterprise” (RA 9054 Art XII(15)). Generating such economic growth is an
extremely difficult task in a region shunned by capital because of its association
with armed conflict, kidnapping, bombings, and limited human resources and
physical capital.
To preside over economic development is a Regional Economic and Devel-
opment Planning Board. It is chaired by the regional governor with members
including provincial governors and city mayors, the Speaker and two members
of the regional assembly, and five people from the private sector. The board’s
purpose is to serve as “the planning, monitoring, and coordinating agency for
all development plans, projects, and programs intended for the autonomous re-
gion” (RA 9054 Art XII(10)). It recommends the annual work programmes and
comprehensive development plan for the ARMM to the regional assembly. Once
these items are approved, the regional governor theoretically must implement
them. Lack of capacity, shortage of funds, and political manoeuvring may prevent
effective implementation. The board must also design and maintain a master plan
for “total development of the region” taking into account the development plans
of provinces, cities, municipalities, and barangays in the ARMM. This master
plan should include “the comprehensive and integrated urban and rural develop-
ment policies, plans, programs, and projects” specified in Article XI of RA 9054.
It has been slotted into the updated medium-term Philippines development plan.
Not only is the ARMM government responsible for regional planning but also it
is given the task of strengthening the planning bodies of lower level subnational
government units.
While the act does acknowledge the private sector as the ‘prime mover’ of
trade, commerce, and industry, it nonetheless gives the regional government a
range of tasks to facilitate the process or even become directly involved in. It has
an important role in supporting the development of entrepreneurial capability.
Cooperatives and cottage industries are also to be promoted and protected by
regional government. The regional government is to encourage the establishment
of banks and financial institutions, including foreign banks, in the region, subject
to the ‘supervision’ of the central bank. The latter is also given the task of estab-
lishing an Islamic bank in the region and authorizing its branches. It must set up
its own Bangko Central Regional Branch with full banking services. This replaces
its ‘regional office’. Away from the bureaucracy of banking, the regional govern-
ment is somehow supposed to regulate the barter trade and counter-trade with
neighbouring Indonesia, Malaysia, and Brunei. Such regulation is an impossible
task given the porosity of the international borders in the Celebes and Sulu Seas.
The ARMM government should also look after consumer interests and promote
the preferential use of local labour. A final financial matter is the rather vague
provision that the regional government will “regulate and exercise authority over
foreign investment within its jurisdiction” (RA 9054 Art XII(34)). The regional
government may itself seek loans from both domestic and foreign sources. While
the region remains such an unattractive destination for investment capital, the
ARMM government will have little foreign investment to supervise and few loans
to administer.
The act devotes a few paragraphs to regional government’s role in agriculture,
fisheries, and aquatic resources. Its primary task is to look after the rights and
welfare of the workers in these activities in the predominantly rural ARMM.
There is the briefest of statements asserting the regional government’s function
of encouraging agricultural productivity, and diversified and organic farming. It
‘may’ enact an aquatic and fisheries code to “enhance, develop, conserve, and pro-
tect marine and aquatic resources” (RA 9054 Art XII(24)). To pursue its primary
industry responsibilities, the regional assembly must create a Bureau of Agricul-
ture and Fisheries and then define its composition, powers, and functions. The
act gives no advice on what these may be. The possibility of a regional law on
agrarian reform is flagged but not prescribed.
Natural resources are seen as an important responsibility of regional govern-
ment. It is awarded “control and supervision over the exploration, utilization,
development, and protection of the mines and minerals and other natural re-
sources within the autonomous region (RA 9054 Art XII(5)). It can grant fran-
chises, concessions, leases, permits, and licenses over mineral, forest, and agri-
cultural lands. Consultation with cultural communities is mandatory where their
lands are involved, but the act does not indicate if the regional government can
overrule opposition from such groups. The fees for exploration and utilization
are set by the regional assembly. Citizens of the autonomous region have prefer-
ence in the exploration, utilization, and development of the natural resources,
although in most instances they would lack both capital and technology. Where
they do sometimes operate is in small-scale mining – an unruly activity, which
176 Mark Turner
the regional government is supposed to regulate. Foreign involvement in natural

resources can be no more than 40% ownership of the organization involved. The
central government retains control over ‘strategic minerals’, such as uranium and
petroleum, all sources of ‘potential energy’, national reserves and aquatic parks,
and existing forest and watershed reservations. As compensation, the regional
government does receive a share of revenues from these strategic minerals, which
it then distributes to other local government units according to a set formula.
Protection of the environment is mentioned in various places in the act. The
regional government is to give priority to environmental protection and sus-
tainable development. More specifically, the regional assembly can pass laws to
provide compensation and rehabilitation to peoples and areas adversely affected
by natural resource development. Corporations operating on ancestral domain
lands are required to safeguard or restore ecological balance. The regional gov-
ernment must ‘immediately’ instigate measures to ensure that at least 50% of the
ARMM is covered with trees. To assist this process, 10% of the internal revenue
taxes of the regional government and all other local government units should
be devoted to reforestation and development of the environment. It is not clear
whether this actually occurs.
A final item of development responsibility for regional government is tour-
ism. Indeed, it is awarded primary responsibility for promoting a tourist indus-
try that upholds, respects, and maintains the diverse cultural heritage and moral
and spiritual values of peoples in the region. Elsewhere in the act, the regional
government is given the job of protecting and promoting culture and given the
option of setting up a bureau of cultural heritage. Similarly, the regional assembly
may set up a regional tourism office. This is unlikely to be a priority, as foreign
governments issue stern travel warnings to their citizens against venturing into
the ARMM. Sporadic clashes between armed Islamic groups and the military,
kidnappings by the Abu Sayyaf and criminal gangs, and other examples of violent
disorder mean that tourists are few and far between.
5 Administration of Justice
The administration of justice has been an important concern for the region’s in-
habitants and is reflected in the act’s attempt to satisfy diverse constituencies
through the creation of a plural system. A tripartite structure was defined and
comprises the national system of courts, the Muslim shari’a courts, and tribal
courts for indigenous cultural communities. The shari’a and tribal courts are
only applicable to members of the Muslim and indigenous cultural communities.
Where conflicts occur between the different types of law, national law prevails,
with the Supreme Court as the overriding authority for the whole regional legal
system.
The act has little to say about the national system except on appointments. It
is deemed desirable but not compulsory to appoint one justice in the Supreme
Court and two in the Court of Appeals from qualified persons in the ARMM. A
consultant to the Judicial and Bar Council is to be appointed by the president
on the recommendation of the regional governor to advise on appointments to
judicial positions in the ARMM. A deputy court administrator for the ARMM is
appointed by the Chief Justice of the Supreme Court, also on the recommenda-
tion of the regional governor.
Although the 1996 peace agreement makes no mention of shari’a courts, the
ensuing act has much to say about their powers and operation. These courts are
not new institutions and the contents of the RA 9054 reiterate and extend earlier
legislation. One innovation is the creation of a Shari’a Public Assistance Office to
provide free legal assistance to poor litigants. A Shari’a Appellate Court is also
created and has exclusive appellate jurisdiction over all cases tried in the existing
shari’a district courts. The act provides details of positions in the Shari’a Appel-
late Court, eligibility, and the appointment process. The Shari’a circuit courts
continue to function as before.
The act also provides for the possibility of creating a system of tribal courts,
including a Tribal Appellate Court. The tribal courts may “determine, settle,
and decide controversies and enforce decisions involving personal and property
rights of members of the indigenous cultural community concerned” (RA 9054
Art VII(19)). They are also awarded authority over minor crimes within indig-
enous communities. The composition and jurisdiction of the tribal courts will
be determined by the regional assembly, which is awarded the daunting task of
codifying the customary laws of Muslim groups and indigenous cultural com-
munities.
The maintenance of law and order is to be given priority status by the regional
government. This is to be expected in a region that has been host to consider-
able political and criminal violence since at least the 1960s. However, ensuring
“peace and the protection of life, liberty, and property of the people in the au-
tonomous region” is an extremely difficult task in this context. The act creates
a Philippine National Police Regional Command known as the Special Regional
Security Force (SRSF) or the Regional Police Force. This is the force into which
the former MNLF forces were integrated in Phase I of the 1996 peace agreement.
The regional assembly is to pass laws governing the SRSF, but they must be con-
sistent with the relevant parts of the Constitution and the provisions of RA 9054.
The functions of the SRSF are unremarkable and include normal police functions
such as enforcing congressional and regional assembly laws; maintaining law and
order and ensuring public safety; effecting arrests, searches, and seizures in ac-
cordance with the Constitution and pertinent laws; detaining persons for lawful
periods and informing them of their constitutional rights; observing the human
rights of all people in the region; and licensing firearms and initiating drives to
secure surrender of unlicensed firearms.
The SRSF is a civilian organization. It operates in the ARMM but its personnel
can be deployed elsewhere in the Philippines if required by the National Police
Commission and authorized by the president. The SRSF is headed by a regional
director and two deputies who are drawn from the ranks of the professional po-
lice force, preferably from within the ARMM. Its offices at provincial, city, and
municipal levels are also to be headed by professional police.
178 Mark Turner
The regional governor is given considerable influence and control over the
SRSF. The governor is deputy of the National Police Commission in the region
and ex officio chair of the Regional Police Commission. Most importantly, the
governor exercises “operational control and general supervision and disciplinary
powers over the Regional Police Force” (RA 9054 Art XIII(8a)). The governor
also has authority over the regional director regarding the deployment of the
SRSF and can impose administrative penalties on police after hearing citizen’s
complaints. Finally, the governor recommends to the president the candidates for
regional director and deputy directors of the SRSF.
The SRSF is not the only official armed force in the ARMM. The act clearly states
that “the defense and security of the autonomous region shall be the responsibil-
ity of the central government” (RA 9054 Art XIII(11)). For this purpose, a regional
command of the Armed Forces of the Philippines (AFP) has been specified in the
act. This regional command remains under central government and AFP control.
The regional governor may ask the president for the deployment of the regional
AFP command if there is invasion, rebellion, or simply lawlessness and violence
where the SRSF cannot cope. But the president may take independent action to
send the AFP into the region in order to deal with violent situations such as the
efforts to destroy the Abu Sayyaf in the province of Basilan in 2001–2002.
6 Human and Minority Rights

One of the major themes of RA 9054 is guarantees for the rights of the differ-
ent communities within the ARMM, whether minorities or not. Early in the act
there is a strong statement recognizing, protecting, and guaranteeing the beliefs,
customs, traditions and freedom of religion for people in the region. The regional
assembly is then charged with adopting the necessary measures to achieve this
state of affairs. Also among the guiding principles and policies is a direction to
the regional government “to uphold and protect the fundamental rights of wom-
en and children,” the right of women to engage in lawful employment, and the
right of women and children to be protected from exploitation, abuse, and dis-
crimination (RA 9054 Art III(10)). Later in the act, the regional government is
given the task of ensuring the representation of women in ‘appropriate’ decision
and policy-making bodies.
The rights of indigenous cultural communities is a recurrent theme in the act.
Two types of such communities are recognized: tribal peoples (Lumads) and the
Islamic Bangsamoro people. As already indicated, the act provides for special
education and legal systems for these communities. They should also be able to
rely on regional government to protect their ancestral domains. Strategic min-
erals and other strategic resources appear to be exceptions to total community
control over ancestral lands.
The act also orders the creation of a regional human rights commission within
the ARMM. It will perform the functions of the national government’s commis-
sion on human rights, although there is allowance for additional functions to
enhance human rights in the region. The composition of the regional commission
should reflect the ethnic distribution of the resident population. It is comprised
of a chair who must be a lawyer and resident of the ARMM and two commission-
ers who are preferably lawyers.
The reference points for human rights in the ARMM are made explicit in the
act. These include the act itself, the Constitution, the Geneva Convention, the
United Nations Charter, the United Nations Declaration on the Rights of Indig-
enous Peoples, and the Universal Declaration on Human Rights. Violations of
human rights are still regularly reported by both international and domestic or-
ganizations.
IV Reaction to the 1996 Peace Agreement
A Autonomy in a Decentralized Policy

The first phase of the peace agreement can be summed up as the delegation of
some executive authority by the president. Legislative authority remained in the
hands of the Congress, the local government units (LGUs) and the ARRM. But
even the delegated executive authority had its limits, as the LGUs were already
responsible for administering many of their own affairs. The Local Government
Code of 1991 involved substantial decentralization of responsibility for local af-
fairs.
The momentum of ‘people power’, which had swept Ferdinand Marcos out of
and Corazon Aquino into the presidential office, led to an ambitious programme
of devolution or political decentralization. The Philippine hierarchy of subna-
tional territories builds from the grassroots barangay, through to the municipal-
ity and city, and on to the province. At each level, there is an elected assembly
to legislate on a variety of matters (Turner 1999; Tapales 1996). The barangay
also has the role of interest articulation. Elections are held every three years and
seats are hotly contested. At municipal, city, and provincial levels, the electorate
are asked to vote for the chief executive (mayor or governor) and members of
the sanggunian (council) The range of functions allocated to LGUs was greatly
enhanced with the Local Government Code of 1991. This provided for the decen-
tralization of basic services in health, agriculture, public works, social welfare,
and environment and natural resources. The personnel, buildings, equipment,
and financial resources for these activities were transferred from central bureau-
cracies to LGUs. The LGUs also have the authority to raise revenue from licenses
and other local levies and are encouraged to seek additional innovative ways of
generating income. A further important change decreed in the Local Govern-
ment Code was the compulsory involvement of NGOs and People’s Organiza-
tions (POs) in local governance, especially through Local Development Councils
at all levels.
This system of decentralized local governance was already operating when
the 1996 peace agreement was signed. The enactment of the region’s own local
government code in 1993 had little effect as it was not implemented. The newly
created SZOPAD, SPCPD, and CA did not appear to contest the powers and op-
erations of the LGUs as specified under the national Local Government Code
180 Mark Turner
1991. According to Paragraph 6 of the peace agreement, “the local government

units in the area including the ARMM, shall continue to exist and exercise their
functions in accordance with existing laws.” Presumably the ARMM’s local gov-
ernment code was one of those laws, even though it was having minimal effect
on local governance. Then, in Paragraph 18, the SPCPD was given the function of
providing “support to local government units as necessary.” Thus, the temporary
institutions established under phase one of the peace agreement had no authority
over the functions of government. The LGUs got on with delivering decentral-
ized services, as did the field offices of central agencies such as the Department
of Education, Culture and Sports (DECS). The SPCPD and CA were supposed
to develop horizontal linkages with the LGUs and the deconcentrated offices of
central government but they did not have authority over them. They were to give
“support”, but how that support should be organized remained unspecified in the
agreement.
The LGUs were under no obligation to accept SPCPD and CA support or ad-
vice. However, most LGUs were poor, characterized by low income and develop-
ment indicators consistently below national averages. This provided the oppor-
tunity for the SPCPD and CA to exert influence and win friends in local places,
as considerable developmental funding was anticipated to rebuild communities
in the SZOPAD. This funding would be channelled through the SPCPD, thereby
making cooperation more alluring for LGUs. Whether the SPCPD and CA had
much autonomy is debateable. The financial lifeblood and any authority they ac-
quired derived from the Office of the President. The SPCPD and the CA might
even be viewed as extensions of the Office of the President, albeit unpredictable
ones.
The signing of the peace treaty did not mark the cessation of international in-
volvement. There was still a place, after more than twenty years of engagement,
for the OIC. According to Paragraph 13 of the agreement, the OIC was to be part
of a Joint Monitoring Committee with representatives of the MNLF and govern-
ment of the Philippines. The task of this committee was “to review and identify
agreements that can be immediately implemented, and monitor the implementa-
tion of this Agreement during Phase I.” No other international body was men-
tioned in the agreement but it was anticipated that under UNDP (United Nations
Development Programme) leadership there would be an influx of multilateral
and bilateral donors anxious to extend assistance in the “land of promise.”
Finally, on the individual level, the peace agreement represented the triumph
of Nur Misuari. He had transformed himself from long-time guerrilla leader into
legitimate politician and won several prizes as a consequence. First, he was ap-
pointed the Chair of the SPCPD. Second, as incumbent of this position he be-
came “the head and presiding officer” of the CA. Finally, the signing of the peace
agreement just before the elections for the Governor of the ARMM allowed Mis-
uari to run unopposed for this position. As chair, presiding officer, and governor
Nur Misuari would be identified as saviour or villain in the quest for peace and
development on Mindanao. Clarification of his multiple roles and the relation-
ships between the SPCPD and existing government agencies would have been of
considerable assistance but the Executive Order 371 to implement the agreement
continued the theme of vagueness.
B The 1996 Peace Agreement: Support and Opposition

During the early years of implementing the peace agreement, the principal sup-
porters remained its chief sponsor, President Ramos, and his executive staff. The
peace was one of the preconditions for the economic development of Mindanao
that the president so urgently desired. Such development would facilitate Philip-
pine efforts to catch up economically with more successful ASEAN neighbors.
Ramos had clearly flagged his intentions on Mindanao immediately on assum-
ing the presidency by creating the first ever Office of the President in Mindanao.
Furthermore, he appointed as its head Paul Dominguez, a businessman who
shared Ramos’s visions, strategies, and penchant for hard work. Dominguez de-
clared his intention to generate “quick starts, early results on infrastructure and
peace and order problems to gain widespread support” (Vitug and Gloria 2000:
253). This executive commitment was important in increasing infrastructure and
private investment in Mindanao, thus producing consistent economic growth.
Throughout his presidency, Ramos emphasized the importance of Mindanao and
the need to have a Mindanao-wide strategy instead of the fragmented and stut-
tering approaches of the past. As Vitug and Gloria (2000: 255) observe, “No other
president gave such consistent and sustained attention to Mindanao.”
His successor, Joseph Estrada, did not continue the trend. Elected with an over-
whelming popular mandate in 1998, the new president had garnered considerable
support in Mindanao. This was odd given his lack of coherent policy on Mind-
anao. It was described as a “clueless executive” in this regard (Vitug and Gloria
2000: 256). The Office of the President in Mindanao was closed – Dominguez
had already gone. In their place emerged a Presidential Assistant for Regional
Concerns (PARECO) to cover the whole country. Mindanao was split into two
regions with the heads of these regions reporting to the Executive Secretary in
Manila. Paradoxically, this represented a re-centralization of Mindanao affairs
where nobody was specifically in charge of the affairs of the island. Mindanao had
slipped down the list of policy priorities. This was strange given Estrada’s election
on a pro-poor platform and Mindanao’s appalling incidence of poverty.
The Congress and elected politicians in local government on Mindanao did
not share a unified view on the SPCPD, the CA, and the future prospect of an
expanded ARMM. The opposition, which had attempted to prevent the peace
agreement continued its anti-SPCPD mantra. In the 1998 elections, some pro-
agreement politicians were voted out of office. Rosalita Nuñez, the Mayor of
General Santos, a booming city in the south, was ousted by the husband of one of
the oppositionist ‘Tres Marias’. In Zamboanga City, another Maria switched from
congressional representative to mayor, leaving her son to occupy the congres-
sional seat. Their election slogan in this largest city of the SZOPAD was “No to
SPCPD.” Even among the Muslim constituency, the MNLF seemed to be losing
support. MNLF candidates who were mostly associated with the political party of
182 Mark Turner
President Ramos polled poorly. There were only two successes among the MNLF
leaders who ran for Congress or important local government posts (Gutierrez
1999).
The MNLF’s leadership of Mindanao’s Muslim population was under threat
from the MILF and the Abu Sayyaf. During the peace negotiations and subse-
quent implementation of the agreement, the MILF had steadily expanded its geo-
graphical coverage, the numbers of its supporters, and its military capacity. The
MILF had been excluded from the peace talks and had not endorsed the agree-
ment. Their position, said leader Hashim Salamat in 1998, remained “the re-es-
tablishment of a sovereign Moro Islamic State” (Zahir 1998). And they seemed to
be to doing this as part of a “transition from a guerrilla force into a ‘semi-autono-
mous army’” (May 2001: 269). Their sprawling Camp Abubukar headquarters in
Maguindanao and the substantial Camp Bushra in adjacent Lanao del Sur were
effectively like autonomous municipalities.
The AFP was concerned by MILF expansion and wished to contain them with-
in their two major camps. Other autonomous military settlements of the MILF
were perceived by the AFP as threats to the territorial integrity of the Philip-
pines. This contributed to a couple of significant breaches of the government-
MILF ceasefire during Ramos’s later years, as intermittent negotiations failed to
produce agreement between the two sides.
President Estrada seemed to enjoy better relations with the MILF. In the run-up
to the 1998 election, the Estrada camp formed an alliance with the MILF. On his
election, Estrada even received a congratulatory letter from MILF leader, Hashim
Salamat, and a formal peace with the MILF began to look possible. But the antici-
pated agreement was not forthcoming. MILF adviser Alunan Glang complained
that “We’re being taken for granted” (Vitug and Gloria 2000: 156). Formal peace
talks stuttered back into life in October 1999 but in early 2000 Mindanao was
once again plunged into full-scale war. The MILF occupied two municipalities.
The AFP retook them and the situation then degenerated into major military
operations against the MILF. The president’s flagging ratings in the opinion polls
rose sharply in response to his uncompromising stand against the MILF. Military
victories were celebrated as MILF camps, including Abubukar and Bushra, were
overrun by the AFP. There was massive dislocation and displacement of popula-
tion – an estimated 900,000 people – but most MILF fighters simply retreated to
continue their 400 years struggle at a later date. The MILF was weakened but the
government still needed its cooperation for peace on Mindanao.
This was not necessarily the case for the Abu Sayyaf. Eradication had always
been the aim of government policy towards this secessionist group. The Abu
Sayyaf came to public notice in 1992 after a series of violent incidents against
Christian targets in Zamboanga City (Turner 1995a). Over the next few years,
there were further Abu Sayyaf bombings, murders, and kidnappings, culminat-
ing in Abu Sayyaf participation in the massacre of fifty-three people in a raid on
the town of Ipil on the Zamboanga peninsula. The Abu Sayyaf was described as
espousing “Islamic fundamentalism” or labelled as “extremist bandits” (Turner
1995a). The prime aim was clear – an independent Islamic state in Mindanao
– and terror was seen as a legitimate way to achieve it. While largely confined to
Zamboanga-Basilan-Sulu, the notoriety of the Abu Sayyaf spread far and wide
and served to reinforce negative Muslim stereotypes and to raise concerns about
whether peace could ever be achieved in Mindanao. Even the MNLF declared
their opposition to this radical force.
By 1996, the Abu Sayyaf was in retreat. The death of its leader and military
reversals led to predictions of its demise or degeneration into yet another ban-
dit group confined to Basilan, one of the Philippines’ poorest provinces. But the
Abu Sayyaf ’s small numerical size (probably not more than 300 prior to 1996), its
lack of formal structure, and its dubious claims to legitimate representation of
the Bangsamoro people had not proven to be handicaps in the business of sow-
ing terror. In the late 1990s, sporadic violent incidents were blamed on the Abu
Sayyaf, but it was in 2000 when the group burst back onto the scene in spectacu-
lar fashion by kidnapping twenty-one people, mainly foreigners, from an island
resort in Malaysia and holding them to ransom in Sulu. Money flowed from the
ransoms and recruitment boomed. Another mass kidnapping from a Philippine
resort in 2001 ensured that the Abu Sayyaf remained in the national and inter-
national consciousness and confirmed the popular view that Mindanao was a
dangerous place, outside of the rule of law, and unsuitable for investment. Many
thought that peace and the Abu Sayyaf were incompatible.
On the international scene, the OIC remained involved and supportive of the
peace process. The OIC assumed the role designated in Paragraph 12 of the peace
agreement of “monitoring the full implementation of this agreement during the
transitional period until the regular autonomous government is firmly estab-
lished and for this purpose, help generate broad international support for the
Zone of Peace and Development.” Other international actors also expressed sup-
port for the peace process through foreign aid. Spearheaded by the UNDP, both
multilateral and bilateral donors sought to invest in development activities that
would support the peace agreement.
C Phase One: The SPCPD Performance

If the peace process on Mindanao was to be successful, then much depended
on the peace and development activities implemented under phase one of the
agreement. The more permanent institutional structures envisaged for phase two
would gain little support if poor performance was perceived to be characteristic
of phase one. There was substantial opposition to and considerable suspicion of
the peace process and the peace agreement among Mindanaoans. Good results
from socio-economic development initiatives might encourage others to join the
expanded ARMM following the termination of the transitional SPCPD.
An early disappointment of the implementation process was the failure of Ex-
ecutive Order (EO) 371 to provide adequate elaboration of the peace agreement.
Instead of clarifying the agreement’s generalizations about relationships between
institutions and the allocation of authorities, EO 371 was unenlightening. One
commentator believed EO 371 had created a “complicated bureaucratic maze” in
184 Mark Turner
which the new institutions would wander about and make little impact (Gutierrez
1999). The MNLF could hardly be expected to demonstrate good developmental
leadership and managerial capacity in the labyrinth of bureaucratic confusion.
“Like a fish out of water,” wrote Vitug and Gloria (2000: 98). The transformation
of guerrilla fighters into functioning bureaucrats is an enormous challenge. They
must alter their organizational structures, patterns of authority and operating
methods, which had been created for armed struggle. For peacebuilding these ar-
rangements are inappropriate. Furthermore, the guerrillas turned peace-builders
were dealing with a government system long associated with a range of bureau-
cratic pathologies (De Guzman 1993: Varela 1996: Turner 2001). When a govern-
ment fails to delineate and delegate in such a context, official performance will
suffer.
The great hope for peace-building was development plans and projects. War
and disorder had contributed to official neglect for over three decades. A post–
agreement survey found that 64% of MNLF personnel were unemployed, “their
communities characterized by lack of potable water, poor sanitation, poor access
roads, inadequate schools, poor community health services, poor reproduction
support facilities, low productivity and displacement” (Aguja 2000). As the gov-
ernment had limited funds to address these enormous problems the international
donors were approached.
There was a slow start, but, by December 1997, approximately usd 500 mil-
lion had been pledged by donors. One of the most significant innovations was
the SPCPD-NEDA (the central government planning body National Economic
Development Agency)-UN/Multi-Donor Program specifically oriented to as-
sisting MNLF members, their families, and their communities. The programme
involved livelihood assistance, small-scale economic projects, vocational skills
training, and construction or rehabilitation of community infrastructure. Par-
ticipation by the beneficiaries was an essential ingredient, as the programme had
a strong confidence-building objective. While the programme has undoubtedly
produced benefits among the affected communities, there have been criticisms
(Aguja 2000). Other stakeholders in local development felt marginalized. This in-
cluded Christians, Lumads, LGUs ,and civil society organizations. Thus, it did not
promote the principal objective of peacebuilding – reconciliation. Paradoxically,
the amounts of money were small (only usd 9.1 million as of the end of 1998) and
so the tangible benefits have been negligible. The programme management also
reinforced the territorial structures of the MNLF as a fighting force. The MNLF
‘states’ (groups of adjacent communities) and their MNLF leaders were used as
the references for the programme. This could undermine the position of the LGU
as a focal point for governance and developmental activity. The MNLF has also
been disturbed by the bureaucratic requirements of the programme which re-
quired considerable documentation of MNLF communities. Should peace fail,
people in these communities may be at considerable risk. Finally, there has been
concern that the SPCPD-NEDA-UN/Multi-Donor Program has created a ‘dole-
out’ mentality rather than a self-help orientation. That is, MNLF members are
entitled to compensation for their years of hardship.
Other funds have been flowing into Mindanao and SZOPAD. The government
calculated that, as of March 2000, more than P30 billion of socioeconomic devel-
opment funds had flowed into the SZOPAD (GRP 2000). The World Bank set up
a Social Fund in October 1997 but ten months later no project had been started.
Applicants needed to comply with bureaucratic requirements, which were not
always understood or which could not be met. Social funds have been used by
the World Bank in many poor countries to provide small-scale infrastructure that
benefits the wider community and to use local paid labour in its construction.
According to Vitug and Gloria, the World Bank rated the Mindanao Social Fund’s
performance as “moderate” but worried about sustainability (2000: 270). Despite
these concerns, a new social fund project is set to start in 2003 to expand and
build on the original project. The line departments of central government agen-
cies have also built or improved infrastructure, promoted community develop-
ment, and engaged in other developmental activities. Whether these initiatives
can be attributed to the peace process is difficult to determine, but suspicions
are that they represent little if any additional investment. But other bilateral and
multilateral donors have been active in promoting a range of projects. There is
even a reported “consensus in Mindanao that the (peace) process would have
failed without donor support” (UNDP adviser, as quoted by Vitug and Gloria
2000: 275).
While the Philippine government has been anxious to paint a picture of great
investment and achievement in socioe-conomic development, other commenta-
tors have been less enthusiastic. In 1998, the OIC monitors noted that: “Develop-
ment is still wanting and is not anywhere near the target to uplift the sufferings of
the people in Mindanao who still live in dire poverty, specially that of the MNLF
members and their families” (OIC 1998). The OIC went on to criticize disburse-
ment of funds, the meagre sums allocated to the ARMM, and the institutional
inadequacies of the SPCPD.
One area for which the OIC reported matters running “quite smoothly” was in
relation to the integration of the MNLF into the AFP and PNP. By March 2000,
5,250 MNLF members had joined the AFP, 91% of the agreed quota (GRP 2000).
The PNP quota of 1,500 was also nearly achieved. The PNP had recruited 1,250
former MNLF personnel and some had already completed training and were allo-
cated to the Special Mobile Battalion in ARMM. However, the integration has not
been without friction. AFP leaders strongly support the process in public state-
ments but are reportedly wary in practice. There have been incidents of MNLF
recruits resigning after complaints of discrimination and ethnic bias (May 2001).
Some existing AFP personnel have let it be known that they do not welcome the
former enemy and resent their exemption from particular requirements.
186 Mark Turner
V Implementing Phase Two of the 1996 Peace Agreement
A The Delayed Transition

The timetable for transition from phase one to phase two had always been im-
practical. Three years was too short a time to persuade people in SZOPAD that
their interests would be best served by joining the expanded ARMM. The prob-
lems encountered in implementing phase one simply worsened the prospects
for encouraging Mindanoans to vote for inclusion into the expanded ARMM of
phase two.
It was no surprise that the plebiscite for the expanded ARMM and new ARMM
elections were postponed. The government, the MNLF, the OIC and the popula-
tion of Mindanao knew that a plebiscite would result in little or no expansion. Ba-
sic developmental problems had not been satisfactorily addressed and ingrained
attitudes of mistrust had not been overcome. Also, when the specified time for
transition to phase two arrived, no bill had been passed in Congress to amend
the original ARMM Organic Act (RA 6734). Three readings in the lower house
followed by another three in the Senate presents a formidable challenge to the
proponents of any legislation. A tortuous and lengthy path is usual, especially for
contested legislation such as that for the ARMM. Even the government described
the democratic process of Congress as “tedious yet rightful” (GRP 2000).
On 15 July 1999, House Bill (HB) 7883 to amend RA 6734 eventually passed
all readings in the lower house and moved on to the Senate. By the end of May
2000, the government was confident to report that the country had reached the
“threshold in the implementation of the Peace Agreement” and to reiterate its
“sincerity and commitment to the full implementation of the Peace Accord” (GRP
2000). The legislative process would be completed in 2001, almost five years after
the signing of the peace agreement and three years behind schedule.
A stumbling block was that 2000 was such a bad year for the peace process.
After starting his administration by being inattentive to Mindanao, President
Estrada switched to a policy of all-out armed conflict with the MILF in central
Mindanao. In the southern islands of Sulu, he permitted protracted negotiations
with the Abu Sayyaf for the release of foreign hostages. For SZOPAD’s Christian
majority, these events confirmed their opposition to the peace agreement, par-
ticularly incorporation within the ARMM. The MILF and the Abu Sayyaf actually
wanted independent Islamic states.
Within the MNLF, all was not well for Nur Misuari, the long-time leader, Chair-
man of the SPCPD, and Governor of the ARMM. Many of his MNLF constituents
felt they had not received the anticipated dividends of peace. They now questioned
Misuari’s performance. He was known to have grandiose visions for Mindanao’s
development – floating airports, high-rise cities, and multi-lane highways. All
of these were far removed from the real, basic needs of the MNLF members.
There were also rumours of self-aggrandizement and accusations that he spent
too much time out of Mindanao, often overseas. He countered by claiming he
was there raising money for Mindanao’s development. But little if any finance ar-
rived as a result of Misuari’s efforts, even from Islamic sources. Misuari had also
lost his strongest ally in government, President Ramos. President Estrada was not
nearly as accommodating. By the end of 2000, the future was looking bleak for
Nur Misuari and the prospects of a strong MNLF.
In late 2000, Mindanao became even less of a policy priority. The nation fo-
cused its attention on sensational accusations of graft and corruption against
President Estrada. A Senate enquiry was broadcast on television and the na-
tion watched. There were popular protests and officials resigned. People power
erupted onto the streets once again and, in January 2001, Vice-President Gloria
Arroyo stepped up to assume the presidency with the backing of the military and
the church.
This change of government signalled a re-invigoration and re-think of Min-
danao policy. The belligerent Estrada approach towards the MILF was replaced
by the order to suspend military operations against the MILF. This would, said
Arroyo, create “the proper environment for the reconstruction of peace talks”
(Philippine Daily Inquirer 21 February 2001). President Arroyo’s attitude towards
the Abu Sayyaf was far less accommodating. She knew that the group was mar-
ginalized, with no support from leading OIC states such as Libya, Saudi Arabia,
and Indonesia, and was disowned by the MNLF and MILF hierarchies. Arroyo
could thus take strong action against the ‘bandits’ who made up the Abu Sayyaf.
Such action was precipitated by another spectacular Abu Sayyaf kidnapping. On
27 May 2001, twenty hostages were taken from a Palawan beach resort and trans-
ported to Basilan, the Abu Sayyaf heartland. The government eschewed negotia-
tions. Instead, a large military force was gathered to pursue the kidnappers and
their hostages. When this failed to secure the release of the hostages, President
Arroyo ordered the arrest and interrogation of known or suspected Abu Sayyaf
supporters on Basilan and in Zamboanga City.
All of this activity was cordoned off in Basilan. The main events for the peace
process were happening elsewhere, and happening rapidly. As soon as she had
assumed office, President Arroyo had approached Hashim Salamat, the MILF
leader, to determine his attitude to peace talks. The feedback was positive. Cru-
cial to this fresh climate of negotiation was a new actor to the peace process – the
government of Malaysia. President Arroyo had begun cultivating a good rela-
tionship with Malaysian Prime Minister Mahathir while she was vice-president.
Malaysian assistance was thus forthcoming in facilitating the early meetings with
the MILF, and Arroyo availed herself of such assistance knowing that Malaysia
was opposed to secession. Thus, in March 2001, the Philippine government met
with the MILF in Malaysia and agreed to:
• Continuous talks until a negotiated peace settlement is reached;
• Immediate formation of an MILF panel;
• Immediate cessation of offensives by the MILF;
• Commitment, by both parties, to respect and implement past agreements
subject to discussion of details of implementation by both sides;
• Holding the first formal meting at a mutually agreed venue within three
months of the signing of the agreement; and
188 Mark Turner
• Undertaking of relief and rehabilitation measures for evacuees and joint

development projects in areas affected by the conflict (Philippine Daily In-
quirer 14 April 2001).
The MNLF and the MILF were also talking. The top body of the MNLF, the Coun-
cil of Fifteen, formally removed Misuari from the MNLF chairmanship on 29
April 2001 on the grounds of “loss of trust and confidence” due to his poor per-
formance (Philippine Star 10 July 2001). In June, an MILF delegation accepted an
MNLF invitation to join the MNLF at the Twenty-eighth Session of the Islamic
Conference of Foreign Ministers held in Mali. The OIC’s Committee of Seven
for Mindanao had now grown to eight with the inclusion of Malaysia. Presiden-
tial peace adviser Eduardo Ermita expressed the view that the MNLF-MILF rap-
prochement “augurs well for the future of the MNLF-MILF unity talks and the
overall prospect [of the government peace talks with the MILF]” (Mable 2001:
19-20). Ermita saw the single representation of a unified MNLF-MILF as more
advantageous for talks and for securing the cooperation of more Muslim groups.
By contrast, the Philippines defence secretary interpreted these events as “a po-
tentially bigger threat.” (Mable 2001: 20). In early August 2001, after further talks
in Malaysia, it was announced that the MNLF and MILF had agreed to unite, al-
though no details of the agreement have ever been released and the two organiza-
tions continued to act independently (Philippine Daily Inquirer 3 August 2001).
The act of Congress required as the foundation for phase two of the 1996 peace
agreement was eventually passed in Congress in February 2001 and signed into
law by the president on 31 March 2001. This meant that President Arroyo could
now move on the long-delayed plebiscite that would usher in phase two of the
1996 peace agreement. The purpose of the plebiscite was to seek new member
territories for the expanded and final ARMM. At the urging of the MNLF and
MILF, the OIC ministers had issued a statement from their Mali meeting urging
the Philippine government to postpone the plebiscite until 2003. The govern-
ment declined to follow this route, saying the 15 August 2001 date for the plebi-
scite and the 26 November 2001 elections for positions in the ARMM were in RA
9054 and were therefore law.
President Arroyo asked the voters to “forge unity out of our diversity.” Nur
Misuari urged a boycott by Muslims, calling the plebiscite “an exercise in futility.”
Celso Lobregat, congressman for Zamboanga City, said his city “already enjoys
local autonomy through the local government code”, while the Muslim gover-
nor of Sultan Kudarat Province thought his people were better off without the
ARMM and said that joining it “would be like bashing our own heads with a rock”
(Philippine Daily Inquirer 13 August 2001). The voting passed off peacefully but
there was a low turnout of about 50%, well below normal electoral contests. The
result was a foregone conclusion – a resounding ‘No’. Of the eleven provinces and
fourteen cities of the SZOPAD that were not members of the existing ARMM,
only one province (Basilan) and one city (Marawi City) chose to join the expand-
ed ARMM. There were no close calls. In some cases more than 95% voted ‘No’.

32% of Mindanao’s population resided in the SZOPAD. But that territory no lon-
ger exists. Neither does the SPCPD, which supposedly oversaw development in
it. The new ARMM contains 2,745,00 people, only 15% of Mindanao’s population.
According to the 1996 peace agreement, this ARMM is the final act in sorting out
the institutional shape of the Autonomous Region of Muslim Mindanao that was
identified in the 1986 Constitution. But it is not a final solution to Mindanao’s
problems. It may not be any solution. The performance of the original ARMM
left much to be desired and there are severe misgivings about the performance
prospects of the latest version. It is an awkward construct satisfying a require-
ment of the Constitution in a way which will not be in the spirit of the Constitu-
tion.
The birth of the extended ARMM was marred by political violence and pro-
vided the least desirable environment for implementing phase two of the 1996
peace agreement. Nur Misuari, the incumbent governor, had no wish to surren-
der his position, but in the November 2001 elections for the ARMM he was not
a serious contender. Both the MNLF and government had dropped him. They
backed other MNLF leaders. Misuari and his supporters on his home island of
Sulu reacted to this rejection and tried to stop the election by attacking police
and military outposts. In the fighting, approximately fifty persons were killed and
Misuari sought sanctuary in neighbouring Malaysia where he was detained for il-
legal entry. The election went ahead as scheduled and Parouk Hussin, the MNLF-
government candidate, won the governorship by a landslide. A legal attempt to
block Hussin from taking up his new position was defeated but Misuari support-
ers in Zamboanga City engaged in further skirmishes with the military and at one
stage seized over 100 hostages in exchange for secure passage to another MNLF
camp (Philippine Daily Inquirer 7 January 2002). Twenty-five Misuari support-
ers were killed in that incident. Other engagements between the Misuari faithful
and government forces have been reported from other parts of the ARMM as has
collaboration with the Abu Sayyaf. Meanwhile, the Malaysian-sponsored peace
negotiations between the government and MILF came to a halt in March 2002
following clashes between government forces and the MILF. Plans for resuming
these negotiations were resuscitated only to be postponed due to new armed
clashes and bombings in 2003. In Basilan, the large government force pursuing
the Abu Sayyaf was joined by US advisers as part of America’s global anti-terror-
ist initiative. Misuari returned from Malaysia to face charges of rebellion but was
held in Luzon, a long way from his support base.
These were inauspicious circumstances for the inauguration of Parouk Hussin
as governor of ARMM and of the new regional members. Hussin vowed to steer
a different course from his predecessor. He stressed “accountability” and that
through his “seriousness, dedication and determination” he would develop the
poor region (Philippine Daily Inquirer 6 January 2002). He faces enormous chal-
lenges. He is taking over a government that has a reputation for inefficiency and
ineffectiveness. Misuari’s removal had been precipitated by his poor performance
190 Mark Turner
in bringing development to the ARMM. While there is little in the way of system-
atic study of earlier ARMM management practice, anecdotal evidence suggests
severe capacity problems. Organizational structures, processes, and personnel
require review and reform. Such basic management problems are not amenable
to quick fixes. They are long-term matters. This generates problems for leaders
who must answer to constituencies that want to see development happening and
to reap some of its benefits as quickly as possible.
Hussin may also face financial problems. At his inauguration, he told reporters
that the ARMM only had php 400,000 in it treasury but obligations of php mil-
lion. It is not yet clear whether the financial arrangements set out in RA 9054 will
provide adequate revenue for ARMM activities. The question of whether funds
actually match functions does not appear to have been fully investigated. Educa-
tion, the major expenditure, seems secure, as there will be an explicit central gov-
ernment transfer for this sector. However, RA 9054 identifies many activities in
which the regional government must be involved. It also lists a variety of organi-
zations that ‘may’ be established. An inherent danger is that these organizations
may be established but with inadequate operating expenditure so that their out-
puts and outcomes are negligible and ARMM income becomes devoted to paying
bureaucrats’ salaries rather than producing measurable development outcomes.
86% of regional government budget is used for salaries, leaving little for urgently
needed development investment.
One answer to Hussin’s prayers is foreign aid. According to President Arroyo,
there is usd 1.3 billion of committed official development assistance that the gov-
ernment will pour into Mindanao (Philippine Daily Inquirer 25 April 2003). A
significant amount will go to the ARMM. There are several potential dangers for
the ARMM. Firstly, governance becomes ‘projectized’ whereby the major flows
of money and other resources are found in externally funded projects. Much
political competition then focuses on how to access project funding. Secondly,
fears have been expressed within the donor and NGO community that hand-out
mentalities might have developed in relation to some ARMM constituencies and
with regard to certain project activities. This is problematic where sustainable
development is the prime objective. Thirdly, the record of aid disbursement to
date is characterized by slowness. Much needed projects and programmes are
held up, sometimes by endemic bureaucratic dysfunction and sometimes by the
turbulence and difficulty of the environment in which such activities are to take
place. Finally, Hussin will need to demonstrate and maintain his declared com-
mitment to accountability, as there is an unfortunate record of corruption in Phil-
ippine politics and administration. Aid funds are prime candidates for ‘leakage’
into private hands.
A further difficulty that the regional government faces is intergovernmental
relations. The ARMM has been inserted into a decentralized polity in which
many basic service delivery functions have been devolved to provinces, cities,
municipalities, and barangays. The question arises as to how far the regional gov-
ernment can or wishes to tell its component subnational territories what to do
in these fields. It has its own act on local level governance but lacks the capa-
bility to enforce it and the incentive to review it. Perhaps these territories have
far greater competence in such activities than the ARMM government. Is the
regional government an advisor or compliance seeker? If the component terri-
tory governments believe that there is too much regional interference, there may
be problems, especially if they feel that revenue is being wasted in unproductive
regional administration, or worse still revenue is being taken away from them. In
such circumstances, the component governments of the ARMM might see the
Department of Interior and Local Government (DILG) as a source of support. As
well as managing downwards in the territorial hierarchy, the regional government
must also manage upwards with national government. This is a neglected aspect
of decentralization in many countries. For the ARMM government, this involves
dealing with a range of central agencies, including the Office of the President,
the DILG, the Supreme Court, and many line agencies. Prime amongst the latter
will be the Department of Education, Culture and Sports, which shares various
educational functions with the regional government and monitors compliance
with national performance standards. For education and other sectors, the lead-
ing question is whether regional government has the organizational capacity for
maintaining and improving performance. If it does not, and this is a distinct pos-
sibility, then there is the challenge of managing relations with national agencies,
which should have the capacity but not necessarily the inclination to assist. The
challenge of managing upwards and downwards within the territorial hierarchy
is immense and carries a high risk. If the regional government is perceived by key
stakeholders as unsuccessful then the utility of the ARMM construct will come
under severe pressure.
Even more ominous is that peace with the MILF is yet to be negotiated. The
MILF claims to control a well-armed force of over 12,000 men (Philippine Daily
Inquirer 13 February 2003). Groups of the MILF regularly clash with the Philip-
pine military. In May 2003, the government cancelled its peace talks with the
MILF and announced that it was considering declaring the MILF a terrorist orga-
nization (Philippine Daily Inquirer 14 May 2003). This course of action followed
a series of bombings officially blamed on the MILF and several deadly clashes
between the MILF and the Philippine military. In all cases, it has been civilians
who have comprised the bulk of the casualties. Sporadic government-MNLF en-
counters are also reported. Abu Sayyaf gunmen still hang on in Basilan and Sulu
despite major operations by government forces and their American advisers to
destroy the rebels. Disaffected MNLF members are easy to find. Development is
still an unfulfilled promise for the majority of Mindanao’s inhabitants. The rec-
onciliation that was supposed to characterize the implementation of the peace
agreement has not happened. There have been great efforts by civil society and
religious leaders, both Christian and Muslim, to work for reconciliation. There
has also been an influx of foreign aid, though not nearly enough. The participa-
tion of Malaysia has also helped in seeking ways to reduce armed conflict. But the
‘Mindanao problem’ has not been resolved. Underlying conditions that give rise
to conflict still persist and are unlikely to diminish significantly in the short-term.
The government, the MNLF, the MILF, local communities, and civil society have
192 Mark Turner
a long way to go before a genuine peace is brought to Mindanao. Such a peace

does not simply mean a cessation of armed hostilities but also entails mutual
respect for culture, religion, and locality, the feeling of security in daily lives, the
expectation of decent services, an ecologically sound development, human dig-
nity, and the capacity to earn a decent living. When these things are achieved,
there will be peace in Mindanao.
Chapter 7
Power-sharing and International Intervention:
Overcoming the Post-conflict Legacy
in Bosnia and Herzegovina
Florian Bieber
I Introduction
Bosnia and Herzegovina, or in short Bosnia, has been the site of an international
experiment in reconstructing a state fractured by a three-and-a-half-year war.
This experiment has sought to re-establish a multiethnic system of governance
after the complete fragmentation of a state and the unsuccessful attempt of Serbs
and Croats to secede from Bosnia.
The efforts by the international community to reconstruct a war-ravaged soci-
ety in order to prevent a recurrence of the conflict have since been replicated in
Kosovo, East Timor, and Afghanistan, as well as in a number of cases with a lesser
degree of interventionism, such as in Macedonia. Most conflicts of the 1990s that
gave rise to this type of comprehensive international intervention were caused
by the mobilization of ethnic or national identity and had a self-determination
dispute at their heart. Unlike inter-state conflicts, resolving these intra-state
conflicts requires, in addition to classical peacekeeping efforts and economic
reconstruction, new political systems which would accommodate the divergent
demands over the country’s future. This has been particularly so, as long as the
international community has generally been loath to recognize claims to self-de-
termination, with the exception of secondary decolonization (East Timor) or the
dissolution of federal states on the basis of uti possidetis juris. The end of a war
does not necessarily imply an end to its underlying self-determination dispute.
Recent post-war peace agreements were thus faced with the challenge of cre-
ating a functioning polity that would satisfy the demands of all its constituent
nations. In the case of Bosnia, the post-war arrangement encompasses three key
components: (a) a long-term international military and civilian presence; (b) a
power-sharing arrangement between the different groups; and (c) a high degree
of decentralization towards ethnically homogeneous autonomies. This case study
seeks to assess the impact and success of the power-sharing arrangement in Bos-
nia, while also incorporating the other two components of the peace-settlement
into the evaluation. It will do so by examining the history of the conflict and the
194 Florian Bieber
background to the Dayton Peace Accords (DPA). It subsequently turns to the re-
sponses to the agreement and its record of implementation since 1996. Contrast-
ing the political structures established in the peace plan with the reality in post-
war Bosnia shows a very mixed picture. While the self-determination dispute in
Bosnia has been largely ‘frozen’ and the disintegration of the country no longer
an immediate threat, the state remains weak and internally fragmented. Ethnicity
permeates political decision-making and the economic and social developments
have offered no distraction from ethnicity as the most significant identifier.
II A Brief History of the War and the International Peace Initiatives

The war in Bosnia-Herzegovina, which came to an end with the Dayton Peace
Agreement (DPA) in late 1995, was the third and most destructive armed conflict
in the course of the dissolution of the Socialist Federative Republic of Yugoslavia.
Fighting first broke out in Slovenia in June–July 1991 (lasting for ten days), fol-
lowed by a more intense armed conflict in Croatia between July 1991 and January
1992. The war in Bosnia-Herzegovina began in April 1992, although there had
been violent clashes earlier that year.
The conflict in Bosnia progressively escalated into violence between the first
free elections in Bosnia in November 1990 and the outbreak of war in April 1992.
The origins of the war in Bosnia lie in the weakening of the communist and fed-
eral authority of Yugoslavia from the mid-1980s. In Serbia, this led to the rise of
Slobodan Milošević, who transformed the League of Communists of Serbia into
a populist nationalist party, while retaining some of its communist facade (Popov
1994: 22-84). The Bosnian League of Communists was considered to be more
dogmatic than its equivalent in most of the other republics, contributing to a loss
of legitimacy. In addition, a series of financial scandals (especially the so-called
Agrokomerc scandal in 1987) weakened the party even further. At the same time,
the Serbian intelligence services became active in Bosnia in attempts to either
promote a new pro-Serbian leadership of the republic or to strengthen Serbian
nationalist forces loyal to the new Serbian republican leadership (Friedman 1996:
192; Andjelić 2003: 51-78). In the election year 1990, several new parties were
founded in Bosnia, mostly, but not exclusively, along national lines. The first such
party to come into existence was the Party for Democratic Action (SDA, Stranka
demokratske akcije), followed by the Serbian Democratic Party (SDS, Srpska de-
mokratska stranka) and the Croat Democratic Community (HDZ, Hrvatska de-
mokratska zajednica). Although these parties campaigned on a national platform,
all three were united in their determination to defeat the League of Communists.
During the first elections in Bosnia, the three national parties together secured
approximately three quarters of the vote and formed a coalition to control parlia-
ment. The seven-member presidency of Bosnia was also exclusively controlled
For a more detailed discussion of governance in post-war Bosnia, see Bieber 2005.
Also known as the General Framework Agreement for Peace in Bosnia and Herze-
govina.
7 Power-sharing and International Intervention: Overcoming the Post-conflict Legacy in Bosnia and Herzegovina 195
by the three national parties (two Serbs from SDS, two Muslims from SDA, two
Croats from HDZ, and one for ‘Others’ from SDA) (Arnautović 1996: 103-107).
In the subsequent power-sharing agreement between the three parties following
the elections, the Muslim member of the presidency, Alija Izetbegović, became
the president of the presidency; a Croat was named prime minister, while a Serb
was elected as president of parliament. Both the League of Communists-Party
for Democratic Change (SK BiH-SDP, Savez Kommunista Bosnia i Herzegovine-
Stranka demokratskih promjena) and the multinational pro-Yugoslav Alliance
of Reform Forces (SRSJ BiH, Savez reformskih snaga Jugoslavije) played no sig-
nificant role and were excluded from office, except at the local level (e.g., Tuzla).
Not until November 2000 would the successor party to the SK BiH, the Social
Democratic Party (SDP, Socijaldemokratska partija), enjoy any electoral success.
The temporary coalition between HDZ, SDA and SDS was pragmatic but not
sustainable, as they represented conflicting concepts on the future of the republic
and its status within Yugoslavia.
The Serbian Democratic Party, with close links to both its sister party in Croa-
tia and the Socialist Party of Serbia (SPS, Socijalistička partija Srbije) led by Slo-
bodan Milošević, originally demanded a decentralization of Bosnia on the basis
of the three dominant ‘nations’, without openly proposing the dissolution of Bos-
nia. At the same time, it vehemently opposed the dissolution of Yugoslavia and
insisted on strong links between Serbia and Bosnia, be it within Yugoslavia or a
new state structure. In the course of 1991, the claim for territorial autonomy was
dropped and substituted by a demand for outright independence for undefined
parts of Bosnia, under Serbian control. Throughout 1991, the SDS established
‘autonomous areas’ in territory that as a result of the elections had come under
their control and thus effectively ceased to be governed by the central govern-
ment. The SDS argued that the right to self-determination, as formulated in the
preamble of the Yugoslav Constitution of 1974, applied to peoples, not just the
republics. In consequence, Muslims and Croats might seek to leave Yugoslavia,
but would not be permitted to pursue the independence of Bosnia as a whole,
but only of ‘their’ respective territories. While the party insisted that it would
separate from Bosnia at the same pace as the Bosnian government severed its ties
with Yugoslavia, the project of an autonomous/independent Serb territorial unit
within Bosnia was consistently ahead of Bosnian steps towards independence; as
such, the Republika Srpska (RS) was proclaimed four months before the Bosnian
declaration of independence (Cohen 1993: 141-143).
The Croat Democratic Community (HDZ) was divided into two wings with
different political agendas and positions on Bosnia. The moderate wing, which
dominated the party early on and was mostly represented by Croats from cen-
tral and northern Bosnia, supported the preservation of Bosnia as a multiethnic
country and sought to take it out of Yugoslavia in coalition with the SDA. The
more nationalist wing, mostly identified with its representatives from Herze-
The League of Communists changed its name shortly before the elections, but did
not undergo a fundamental programmatic change.
196 Florian Bieber
govina, also campaigned for an independent Bosnia, but simultaneously sought

territorial autonomy for Herzegovina, with the aim of joining this territory with
Croatia. This nationalist wing dominated the party from 1991–1992 onwards and
openly pursued secession of Western Herzegovina with the establishment of Her-
ceg-Bosna in 1993 (Burg and Shoup 1999: 65-66).
The dominant Muslim party, the SDA, sought to preserve Bosnia as a unified
state and to obtain independence from Yugoslavia. Independence, however, was
not the only policy option pursued: Bosnia and Macedonia together proposed an
asymmetric Yugoslav Federation in 1991, in an attempt to strike a compromise
between the Slovene, the Croat, and the Serb positions (“Die Plattform des Prä-
sidiumsvorsitzenden...” 1991: 22-24). The party wanted to preserve a centralized
republic and opposed any decentralization towards ethnic autonomy. While be-
fore the war some SDA representatives had sought a compromise with the Ser-
bian government, during the war some members of the nationalist/religious wing
supported the creation of a Muslim state in parts of Bosnia, thus advocating the
right to self-determination of parts of the country. However, proponents of such
a policy did not dominate the party or government policy, mostly due to the lim-
ited territorial space such a state would occupy (Friedman 1996: 212-213).
The early stages of the conflict in Bosnia included two separate, yet interlinked
self-determination disputes, as outlined above. The first dispute focused on the
structure of Yugoslavia, while the latter concerned Bosnia. In the dispute over
Yugoslavia, the HDZ and most Bosnian Croats supported independence for Bos-
nia from Yugoslavia, following the line taken by the HDZ-government in Croatia.
While the SDA also favoured independence, its stance was not as firm and it was
willing to compromise. Such a compromise on a looser federation was, however,
opposed by the Serbian government and its allies in Montenegro (and Vojvodina
and Kosovo) and by the SDS. The SDS finally rejected any move by the Bosnian
government towards independence. It threatened to follow every step of the Bos-
nian government to seek independence with the equivalent step within Bosnia
for the Serb-controlled territories.
The concept of Bosnia as a unified state was only supported by the SDA and
the (marginalized) multinational parties. Since its foundation, the SDS had de-
manded territorial autonomy, but after the elections it went even further and
sought outright self-determination of ‘Serb territories’. With the outbreak of war,
the HDZ also started to campaign for self-determination for Herceg-Bosna. As
neither the HDZ nor the SDS had campaigned for the 1991 elections on a self-
determination platform, the degree of popular support among their constituency
for secession from Bosnia is unclear. While the political polarization in Bosnia
and the war in Croatia strengthened support for Croat and Serb secessionist par-
ties and policies, a significant number of Serbs and Croats continued to support
a unified Bosnian state, as was evidenced in 1991–1992 by numerous peace rallies,
opinion polls, and the departure/marginalization of leading HDZ and SDS politi-
cians (Wieland 2000: 261-263).
In October 1991, however, a parliamentary debate sharply revealed the weak-
ness of the Bosnian state institutions. While Croat and Muslim deputies voted
for the sovereignty of Bosnia, Serb members of parliament boycotted the session
(Reuter 1992: 672). In response, the SDS organized a referendum on 9–10 No-
vember 1991 among the Serb population in areas under its control, in which a ma-
jority voted in favour of remaining within Yugoslavia. On 9 January 1992, the SDS
proclaimed the Republika Srpska i Bosna-Hercegovina (which since August 1992
has been called Republika Srpska) (Buha 1999: 8-9). The statelet was included in
plans for a reconstructed ‘Yugoslavia’ that were circulated in Serbia at the time.
This new Yugoslavia would have been composed of Serbia, Montenegro, and the
two Serb para-states in Croatia and Bosnia (Schwarm 1994: 3)
The connection between the Bosnian self-determination claims within Yugo-
slavia and the self-determination disputes within Bosnia were highlighted by a
referendum on the independence of Bosnia. The referendum was held 29 Febru-
ary–1 March 1992 at the recommendation of the European Community (EC),
which followed the conclusions of the Badinter Arbitration Committee. The
SDS immediately launched a boycott of the referendum and most Bosnian Serbs
refrained from participating. A majority of 62.68% of voters favoured indepen-
dence. However, as one community had largely boycotted the referendum, it
lacked the cross-communal support necessary for a far-reaching change to the
status of the country, such as independence. While it fulfilled the criteria laid
out by the Badinter Committee, it lacked the necessary legitimacy within Bosnia
(Bougarel 1996: 56; Burg and Shoup 1999: 117).
The war began in April 1992, after Bosnia received international recognition
as an independent state, when military and paramilitary forces under the control
of Republika Srpska (RS), with strong logistical support from the Yugoslav army,
began conquering large parts of the country and besieging major cities, includ-
ing Sarajevo. In a well-documented campaign of ‘ethnic cleansing’, which was
marked by territorial conquest, mass expulsion, and the brutal murder of people
from differing ethnic backgrounds, Serb forces brought 70% of Bosnian territory
under their control within the first few months. (Cigar 1996: 47-61). Throughout
the war, RS, which was ruled by the Bosnian Serb army, declared its aim to join
with Serbia and other ‘Serb’ states, such as Montenegro and the Serb Republic in
Krajina (RSK, Republika Srpska Krajina) to form a Union of Serb States. This aim
was not formally carried out, as Serbia and the newly-created Federal Republic
of Yugoslavia sought to retain some international support by playing down the
strong links between the RS and the FRY. At the same time, the ideological back-
ground of the SDS and the SPS differed considerably, which eventually led to a
split in 1994–1995. The Serb para-states in Croatia and Bosnia sought to cooper-
The Badinter Committee was an ad hoc body appointed by the EC to evaluate the
self-determination claims of some of the republics of Yugoslavia. It was chaired by
the president of the French Constitutional Court, Robert Badinter.
The SDS strongly adhered to the Četnici tradition of Serbian nationalism, including
a staunch anti-communist position. In addition to the creation of ‘Greater Serbia’, it
advocated a prominent role for the Serbian Orthodox Church. The SPS, on the other
hand, never gave full support to the idea of ‘Greater Serbia’. Support for Serbs outside
198 Florian Bieber
ate, and at several times during the conflict proclaimed a union. This cooperation
was always limited, as was revealed when the RS did not assist the Croatian Serbs
during the Croatian army offensives of May and August 1995. This action also
highlighted the limited support from Serbia, which did not intervene during the
Croatian offensive. Both economically and militarily, the Bosnian RS was heavily
dependent on Yugoslavia and thus very limited in its political choices, as was evi-
denced during the peace negotiations leading to the Dayton Peace Agreement.
The Croatian intention to take Herceg-Bosna out of Bosnia became obvious
during the first months of the war, although both secret and open negotiations to
this effect had taken place between Franjo Tudjman and Slobodan Milošević, and
between the President of the RS Radovan Karadžić and Mate Boban, the leader
of the HDZ in Bosnia, before and during the early phase of the war (Hedl 2000;
Burg and Shoup 1999: 198). The conflict between the Bosnian army and the Croat
military units intensified in late 1992, leading to an outright war, which lasted
for a year from spring 1993 to spring 1994. During this war, the HDZ established
Herceg-Bosna in Western Herzegovina and openly sought to incorporate the ter-
ritory into Croatia. The territory had already been linked closely to Croatia in
military, economic, and political terms and remained so even after the end of the
Muslim-Croat conflict.
Croat and Serb territorial claims taken together amounted to almost the entire
Bosnian territory. Serb autonomous regions laid claim to 63% of the territory,
while the Croat community claimed 21%. Only 13% were disputed between the
two, with little more than 3% claimed by neither the SDS or HDZ (Begić 1997:
68–69).
International involvement in the conflict had begun before the actual outbreak
of war. However, both the international community and the European Commu-
nity (EC), which was particularly active during the early stages of the war in the
former Yugoslavia, were taken by surprise, which led to a number of well-identi-
fied policy failures. Early diplomatic efforts aimed at the preservation of Yugosla-
via’s territorial integrity and tried to broker a new constitutional framework for
the country after Slovenia and Croatia had left the federation. By the time war
broke out in Bosnia in April 1992, the preservation of Yugoslavia was no longer
a policy option pursued by the United Nations or the EC. Instead, international
efforts started to focus on the preservation of Bosnia. In a statement that marked
a fundamental change of mind, the committee set up by the EC and headed by
the French judge Robert Badinter came to the conclusion that Yugoslavia was
in the process of disintegration and thus any of its constituent republics could
seek international recognition, provided they fulfilled certain criteria, such as
adequate minority protection or human rights standards (Cohen 1993: 236-237;
Pellet 1992). A referendum was a further requirement for recognition that the
Badinter committee added, although it was not explicitly stated in the EC’s ‘Dec-
of Serbia declined among the Serbian leadership as a result of sanctions and the eco-
nomic decline in Serbia during 1993–1994. In addition, the SPS always maintained its
communist heritage, which inherently brought it into conflict with the SDS.
laration on the “Guidelines on the Recognition of New States in Eastern Europe

and in the Soviet Union” and “Declaration on Yugoslavia” from 16 December 1991
(“The Opinions of the Badinter…” 1992).
• The first concerted international mediation effort was undertaken by the EC
under the Portuguese presidency in February–March 1992. This first peace
plan, named after the EC mediator Jose Cutileiro, set the tone for all sub-
sequent international attempts at resolving the conflict: It insisted on the
invariability of the Bosnian borders and on finding a solution within Bos-
nia. At the same time, it considered territorial autonomies within Bosnia
to accommodate claims by the three parties. This view was shared by the
international community, which in addition to the negotiators from the EC/
EU, included the UN, and later also included the contact group (Russia, the
United States, France, Germany, Italy, and Great Britain). Within this frame-
work, the main developments in the peace plans from the Cutileiro plan to
the Dayton Peace Accords can be mapped along three parameters: size and
contiguity of the ethnic units;
• power of the centre; and
• shifting borders of the territorial units depending on the balance of power
on the ground.
All the internationally negotiated plans had at some point appeared close to be-
ing signed by the parties to the conflict. It can be assumed, however, that neither
the consent of the parties nor their participation in the negotiations were alto-
gether in good faith, especially that of the Bosnian Serbs. Consequently, parties
agreed to proposals in the knowledge that other parties would reject them, and
that they could thus gain kudos for supporting a peace agreement that would
never be implemented. In addition, many peace agreements were deliberately
misinterpreted at the local level. Despite the ‘ethnic’ definition of the territorial
autonomies applied in the different peace plans, the early proposals at least in-
sisted on some degree of political representation for the non-dominant nation in
each territory. At the local level, however, this provision of any agreement was
systematically ignored. Furthermore, internationally proposed maps helped at
times to justify the conquest of territories as part of a ‘pre-emptive implementa-
tion’. This was the case for the Vance-Owen Plan and for some Croat conquests in
1993 at the expense of Muslims.
The Cutileiro plan proposed the creation of a varying number of non-contigu-
ous national regions. Although the parties initially signalled their support for the
agreement, disputes arose over the size of these cantons. Certain crucial policy
areas such as the control of the police and the army were not even addressed in
this plan. As no consensus over the distribution of territory could be reached,
the agreement collapsed. Less than a month later war broke out (Burg and Shoup
1999: 108-117, Begić 1997: 83-108). The EC negotiator Lord Owen and UN envoy
Cyrus Vance drew up the next peace plan during the first months of 1993. The
plan strengthened the cantons, by creating fewer and larger cantons vested with
greater political powers. The composition of the cantons was based on ethnic-
200 Florian Bieber
ity, but there was a notable difference to the previous plan: ethnic composition
was no longer based on census figures but increasingly on lines of control. This
meant the de facto acceptance of the results of ethnic cleansing. This implicit in-
ternational recognition of territorial conquest and ethnic cleansing endorsed by
the Vance-Owen plan gave even further impetus to ethnic cleansing. Especially
in parts of Herzegovina, Croat armed forces felt encouraged to conquer terri-
tory that according to the Vance-Owen plan would have formed part of a Croat
canton. The Bosnian Serb ‘parliament’ rejected the amended Vance-Owen plan,
after Milošević had forced Radovan Karadžić to sign the agreement. While the
HDZ supported the agreement, the SDA and the central government only lent
it wavering support, as they felt it offered the Muslims too little (only 26% of
the territory would have been under Muslim control, as the canton of Sarajevo
would have been under international administration). Further, the plan reduced
the powers of the central government to a minimum. Another controversial as-
pect of the agreement was that it sought to prevent the secession of the cantons,
by denying the western Serb canton a link to Eastern Bosnia and by refusing to
create a contiguous Serb canton in eastern Bosnia that could have easily joined
Serbia. The three predominant Croat cantons, however, were all bordering Croa-
tia. The cantons were unable to establish bilateral ties with Croatia or Serbia, as
only Bosnia as a state was vested with such powers. Although the Vance-Owen
plan never received full US endorsement, which was partly due to a change in
administration during the early stages of plan, the main reason for its eventual
failure was that it was the last internationally proposed plan that did not explic-
itly endorse the results of ethnic cleansing (Burg and Shoup 1999: 189-262, Begić
1997: 109-125).
Both the subsequent Owen-Stoltenberg plan, proposed in August 1993, and
the Contact Group plan, put forward in July 1994, essentially abandoned the idea
of creating small, non-contiguous cantons in favour of three large ethnic territo-
ries, with little attempt to ensure some degree of minority representation within.
While the territories allocated to the different parties by international mediators
during the negotiations changed between 1993 and 1995, the borders principally
reflected the territorial control by the three parties, with the Serb party being
forced to return some of the territory under its control. Starting with the Contact
Group plan, the proposed settlements foresaw a division of Bosnia between the
Serb-controlled parts and the Croat-Bosniak Federation, according to the for-
mula of 49 to 51% of the territory of Bosnia.
During this phase of the conflict, the second entity of current Bosnia was es-
tablished under US auspices: the Federation of Bosnia and Herzegovina, which
included Croat and Muslim (now Bosniak) territories. The agreement for this
Federation ended the Croat-Bosniak war and made the Bosnian Croat and Croa-
In this study, ‘Federation’ denotes the Federation of Bosnia-Herzegovina, which con-

stitutes 51% of Bosnia and includes the war-time Croat and Bosniak territories. Al-
though the state of Bosnia-Herzegovina can be described as a federation, this term
will not be used to describe the entire Bosnian state to avoid confusion.
tian leadership formally renounce claims to self-determination of Herceg-Bosna

(Burg and Shoup 1999: 292-298). The implementation of the agreement was not
achieved even after the end of the war, as many vestiges of the independence of
Herceg-Bosna remained intact until 2000. As a result of the Washington agree-
ment, which established the Federation, the Bosnian Serb side came under in-
creased pressure, as all subsequent plans provided for the division of Bosnia into
two parts, the Republika Srpska (RS) and the Federation, with both territorial
units defining themselves as territories of their respective national group. The es-
tablishment of the Federation also helped turn the tide in the conflict, as Bosniak
and Croat forces in the summer of 1995 launched a joint offensive that conquered
large parts of the Serb Republic. Against this backdrop, the RS agreed to a cease-
fire and to beginning of peace talks leading to the Dayton Peace Accord.
III The Origins of the Peace Agreement

The Dayton Peace Accords (DPA), initialled on 21 November 1995 in Dayton,
Ohio and signed in Paris on 14 December 1995, eventually brought the three-
and-half-year conflict to an end. The war had left over 200,000 Bosnians dead or
missing, more than half the population displaced, either within Bosnia or in third
countries, had destroyed most of the country’s infrastructures and left much of
Bosnia’s cities and countryside scorched. The peace agreement was based both
on the previous efforts by the international community to end the war and on
four significant events in Bosnia and Croatia during the summer of 1995:
1. The fall of the two UN Safe Havens, Žepa and Srebrenica, and the subse-
quent mass murder of most of the male citizens of Srebrenica by the Bos-
nian Serb army (VRS, Vojska Republike Srpske), with the active support of
the Yugoslav Army (VJ, Vojska Jugoslavije), strengthened the resolve of the
international community to end the war even against Serb resistance. It also
paved the way for NATO bombardments of Serb positions, which signalled
a shift in international policy towards active intervention in Bosnia.
2. The conquest of western Slavonia in May 1995 and the Krajina in August
1995 by the Croat army and the collapse of the RSK weakened the Bosnian
Serb position and led to the rapid military advance of Croat forces in Bosnia
in cooperation with Bosniak forces. The territorial division on the ground in
Bosnia thus approached the 51:49 formula established in the peace plans of
the previous year.
3. Slobodan Milošević, determined to end the war to end the isolation of Yu-
goslavia, applied pressure on the Bosnian Serb leadership to accept a peace
plan negotiated by him on their behalf. This policy shift of Yugoslavia pro-
gressively took place in 1994 and 1995 and was the consequence of the effects
of sanctions and the rift between Milošević and the Bosnian Serb leadership
over war aims.
4. After three and a half years of war, all parties, especially the Bosnian Serb
side, suffered from low morale and the continuation of armed conflict was
202 Florian Bieber
more difficult to sustain, especially in terms of securing enough manpower

to maintain the long front lines across Bosnia.
This mixture of a change of fortune between the warring parties, the stronger
military Western (i.e., US) commitment to ending the war and overall war-wea-
riness enabled the conclusion of a peace agreement, which ended the war and
provided for a complex system of post-war governance and reconstruction of
Bosnia.
Despite the exhaustion of the three parties and the reversal of fortunes in the
war, the peace agreement was not based on the defeat of any of the three parties,
nor did any of the three communities perceive the agreement as the only alterna-
tive to a defeat. As a peace between three ‘undefeated losers’ of the conflict, the
agreement solicited little support from any of the three parties during its imple-
mentation process, unlike some of the other agreements in former Yugoslavia
(e.g., the Erdut Agreement on Eastern Slavonia concluded simultaneously with
the Dayton Peace Accords, where Croatian Serbs had clearly lost any leverage to
block the agreement’s implementation).
Although an agreement between the RS, the Croat territories, and the pre-
dominantly Bosniak central government, the accord was negotiated between
three countries, Bosnia, Croatia, and Yugoslavia/Serbia. The strategy of negotiat-
ing with Croatia on behalf of Bosnian Croats and Yugoslavia/Serbia for the Bos-
nian Serbs was based on the pragmatic assessment that the main power-holders,
Tudjman and Milošević, were more likely to assure the implementation of the
agreement than their local proxies. Secondly, local political actors had become
unacceptable partners to the international negotiators due to their intransigence
and direct implication in war crimes, especially after the indictment of Radovan
Karadžić, the president of the RS, and Ratko Mladić, the chief of staff of the Bos-
nian Serb army, by the International Criminal Tribunal for the Former Yugoslavia
in The Hague (Holbrooke 1998: 105-108). Thirdly, the Bosnian Serb leadership
had repeatedly withdrawn initial support from peace agreements by submitting
the proposed agreement to the parliament of the RS, which rejected it, or by
initiating a ‘referendum’ in the areas under its control. Such a delay and obstruc-
tion was this time excluded by confining representation of the Bosnian Serbs in
the negotiations to Slobodan Milošević as president of Serbia. Subsequently, the
parliaments of Croatia, Bosnia, and Yugoslavia approved the agreement within
the three weeks between the initialling and signing of the agreement. The ab-
sence of democratic scrutiny of the agreement reflected both the degree of au-
thoritarianism in the three countries and Western scepticism towards an abuse
of ‘democratic’ procedure for the benefit of an uncompromising political course.
This scepticism was based on the usage the Bosnian Serb leadership had made
of the parliament of the RS to veto earlier peace plans and thus circumvent in-
ternational pressure (Schneider 1996: 1-2). The reliance of the peace agreement
and its implementation on the authoritarian leaders of Croatia and Serbia/Yugo-
slavia limited the agreement’s initial implementation, as both neighbouring gov-
ernments sought to continue their intervention in Bosnia’s affairs and supported

nationalist political elites in the country.
IV The Bosnian Power-sharing Arrangement

The DPA has produced a highly complex power-sharing arrangement, both in
terms of the role of the international community and the different layers of gov-
ernance and their competence. It is important to note that the DPA governs a
country that lacks some key attributes of sovereignty. As the peace accords both
recognize existing political institutions and institute additional international su-
pervision, it has been difficult to adequately characterize the degree of sover-
eignty of the country. Security aspects are largely under the control of the inter-
national military forces in the country (SFOR or Stabilisation Force, later IFOR or
Implementation Force), while the civilian administration is supervised and partly
conducted by a host of international organizations. Additionally, a number of
domestic institutions (i.e., the Central Bank, Constitutional Court) have foreign
members, making them only partially domestic institutions (Nowak 1999: 287).
As a result, it is important to bear in mind that, when considering power-sharing
in Bosnia, a considerable degree of power is vested with international institutions
and individuals and thus cannot be ‘shared’ among domestic political actors.
The system of government set up by the Bosnian peace arrangement has many
attributes of the system of power-sharing described by Arend Lijphart in his
theory of consociationalism, including the participation of all major groups in
government, veto rights for the communities, a high degree of segmental auton-
omy for the nations and proportional representation in the state administration
(Bieber 1999b; Lijphart 1984: 23-32). At the same time, the Bosnian consociational
arrangement rests on weak foundations, given the absence of most of the criteria
that are generally considered conducive to such a system. Of the structural and
actor-oriented conditions identified by Ulrich Schneckener, Bosnia only fulfills
three, considerably less than either Northern Ireland, South Tyrol, or Belgium, all
classic examples of consociational arrangements (Schneckener 2000: 10-18). Un-
like conventional approaches to power-sharing, the arrangement in Bosnia em-
phasizes autonomy over power-sharing at the centre. As will be described, very
little power has been vested in the central institutions and the most contentious
powers (defence, education) have been decentralized to such an extent that deci-
sions are essentially taken without cross-communal consultation. Some elements
of power-sharing can, however, be found at the level of the entities, and even with
the cantons and communes in the Federation. The Bosnian system of governance
differs most substantially from classical approaches to power-sharing in its inter-
nal logic. In most other power-sharing cases, the autonomy granted to communi-
ties in the form of decentralization and/or cultural autonomy is a function of the
power-sharing arrangement at the centre. In Bosnia, the opposite has been the
case. Decentralization, resulting from the war, was a given, whereas some degree
of central coordination and cooperation had to be instituted to maintain at least
the resemblance of a country rather then two or three separate statelets. These
204 Florian Bieber
origins of the system of government have dictated a power-sharing at the centre

that has been marked by the absence of power – as it is vested with the entities or
cantons – and an absence of ‘sharing’ – as the arrangement in Bosnia was less the
result of a domestic compromise, but rather was imposed by both international
actors and the kin-states of the two secessionist communities. Whereas some
scholars of power-sharing have generally identified decentralization and ethnic
federalism as one promising aspect of power-sharing (e.g., Sisk 1996: 49-53), the
Bosnian experiences suggest that excessive decentralization on the basis of eth-
nicity reduces the willingness of political actors to engage in power-sharing and
increases the fragmentation of governance.
In addition to these factors, which reduce the strength of power-sharing mech-
anisms, some of the inherent assumptions of power-sharing in Bosnia are prob-
lematic. By definition, consociationalism rests on the concept of elite cooperation,
by relying on the willingness of political elites to work together in administering
the country. The elite-driven component of Lijphart’s theory is problematized in
Bosnia, where political elites, especially of the three nationalist parties, have been
responsible for the war (Bose 2002: 246-252): In fact, it could be argued that the
brief pre-war interlude of ‘democracy’ in Bosnia from 1990 to 1992 constituted
a power-sharing arrangement that failed (Bieber 1999a: 203-211). The reliance
on cooperative elites led to a conceptual shift in the international community’s
policy towards Bosnia. While the formal system of government can be best de-
scribed as a variation of the Lijphartian model of consociational power-sharing,
the approach of the international community in recent years is more akin to the
more integrative approach of Donald Horowitz, however without altering the
institutional set-up (Horowitz 1985). The two key aspects of this approach have
been the changes and developments in the elections laws and regulations, and
the active support given to moderate politicians. While a number of different
electoral systems were used for the elections in post-war Bosnia, the tendency
has been to promote the development of cross-ethnic votes so as to promote can-
didates who adopt a more moderate political stance (Bose 2002: 215-41). While
the outcome of the elections has not been able to demonstrate the immediate
impact of the different electoral systems used on the overall result, it has been ar-
gued that the impact of electoral systems on the positions of elites is incremental
(Reilly 2001: 167-93). The informal support for moderate politicians through the
international actors has largely failed, and in a number of cases back-fired. The
measures which resemble – albeit through international intervention – aspects
of integrative power-sharing had only limited impact in light of the unaltered
In addition to elitism, they include bureaucratization, segregation, and blocked deci-

sion-making (Bieber 1999b: 88-90).
For a list of Horowitz’s key recommendations for moderation, see Horowitz 1985:
646-652.
Support included post-election coalition-building through international agencies,
implicit endorsement of candidates by international agencies, public awareness cam-
paigns advocating ‘vote for change’, and political pressure against nationalist parties.
institutional arrangement, which grants considerable power to nationalist elites

and provides little incentive for cooperation.
Returning to the institutional aspects of the Bosnian power-sharing arrange-
ment, two mechanisms of governance can be identified:
1. The central institutions of Bosnia, as devised in the Constitution of Bos-
nia-Herzegovina (Annex 4 of the DPA), are established as a consociational
democracy; with the inclusive governance of all major national communities
through broad representative governments and a system of veto powers.
2. The country is decentralized to two entities, which primarily define them-
selves as territorial autonomies of the respective national groups. This eth-
nic federalism grants high autonomy to the three major national groups in
the conduct of their affairs.
While these two levels encapsulate the mechanisms of multinational governance

used in Bosnia, they fail to reflect the complexity of the Bosnian institutional
system. In the following, we shall discuss the five relevant levels of government
in Bosnia:
1. International Administration;
2. Central Government;
3. Entities;
4. Cantons (only in the Federation);
5. Communes.
In analyzing the different levels, it is important to highlight the role of the DPA,
as well as other laws and regulations governing these. Furthermore, each lev-
el should be evaluated as to whether it is primarily a tool of power-sharing or
whether it serves largely as a safeguard for the autonomy of the national groups.
Determining the function of the different levels in regard to autonomy or power-
sharing will allow us to evaluate its success and failures.
The power-sharing structure in Bosnia is the result of a compromise between
representatives of two of the three nations – Serbs and Croats – which sought to
secede from Bosnia and representatives of the largest community – Bosnia’s Bos-
niaks – that sought to preserve the state. Power-sharing is thus not based on the
willingness of the national groups, or rather their political representatives, to de-
volve power in the interest of subsidiarity or increasing the segmental autonomy
of the national communities, but constitutes the lowest common denominator
on which the three communities agreed to re-establish Bosnia after the end of
the war. While Dayton reaffirmed the legal continuity of the ‘Republic of Bosnia-
Herzegovina’ as it was recognized in April 1992, in reality the Bosnian state had
ceased to exist during the war years 1992–1995. The constitutional arrangement of
the DPA thus established entirely new political institutions and provided little in-
stitutional continuity to the previous central government. The war-time Bosnian
government and institutions legally transformed themselves into the post-war
Dayton state institutions. De facto, however, the Bosnian government institutions
and personnel were largely integrated into the Federation of Bosnia-Herzegovina
206 Florian Bieber
formed in 1994, as the Bosnian government’s reach was limited to the Bosniak
majority areas at the time they were incorporated into the Federation.10
A The Joint Institutions

The central institutions of Bosnia are vested with extraordinarily few powers,
while most competencies rest with the two entities11 or their sub-structures. Ar-
ticle III of the Bosnian constitution specifically lists ten areas of competence:
• Foreign policy;
• Foreign trade policy;
• Customs policy;
• Monetary policy;
• Finances of the institutions and for the international obligations of Bosnia;
• Immigration, refugee, and asylum policy and regulation;
• International and inter-entity criminal law enforcement, including relations
with Interpol;
• Establishment and operation of common and international communications
facilities;
• Regulation of inter-entity transportation; and
• Air traffic control.
These fields can be described as competencies in the international relations of

the country and basic control over communication and borders. In addition to
these powers, the highest human rights protection mechanisms are located at the
level of the joint institutions with the Human Rights Chamber. Additional com-
petencies of the joint state can be recognized, if one includes Annexes 5 through
8 of the DPA (dealing with Arbitration, Human Rights, Refugees and Displaced
People, and National Monument Preservation, respectively). Furthermore, the
entities are free to delegate additional competencies to the common institutions,
while the central authorities can establish additional institutions as follows:
Bosnia and Herzegovina shall assume responsibility for such other matters as are agreed
by the Entities; are provided for in Annexes 5 through 8 to the General Framework
Agreement [DPA]; or are necessary to preserve the sovereignty, territorial integrity,
political independence, and international personality of Bosnia and Herzegovina, in
accordance with the division of responsibilities between the institutions of Bosnia and
Herzegovina. Additional institutions may be established as necessary to carry out such
responsibilities. (Constitution Art III (5)(a))
10 As the Federation came into existence more than a year prior to the joint state estab-
lished at Dayton and as the Federation agreement was concluded between the Croat
leadership of ‘Herceg-Bosna’ and the Bosnian government, the Federation agreement
constituted a de facto recognition of the fact that the Bosnian government no longer
represented all of Bosnia, but rather the Bosniak-controlled parts of the country.
11 The District of Brčko exists in addition to the two entities and is a case sui generis.
While this broader interpretation has some support among ‘Bosnia’s’ legal schol-
ars, the Bosnian political elites, especially Croats and Serbs, have generally opt-
ed for the minimalist interpretation of the common institution’s competencies
(Ibrahimagić 1999: 27-28). The possibility to extend the competencies of the joint
institutions has been used increasingly frequently, however, by the international
community, starting with the establishment of the State Border Service in 2000,
followed by numerous other new state institutions and ministries. Unlike other
federal states, the central state institutions in Bosnia originally had no control
over the two entity-based armies, but, by 2005, the two armies had merged and
were placed under control of a state-level ministry of defence whereas the entity
ministries were dissolved. Although the units of the army continue to follow eth-
nic lines, the establishment of a Bosnian army constituted a major step towards
the development of a fully fledged state government.12
The powers of the central authorities cannot only be enhanced by the enti-
ties or a re-interpretation of the mandate of the institutions, but also through
parliamentary procedure. Article X of the Constitution allows for constitutional
changes with a two-thirds majority in the House of Representatives and a simple
majority in the House of Peoples.
It is important to note that during most of the post-Dayton period the powers
of the joint state have been limited, but that the state has also lacked enforcement
mechanisms towards the entities, for instance, as Manfred Nowak observed in
regard to human rights protection: “[T]he State of BiH and its so-called com-
mon institutions … lack the power to enforce international obligations upon the
two entities and their substructures” (Nowak 1999: 287). The weakness of the cen-
tral institutions extends to their financing, as the budget is (nearly) exclusively
financed by the entities. Reforms in 2004/2005 aimed at introducing a Value Add-
ed Tax have since provided the state institutions with an independent source of
income. Despite these inherent weaknesses of the central state institutions, some
agencies, especially those established originally as quasi-international institu-
tions, such as the Independent Media Commission, later the Communications
Regulatory Agency, have performed relatively well and since 2000 considerably
added to the weight of the state.
As previously mentioned, the joint institutions of Bosnia have been established
by the DPA, despite some similarities to pre-war institutions. The main institu-
tions at the state-level are:
• The Presidency;
• The Council of Ministers;
• The Parliamentary Assembly, consisting of the House of Peoples and the
House of Representatives;
• The Constitutional Court;
• The Human Rights Chamber; and
• The Central Bank of Bosnia
12 As the case study was being completed, negotiations over establishing a joint com-
mand of the armies under state-level control were ongoing.
208 Florian Bieber
In addition to these institutions, a number of smaller institutions and agencies

have emerged, reflecting a general trend towards strengthening the weak state-
level.
1 Presidency
Bosnia could be described as a semi-presidential system, whereby there is not one
president, but three. The institution of the presidency is not common in political
systems around the world, especially among other multinational states. It could
be argued that the presidential system is generally not conducive to governance
in divided societies. Conventional presidential or semi-presidential systems re-
duce the effectiveness of grand coalitions, one of the characteristics of power-
sharing arrangements, as considerable executive power is concentrated in one
hand (Ljiphart 1977: 34). While this challenge is addressed somewhat through the
three-member presidency, it reduces the ability of a president to act as an arbiter
in conflict. This ability – as had been the case under Tito in Yugoslavia – is limited
in Bosnia due to the high national polarization and the limited number of politi-
cal actors who would be acceptable to all communities. The main problem with
the presidential system in diverse societies is the limitation on coalition-building,
as the members of the presidency are directly elected.
The system of a multi-member presidency is a distinct Yugoslav heritage. Yugo-
slavia, according to its 1974 Constitution, had an eight-member presidency, rep-
resenting all six republics and the two autonomous provinces equally, with the
chairmanship rotating on an annual basis. Most republics, including Bosnia, had
a multi-member presidency. Between 1990 and 1996, the presidency had seven
members: two from each of the three nations and one for ‘others’, i.e., minorities
and ‘Yugoslavs’. The Dayton presidency includes only three members, one Bosniak,
one Serb, and one Croat, and additionally incorporates an element of territoriality
by requiring that the Serb member be elected in the RS, while the Bosniak and
Croat members are elected in the Federation. As a consequence, the presidency
contains one representative from each nation, and two from one and one from
the other entity. This twin-definition of membership means that other minorities,
non-ethnically defined citizens (mentioned in the preamble of the Constitution),
as well as members of the respective non-dominant national community (Serbs in
the Federation, Bosniaks and Croats in the RS) are precluded from being elected
to the presidency. Most past members of the presidency acted on behalf of ‘their’
ethnic community in ‘their’ entity, thus de facto depriving the above mentioned
groups not only of running for the presidency, but also of being represented by it.13
It also reveals a key feature of the Bosnian system of power-sharing: all national
representation and autonomy is territorial. Even if a particular position or prerog-
13 During the discussions over a permanent election law in the summer of 2001, some of
the governing non-nationalists advocated abolishing this dual constriction on presi-
dency membership. As such a change would have required a constitutional change,
which could not have secured a majority in the parliamentary assembly, the proposal
was abandoned.
ative is defined in national terms, it is supplemented with a territorial dimension.

As a consequence, non-territorial or cultural autonomy is not part of the current
Bosnian system of power-sharing. This is surprising, considering the traditions of
non-territorial autonomy, such as the Millet system under the Ottoman Empire,
which granted autonomy to the religious communities of the Empire, as well as
similar arrangements under Austro-Hungarian rule (Hanf 1991).14
The presidency is chaired in annual rotation by one of the three members. The
chairmanship of the presidency does not, however, carry any additional powers.
In terms of competences, the presidency is primarily charged with conducting
the foreign policy of the country, one of the key powers vested in the joint in-
stitutions. Nevertheless, the presidency has been generally a weak institution,
as it has lacked cohesion betweens its members. The fact that both entities are
entitled to special relations with neighbouring states further undermines the for-
eign policy leverage of the joint institutions towards Croatia and Yugoslavia. In
addition, presidency members have represented their country differently, accord-
ing to their national background.15 As a result, the foreign policy of Bosnia has
been weak and incoherent. In addition to the inability of the presidency to shape
Bosnian foreign policy, the institution has been weak in acting as an arbiter in do-
mestic politics, especially between the entities. Between 2000 and 2002, Bosnia
experienced some co-habitation, with presidency members lacking parliamen-
tary support or the backing of the respective entity.
2 Parliamentary Assembly
At the state-level, Bosnia is a two-chamber parliamentary system, reflecting
(con)federal arrangements and the need to grant representation to both the three
nations and the two entities. The House of Peoples (fifteen members) gives equal
representation to all three national groups but stipulates that the members are
elected from the entities where they constitute the dominant group (Serbs from
RS, Bosniaks and Croats from the Federation). The members of this chamber of
parliament are not directly elected, but rather are chosen by the respective entity
parliaments.
The House of Representatives (forty-two members) is directly elected in the two
entities, with two thirds of the seats reserved for the Federation and one third for
the RS (Constitution Art IV). While this could allow for representation of the non-
14 The Ottoman Empire recognized non-Islamic communities and granted them reli-
gious autonomy that included education. Although not based on equality between
Muslims and other groups, it created a non-territorial autonomy that was emulated
by Austria-Hungary.
15 For example, in 2001 the Chair of the Presidency and its Serb members, Živko
Radišić, sought to withdraw the case Bosnia brought against Yugoslavia before the
International Court of Justice. This case was launched by Bosnia in 1993 against Yu-
goslavia, accusing it of genocide and aggression against Bosnia. Application of the
Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia
and Herzegovina v. Yugoslavia), ICJ, 22 March 1993
210 Florian Bieber
dominant communities, they have been largely underrepresented. Of the twenty-

eight deputies elected from the Federation in 2000, approximately twenty were
Bosniak candidates, some six Croats and only one or two Serbs. In the RS, more
than ten of the fourteen deputies are Serbs.16 The representation in the House of
Representatives has thus similarly excluded non-dominant groups from both enti-
ties, but has given a more representative reflection of the three nations’ numerical
strength than the House of Peoples, where all three are represented equally. In the
House of Representatives, about half of the deputies are Bosniaks.
An elaborate system of quorums and vetoes prevents the assembly from tak-
ing any decision against the will of any of the three national communities repre-
sented in parliament. For any law or decision to be passed, at least a third from
each entity has to support the law. If this is not the case, the chair of the chamber
and his/her deputies (from all three national communities) are charged with se-
curing such a majority. In addition, any of the three community caucuses can
(by its majority) object to a decision or law, if it is “destructive of a vital interest”
(Constitution Art IV(11e)) of the respective community. Only a majority of all
nations’ delegates in the House of Peoples can override such an objection, which
makes this veto power de facto absolute. Other national delegates in the House
of Peoples can, however, by majority vote object to the invocation of the vital-in-
terest clause, which leads to the establishment of a joint commission of all three
community’s representatives to find a compromise. In case no compromise is
found, the Constitutional Court is charged with reviewing the process. As the
Constitutional Court currently contains three international judges, their partici-
pation can prevent an impasse, which would be likely to arise as each nation is
represented by two judges and past court decisions suggest national voting-pat-
terns. As a consequence, both entities and peoples have a veto right to block any
decision they perceive to be against their interest. Despite the establishment of
an ad hoc commission to forge compromises, there are few institutional mecha-
nisms to overcome possible blockage of parliament. In particular, there is no defi-
nition of what constitutes a ‘vital interest’ of the national community, allowing it
to be used in practically all areas of parliamentary work. A general veto right, as
in Bosnia, has in fact contributed to the permeation of ‘ethnic interest’ in all fields
of politics, and not to the depolitization of the decision-making process, as had
been proposed by Nordlinger and other scholars of power-sharing arrangements
(e.g., Sisk 1996: 47). Lijphart’s considerations as to why the veto power is unlikely
to result in blockage of the decision-making process have proven not to work in
Bosnia. His observation that “the very fact that the veto is available as a poten-
tial weapon gives a feeling of security which makes the actual use of it improb-
able” (Lijphart 1977: 37) has in particular not held true for Bosnian politics. In
Bosnia, where the strength or weakness of the state-institutions is contested, the
veto right has been an instrument for political representatives with an interest in
keeping the centre weaker than foreseen in the constitution to limit the function-
16 As the official election results do not list the national background of the elected can-
didates, the numbers are based on party-affiliation and name, where possible.
ing of state-institutions, above all parliament. This would suggest for power-shar-
ing theory that veto rights can only be effective in cases where the power of the
institutions that they govern are not fundamentally contested by political actors
partaking in the process. In addition, the ability and willingness of the interna-
tional civilian agencies overseeing the peace-process in Bosnia to take decisions
has weakened the negative consequence of the (mis)use of veto powers.
The Parliamentary Assembly can be active in the areas of competence accorded
to the joint institutions, especially in deciding on the financing of the institutions,
passing the budget, ratifying international treaties, and a more general compe-
tence to enact “legislation as necessary to implement decisions of the Presidency
or to carry out the responsibilities of the Assembly” (Constitution Art IV(4)). In
addition, the parliament is able to change the constitution – with the exception
of the Human Rights catalogue – with a two-thirds majority in the House of Rep-
resentatives (Constitution Art X). This procedure in fact makes the Constitution
the only part of the Dayton Peace Accords, that can be altered exclusively by do-
mestic political actors. The system of vetoes does, however, make constitutional
change extremely difficult, and has so far prevented any substantive discussion
on such changes (Pajić 2001: 40-41).
3 Council of Ministers
The Council of Ministers – the government of Bosnia – is constitutionally weak,
as evidenced by the fact that its powers and competencies are described in the
article on the presidency and not separately, as, for example, is the Parliamentary
Assembly. The chair of the Council of Ministers, the de facto prime minister, is
nominated by the presidency, but must be approved by the House of Represen-
tatives. The constitution prescribes only two ministries of the joint institutions,
a foreign ministry and a ministry of foreign trade. Between 1996 and 2000 only
these two ministries and the Ministry of Civilian Affairs existed. Additional min-
istries can be created with the approval of the House of Representatives. Only in
2000 was the number of ministries extended. The territorial distribution of the
ministers is prescribed in the Constitution, stipulating that no more than two-
thirds can be from the Federation (Constitution Art V(4)). In practice, until 2002,
every minister had two deputy ministers from the other two nations, and since
then a single deputy from a different nation. The positions of both ministers and
their deputies are distributed on a parity basis, which is a legal requirement. In
the past, the chairmanship of the Council of Ministers rotated between the differ-
ent ministers. This system has been replaced by a permanent chair of the Council
of Ministers, thus establishing a more conventional cabinet system. Despite these
steps to strengthen the government and overcome the excessive weaknesses of
the immediate post-war government, it remains constrained by the elaborate
power-sharing mechanisms and the strength of the entities (Bieber 2005: 52-54).
4 Constitutional Court
The Constitutional Court is an extremely powerful institution in Bosnia. Besides
being called upon to resolve decisions blocked by veto in the parliamentary as-
212 Florian Bieber
sembly, it has jurisdiction over some of the most contested areas of governance
in Bosnia, including any “dispute that arises … between the Entities or between
Bosnia and Herzegovina and an Entity or Entities, or between institutions of Bos-
nia and Herzegovina, including special parallel relationship with a neighboring
state” and the consistency of “any provision of an Entity’s constitution or law”
with the Bosnian (Constitution Art VI(3a)). As such, the court has the arbitra-
tional role common to Constitutional Courts in Federal systems (Pobrić 2000:
476-481). Its importance is further enhanced by the absence of other domestic
mechanisms for arbitration between the three communities. The court itself is
composed of nine members, two from the RS and four from the Federation. In
practice, the four members from the Federation have been two Bosniaks and two
Croats. Furthermore, three members were named by the President of the Eu-
ropean Court of Human Rights for a transitional five-year period. In July 2002,
three new international judges were appointed to the court. These international
members have proven critical in key decisions of the court. Especially the deci-
sion that stipulated that the entities’ constitutions contravened the preamble of
the Bosnian Constitution – according to which all three peoples and ‘others’ are
constituent people on the whole territory of the country – was only passed due
to the votes in favour by the international judges (Constitutional Court of Bos-
nia-Herzegovina 2000). Even when excluding the international component, the
court is the only institution where decisions are taken by a simple majority of the
judges, thus lacking any veto rights.
At the level of the common institutions, power-sharing mechanisms are pro-
vided in all areas. The Presidency and the House of Peoples assure the representa-
tion of both the national groups and the entities. In the House of Representatives
and in the Council of Ministers only the territorial representation is prescribed
in the constitution. While both entities are primarily constituted by one or two
dominant nations, here the three national groups are guaranteed representation.
Even in the few fields of competence accorded to the common institutions, veto
rights and guaranteed representation assure that no decision can be taken with-
out the consent of a majority of the representatives of each of the three national
communities, with the Constitutional Court being the only exception.
B Entities
A key compromise contained in the DPA is the recognition of two ethnically-de-
fined units, which are vested with considerable powers: the RS and the Federation
of Bosnia and Herzegovina. The entities differ greatly in their governance and in
the degree of power-sharing they provide. The Federation Constitution between
1994 and 2002 provided for a power-sharing arrangement between Bosniaks and
Croats, while the RS defined itself as the nation-‘state’ of Serbs.
The Federation is highly decentralized, with most powers at the level of the
ten cantons, whereas the RS is centralized, with little competencies at the local
level. In terms of power-sharing mechanisms, the RS lacked any assured repre-
sentation or participation for non-Serbs until the decreed constitutional changes
in early 2002 that will be discussed below. The Federation, on the other hand,
has an elaborate system of power-sharing, mirroring the joint institutions. The
parallel between Federation and state institutions is in fact hardly surprising,
considering that they are both the outcome of US-brokered negotiations with
considerable input of constitutional drafts from the US negotiators. In a range of
fields, the Dayton Constitution copies the Federation Constitution, including the
main institutions (presidency, two-chamber parliament, and government) and
the high degree of decentralization. Unsurprisingly, the Federation has suffered
from many of the same weaknesses of the central authorities.17
While the entities were originally mono- or bi-national autonomies, which ex-
ercised most power and minimized the degree of interethnic cooperation in Bos-
nia, a change to the entities’ constitutions, in April 2002 changed the nature of
the entities (Bieber 2005: 121-132). In 1998, Alija Izetbegović, at the time a mem-
ber of the state presidency, brought a case to the Constitutional Court regarding
the constitutionality of the entity constitutions. The Constitutional Court deter-
mined in several partial decisions in 2000 that, in fact, the entity constitutions
were partly unconstitutional as they did not grant the three nations equal rights
on the whole territory of the country as stipulated in the Bosnian Constitution.
The Court required constitutional amendments from both entities, which would
correct both the symbolic aspects of the constitutions and the institutional ar-
rangements, which excluded both Bosniaks and Croats in the RS or Serbs in the
Federation (Constitutional Court of Bosnia-Herzegovina 2000). In the absence
of any activity by the two entities, the High Representative Wolfgang Petritsch
first appointed two constitutional commissions, one for each entity, to develop
constitutional amendments and temporarily address the Constitutional Court
judgement (Perry 2002: 2-3). Despite intensive negotiations and near-compro-
mises over the amendments to the entity constitutions, the High Representative
eventually imposed amendments to both entities’ constitutions in April 2002.
According to these amendments, Bosniaks and Croats have to be proportion-
ally represented (according to the 1991 census) in the RS at all levels of govern-
ment, as do Serbs in the Federation. In addition, ‘others’ (e.g., Roma or Jews) are
also no longer to be excluded from political representation in the entities (OHR
2002a; OHR 2002b). For example, in addition to the president of the RS, the en-
tity will have two vice-presidents, elected from the other two constituent peoples.
Whereas before April 2002 no ethnic requirements were linked to the position
of the president and vice-president of the RS – considering the overwhelming
Serb majority of the entity the ethnicity seemed a forgone conclusion – the new
constitution requires the election of two vice-presidents of different nations than
the president (OHR 2002c).
While the adding of Serb office-holders in the Federation further likens the en-
tity to the state-institutions, the constitutional changes fundamentally changed
the nature of RS, which had previously defined itself in mono-ethnic terms as the
17 A highly instructive study of this is the case of Mostar, the focal point of the Federa-
tion, as it is the only town divided between Croats and Bosniaks (Bose 2000: 95-148).
214 Florian Bieber
nation ‘state’ of Serbs. It is, however, noteworthy that the constitutional changes
in fact enhanced the ethnic dimension of the Constitution of the RS, which be-
forehand – based on the assumption of Serb dominance – had less explicit refer-
ences to the national background of office holders.
Although the constitutional changes do not provide Serbs in the Federation
or Bosniaks and Croats in the RS with some kind of autonomy within the entity,
three key aspects of power-sharing – veto rights, proportional representation,
and participation in government – now form part of the institutional structure of
the entities as well. Whereas the changes are based on a court decision, and for-
mally do not constitute a change of the DPA, the amendments change the nature
of post-Dayton governance from mono- and bi-national autonomy with weak
power-sharing towards an extension of some elements of power-sharing to the
autonomous regions. The entities remain the main aspect of ethnic self-govern-
ment in Bosnia, but have lost their ethnic exclusiveness.
C The Role of the International Community

The peace agreement for Bosnia has been paradigmatic in terms of the powers it
vests in international institutions and personnel with the consent of the local ac-
tors. Moreover, the role of the international community, the High Representative
in particular, has been considerably broadened during the implementation pro-
cess. When examining the degree and structure of international involvement in
post-conflict Bosnia, some important distinctions need to be made. Firstly, differ-
ent organizations with unequal powers were charged with overseeing the military
and the civilian aspects of the implementation of the peace agreement. The mul-
tinational peacekeeping force – IFOR (Implementation Force), later SFOR (Stabi-
lization Force) and now Eufor (European Union Force) – under first NATO and
later EU leadership, is the international force charged with the implementation
of the military aspects of the agreement (DPA 1995: Annex 1). The International
Police Task Force (IPTF), as part of the United Nations presence in Bosnia, was ad-
ditionally established to oversee the local police forces (DPA 1995: Annex 11), to be
replaced by the EU Police Mission (EUPM) in 2003. The institution established to
supervise the civilian implementation of the DPA is the Office of the High Repre-
sentative (OHR), supported in certain aspects by the OSCE, the EU, and the Coun-
cil of Europe, as well as a host of other international organizations. In addition to
these institutions, be they already established international organizations or set up
specifically for Bosnia, a number of foreign officials, nominated by international
organizations, form an integral part of the Bosnian institutional structure at the
joint level (Constitutional Court, Central Bank, Human Rights Chamber, etc.), pro-
viding for an additional layer of international involvement in post-Dayton Bosnia.
1 International Military Presence

A robust international military presence under the leadership of NATO has been
a key factor in ending the war in Bosnia. IFOR, consisting of 60,000 troops from
seventeen NATO and fourteen non-NATO countries, was originally tasked with
overseeing the disengagement of the military forces in Bosnia, the establishment

of a ceasefire along the Inter-Entity Boundary Line (IEBL) and the withdrawal of
foreign forces (both from neighbouring states and other third countries) from
Bosnia (DPA 1995: Annex 1a). After these tasks were implemented relatively
swiftly and, with some exceptions, comprehensively (International Crisis Group
1999c: 3-7), the task of the international military presence turned towards ‘sta-
bilization’, i.e., the prevention of new conflicts through their presence and the
supervision of the local military forces, as reflected in the change of names from
IFOR to SFOR in December 1996. The military forces were, however, reluctant to
actively support civilian aspects of the DPA implementation. While SFOR began
arresting some indicted war criminals after 1997 and protected civilian officials
when carrying out their duties, i.e., during the raid on the Hercegovačka Banka
in April 2001 or on Bosnian Serb television transmitters in August 1997, military
support has been partial and only provided hesitatingly, with frequent citing of
the limitation of the mandate to military matters only. In 1999, US army personnel
within SFOR were for the first time permitted to assist the efforts of the OHR in
supporting refugee return. Despite these steps to augment the mandate of SFOR,
the troops were not able to prevent the riots in Trebinje and Banja Luka in May
2001, nor could they provide adequate support during the first attempt to take
over the Hercegovačka Banka in late March 2001. These incidents, as well as the
continued presence of indicted war criminals in Bosnia, have drawn considerable
criticism of the international military presence in Bosnia (International Crisis
Group 2000). With a shift in the role of SFOR from traditional peacekeeping to
more specific tasks, such as the prevention of riots or attacks against returnees,
the total strength of the force was reduced to approximately 20,000 personnel by
2001, while a specialized unit for preventing civil disorder and rioting was cre-
ated in 1999. In December 2004, the Eufor Mission Althea replaced the NATO-
led peacekeepers, but maintained the 7,000 strong presence in the country. This
transfer has been part of the progressive ‘Europeanization’ of the international
presence in light of the perspective of European Integration for Bosnia.
The criticism of the international military presence in Bosnia notwithstanding,
the implementation of the military aspects of the DPA have been widely iden-
tified as being most successful. The success can be attributed to a number of
factors: first, the mandate and powers of SFOR/IFOR were clearly outlined and
detailed, unlike the role of civilian agencies. Second, the mandate was clearly
limited and in many areas did not affect the monopoly of nationalist parties over
territory under their control. Third, the military aspect of peace implementation
focused on securing the status quo, rather than changing the balance of power
on the ground, unlike civilian aspects of the agreement.18 Fourth, the military
aims of the DPA have been less ambitious than other aspects, such as the refugee
18 In areas where a change of the status quo took place, they were far less successful,
as evidenced by the handing over of Serb-held suburbs of Sarajevo to Federation
control in early 1996, leading to the mass departure of Serbs living in the area and
large-scale destruction of homes. See Sell 2002: 179-202.
216 Florian Bieber
return. These more achievable goals allowed for a swift and relatively successful
implementation. Finally, it needs to be noted that the financial resources at the
disposal of the international military presence by far outweighed the means of
the civilian agencies in Bosnia.
2 Role of International Civilian Agencies

The primary civilian international agency in Bosnia is the Office of the High Rep-
resentative (OHR), established in the DPA as “the final authority in theatre re-
garding interpretation of this agreement on the civilian implementation of the
peace settlement” (DPA 1995: Annex 10, Art V). The original powers of the High
Representative, an office held by Carl Bildt (1996–1997), Carlos Westendorp
(1997–1999), Wolfgang Petritsch (1999–2002), and currently Paddy Ashdown
(2002–present) were limited to “facilitate,” “co-ordinate,” and “report” (Annex
10, Art II), resulting in the parties to the conflict mostly ignoring the work of
the High Representative (Petritsch 2000: 300). The powers were augmented in
the meeting of the Peace Implementation Council (PIC) in Bonn in December
1997. The PIC is tasked with supervising the work of the High Representative
and brings together 55 countries and international organizations involved in the
peace process. The session of the council granted the High Representative the
power to dismiss officials (both civil servants and elected officials) for obstruct-
ing the DPA implementation and to decree legally binding ‘decisions’ in all areas
where the parliamentary assembly or the entity parliaments are unable to pass
laws. The High Representative has made extensive use of this competence. Since
the broadening of the OHR’s mandate, over 180 officials have been dismissed
and the High Representative took more than 100 decisions. The number of deci-
sions has increased markedly since the introduction of the extended powers of
the High Representative in 1997. In 1998, for example, the High Representative
took only 28 decisions, whereas in 2002 the number had increased to 153.
Most dismissals affected local administrations, but numerous high-ranking
officials were dismissed, including the president of the RS, Nikola Poplašen, in
March 1999; the Croat member of the Bosnian presidency, Ante Jelavić, in March
2001; and the Croat member of the presidency, Dragan Čović, in March 2005.
The decisions taken by the OHR range from areas less pertinent to interethnic re-
lations, such as pension payment plans, to key areas of re-establishing interethnic
cooperation and Bosnian statehood, such as the decision on Bosnia-wide license
plates,19 or the flag and hymn of Bosnia.
This considerable power vested in an international organization – without
clear democratic scrutiny – was conceived as a ‘corrective’ for both the unwilling-
ness of the political elites’ in power after the war to engage in the implementation
of the DPA and for their attempt to maintain the divide between the nations in
Bosnia in order to preserve their own monopolistic hold on their respective con-
stituencies. This policy switch from monitoring to a more assertive role followed
19 The introduction of neutral license plates in 1998, which do not reveal the origin of
the car, has been widely identified as a factor in increased cross-entity traffic.
the demands of international think tanks, such as the International Crisis Group,
as well as a number of Bosnian intellectuals, who demanded the establishment
of an outright protectorate for Bosnia immediately following the end of the war
(Sokolović 2001: 93-106). The lack of democratic control and the proliferation of
international intervention in Bosnian governance did, however, lead to criticism
(Chandler 2001; 1999). In addition, the willingness of the High Representative to
make decisions in areas were parliament was not able to agree did in fact increase
the unwillingness of some political leaders to also make (unpopular) decisions
in areas not pertaining to interethnic relations (Cox 2001: 12-15). The reduction
of joint decision-making as a result of the HR interventions was acknowledged
as a problem even by Wolfgang Petritsch: “In the long run the usage of the far-
reaching powers lead to a type of dependency syndrome. Local parties began to
opportunistically rely too much on the political intervention of the High Repre-
sentative, especially when it came to unpopular measures. They can behave, de-
spite being in government, as if they were in opposition and defend their ethno-
nationalist goals without need to compromise.” (Petritsch 2000: 301).
In addition to the impact of the OHR’s (and international agencies) executive
and legislative powers on power-sharing arrangements, it can be argued that the
interventionist approach of the international implementing agencies has had a
negative impact on the larger process of democratization in the country (Chan-
dler 2001). A number of decisions taken by the High Representative during the
past years have suspended or opposed decisions – election results or votes in
parliament – that had been democratically endorsed by the Bosnian voters or
elected officials. While in most of these decisions the argument has forcefully
been made that they sought to correct nationalist, obstructionist, or corrupt
practices and decisions, they undermine the nascent democratic process in the
country. In addition, the nature of decision-making in international organiza-
tions limits their credibility as agents of democratization. The decision-making
process of the OHR, as that of other international agencies, is inherently secretive
and ‘intransparent’. Additionally, relatively small groups of policy makers are able
to exert disproportionately large influence. This development has been widely
recognized by a range of domestic actors who have sought to influence policy
less through parliament or other democratic institutions, but rather through lob-
bying the OHR and the other agencies, effectively extending the role of these
international agencies in terms of their duration and breadth. These critiques
gained additional weight with the parliamentary assembly of the Council of Eu-
rope criticizing the powers of the OHR. A detailed report, in particular noting
the lack of legal recourse for dismissal, was published by the Venice Commission
in early 2005 (Venice Commission 2005).
After the OHR, the OSCE has been the international organization with the
most significant role in the implementation of the DPA. Besides a number of
more general projects in the field of democratization and media, the OSCE was
charged with carrying out the first elections in Bosnia. Elections have been a
key pillar in the international community’s attempt to further reconciliation and
democratization in Bosnia. Through the Provisional Election Commission, the
218 Florian Bieber
OSCE determined the election rules. Despite this high degree of international
intervention into the electoral process – from party registration, partial funding
of parties, to the dismissal of official and candidates – the election results did not
favour candidates or parties that were supported by or supportive of the interna-
tional community in Bosnia (Shoup 1997: 10-14).
In addition to the OHR and the OSCE, the United Nations contributed sig-
nificantly to the implementation of the peace agreement. The organization was
charged with the training and monitoring of the local police forces through the
creation of the International Police Task Force (IPTF). The 1,500 members of the
IPTF were charged with monitoring the police forces with the mandate to pro-
fessionalize and de-‘ethnicize’ them. Furthermore, the UN has been instrumental
in the creation of the State Border Service, the only security structure under the
control of the joint institutions, tasked with controlling the external borders of
Bosnia (Report of the Secretary General 2000). The EU Police Mission succeeded
the UN presence in 2003 with similar activities. Nevertheless, a more substantial
police reform, including the abolition of the entity and cantonal control over the
police have been under discussion since 2004. Generally speaking, the interna-
tional presence has been scaled down in size and in the number of actors since
1996. In addition, the EU has taken on key responsibilities from NATO and UN.
Furthermore, the OHR is at the same time the Office of the EU Special Represen-
tative in Bosnia, with the latter position to eventually eclipse the OHR.
In the sphere of economic reconstruction, the EU again and international fi-
nancial institutions, such as the World Bank, lead the way. Their tasks do, how-
ever, not fall within the narrow range of the DPA itself.
Finally, a number of individuals from outside Bosnia, Croatia, and Yugoslavia
have played a key role in the institutional framework of Bosnia. As mentioned ear-
lier, three of the nine members of the Constitutional Court are foreigners. Similar-
ly, the Central Bank contains an international component. The first governor of the
bank, chosen for a six-year term by the International Monetary Fund (DPA 1995:
Art VII), contributed significantly to the acceptance of the Bosnian currency, the
Konvertibilna Marka. In addition, the joint institutions charged with overseeing
human rights in Bosnia contain a strong international component. The Ombud-
sperson for Human Rights is nominated for a five-year term by the Chairman-in-
Office of the OSCE (DPA 1995: Annex 6, Art IV), while eight of the fourteen mem-
bers of the Human Rights Chamber are nominated by the Committee of Ministers
of the Council of Europe (DPA 1995: Annex 6, Art VII), for a five-year term.
These positions – held by citizens of neither Bosnia nor its two neighbours –
are to eventually revert to domestic office holders after the transition period. The
original transition periods of five or six years have been extended, however, due to
the lack of political progress at the state level. As result, a number of international
organizations, the IMF, the OSCE, the Council of Europe, and the European Court
of Human Rights have been granted significant influence on the domestic insti-
tutional system. The creation of Bosnian institutions with international participa-
tion has been largely more successful than the outright separation of international
and Bosnian institutions. Both the Central Bank and the Constitutional Court did
emerge as key institutions of the joint state in recent years. Their independence
from the political process and their decisions in strengthening the Bosnian state
(such as the introduction of the convertible Mark and the Court decision on the
entity constitutions) could be considered as successes of mixed international-do-
mestic institutions. The eventual phasing out of international participants in the
institutions also will provide for complete local ownership of these institutions.
In addition to the formal powers of the international organizations in Bosnia,
one has to note the strong informal influence of these organizations and of West-
ern governments (represented through their embassies in the country). For ex-
ample, the Alliance for Change, comprising a motley group of ten parties centring
on the SDP, was established following the November 2000 election with the active
involvement of the OHR and the US embassy in Sarajevo (Domi 2001). Only this
post-election engineering enabled moderate political parties to take power for the
first time in the Federation and at the state level. Both the economic and diplomat-
ic power of the international community and Western governments in Bosnia and
the extensive powers accorded to the OHR have yielded the international com-
munity leverage in the political system of Bosnia well beyond the power granted
in the DPA or by the Peace Implementation Council. At times, the intervention
of different countries has even been detrimental to the peace process. The most
notable example was the arrest by American agents of six Algerians suspected of
planning terrorist attacks in early 2002. The Bosnian authorities had previously
released them, after not receiving sufficient evidence from the US authorities. This
clear breach of Bosnian laws by foreign countries was widely criticized in Bosnia
by human rights organizations and undermined the claim by international agen-
cies to further the establishment of the rule of law in Bosnia (Simić 2001).
Despite the length and the detail of the Dayton Peace Accords in many of its as-
pects and the significant international presence in Bosnia since 1995, a surprising
number of areas of governance are either not addressed at all, or only in passing.
The entities, wherein most power is vested according to the constitution, par-
ticularly have had little formal regulation of their policies through the DPA. The
Constitutional Court decision of 2000 forced the entities to significantly change
their institutions and structures so as to conform to the state constitution. Fields
that were not deemed to fall into the narrow scope of interethnic relations, such
as economic reform or social entitlements, were excluded from the agreement.
At the same time, the international community has exercised its mandate in Bos-
nia very broadly, which has meant that it has become active in fields far beyond
the narrow scope of the agreement itself.
The peace agreement, with its 11 Annexes, maps and ‘side-letters’, was an attempt
to broaden traditional conceptions of peace treaties or ceasefire agreements, while
also providing a blueprint for the post-war reconstruction of Bosnia. In doing so,
however, it created a structure that was so weak that it gave Bosnia little institu-
tional hold at the centre. The agreement, written with the concerns of the time in
mind, focused rather narrowly on key issues such as central institutions, human
rights and the status of refugees, and, most importantly, military disengagement.
220 Florian Bieber
Other aspects, such as the institutions of the entities, economic reform, and reform
of the police and judiciary lay outside the scope of the agreement.
V Acceptance and Obstruction of the Dayton Peace Accords

Only the Republic of Croatia, the Federal Republic of Yugoslavia and the Republic
of Bosnia-Herzegovina formally negotiated the Dayton Peace Agreement. The ex-
clusion of both the Bosnian Croats and Bosnian Serbs from the negotiation process
bore the risk of rejection by either of the two groups’ political leadership. In the
case of Bosnian Croats, however, the close party links with Croatia and clear con-
trol exerted by the Croatian government and ruling party over the Bosnian Croat
leadership20 ensured that Croatian acceptance would translate into Bosnian Croat
acceptance. In the absence of such a direct link between Bosnian Serbs and Ser-
bian/Yugoslav authorities after 1993–1994, the danger of rejection was considerably
larger. The Bosnian Serb leadership had only conceded that Milošević could ne-
gotiate on their behalf after considerable pressure was exerted on the leadership
of the RS. Furthermore, the Bosnian Serb leadership had successfully undermined
previous peace agreements, most notably the Vance-Owen plan, which was first
signed by Radovan Karadžić, but later renounced by the Bosnian Serb parliament
and submitted to a ‘referendum’ in which the plan was rejected (Burg and Shoup
1999: 248-249).
The agreement was nevertheless accepted – albeit reluctantly – by all the main
political forces in Bosnia. The acceptance by the nationalist forces in Bosnia was
largely tactical with little commitment to any significant implementation of the
agreement.
The most significant resistance to the implementation of the agreement arose
from local authorities under the control of the nationalist parties, most notably
the SDS and the HDZ. The rejection of the agreement’s implementation was par-
ticularly evident from the unwillingness to accept the return of non-dominant
groups to their former homes. With the widening of the High Representative’s
powers in 1997, a number of local officials were dismissed for hindering the re-
turn of refugees or internally displaced persons (IDPs), or for failing to prevent
acts (at times even the organization) of violence against returnees (International
Crisis Group 1999b).
While the three nationalist parties remained at least formally committed to
the DPA, largely because they were able to prevent and/or delay any substantial
implementation and also out of fear of international reprisals, the most important
political force that openly opposed the agreement was the Serbian Radical Party
(SRS Srpska Radikalna Stranka). The party, a branch of the extremist national-
ist party in Serbia under the leadership of Vojislav Šešelj, won the presidential
elections in the RS in 1998 in coalition with the Serb Democratic Party. Nikola
20 For example, both the moderate HDZ leader in Bosnia Stjepan Kljuić (in 1992) and
the hardliner Mate Boban (in 1994) were ousted as a result of HDZ intervention from
Croatia.
Poplašen exercised his office as president of the RS for only a few months, as he
was dismissed by the High Representative for obstructing the DPA in March 1999
(International Crisis Group 1999a). While the political programme and the behav-
iour of Poplašen in office suggest that the party opposed the DPA, even the SRS
and its president formally endorsed the peace agreement (Glas Srpski 1998).
The most controversial aspect of the agreement was the right of refugees to re-
turn to their pre-war homes, as stipulated in Annex 7 of the DPA. While the peace
agreement in its other aspects did not threaten the demographic homogeneity
and political dominance created by the nationalist parties, the principle of refugee
return provides a serious challenge to the results of ‘ethnic cleansing’. Significant
parts of the governing elite in the RS and in Croat controlled areas unofficially
continue to support separation from Bosnia and an eventual union with their kin-
states. Nationalist political elites in Bosnia and neighbouring Croatia and Yugosla-
via have, however, realized that this goal had to be postponed in favour of main-
taining the entities or territories with far reaching autonomy (Koštunica 1999).
After the challenge to the DPA by Nikola Poplašen, the decision of the HDZ
leadership in late 2000 and into early 2001 to withdraw from the joint institutions
and form a system of Croat self-government constituted the next most serious
challenge to the peace agreement on the state-wide level. While the HDZ under
the leadership of Ante Jelavić sought to maintain a degree of formal support for the
DPA while criticizing post-Dayton international policy towards Bosnia, especially
some last-minute election rules for the November 2000 elections, their position
led to the dismissal of Jelavić from the Bosnian presidency by the High Represen-
tative (Bieber 2001). After a six-month boycott, the HDZ returned to the state and
Federation institutions, thus ending that challenge to the peace process.
In the course of DPA implementation, the support and rejection of the agree-
ment has shifted considerably. While the moderate non-nationalist parties, such
as the social democratic parties, supported the agreement in its early phase, they
emerged as the main critics of the accords by 2000. This was particularly true of
the Party for Bosnia-Herzegovina (SBiH, Stranka za BiH), founded by the former
Bosnian prime minister and leading SDA official Haris Silajdžić, which had vo-
cally advocated a change of the DPA and the abolition of the entities (especially
the RS). Despite the party’s opposition to the agreement, its position was not
criticized by the international community, and in fact it became, together with
the SDP, a main pillar of the international community’s strategy to marginalize
nationalist forces in Bosnia.
The fact that the DPA is itself inherently contradictory and encompasses the
concept of Bosnia’s re-integration, as well as those of decentralization and ethnic
separation, has made it difficult to assess the level of rejection or support for the
DPA. The High Representative has dismissed a number of elected officials in the
past years for obstructing the peace accords’ implementation or for openly reject-
ing the agreement. These dismissals did, however, only affect officials who had ar-
gued for a lower degree of re-integration of Bosnia than foreseen in Dayton, while
support for a stronger state than prescribed in the DPA was deemed acceptable.
Since the challenge to the agreement in 2001, the preferred strategy of opponents
222 Florian Bieber
has become obstruction, i.e., formal support coupled with non-implementation.

As the political and security situation in Bosnia evolved fundamentally since 1995,
the key challenge has become the interpretation of the Dayton Accords in the
light of the time eclipsed. Whereas the international actors sought to integrate
the state at a higher degree of centralization than foreseen in the text of the peace
accords, the Bosnian Serb political forces primarily sought to limit the evolution
of the state to the minimum requirements foreseen in the peace accord.
VI The Implementation Record of the Agreement

The Dayton Peace Accords have been widely credited with ending a war, while
both the agreement itself and its implementation have been severely criticized for
not addressing some of the negative effects of the war and failing to substantially
reduce interethnic tensions. Assessing the success of the implementation of the
DPA has been difficult, due to the complexity of the institutional structures, the
multiple domestic and international actors charged with its implementation and
the internal contradictions of the agreement itself.
In evaluating the implementation, one can divide the process since the signing
of the accords into three phases, which mark a progressive increase in the imple-
mentation process, both quantitatively and qualitatively.
During the first phase (December 1995–December 1997), when the High Rep-
resentative was given increased competencies, the so-called Bonn powers, the
implementation was restricted to Annex 1 A, i.e., the establishment of a ceasefire.
The reluctance of IFOR/SFOR to act in areas outside the narrow military imple-
mentation process and the lack of power given to the civilian agencies meant
that joint institutions existed only pro forma, freedom of movement across inter-
entity boundaries was impossible, and the parallel power structures of the three
nationalist parties remained largely undisturbed.
The second phase (December 1997–early 2001), marked an incremental in-
crease in international intervention in the implementation process and a weak-
ening of the parallel structures. While the joint institutions remained frail due
to the continued electoral success of nationalist and obstructionist politicians,
the international community sought to bolster moderate, or – more often – the
slightly less compromised figures in both entities, such as Biljana Plavšić or Milo-
rad Dodik in the RS. The lack of popular support meant, however, that the ac-
celeration in the implementation process was internationally driven and lacked
roots in a popular consensus, even if members of all three communities eventu-
ally welcomed some of the measures, such as the introduction of the convertible
Mark or the establishment of Bosnia-wide license plates.
The third and final phase began in May 2000, when for the first time since the
introduction of multi-party democracy in Bosnia non-nationalist parties could
make significant gains at the expense of the three nationalist parties. Although
the November 2000 elections did not bring about a significant shift of public
opinion away from the nationalist parties, with the exception of the Bosniak-
dominated part of the country,21 it did yield an election result (Table 1) that en-
abled a fragile coalition of multinational and moderately nationalist parties to
take over from the ‘big three’ in the state government and parliament, and in the
Federation. While the record of the new government has been mixed, it provided
the international community with a local interlocutor in the implementation pro-
cess. The emergence of a local partner coincided with a shift of the international
community’s policy, which recognized the need to strengthen the state institu-
tions. A number of individual events, such as the aforementioned Constitutional
Court decision, the change of regimes in Croatia and Yugoslavia, as well as the
isolation of Croat extremists due to the proclamation of Croat self-government in
March 2001, furthered the demise of parallel power-structures.
Table 1 Results of the Bosnian Elections, November 2000

(House of Representatives)
Party Percentage of Seats won Party orienta-
vote tion
Federation of Bosnia-Herzegovina
SDP 27.3 8 Moderate
SDA 27.0 7 Nationalist
HDZ BiH 19.3 5 Nationalist
Stranka za BiH 15.6 4 Moderate
DNZ BiH 2.2 1 Moderate
NHI 2.0 1 Moderate
Others 6.6 1
Republika Srpska
SDS 39.7 6 Nationalist
PDP 15.2 2 Moderate
SNSD/DSP 10.6 1 Moderate
SDA 7.4 1 Nationalist
SP RS 5.7 1 Moderate
Stranka za BiH 5.4 1 Moderate
SDP 5.2 1 Moderate
SNS RS – Biljana Plavsic 4.5 1 Moderate
Others 6.3 0
Source: OSCE Mission to Bosnia and Herzegovina website www.oscebih.org.
On the fifth anniversary of the signing of the DPA, the International Crisis Group
concluded in their comprehensive review of the implementation process that
21 An ESI report points out that in fact the November 2000 election results largely
mirror the November 1990 elections that brought the nationalist parties to power
(European Stability Initiative 2001: 7).
224 Florian Bieber
“[t]he effect has been to cement wartime ethnic cleansing and maintain ethnic
cleansers in power within mono-ethnic political frameworks … Indeed, the only
unqualified success has been the four-year absence of armed conflict.” (Inter-
national Crisis Group 1999c: i). Since this assessment, some progress has been
made, as aptly described by the European Stability Initiative (ESI):
Over the last two years, Bosnia has changed dramatically. The ethnic power structures left
over from the war have begun to crumble. Political changes in Croatia and Serbia have cut
external revenues, forcing the political elites in Bosnia to reorient themselves towards the
international community and participate in the Bosnian state. Although the nationalist
parties continue to enjoy widespread electoral support, the post-war nationalist regimes,
built on the ideology of ethnic cleansing and dedicated to the objective of dividing Bosnia,
have gone. Determined international efforts to support the return of displaced persons,
arrest indicted war criminals and challenge the security structures, including paramilitary
and intelligence services, have yielded results (European Stability Initiative 2001: 2).
Any assessment of the implementation of the agreement thus far has to take into
account the gradual process in which implementation has taken place. In assess-
ing the implementation of the accords, one can divide the degree of success (or
the lack thereof ) into four categories:
• no or little implementation (1);
• partial, yet incomplete implementation (2);
• significant implementation (3);
• full implementation (4).
Table 2 shows that the degree of implementation varies greatly between the differ-
ent annexes of the peace agreement. While some have been largely implemented,
with others, implementation remains in the early stages.
Table 2 Implementation record by Annex, 1996–200522

Implementation level 1 2 3 4
Military Aspects (Annex 1a) *
Regional Military Aspects and Stabilization *
(Annex 1b)
Inter-entity Boundary Line (Annex 2) *
Elections (Annex 3) *
Constitution (Annex 4) *
Human Rights (Annex 6) *
Refugee (Annex 7) *
Monuments (Annex 8) *
22 Annex 5 on Arbitration and Annex 9 on Corporations have been excluded, as their

implementation cannot be evaluated.
Evaluating the different aspects of the peace accords according to degree of imple-
mentation allows for an analysis of the most problematic fields in the accords and
helps identify agencies that have proven to be particularly weak in implementing
the peace accords. In the following, I will examine the crucial issues of Refu-
gee Returns (Annex 7), Joint Institutions (Annex 4), and Human and Minority
Rights (Annex 4 and 6), which are directly related to the Dayton Peace Accords.
Subsequently, I turn to fields that are of particular importance in post-conflict
settlements for divided societies, such as education, communications and media,
administration of justice, and cross-border cooperation, but which are not or are
only marginally addressed by the DPA.
A Refugee Return
The war in Bosnia saw the displacement of over half of the population. 1.1 mil-
lion people were internally displaced, while 1.2 million sought refuge outside of
Bosnia in Croatia, Yugoslavia and other third countries. The peace agreement
sought to redress this massive human displacement for both moral and practical
reasons. Accepting the permanent displacement of over 50% of the population
would have been interpreted as rewarding the political forces who sought to ‘eth-
nically cleanse’ territory under their control in order to establish mono-ethnic
statelets. From a practical perspective, the high number of refugees placed a sig-
nificant burden on many countries, not to mention the two neighbouring coun-
tries and presented the risk of a permanent source of insecurity, as has been the
case with Palestinian refugees in countries neighbouring Israel.
Annex 7 thus stipulated that:
all refugees and displaced persons have the right freely to return to their homes of ori-
gin … [and] have the right to have restored to them property of which they were de-
prived in the course of hostilities since 1991 and to be compensated for any property
that cannot be restored to them. (DPA 1995: Annex 7, Art I, 1).
The focus of the return process has been on so-called minority returns, i.e., the
return of displaced people to their original place of residence in which they now
constitute a minority, be it at the level of the entity for the RS, or at the level of
cantons for the Federation.23 While the return to majority areas has also been
frequently problematic, the minority returns have posed the most significant
problems and were often obstructed by local authorities. Most of the returns in
the first phase after the war were returns to areas of the majority of the refugees
and displaced people, while the later and slower returns were minority returns.
By mid-2000, the UNHCR concluded that “return movements to ‘majority’ areas
have largely been completed” (UNHCR 2000b: 205).
23 For the two ‘mixed’ cantons in the Federation, the minority status is determined at
the level of the municipality.
226
3.
Table 3 Return of Refugees and IDPs (January 1996– July 2005)
Florian Bieber
Refugees Internally Displaced Persons

Bos- Croat Serbs Other Total Bos- Croat Serb Other Total
niaks niaks
1996 76,385 3,144 8,477 33 88,039 101,402 505 62,792 42 164,741
1997 74,756 33,568 11,136 820 120,280 39,447 10,191 8,452 205 58,295
1998 78,589 23,187 6,765 1,459 110,000 15,806 4,325 9,139 300 29,570
1999 18,440 6,299 6,332 579 31,650 24,907 6,760 11,315 403 43,385
2000 7,633 4,834 5,303 837 18,607 36,944 7,779 14,175 449 59,347
2001 4,642 4,244 9,155 652 18,693 48,042 5,960 25,734 436 80,172
2002 12,592 5,933 18,220 389 37,134 41,511 5,319 23,215 730 70,775
2003 5,257 2,852 5,482 421 14,012 21,861 2,267 16,023 152 40,303
2004 976 450 942 74 2,442 12,976 1,028 3,888 56 17,948
2005 518 127 195 3 843 2,470 302 650 6 3,048
Total 279,788 84,638 72,007 5,267 441,700 44,436 175,383 2,779 567,964
Source: UNHCR 2005a.
cal system. Even more pertinent is the lack of economic prospects for returnees,
which are the result of both discrimination and the absence of overall economic
ical representation and inclusion of the non-dominant returnees into the politi-
On a broader scale, the return of refugees is closely interconnected with the polit-
Bosnia remains drastically altered in comparison to 1991, as is illustrated in Table
Altogether, only about a third of all refugees and IDPs had returned in the six
years since the end of the war, with the result that the current ethnic structure of
development in Bosnia (UNHCR 2000a). In addition, the communes to which

refugees seek to return have to be able to provide certain services to the return-
ees, such as schooling, social and health care, and the prospect of finding a job.
The absence of these basic conditions for return, often not based on discrimina-
tion but on the overall poor state of many communes, has been an important
obstacle to refugee return.
The acceleration of the return of refugees to minority areas began in 1999 (Fig-
ure 1) and has since increased dramatically. This even included areas where re-
turn had been prevented for a long time by nationalist forces, especially in east-
ern Bosnia and in Herzegovina by local Serb and Croat authorities respectively
(International Crisis Group 2000: 3-6).
Figure 1 Minority Returns in Bosnia, 1996-2004

120,000
100,000
80,000
60,000 Line 1
40,000
20,000
0
1996/7 1998 1999 2000 2001 2002 2003 2004
Source: UNHCR 2005
The return of refugees to minority areas is connected with a number of factors

and can be seen as a litmus test for the overall progress in stabilizing Bosnia.
In the narrowest sense, the return of refugees is dependent on the immediate
security for those who wish to return. Secondly, the return process is largely con-
ditional on a clarification of the ownership of property and the restitution of real
estate to pre-war owners.
In an overall assessment, the return of refugees has only been partially success-
ful. The acceleration of minority returns and the de facto completion of major-
ity returns point to the possibility of success in this most ambitious part of the
Dayton Peace Accords. At the same time, the fact that approximately half of all
refugees and IDPs continue to live in their new areas of residence ten years after
the end of the war and that the return process has sharply declined since 2003
renders the assessment of the implementation of the peace accords less optimis-
tic in this aspect.
With the passage of time, the eventual return of many refugees and IDPs will
become less likely, as families establish new homes with professional and personal
ties and might be reluctant to give them up again. The reason for the lack of suc-
cess in the implementation of refugee return can be attributed to the obstruction
to the return process, mostly at the local level. The RS, particularly, even under its
more moderate governments, has hindered returns to a considerable degree, with
228 Florian Bieber
the clear intent of preventing a reversal of the population structure to a more di-
verse society where Serbs could lose political, socia,l and economic dominance.
In addition to political obstruction, the lack of economic opportunities and the
problem of refugees living in occupied property have further hindered return. As,
in 2000, Bosnia took in nearly 40,000 Serb refugees from Croatia and both Croa-
tia and Yugoslavia had over 220,000 refugees from Bosnia (UNHCR 2000b: 205),
the return process has a regional dynamic and is closely linked with the policy
towards return in the neighbouring states.
B Building Joint Institutions

A key endeavour of the Dayton Peace Accords has been the re-establishment of
joint institutions, which would maintain a degree of territorial integrity for Bos-
nia. The institutions established in the DPA are, as pointed out above, weak and
largely dependent on the entities. The slow implementation process further un-
dermined these institutions. The predominance of the three nationalist parties
in the post-war elections until November 2000 had meant that, at the level of the
joint institutions, an informal coalition of the three parties prevented the creation
of any viable institutions. In addition, the international community in Bosnia did
not focus on strengthening the institutions until 2000. The European Stability Ini-
tiative has pointed out that this delay was caused by the focus of the international
agencies on creating a sound security environment, which enforced cooperation
with existing power-structures at the expense of newly created institutions (Euro-
pean Stability Initiative 2000). Only three joint ministries (foreign affairs, foreign
trade and economic relations, and civil affairs and communication), the minimum
number, existed before 2000 and even these were seriously understaffed. The
Council of Ministers was weakened by a regular rotation (every eight months) of
the chairman of the council. Political representatives from the RS also refused to
upgrade the council to government, if only in name, as this would have implied an
additional quality of statehood vested in the joint institutions.
The elaborate veto-system of the joint institutions provided an additional
disincentive to effective decision-making. Although the ‘vital interest’ veto was
rarely invoked, the implicit or explicit threat of its usage effectively blocked deci-
sion-making. The rigidity of the veto rights accorded to the three communities in
the DPA has been problematic as it scarcely offers any mechanism for reaching
compromise. In the presidency, vetoes by its members can only be overridden if
two thirds of the national assembly of the RS or the respective Croat or Bosniak
delegates of the House of People of the Federation endorse the original deci-
sion vetoed by the presidency members (Constitution, Art V(2)(d)). As such the
mechanisms only provide for the over-riding of the veto if the presidency did not
‘represent’ the interests of the respective community adequately, while neither
‘national interest’ is defined, nor does a mechanism exist which seeks to medi-
ate between the different interests. The veto mechanisms in the parliamentary
assembly, on the other hand, do have a mediating component. In case of veto by
either community, a committee of three members from each community can be
constituted to find a compromise. If this committee does not, however, arrive at

a solution, the matter is then referred to the Constitutional Court (Constitution,
Art IV(3)(f )). Thus, there are only limited incentives and mechanisms for media-
tion, in contrast to the complex procedures in the current Constitutional Frame-
work in Kosovo, which does not grant an explicit veto to minorities but where the
different communities can delay parliamentary decisions.
In addition to institutional reasons for the weakness of the joint institutions,
the financial and fiscal dependency of the state on the entities exacerbated the
problems at the state level. With no substantial resources of its own and only spo-
radic payment by the entities, the joint institutions lacked the financial viability to
carry out the tasks accorded to them in the DPA. International financial support
was largely directed at the entities, further weakening the leverage of the joint
institutions (European Stability Initiative 2000).
A change to the weakness of the joint institutions came in a number of steps
occurring in late 1999 and 2000. In late 1999, the Constitutional Court declared
parts of the Law on the Council of Ministers of Bosnia and Herzegovina and the
Ministries of Bosnia Herzegovina unconstitutional. The law had further weak-
ened the council through the establishment of a vice-chair of the Council and
additional control through the presidency. The court’s decision emphasized the
role of the prime minister, i.e., the chairman of the Council of Ministers, thereby
enhancing the role of this joint institution (Constitutional Court 1999). The meet-
ing of the Peace Implementation Council in Brussels in May 2000 further shifted
the attention to the strengthening of the joint institutions:
Effectively functioning State institutions are a prerequisite for a modern European State
and for progress towards BiH’s entry into European and Euro-Atlantic structures. Ensur-
ing that BiH has such institutions remains a key strategic priority for the Council. Many
public institutions at all levels, but in particular State institutions, continue to fail the
citizens of BiH, due to lack of political will on the part of the ruling political parties and
the continued existence of parallel institutions. (Peace Implementation Council 2000).
The focus of the High Representative and the international community in institu-
tion-building has been the strengthening of existing institutions and the creation
of new institutions (ESI 2000).
An example for this has been an EU-funded project that seeks to strengthen
and professionalize the civil service at the state level. In addition to strengthening
existing institutions, the HR representative initiated the creation of a number of
new institutions, such as the establishment of a Bosnian State Court to super-
vise the implementation of state laws or the State Border Service. Additionally, a
number of regulatory agencies were created at the state level, such as the Com-
munications Regulatory Agency (CRA). Other bodies were created by the OHR
to supervise and advise on institutional reform in Bosnia, for example, the Inde-
pendent Judicial Commission (IJC). The activities on behalf of the international
community could not overcome the fact that the institutions themselves were
too weak to establish their supremacy over parallel power-structures and the en-
230 Florian Bieber
tities. The fact that the High Representative imposed most of these institutions
reflects the lack of consensus and the degree of obstruction, especially by most
Serb members of parliament, to enhancing the joint institutions.
The change of government at the level of joint institutions in January 2001 pro-
vided an opportunity for new authorities with a stronger commitment towards
the joint institutions. In some areas, the functioning of the institutions has been
enhanced. In fact, parliament and government worked more effectively in 2001
than before. Moreover, the limited mandate of chairmanship of the Council of
Ministers, which was limited to eight months, was dropped. Even after the na-
tionalist parties took again power in 2002, the institutions continued to operate
more effectively than earlier.
The fragility of the coalition government and resistance of Serb members of
the new joint government and parliament to any steps strengthening the state
institutions prevented any substantial steps towards a full implementation of the
DPA with regard to the joint Bosnian institutions. As is the case with the return
of refugees and all other aspects of the peace accords that have seen a full or
partial implementation, the process has been largely driven by the international
organizations and only marginally by local political actors.
C Human and Minority Rights

The constitution set forth in the DPA commits Bosnia to a comprehensive list
of international human rights agreements.24 The human rights situation on the
24 The constitution lists:

1. 1948 Convention on the Prevention and Punishment of the Crime of Geno-
cide;
2. 1949 Geneva Conventions I–IV on the Protection of the Victims of War, and the
1977 Geneva Protocols I–II thereto;
3. 1951 Convention relating to the Status of Refugees and the 1966 Protocol there-
to;
4. 1957 Convention on the Nationality of Married Women;
5. 1961 Convention on the Reduction of Statelessness;
6. 1965 International Convention on the Elimination of All Forms of Racial Dis-
crimination;
7. 1966 International Covenant on Civil and Political Rights and the 1966 and 1989
Optional Protocols thereto;
8. 1966 Covenant on Economic, Social and Cultural Rights;
9. 1979 Convention on the Elimination of All Forms of Discrimination against
Women;
10. 1984 Convention against Torture and Other Cruel, Inhuman or Degrading
Treatment or Punishment;
11. 1987 European Convention on the Prevention of Torture and Inhuman or De-
grading Treatment or Punishment;
12. 1989 Convention on the Rights of the Child;
13. 1990 International Convention on the Protection of the Rights of All Migrant
Workers and Members of Their Families;
ground, however, remains problematic in large parts of Bosnia. Human rights

violations against national minorities are frequent. In addition, the dominant na-
tionalist parties use their predominance against political opponents and the free
media (Helsinki Committee for Human Rights in BH 1999a). Cases of politically
motivated violence by authorities and individuals has declined in the past five
years, but remains a problem, especially in association with minority return. Fur-
ther serious human rights violations are connected with the rights of minorities,
ranging from political to religious rights (Department of State 2000).25 Further-
more, war crimes committed between 1992 and 1995 are only addressed slowly
and many major suspects remain at large, especially in the eastern RS (Helsinki
Committee for Human Rights in BH 1999b). Difficulties in implementing the Hu-
man Rights protection foreseen in the Dayton Accords are largely caused by lo-
cal authorities and the three nationalist parties (in varying degrees). The Human
Rights Ombudsman and the Human Rights Chamber are charged with investi-
gating human rights violations, as outlined in Annex 6 (Ch 2) of the DPA, and
have seen their work expand in the past years. The Human Rights Chamber, with
fourteen members (eight nominated by the Council of Ministers of the Council
of Europe, four by the Federation and two by the RS), hears cases of human rights
violations brought by individuals or organizations. Similarly, the Ombudsman,
who is a foreign national, takes on individual complaints and tries to investigate
them. The Human Rights Commission, a part of the Constitutional Court, took
over from the Human Rights Chamber after it was shut down. The massive chal-
lenge of addressing human rights concerns in Bosnia is visible when considering
that the Commission had a backlog of over 5,000 cases in January 2005.
Table 4 Cases at the Human Rights Chamber, 1996–1999

Year Number of applications Number of
received decisions issued
1996 31 0
1997 83 19
1998 1,387 67
1999 1,953 206
Total 3,449 292
Source: Human Rights Chamber 2000.
While these institutions have worked and addressed human rights concerns for
a number of years, they have been frequently overburdened with the number of
cases (Table 4). In addition, the central human rights problem of refugee returns
easily escapes these institutions, as the insecurity at the original place of residence
14. 1992 European Charter for Regional or Minority Languages;

15. 1994 Framework Convention for the Protection of National Minorities.
25 See also the regular Faxletter published by the Helsinki Committee for Human Rights
in BH, available at: http://www.bh-hchr.org/.
232 Florian Bieber
is only in part due to discrimination by local authorities and is very difficult to

grasp. Despite the heavy foreign involvement in these institutions, they contain
a strong Bosnian component and are rooted in the institutional infrastructure of
Bosnia, making them more likely to outlive the international presence than other
structures.
In addition to the aforementioned problems regarding the implementation of
human rights, the DPA is surprisingly weak in regard to minority rights. While
the Constitution obliges Bosnia to adhere to the European Charter for Regional
or Minority Languages and the Framework Convention for the Protection of Na-
tional Minorities, the international standards are inadequate for protecting mi-
norities in the mono-ethnic autonomies of the entities. Neither of the two entities
was obliged to grant minority rights to either members who belong to none of
the three constituent peoples or to citizens from the non-dominant nation. Thus,
while the right to return and other human rights were (inadequately) protected,
key group rights to counterbalance the ethnic autonomy of the entities – for ex-
ample, providing for education of minorities in the non-dominant history and
culture – are lacking. Arguably, some aspects of minority rights are less relevant
in Bosnia, as the three nations of Bosnia speak the same language, despite the
recognition of Croatian, Serbian, and Bosnian, only differing in the official use of
the alphabet.26
As a consequence of the absence of a comprehensive minority rights system,
the ethnic autonomy created through and in both entities has been particularly
problematic. Only the changes to the entity constitutions in 2002 and the efforts
at passing a minority law have recently initiated a process which can offset some
of the deficiencies of the original DPA.
D Education
During the war the educational system of Bosnia was degraded to a considerable
degree. Nevertheless, schools and universities continued to operate throughout
the war. Educational institutions were divided into three separate systems dur-
ing the war. The Croat and Serb educational systems were integrated with their
counterparts in Croatia and Yugoslavia, respectively, during the war. The Dayton
Peace Accords vests educational policy in the entities. While in the RS education
is governed at the level of the entity in a highly centralized fashion, in the Fed-
eration the cantons and their ministries administer education. In cantons with a
mixed population, education has even been partly devolved to the municipali-
ties. As a result, there are no institutional mechanisms for the re-integration of
the divided education system in Bosnia. The peace agreement did not seek to
re-integrate the education sector in Bosnia. This segmentation has been assessed
as being largely detrimental to the quality of the educational system in Bosnia,
as a Council of Europe report asserts: “As a result of Dayton’s legal mandate, the
26 While all official communication of the Republika Srpska is in Cyrillic, most print
media is produced in Latin script.
educational sector is dominated by politics” (Council of Europe 1999: 9). In ad-

dition to the segregation of the education system, the content of curricula and
textbooks has been a major problem. Textbooks made no or little reference to the
history or culture of the other nations in Bosnia and interpreted recent events,
such as the war in Bosnia, according to the political interpretation of the respec-
tive community. In the RS and in Croat-controlled areas of the Federation, most
textbooks were from Yugoslavia and Croatia, respectively (Low-Beer 2001: 215-
216; Donia 2000).
Attempts at some degree of coordination in the education sector were only
undertaken by the OHR in 2000 in light of the extension of the international
mandate for Bosnia, more than four years after the end of the war. In an agree-
ment brokered by the OHR in May 2000, the educational ministers of the en-
tities, including a Bosniak and a Croat representative from the Federation, the
entities committed themselves to coordinating education policy:
Education must no longer be used to divide and fragment the communities of Bosnia
and Herzegovina; on the contrary, it should be used to bring them together and live in
tolerance with one another. Any existing forms of segregation must be removed from
the parallel education systems in the Federation and Republika Srpska, and co-ordina-
tion assured in order to facilitate the return of refugee families throughout the whole
of Bosnia and Herzegovina. (Meeting of the Conference of the Ministers of Education
2000a).
The agreement concluded at the meeting committed the education ministers to

remove offensive materials from the curricula and textbooks and to emphasize
the common cultural heritage in future textbooks, including the usage of both
Latin and Cyrillic script. As a consequence, most textbooks used in Bosnia today
are authored in Bosnia, rather than imported from Croatia and Yugoslavia. The
ministers also created a ‘Curriculum Harmonization Board’ to bring the entities’
curricula in line with the goals of the agreement and also committed themselves
to including minority-specific education, mutual recognition of diplomas, and
country-wide education about Bosnia’s main religious communities (Meeting of
the Conference of the Ministers of Education 2000b).
While the agreements provided for a major improvement of the education sec-
tor, the system remains segregated, reflecting the high degree of ethnically-based
decentralization in the country. This segmentation extends to the university sys-
tem as well. In Mostar, two universities, one in the Croat and one in the Bosniak
part of the town, continue to exist after efforts to unify them failed (Bose 2002:
134-138).
Since these agreements have been signed, a number of additional reform ini-
tiatives have been undertaken by the OSCE and other international organizations
with the aim of harmonizing the curricula of the different education systems in
Bosnia. In Brčko, in particular, schools have been entirely reintegrated and stu-
dents only study key national subjects (history, language) in separate classes (Per-
ry 2003). Although the international community in Bosnia has been constrained
234 Florian Bieber
by the lack of a mandate in the DPA, the attention education has received in
recent years has allowed for a reform process to be initiated. Even if a full reinte-
gration of the educational system seems unlikely, a reduction of the excesses of
segregation and ‘hate-speech’ in textbooks has been achieved since 2000.
E Communications and Media

Most observers of the war in Bosnia agree that the media played a pivotal role in
inciting hatred against members of other nations and in spreading fear and stereo-
types (Thompson 1994). Throughout the war, the main media outlets were under
the control of the three nationalist parties, with only very limited activity by inde-
pendent media outlets during the war in Sarajevo. Despite the significance of the
media in Bosnia, it is not mentioned in the DPA. However, communication, in a
broader sense, is a competence vested in the central institutions of the state (DPA
1995: Art III/1h). Only with the establishment of the Independent Media Com-
mission (IMC) in 1998 was this competence addressed. This institution, which
was later transformed into the Communications Regulatory Agency (CRA), re-
named Regulatary Agency for Communication (RAK) in 2004, was charged with
supervising the electronic media (Babić 2001). The international supervision of
the electronic media has been instrumental in curbing hate speech in the media,
as media contravening the IMC’s regulations could be fined, suspended, or shut
down. The CRA has also been licensing the electronic media. This process, com-
pleted in 2001, has led to a consolidation of the media with a large number of
small, non-viable stations not obtaining permanent operating licenses.
The CRA has been identified as one of the more successful institutional experi-
ments in post-Dayton Bosnia. The agency is currently well-established and one
of the most effective Bosnian institutions at the state-wide level. It was created as
an internationally-run agency and gradually transformed into a Bosnian institu-
tion, both in terms of personnel and in its structural development, to eventually
become a local (and permanent) institution. As such, the RAK has been running
somewhat counter to the general trend of expanding international involvement
in the Bosnian process. However, since this experience, this model has been used
with other state agencies and institutions, such as the Civil Service Agency, the
State Border Service, and the High Judicial Councils.
F Economic Policies
Bosnia’s post-war development was shaped not only by the consequences of the
conflict, but also by the legacy of communism, especially in the sphere of eco-
nomics. Bosnia had been one of the least-developed regions of Yugoslavia, with
a significant agricultural sector. Despite the victory of anti-communist parties in
the first democratic elections in 1990, interethnic tensions prevented any steps
towards economic reform or privatization. During the war, the three nationalist
parties established a close symbiosis between the state and party, reminiscent of
the League of Communists’ relationship to the state. State-run enterprises pro-
vided funding for the parties, and the parallel power structures they maintained
offered employment in return for political loyalty.
Only with the establishment of the Central Bank in 1997 and the issuing of
the new currency, the Kovertibilna Marka (KM) in 1998 did the state assume its
responsibility in the area of monetary policy. The success of the currency, despite
some setbacks early on, managed to sideline the two other main currencies used
in Bosnia, the Croatian Kuna in Croat-controlled areas and the Yugoslav Dinar
in the RS.
Economic policy, except for foreign trade and monetary policy, are the sole
responsibility of the entities, with the competence lying with the cantons in the
Federation. As a result, economic policy was determined mostly by the three na-
tionalist parties in the areas of their influence (Donais 2002). While the economy
has been largely untouched by the DPA, especially in the areas that remained
under entity administration, some steps have been taken to reduce the overlap
of corruption and nationalist parties. The most significant step here was the clo-
sure of the Payment Bureaus, the remnants of communist administration that
served as clearing houses for all commercial transfers. These centralized institu-
tions were controlled by the nationalist parties and provided important sources
of income. Their closure, as proposed by the Peace Implementation Council, was
carried out by the High Representative (OHR 2000a). This step has been part of
a broader strategy of the international community to engage in a reform of the
economy, as little privatization has taken place and unemployment remains high
(Petritsch 2000). Since 2002, the economy received more extensive attention by
the High Representative, in particular in regard to removing administrative ob-
stacles to private business, such as the Bulldozer Initiative. However, the eco-
nomic situation in Bosnia has remained precarious due to high unemployment
and little foreign investment.
G Administration of Justice
Excluding the area of human rights protection and the institution of the constitu-
tional court, the DPA did not foreshadow any mechanisms for either the recon-
struction or reform of the judicial system or for its international monitoring. The
High Representative has become active in the field of judiciary reform, however,
as the rule of law, or rather the lack thereof, was identified as a major hurdle to the
success of democratization and interethnic reconciliation. The line of argument
for extending the High Representative’s mandate to the judiciary – making it de
facto part of the DPA – can be traced in the decision on the establishment of a
State Court for Bosnia:
In the exercise of the powers vested in me by Article V of Annex 10 … according to

which the High Representative is the final authority in theatre regarding interpretation
[of the DPA] …; and considering in particular Article II.1. (d) of the last said Agreement,
according to the terms of which the High Representative shall ‘Facilitate, as the High
Representative judges necessary, the resolution of any difficulties arising in connection
236 Florian Bieber
with civilian implementation’; Recalling paragraph XI.2 of the Conclusions of the Peace
Implementation Conference held in Bonn on 9 and 10 December 1997, in which the
Peace Implementation Council welcomed the High Representative’s intention to use his
final authority in theatre regarding interpretation of the Agreement on the Civilian Im-
plementation of the Peace Settlement in order to facilitate the resolution of any difficul-
ties as aforesaid ‘by making binding decisions, as he judges necessary’ on certain issues
including (under sub-paragraph (c) thereof ) ‘measures to ensure implementation of the
Peace Agreement throughout Bosnia and Herzegovina and its Entities’; Recalling fur-
ther paragraph 12.1 of the Declaration of the Peace Implementation Council which met
in Madrid on 15 and 16 December 1998, which made clear that the said Council consid-
ered that the establishment of the rule of law, in which all citizens had confidence, was a
prerequisite for a lasting peace, and for a self-sustaining economy capable of attracting
and retaining international and domestic investors (OHR 2000b, emphasis added).
This example demonstrates the capacity of the High Representative to become

active in any area that impacts on either the implementation of the DPA or on the
establishment of lasting peace in the country. The reform of the judiciary in Bos-
nia was placed under the supervision of the Independent Judicial Commission
(IJC), established by the OHR in November 2000 and closed in March 2004 after
completing its work. The commission, consisting of Bosnian and international
legal experts, reports to the High Representative, and monitors and intervenes
in the work of the judicial reform commissions of the entities and cantons. A
Memorandum of Understanding between the IJC and the Ministries of Justice
of both entities, signed in July 2001, spelled out specific procedures for hiring
judges and prosecutors to ensure both a professionalization of the legal system
and better representation of minorities (OHR 2001c). The IJC also monitors the
work of courts and prosecutors (OHR 2001a). Since then, the High Judicial and
Prosecutorial Council, established in 2002 (in 2004 the two entity councils were
merged) is responsible for the appointment of all judges (except for the Con-
stitutional Court) and prosecutors and may remove judges and prosecutors or
discipline them.
In addition, the High Representative has taken a number of other decisions
since 1999 aimed at reforming the judicial system. At the same time, the system is
still marred by lack of professionalism, lack of resources and national bias (Huit-
feldt 2000). Laws were imposed in both the Federation and the RS by the High
Representative to overcome delays in the reform process of the judiciary. These
decisions include amendments to the laws to further the independence of the
judiciary, especially in regard to the nomination of judges and prosecutors.
H External Relations and Trans-border Cooperation

One of the few domains in which the central Bosnian government is equipped
with powers is in the field of foreign relations (DPA 1995: Art III(1)). This compe-
tence has, however, been undermined by the fact that the two entities are allowed
to establish “special parallel relationships with neighboring states” (DPA 1995:
Art III(2)), thus allowing the Federation to form close links with Croatia, and
the RS to associate itself with the Federal Republic of Yugoslavia. While these
relations have to be consistent with the “sovereignty and territorial integrity” of
Bosnia, the formal acknowledgement of close links between the entities and the
respective ‘mother state’ of two nations of Bosnia further eroded the already lim-
ited powers of the central institutions. As result of the weakness of the joint in-
stitutions and the limited basis for an agreement on foreign policy, the country’s
external relations have been largely muted, both in the international and regional
contexts.
The Federation and Croatia concluded the Special Parallel Relations Agree-
ment (SPRA) in 1998, while a similar agreement between the RS and the Fed-
eral Republic of Yugoslavia (FRY) was not concluded until spring 2001. The latter
agreement was then vetoed by the Croat and Bosniak members of the Constitu-
tional Commission of the RS. This veto was, however, overridden by the HR, who
decided that the special relations agreement between RS and Yugoslavia did not
constitute a threat to the vital interests of the other communities (OHR 2001b).
The special relationships of the two entities with neighbouring states include in-
stitutions, such as in the case of RS-FRY relations, a Council for Cooperation and
a Standing Committee. That both agreements were concluded a number of years
after the end of the war highlights the fact that they constitute the formal tip of
the iceberg in relations between the entities and the mother states. Indeed, most
relations between either entities or Croat and Serb formal and informal insti-
tutions were conducted through non-formalized cooperation with Croatia and
Yugoslavia in the post-Dayton period. For example, only on 1 March 2002 did the
Yugoslav Army end its financial aid to the army of the RS, despite the fact that this
type of support formed neither part of the special relationship, nor was endorsed
by the DPA. Allowing for the creation of special parallel relations was not only a
compromise with the nationalist demands of the Croat and Serb side during the
Dayton negotiations. Establishing regulated relations with neighbours allows for
the possibility of greater international control than during the shadowy wartime
period. After the end of the Milošević and Tudjman regimes, respectively, the op-
portunity emerged for more substantial bilateral relations that were not necessar-
ily detrimental to the internal stability of Bosnia. Both Croatia and Serbia ceased
illicit funding to either entity or cantons, but political support among the political
elites for their respective community has remained problematic, in particular on
behalf of Serbia.
I Record of Implementation
The record of implementation has been mixed at best. While in some areas the
implementation process did not progress beyond the formal creation of the in-
stitutions or structures foreseen in the DPA, in other areas international involve-
ment in the peace process has extended far beyond the original text of the agree-
ment. This seemingly contradictory process is a reflection of the inadequacies
of the agreement. It is also an expression of the link between democratization
238 Florian Bieber
and conflict prevention. The original agreement, as has been observed, focused
on preventing a continuation or a near-immediate resumption of the conflict. It
lacked, however, tools to democratize the country and to reduce the degree of
institutionalized ethnic antagonism. Only the broad mandate of the High Rep-
resentative through the so-called ‘Bonn Powers’ and a broadened interpretation
of the agreement itself after 1997–1998 provided the tools to progress with the
implementation of the agreement and the tentative stabilization of the country.
The implementation process is thus nearly exclusively driven by the interna-
tional community. During the early years, there was no alternative to an enhanced
role for the international community, as domestic political actors either lacked
interest in the implementation process or sufficient support from the population.
This, however, involved the risk that the degree of intervention of the interna-
tional community, which has been termed as being a quasi-protectorate, created
a disincentive for local actors to agree on compromises and engage in a process
of power-sharing (Bieber 2002).
In addition, both the agreement and the international implementing agencies
largely ignored the issue of reconciliation and the construction of a democratic
political culture. Such processes still remain in their infancy and are mostly driv-
en by nongovernmental actors rather than by domestic or international organiza-
tions.
VII Current Status and Future Prospects
A Power-sharing Theory and the Case of Bosnia

After ten years of post-war reconstruction of Bosnia, it remains to be seen wheth-
er the arrangement that sought to resolve the overlapping self-determination
disputes will provide for the long-term viability of the state and its institutions.
There can be little doubt, however, that the Dayton Peace Agreement and its sub-
sequent implementation and development ushered in a new type of power-shar-
ing arrangement which extends beyond classical theories of power-sharing. Five
key aspects of multiethnic governance characterize post-Dayton Bosnia:
1. Power-sharing at the level of state-institutions;
2. A high degree of ethnic autonomy;
3. Power-sharing in sub-state units, down to the level of communes;
4. Substantial institutionalized and informal international intervention, me-
diation and arbitration; and
5. Weak provisions for minority rights.
Whereas the institutional origins of the settlement suggested a minimal degree of

interethnic cooperation in multinational institutions with most powers vested in
ethnically homogeneous institutions, the trajectory of the implementation pro-
cess suggests a more integrative approach towards governance and the extension
of power-sharing towards the entities through the inclusion of institutional rep-
resentation of the three communities in the Federation and the RS.
As described previously, the degree of power-sharing in post-war Bosnia de-

rived from the autonomy of the entities, whereas most power-sharing theory
considers autonomy to be a function of the power-sharing arrangement at the
centre (Lijphart 1977: 41-44; Horowitz 1985: 619-628). As Yash Ghai observes,
‘“[a]utonomy, particularly federal autonomy, is built around the notion that the
people of a state are best served through a balance between the common and
the particular … The secret of autonomy is the recognition of the common; cer-
tainly it seems to be the condition for its success”, (Ghai 2000: 24). In Bosnia, the
particular continues to prevail over the common, both in terms of the division
of powers and in terms of domestic political support. Whether the attempt to
offset this trend by strengthening the centre through integrative measures and
by promoting power-sharing in the entities also, as evidenced by the institutional
changes at the entity level in spring 2002, will be successful it is still too early to
judge.
Two key observations or even lessons can be drawn on the basis of the Bosnian
case in understanding internationally-mediated power-sharing arrangements.
First, a high degree of ethnic autonomy has to be balanced by comprehensive mi-
nority rights mechanisms to prevent ethnically exclusive governance in parts of
the country – even if a power-sharing arrangement exists at the centre. Second,
power-sharing arrangements have to be weary of excluding citizens without eth-
nic affiliation and smaller minorities. Post-conflict power-sharing arrangements,
such as the one in Bosnia, are seeking to prevent the recurrence of armed conflict,
frequently ignoring other groups and interests that have not been party to the
original conflict or dispute. While their inclusion might not be required to pre-
vent future conflict, the democratic quality of the arrangement suffers from the
exclusion of such constituencies.
B Freezing or Resolving Self-determination in Bosnia?

The general observation regarding the Bosnian experience of power-sharing and
post-war reconstruction applies to the country itself as well. While most observ-
ers of the peace process would agree that there has been some degree of im-
provement in terms of implementation and ‘normalization’ in recent years, most
structural problems remain acute. The acceptance of the Bosnian state by two of
its nations, namely Croats and Serbs, remains very low. In the first of half of 2003,
a majority of Croats and Serbs opposed the current structure of the state. Among
Serbs surveyed, 63.3% supported either the entity as an independent state or its
accession to Yugoslavia. 41.9% of Croats favoured the creation of a third Croat
entity (UNDP 2003: 27).
Despite the wide-spread opposition of significant parts of the population to
Bosnian statehood and its institutions, it would be simplistic to consider the
attempt at defusing the self-determination dispute in Bosnia through complex
power-sharing as a failure. As outlined in this chapter, some progress has been
made in recent years. Particularly during the term of the moderate government
between January 2001 and the general elections in October 2002, some of the
240 Florian Bieber
power-sharing institutions functioned relatively effectively. However, every step

to enhance the capacity of the joint state institutions, such as the creation of a
joint border service or the discussions on the establishment of a joint army or a
joint command, is met with fierce resistance by the dominant political forces of
Croats and Serbs. The political leadership of the RS particularly resists any trans-
fer of power to the joint level. But opposition to strengthening state institutions
is only partly based on ethnic politics, i.e., the fear of marginalization and/or the
project of ethnic self-government.
Although the level of violence has decreased significantly since the end of the
war, a number of incidents in 2001 confirm the danger of nationalist mobilization
by political parties and organizations that see their own or their community’s
interest threatened by political processes either aimed at dismantling parallel
power-structures or at undoing some of the effects of ethnic cleansing. The riots
in Banja Luka and Trebinje, as well as the organized resistance to the OHR and
SFOR takeover of Hercegovačka Banka, the main financial centre of the HDZ,
highlighted the ease with which violence can flare up. As a result 40.1% of the
Bosnian population – distributed relatively equally across the three nations – fear
the outbreak of a new war following the withdrawal of the international military
presence (UNDP 2001: 33). Despite these fears, the trajectory of the peace pro-
cess suggests that the gravest danger for Bosnia following the withdrawal of the
international military and civilian presence is not so much the renewed outbreak
of the war, but the breakdown of the political system and failure to implement
sensitive areas of the accords, such as the return of refugees. The Bosnian power-
sharing system owes its lack of viability only to some degree to the unwillingness
of the nationalist parties to engage in interethnic cooperation. The dysfunctional-
ity of the institutions and a dependency on international financial and political
intervention has seriously exacerbated the weakness of power-sharing arrange-
ments in Bosnia. While the reform of the current power-sharing arrangement has
been contemplated by both domestic and international actors, a comprehensive
reform is unlikely, as it would neither muster the necessary domestic support re-
quired for constitutional changes, as stipulated in the constitution (DPA 1995: Art
X), nor sufficient encouragement from the international actors to enter into re-
newed negotiations under international auspices, sometimes referred to as ‘Day-
ton 2’. As a consequence, the international presence seems likely to continue until
new political elites emerge (despite the current system) that might be more ready
to consider a thorough institutional reform. A key (or rather the only) incentive
for such a process might prove to be European integration (Nowak 1999: 285-
289), as integration into European structures is not disputed by any of the three
nations or their representatives.27 However, the possible integrative dynamic of
27 This was exemplified by the general support for the country’s accession to the Coun-
cil of Europe in spring 2002. The conditionality of membership in international or-
ganizations, such as the creation of an integrated system of defence for accession to
Partnership for Peace (PfP), has dampened support for integration into some inter-
national organizations, especially among the Bosnian Serb leadership.
European integration has not yet been seized upon in Bosnia. In fact, divisive
ethnic politics remain more frequently utilized by political elites than those that
could contribute to reducing the interethnic divide.
Chapter 8
Interim-governance for Kosovo: The Rambouillet
Agreement and the Constitutional Framework
Developed under UN Administration
Marc Weller
Kosovo represents a unique example of international administration. Its ini-

tial phase, internationally-directed mediation, backed by the threat of the use
of force, was meant to achieve a settlement for Kosovo that would balance the
continued territorial integrity of Yugoslavia (Serbia and Montenegro) with sub-
stantial self-government for Kosovo. A draft settlement was generated at the 1999
Rambouillet Meeting, aiming to square that particular circle. The draft was even-
tually accepted by the political leadership of the mainly ethnic Albanian parties
in Kosovo but it was rejected by Belgrade. Armed hostilities with NATO ensued.
These were terminated under a compromise formula, which foresaw a UN-led
interim administration that would bring the territory towards self-governance
until such time as a definite settlement might be achieved.
As this review will indicate, the Rambouillet text was highly complex indeed.
It featured a very elaborate layering of public authority, involving what was then
still known as the Federal Republic of Yugoslavia, one of its constituent repub-
lics (Serbia), Kosovo as an entity, local authorities, and even ethnic communities
endowed with their own legal identity and some institutions. After the armed
conflict, the international administration developed in three major phases. First,
there was the immediate post-conflict environment, where UNMIK, the Security
Council mandated UN presence in the territory, exercised full powers in relation
to governance. During a far longer, second phase, UNMIK devolved authority to
emerging Kosovo institutions and an attempt at co-governance was made. The
third phase, which is still ongoing at the time of this writing, was focused on gen-
erating the right conditions for a final settlement and transferring the remaining,
formerly reserved powers held by UNMIK to local institutions. Once a settle-
ment has been achieved – if one is achieved – a fourth phase of transition is to
be expected.
244 Marc Weller
I Background
Kosovo is a territory occupying an area of some 10,887 square kilometres. Ac-
cording to a 1990 census, it featured a population of roughly 2 million inhabit-
ants, of whom some 88% were estimated to be ethnic Albanians, 8% Serbs, and
the remaining number divided among small groups of Turks, Vlachs, Goranies,
Roma (Gypsies), and others. As a result of poverty and repression, during the
1990s between 300,000 and 400,000 ethnic Albanians left the territory, even be-
fore the hostilities and the dramatic Yugoslav campaign of forced displacement
began in earnest in 1998/1999. After the conclusion of hostilities involving NATO
in June 1999, over 100,000 ethnic Serbs left the territory; few of whom have re-
turned as of yet.
The Yugoslav Federal structure had been fundamentally revised in 1974, es-
tablishing a complex balancing of public powers among the republics and also
the two autonomous provinces that were contained within Serbia: Kosovo and
Vojvodina. The latter enjoyed a quasi-federal status. While their continued ap-
purtenance to the republic of Serbia was confirmed, at the same time the two
entities enjoyed all the substantive rights also granted to full republics under the
1974 SFRY constitution, including equal representation in the collective federal
presidency. Like the republics, the autonomous provinces were entitled to a very
wide measure of self-government, extending even to limited external affairs pow-
ers and the running of a central bank. The Tito design for Yugoslavia was there-
fore already one of power-sharing, combining self-government or autonomy of
constituent entities with joint representation in collective federal organs. This
was underpinned by a doctrine of ‘unity and brotherhood’, which was meant to
reflect the equality of nations, nationalities, and minorities.
The doctrine of nations and nationalities is one that has retained some rel-
evance in the post-communist transition, as it still informs the thinking of some
actors. Ethnic populations that do not possess a kin-state elsewhere (Serbia, Cro-
atia, Montenegro, Macedonia, Bosnia, Slovenia) were perceived to be nations.
They were entitled to full self-government through ‘their’ own republic and they
enjoyed an express right to self-determination, to be exercised through the re-
publics. Nationalities were other very large groups that did possess external kin
states and could therefore be accommodated through autonomy (the ethnic Al-
banians of Kosovo and the ethnic Hungarians of Vojvodina). Finally, the smaller
groups of Turks, Vlachs, Goranies, Roma, etc., were considered minorities – an
odious term within this system, as it implied virtual disenfranchisement.
The conflict in Kosovo progressed from a struggle for identity and control of
political power within an autonomous province as part of a federal structure to a
genuine self-determination conflict aiming towards secession. This transforma-
tion occurred in parallel with the attempts of Serbia, one of the six constituent
republics of the Socialist Federal Republic of Yugoslavia, to gain political domi-
nance within the Federation as a whole that commenced in the second half of the
1980s. While it still seemed possible to reform the Federation, Kosovo agitated
8 Interim-governance for Kosovo: The Rambouillet Agreement and the Constitutional Framework 245
for promotion to the status of a Republic. When the Federation dissolved in 1991,
Kosovo declared itself independent.
II Rambouillet
Between 1991 and 1995, Kosovo enjoyed a period of uneasy calm. The Kosovo
leadership had observed the violence of the conflicts that had raged in Croatia
and Bosnia and Herzegovina and was unwilling to take similar risks in provok-
ing armed action by the Yugoslav military and paramilitary forces. The situation
changed when the Bosnian issue was settled late in 1995 with the adoption of
the Dayton accords that followed upon the limited force by NATO in relation
to Bosnian Serb forces. It then became clear that the Kosovo issue would not be
addressed through similarly decisive measures. The moderate Kosovo leadership
under Ibrahim Rugova that had built up a peaceful parallel administrative struc-
ture in the territory was increasingly sidelined by the KLA, an armed opposition
movement that engaged in attacks against the Yugoslav infrastructure and was
also accused of having engaged in acts of terrorism. Over the years that followed,
these armed actions were answered by increasingly repressive operations by Bel-
grade. By 1998, these tactics appeared to resemble aspects of the conflict in Bos-
nia and Herzegovina, with increasingly large numbers of ethnic Albanians being
forcibly displaced and a humanitarian emergency looming.
From the summer of 1998 onwards, an attempt was made under the leadership
of US mediator Chris Hill to achieve a political settlement, or interim political
settlement, for Kosovo. A number of drafts were presented to the Yugoslav/Serb
authorities and to a team of ethnic Albanian delegates, nominated by the unof-
ficial government of Dr Ibrahim Rugova and headed by Dr Fehmi Agani. Yugo-
slavia took the view that the matter was one relating to the domestic jurisdiction
of the SFRY. If a settlement with international involvement could be achieved
at all, it would need to respect the territorial integrity of the SFRY. In fact, such
an arrangement could only be concluded at the lower level of relations between
Kosovo and the Serb republic (a principle maintained to the last and evidenced
by the fact that the Yugoslav delegation at the Rambouillet and Paris talks was
presented as a Serb delegation). While it would be possible to strengthen au-
tonomy arrangements, this would need to be balanced by special rights of minor-
ity populations. Moreover, all of the ethnic groups in Kosovo should be treated
principally equally, enjoying equal representation in political bodies.
The Kosovo delegation took the view that it could accept an interim arrange-
ment, so long as it did not prejudice its claim to self-determination and provided
there would be a process allowing for the ultimate option of independence after
the expiry of an agreed-upon interim period. While it was a strong advocate of
equal human rights for all groups, it objected to special regimes for minority
populations. Kosovo, in particular, was strongly opposed to the assignment of
equal political representation of all ethnic groups in the territory, which would
have given the 90% ethnic Albanians the same level of representation as each
of the minority groups. It argued that in this way the overwhelming majority of
246 Marc Weller
ethnic Albanians would only control one voice out of six or seven, turning the
majority into a minority in terms of representation.
The Hill drafts that were produced slightly oscillated over time on the issues
of the status of Kosovo. At one stage, the express reference to the right to ter-
ritorial integrity of the FRY was dropped. On the other hand, there was never an
affirmation of the right to self-determination for Kosovo. The question about the
timeframe for the ‘interim’ agreement was left unaddressed. While there was pro-
vision for a ‘comprehensive assessment’ of the agreement after three years, any
changes to the agreement would be subject to a veto of either party. Hence, the
situation would have been effectively frozen in time forever. There was also the
question of the precise nature of a mechanism to review the situation after the
three-year period. In the Hill draft, ‘the sides’ would undertake the review, thus
apparently removing the internationalized element of the negotiation process at
the review stage.
When the Holbrooke agreement collapsed in January 1999 in the wake of the
Racak massacre, the Contact Group, acting with the support of the UN Security
Council, summoned the parties to a conference (in fact, a ‘meeting’, to downplay
expectations) where they were expected to reach an interim settlement within a
period of two weeks at most.
A The Issue of Status and the Basic Distribution of Powers

Before the Rambouillet conference, the Contact Group had established a list-
ing of so-called non-negotiable principles. These reflected the FRY/Serb demand
that the territorial integrity of the FRY and its neighbouring countries had to be
respected. The government of Kosovo had responded that it would be willing to
attend the Rambouillet talks, provided a proposed settlement would not preju-
dice the status of Kosovo, in accordance with the approach that it had adopted in
the initial Hill proposals. The first draft of the agreement presented to the Con-
ference was in accordance with this idea of leaving out express statements con-
cerning issues on which no agreement could be achieved. It contained, in what
started out as the Draft Framework Agreement, merely a preambular paragraph
that recalled the commitment of the international community to the sovereignty
and territorial integrity of the FRY. Hence, in signing this text Kosovo itself would
not have had to take an express view on this matter. The Agreement also made
reference to United Nations and OSCE principles. As these contain both the rule
of territorial unity and the principle of self-determination, this reference was also
acceptable to Kosovo, while it provided reassurance to the FRY/Serbia. The Con-
stitution (initially Annex 1) did not contain a preamble and, instead of addressing
the status of Kosovo and the legal quality of its relations with the FRY or even
Serbia, it focused on a reasonable division of competences.
When the second draft of the agreement was presented on 18 February, a pre-
amble had been added which referred to “democratic self-government in Kosovo
grounded in respect for the territorial integrity and sovereignty of the Federal
Republic of Yugoslavia, from which the authorities of governance set forth herein
originate.” While the Contact Group attempted to assert that this was an insub-
stantial addition, it did in fact fundamentally change the nature of the entire in-
terim settlement. To avoid a failure of the Rambouillet process on account of
this unilateral change made without explanation two days before the deadline for
signature, the following compromise was in the end adopted:
Desiring through this interim Constitution to establish institutions of democratic self-

government in Kosovo grounded in respect for the territorial integrity and sovereignty
of the Federal Republic of Yugoslavia and from this Agreement, from which authorities
of governance set forth herein originate.
Even as amended, this provision almost led the Kosovo delegation to reject the
agreement. Still, the formulation permitted Kosovo to argue that acceptance of
the territorial integrity and sovereignty was limited to the interim period. In any
event, this commitment and the legal personality and powers of Kosovo were
rooted in the Agreement, rather than in a grant of autonomy by the FRY.
Another important change that had been made in the draft of 18 February re-
lated to the assignment of powers. According to Article I(2) of the initial draft of
what then was Annex 1, Kosovo as an entity would enjoy responsibility for “all
areas” other than those where authority was expressly assigned to the FRY. Those
areas of authority were enumerated exclusively and their exercise was subjected
to important restrictions and safeguards for Kosovo. In the draft of 18 February,
the express presumption in favour of Kosovar authority had been abandoned.
Instead, a new paragraph had been added, indicating that Serbia, too, would ex-
ercise competence in relation to Kosovo as specified in the Agreement.
The FRY/Serbia strongly insisted on the inclusion of a further provision in the
introductory section of the Agreement (formerly the Framework Agreement),
stating that the parties would only act within their powers and responsibilities in
Kosovo as specified by this Agreement. Acts outside those powers and respon-
sibilities would be null and void. Kosovo would have all rights and powers set
forth in the Agreement, in particular as specified in the Constitution. While the
FRY/Serb delegation might have intended this provision as a safeguard against
creeping jurisdiction by the Kosovo organs, the delegation of Kosovo interpreted
it as a helpful confirmation that the powers of Kosovo were indeed based in the
Agreement, and not in a sovereign grant of rights by the FRY. In addition, this
formulation supported the view that FRY/Serb exercises of powers in relation to
Kosovo would be strictly limited to competences that have been expressly grant-
ed to them.
A proposal to entitle Chapter 1 “Interim Constitution” was not adopted. However,

given the overall title of the Agreement, the specific reference to the interim period
in this preambular provision, and the concluding provisions, this was not seen by the
Kosovo delegation as a significant setback.
Federal functions were still expressly, and in the view of Kosovo, exhaustively listed:
territorial integrity, maintaining a common market within the Federal Republic of
248 Marc Weller
The issue of legal personality for Kosovo as a whole was also clarified in some
measure through a provision concerning the communes. One of the difficulties
with the initial Hill proposal was its insistence that Kosovo communes be the
basic unit of self-governance in Kosovo and that they exercise all authority not
assigned to Kosovo and not assigned to other Kosovo organs. While the latter ele-
ment was retained, it was clarified in what was to become Article I(8) of Chapter
1 of the final text that the communes were merely the basic unit of local self-gov-
ernment. The insertion of the word ‘local’ ensured that this provision no longer
diluted the overall legal personality of Kosovo as a whole.
B The Kosovo Institutions and Their Powers

The principal organs of Kosovo according to the agreement were the Assembly,
the President of Kosovo, the Government and Administrative Organs, judicial
organs, and the communes. The Assembly was to be composed of 120 Members,
of which 80 would be directly elected; the other 40 members would be elected by
representatives of qualifying national communities. Communities whose mem-
bers constitute more than 0.5% of the Kosovo population but less than 5% were
to divide ten of these seats amongst themselves. Communities whose members
constituted more than 5% of the Kosovo population (in fact, only the ethnic Alba-
nians and Serb) would divide the remaining thirty seats equally.
The draft of 18 February had introduced as an additional feature a second cham-
ber of the Assembly as a result of FRY/Serb pressure. In that Chamber of 100
seats, the Turks, Goranies, Romanies, Egyptians, Muslims and any other group
constituting more than 0.5 per cent of the population (hence also including the
ethnic Albanians and Serbs) would be equally represented. The Chamber would
have had the right of consultation in relation to legislative acts of the Assembly,
and any of the groups represented within it could have initiated so-called vital
interest motions, which would amount to an attempted veto of legislation.
The FRY/Serb delegation itself abandoned the concept of the second chamber
and instead focused on attempting to strengthen the veto power of national com-
munities in the Assembly. According to the initial draft and the subsequent draft
of 18 February, the decision as to which legislative acts would violate the vital
interests of a national community, and would thus be null and void, would have
been taken by the Constitutional Court of Kosovo. The Kosovo delegation had
grave reservations about the very concept of special powers for ethnic groups, in-
cluding separate elections according to ethnic criteria. While strongly endorsing
the notion of equal rights for members of all ethnic groups, the idea of separate
representation appeared to grant to very small groups broad rights of co-decision,
Yugoslavia, monetary policy, defence, foreign policy, customs services, federal taxa-
tion, federal elections, and other areas specified in this agreement.
See also Article VIII(5) of Chapter 1, and the important reference contained therein
to Article II(5)(b) of Chapter 1.
On the national communities and their institutions, see the following section.
which were unrepresentative and hence undemocratic. Moreover, the example of

ethnic politics in Bosnia and Herzegovina had demonstrated the divisive nature
of a political system organized along ethnic lines. Finally, a legislative system that
was subjected to the constant threat of veto by any ethnic group would result in
perennial paralysis.
Despite these concerns, and in view of the strong position of the FRY/Serbia
in this matter, the Kosovo delegation endorsed, albeit reluctantly, the concept of
special representation for ethnic groups for the interim period. However, this
concession was dependent on a judicial process to check vital national interest
motions in the Assembly in order to avoid an arbitrary use of this procedure lead-
ing to constant deadlock in the legislature. The Contact Group disregarded this
view and gave way to a Serb/FRY proposal of settling disputes about vital national
interest motions outside of the judicial system. According to Article II(8)(c) of
Chapter 1, the final agreement provided that the decision about such motions
would be rendered by a panel comprising three members of the Assembly: one
Albanian and one Serb – each appointed by his or her national community – and
a third Member of a third “nationality” to be selected within two days by consen-
sus of the Presidency of the Assembly. As the Serb national community was guar-
anteed a member of the Presidency of the Assembly, it appears that this nominat-
ing process itself could be blocked by a factual veto.
Decisions of the Assembly that had been challenged according to the vital na-
tional interest procedure were to be suspended in response to the national com-
munity having brought the challenge, pending completion of the dispute settle-
ment procedure. Hence, it might appear as if a Serb veto in relation to Assembly
decisions was introduced through the back door. However, as this veto would de-
pend on an abuse of process in frustrating the nomination of the third member of
the arbitration panel, the general dispute settlement mechanism attaching to the
agreement as a whole, or the general powers of the Constitutional Court, would
probably be brought to bear on a matter of this kind. It should also be noted that
the decision on the merits of a vital interest motion, while conduced by a political
body, was to be made according to legal criteria.
The substantive powers of the Assembly were reasonably wide, covering most
aspects of governance. Importantly, this included the power to set the framework
of, and to coordinate, the exercise of competences assigned to the communes.
The first elections in Kosovo were to be held within nine months of the entry into
force of the Agreement under international supervision.
The President of Kosovo was to be elected by the Assembly by a majority vote.
His or her functions included representation before international, FRY, or re-
public bodies, the conduct of foreign relations consistent with the authorities
of Kosovo institutions, proposing candidates to the Assembly for Prime Minis-
ter and for the principal courts of Kosovo, etc. The government, also to be ap-
A vital interest motion shall be upheld if the legislation challenged adversely affects
the community’s fundamental rights as set forth in Article VII, or the principle of fair
treatment.
250 Marc Weller
proved by the Assembly, would have general authority for implementing the laws
of Kosovo, etc. At least one minister would have been a member of the Serb
national community.
The powers reserved for the communes had been narrowed down somewhat
in comparison with earlier drafts. However, while there was provision for coor-
dination on a Kosovo-wide basis, the police was to be organized on a communal
basis and limited to a ceiling of 3,000 active law-enforcement officers throughout
Kosovo. In addition, there was authority in relation to education, child care, the
communal environment, local economic issues, etc.
The judiciary consisted of a Constitutional Court composed of nine judges. At
least one judge would have been a member of the Serb national community and
five other judges would have been selected from a list drawn up by the President
of the European Court of Human Rights. The powers of review of the Constitu-
tional Court were quite wide. They included, but were not limited to, determin-
ing whether laws applicable in Kosovo, decisions or acts of the president, the
Assembly, the government, the communes, and the national communities were
compatible with the Constitution.
The Supreme Court, composed of nine judges, including one member of the
Serb national community, would hear final appeals from subordinated courts in
Kosovo, including communal courts.
A special feature related to the rights of citizens in Kosovo is the option to
have civil disputes partly adjudicated by other courts in the FRY, which would
apply laws applicable in Kosovo. In criminal cases, a defendant would be entitled
to have a trial transferred to another Kosovar court designated by him or her. In
effect, this meant that a defendant could opt to be tried in the local court of a spe-
cific commune, which would be principally composed of members of his or her
ethnic appurtenance. In criminal cases in which all defendants and victims were
members of the same national community, all members of the judicial council
would be from the national community of their choice if any party so requested.
A defendant in a criminal case could also insist that one member of the judicial
council hearing the case be from his or her national community. This might in-
clude judges of courts in the FRY serving as Kosovo judges for these purposes.
C Human Rights and Additional Rights of National Communities

The provisions on human rights were strangely short and undeveloped in the
Constitution and throughout the Agreement. There was no listing of fundamen-
tal human rights to be applied in Kosovo. Instead, Article VI(1) of the Constitu-
tion stated rather generally that all authorities in Kosovo must ensure interna-
tionally recognized human rights and fundamental freedoms. As opposed to the
Dayton agreement, which included a long list of human rights instruments iden-
tifying what internationally recognized human rights and fundamental freedoms
are, the agreement in Article VI(2) rather ingeniously incorporated by reference
the terms of the European Convention for the Protection of Human Rights and
Fundamental Freedoms and its Protocols, which “shall apply directly in Kosovo.”
In this way, a very sophisticated body of human rights law, refined in decades of
jurisprudence by the European Court and Commission of Human Rights, would
have been instantly available in Kosovo. The Kosovo Assembly also had the pow-
er to enact other internationally recognized human rights instruments into law.
The rights and freedoms established in this way would have priority over all
other law. Interestingly, all “courts, agencies, governmental institutions, and oth-
er public institutions of Kosovo or operating in relation to Kosovo shall conform
to these human rights and fundamental freedoms” (Article VI(3)). This means
that FRY and republic authorities would also have had to exercise their compe-
tences in relation to Kosovo and in accordance with these standards. As the FRY
was not a party to the European Convention and its Protocols, this would have
placed it in an unusual position.
While the human rights provisions were compact, the additional rights grant-
ed to national communities were extensive, but not unlimited. Firstly, these rights
were tied to the specific purpose of preserving and expressing their respective
national, cultural, religious, and linguistic identities. This was to be done in ac-
cordance with international standards and in accordance with human rights and
fundamental freedoms.
More controversially, each national community could elect and establish its
own institutions – a feature which was feared would give rise to a parallel state
structure within Kosovo. However, national community institutions would have
to act in accordance with Kosovar law and not take discriminatory action. Na-
tional communities could arrange for the inscription of the local names of towns
and villages, etc., in the language and alphabet of the respective community; issue
information in that language; provide for education and schooling in that lan-
guage and in national culture and history, reflecting a spirit of tolerance between
communities and respect for the rights of members of all national communities;
display national symbols, including those of the FRY and Serbia; protect national
traditions on family law; arrange for the preservation of sites of religious, his-
torical, or cultural importance in cooperation with other authorities; implement
public health and social services on a non-discriminatory basis; operate religious
institutions in cooperation with religious authorities; and participate in nongov-
ernmental organizations.
National communities could also enjoy unhindered contacts with representa-
tives of their respective national communities within the FRY and abroad. They
must be guaranteed access to and representation in the media and may finance
their activities by collecting contributions from their members. Importantly, ev-
ery person has the right freely to choose to be treated or not to be treated as
belonging to a national community.
D Final Status
The draft presented to the parties at the outset of the Conference restated the
concluding provision from previous Hill proposals providing for amendments to
the Agreement to be adopted by consensus of all the parties. Each party was to
252 Marc Weller
be entitled to propose such amendments at any time. However, after three years,
a comprehensive assessment of the Agreement would occur under international
auspices with the aim of improving its implementation and determining whether
or not to implement proposals by either side for additional steps. The means of
undertaking this assessment and the procedure to be adopted were left unclear.
The Kosovo delegation argued strongly that, in accordance with the interim
character of the agreement, a provision would need to be made for a further
international conference on a final settlement for Kosovo. The decisions of that
conference should be based on the will of the people of Kosovo, made manifest in
a referendum. The negotiators pointed out that they were not authorized by the
Contact Group to adopt language on a referendum. However, even the non-ne-
gotiable principles had at least provided for “a mechanism for a final settlement
after an interim period of three years.” In the dramatic final phase of the confer-
ence, it became possible to obtain significant changes to the final provision, in
reflection of this wording. The final text of what became Article I(3) or Chapter
8 reads:
Three years after the entry into force of this agreement, an international meeting shall
be convened to determine a mechanism for a final settlement for Kosovo, on the basis
of the will of the people, opinions of relevant authorities, each Party’s efforts regarding
the implementation of this Agreement, and the Helsinki Final Act, and to undertake a
comprehensive assessment of the implementation of this Agreement and to consider
proposals by any Party for additional measures.
This wording stops short of actually establishing a mechanism for a final settle-
ment, contrary to what might have been expected in view of the language con-
tained in the non-negotiable principles. However, startlingly, it was accepted that
this mechanism shall be established and/or operate, inter alia, on the basis of the
will of the people. On the other hand, this mechanism would be created by “an
international meeting,” the composition, remit and authority of which was not
defined. All organs of the international community had consistently ruled out the
possibility of independence and it would not be surprising if the “international
meeting” might hold a similar view. These uncertainties were not necessarily en-
tirely removed by the unilateral interpretation offered by the US in a draft side-
letter which confirmed that Kosovo could hold a referendum on independence
This formulation actually represented a slight retreat from the final Hill draft, put
forward on 27 January, which had referred to a ‘procedure’ for considering such addi-
tion steps to be determined taking into account the parties’ roles in and compliance
with this agreement.
In a dramatic night negotiating session towards the very end of the conference, the
Kosovo delegation was able to extract from the negotiators an even better formula-
tion, referring to the ‘expressed will of the people’. However, this concession was lost
when the delegation was not immediately able to sign the final text and when the
Contact Group overruled it afterwards.
after three years. In any event, that side-letter was in the end not formalized,
given the failure of the Kosovo delegation to sign the agreement by the stipulated
deadline at Rambouillet.
E Implementation
The provisions for implementation contained in the Rambouillet text were com-
plex and distributed throughout the interim agreement. They consisted princi-
pally of the introductory section of the Agreement entitled Framework; Chapter
2 on Police and Public Security; Chapter 3 on the Conduct and Supervision of
Elections; Chapter 4(a) on Humanitarian Assistance, Reconstruction and Devel-
opment; Chapter 5 on the Civilian Implementation Mission in Kosovo; Chapter 6
concerning the Ombudsman; and Chapter 7 on ‘Implementation II’, that is to say,
military implementation. For reasons of space, it will only be possible to review
some of the principle features of this implementation structure here.
F Confidence-building
Upon signature of the agreement, a ceasefire was to come into force immediately.
Alleged violations of the ceasefire were to be reported to international observers
and could not be used to justify use of force in response. The status of police and
security forces in Kosovo, including withdrawal of forces, was to be achieved ac-
cording to Chapter 7. Paramilitary and irregular forces in Kosovo were deemed
incompatible with the terms of the agreement. The latter provision gave rise to
some difficulty, inasmuch as the KLA did not consider itself a paramilitary or ir-
regular force. However, it was clear that it, too, was addressed through Chapter
7 of the agreement.
All abducted persons or other persons detained without charge were to be
released, including persons held in connection with the conflict in Kosovo. No
one was to be prosecuted for crimes related to the conflict, except for persons ac-
cused of having committed serious violations of international humanitarian law.
Persons already convicted for committing politically motivated crimes related to
the conflict were to be released, provided these convictions did not relate to seri-
ous violations of humanitarian law obtained in a fair and open trial conducted
pursuant to international standards.
The agreement confirmed the obligation, already contained in mandatory Se-
curity Council resolutions, to cooperate with the Hague International Criminal
Tribunal for the Former Yugoslavia. This included the obligation to permit com-
plete access to tribunal investigators and compliance with the orders of the Tri-
bunal. This provision was somewhat contested at Rambouillet. Kosovo attempted
to strengthen its scope, as did, indirectly, the Tribunal itself. However, in the face
of determined opposition from the FRY, a rather short paragraph was adopted,
which did not greatly improve upon the obligations already contained in the de-
mands of the Security Council.
254 Marc Weller
The parties also recognized the right to the return of all persons to their homes,
including those who have had to leave the region. There was to be no impediment
to the normal flow of goods into Kosovo, including materials for the reconstruc-
tion of homes and structures. The FRY would not require visas, customs, or li-
censing for persons or things connected with international implementation.
G NATO-led Implementation and the Withdrawal of Forces

The military implementation chapter was the most detailed element of the entire
accords. The parties would have agreed that NATO would establish and deploy a
force (KFOR) operating under the authority and subject to the direction and the
political control of the North Atlantic Council (NAC) through the NATO chain
of command. However, contrary to much speculation afterwards, a Chapter VII
mandate was to be obtained from the Security Council from the beginning. Oth-
er states would be invited to assist in military implementation. While this is not
spelt out, it was envisaged that KFOR would be of a strength of approximately
28,000 troops. This would include a sizeable Russian contingent, according to the
precedent set by SFOR/IFOR in Bosnia and Herzegovina. In accordance with that
precedent, KFOR would be authorized to take such actions as required, includ-
ing the use of necessary force, to ensure compliance with Chapter 7. As opposed
to the arrangements of Dayton, it was made clear at the beginning that KFOR
would not only be available to ensure compliance with the military aspects of
the agreement, but that it would also actively support civilian implementation
by the OSCE and others as part of its original mandate. As in the Dayton agree-
ment, the mandate of KFOR could have been broadened through further action
by NATO – in this instance acting through the North Atlantic Council.
A Joint Implementation Commission would have been established to consider
complaints by the parties and other matters. It would have been composed of
FRY military commanders and FRY/Serb officials, Kosovo representatives, and
representatives of the military and civilian implementation missions. The agree-
ment envisaged that final authority to interpret the provisions of Chapter 7 would
rest with the KFOR military commander.
The regular armed forces of the FRY (VJ) would have been subjected to a
rigourous regime of redeployment and withdrawal according to fixed deadlines.
This included the removal of assets such as battle tanks, all armoured vehicles
mounting weapons greater than 12.7mm and all heavy weapons of over 82mm.
Within 180 days of the coming into force of the agreement, all VJ units, other
than 1,500 members of a lightly-armed border guard battalion deployed close to
the border would have had to be withdrawn from Kosovo. An additional 1,000
support personnel would be permitted in specified cantonment sites. The border
guards would have been limited to patrolling the border zone and their travel
through Kosovo would have been subjected to significant restrictions. Moreover,
the air defence system in Kosovo would be dismantled and associated forces
withdrawn, as would other FRY or Serb forces, including the Ministry of Interior
Police (MUP). The MUP would initially be drawn down to a size of 2,500 and be
entirely withdrawn upon the establishment of a Kosovo police force within one
year. Upon entry into force of the Agreement, all other forces would have had to
commit themselves to demilitarization, renunciation of violence, to guaranteed
security of international personnel, etc. The definition of the term ‘demilitariza-
tion’ has been subject to some discussion, especially as it applies to the KLA. It
would have included the surrender of heavy armaments and some small arms.
The military chapter was accompanied by two appendices, which were both
published along with the agreement. The rather extravagant claims that they re-
veal a secret agenda by NATO for the virtual occupation of all of Yugoslavia made
some time after the conclusion of the conference are entirely without substance.
Appendix A established cantonment sites for FRY/Serb forces. The famous Ap-
pendix B established what in other contexts would be the standard terms of a
status of forces agreement for KFOR, very much in line with the precedent of
IFOR/SFOR in Bosnia and Herzegovina and United Nations peacekeeping opera-
tions. A provision that permits transit through Yugoslavia for NATO and affili-
ated forces falls within these standard terms, although it may have been phrased
slightly more broadly than would have been usual. If, as was subsequently
claimed, the terms of the Appendix, or this particular provision, were the reasons
that rendered Rambouillet unacceptable, it remains to be explained why the FRY
did not seek clarification or even modification of this provision at the Paris fol-
low-up conference. That conference was exclusively dedicated to negotiations on
the implementation aspects of the agreement, which had not been available at
Rambouillet itself.
H Civilian Implementation
The OSCE would have been charged with principal responsibility over the civil-
ian elements of implementation, operating under a Chief of the Implementation
Mission (CIM). The implementation mission would monitor, observe, and in-
spect law enforcement activities in Kosovo, which would be established princi-
pally at communal level. The Kosovo police force of around 3,000 was to be only
lightly armed. The authority of federal and Serb police would have been very
significantly restricted. Importantly, the CIM would have had final authority to
interpret the provisions of the agreement in relation to civilian implementation.
All aspects of civilian implementation would have been coordinated and moni-
tored by a Joint Commission, including Federal, Republic, and Kosovo represen-
tatives and others, and chaired by the CIM exercising a final right of decision in
this rather powerful body. In addition, an ombudsman would monitor the re-
alization of the rights of members of national communities and the protection
of human rights and fundamental freedoms. Elections were to be held at com-
munal and Kosovo level within a period of nine months from entry into force of
the agreement, once the OSCE certified that conditions had been established for
a free and fair ballot. Finally, the agreement provided for the administration of
humanitarian aid and reconstruction, principally through the organs of Kosovo,
with strong involvement by the European Union. In fact, throughout the confer-
256 Marc Weller
ence, the European Union, through its negotiator and through representatives of
the Commission, exercised considerable influence, especially in relation to this
issue. Great emphasis was placed on careful planning, rapid and unbureaucratic
deployment of resources once needs had been identified, and close cooperation
with the beneficiaries of such aid.
III Resolution 1244 (1999), UN Administration and the

Constitutional Framework
A The Remit of Resolution 1244 (1999)

Security Council Resolution 1244 (1999) was adopted on 10 June 1999 by fourteen
votes to none (China abstaining). The entire text of the resolution is subject to
Chapter VII of the United Nations Charter. Preambular paragraph 10 reaffirms
the commitment of all UN member states to the sovereignty and territorial in-
tegrity of the Federal Republic of Yugoslavia and the other states of the region.
In its very first operative paragraph, it addresses the issue of a political settle-
ment, “deciding” that a political solution to the Kosovo crisis shall be based on
general principles established in two annexes to the resolution. In addition, the
Council authorized the deployment under UN auspices of “international civil and
security presences.” The civil presence was to be “controlled” by a Special Repre-
sentative of the UN Secretary-General. While the military presence established
by “member states and relevant international organizations” was to coordinate
closely with the civil presence, it is not subject to control by the Secretariat. It
enjoys a mandate to enforce the ceasefire, deter hostilities, ensure withdrawals,
and prevent the return of certain forces. It is also empowered to demilitarize the
KLA; ensure public safety and order in the interim; conduct border monitoring;
ensure freedom of movement for itself, the civil presence, and other international
organizations; and support, as appropriate, the civil presence. The civil presence
is charged with:
a. Promoting the establishment, pending a final settlement, of substantial au-
tonomy and self-government in Kosovo, taking full account of annex 2 of the
Rambouillet accords (S/1999/648);
b. Performing basic civilian administrative functions where and as long as re-
quired;
c. Organizing and overseeing the development of provisional institutions for
democratic and autonomous self-government pending a political settle-
ment, including the holding of elections;
d. Transferring, as these institutions are established, its administrative respon-
sibilities while overseeing and supporting the consolidation of Kosovo’s local
provisional institutions and other peace-building activities;
e. Facilitating a political process designed to determine Kosovo’s future status,
taking into account the Rambouillet accords (S/1999/648);
f. In the final state, overseeing the transfer of authority from Kosovo’s provi-
sional institutions to institutions established under a political settlement;
g. Supporting the reconstruction of key infrastructure and other reconstruc-

tion;
h. Supporting, in coordination with international humanitarian organizations,
humanitarian and disaster relief aid;
i. Maintaining civil law and order, including establishing local police forces,
and, meanwhile, through the deployment of international police personnel
to serve in Kosovo;
j. Protecting and promoting human rights; and
k. Assuring the safe and unimpeded return of all refugees and displaced per-
sons to their homes in Kosovo.
Annex I, to which reference is made, consists of the Petersberg G-8 declaration

of 6 May 1999, which effectively contained the peace terms that Yugoslavia was
invited to accept before a termination of hostilities with NATO could take place.
That document refers to an interim administration of Kosovo to be decided by
the Security Council and “a political process towards the establishment of an in-
terim political framework agreement providing for a substantial self-government
for Kosovo, taking full account of the Rambouillet accords and the principles of
sovereignty and territorial integrity of the Federal Republic of Yugoslavia and the
other countries of the region and the demilitarization of the KLA.”
Annex 2 consists of a ten point statement presented to the FRY on 2 June 1999
and expressly accepted by it as part of the cease-fire negotiations. It provides
for:
…
5. Establishment of an interim administration for Kosovo as part of the international
civil presence under which the people of Kosovo can enjoy substantial autonomy
within the Federal Republic of Yugoslavia, to be decided by the Security Council of
the United Nations. The interim administration is to provide transitional admin-
istration while establishing and overseeing the development of provisional demo-
cratic self-governing institutions to ensure conditions for a peaceful and normal
life for all inhabitants in Kosovo.
…
8. A political process towards the establishment of an interim political framework
agreement providing for substantial self-government for Kosovo, taking full ac-
count of the Rambouillet accords and the principles of sovereignty and territo-
rial integrity of the Federal Republic of Yugoslavia and the other countries of the
region, and the demilitarization of the UCK. Negotiations between the parties for
a settlement should not delay or disrupt the establishment of democratic self-gov-
erning institutions.
B Interim Administrative Practices

While there was initially some confusion as to the precise extent of the mandate
and to the progression of events foreseen in the three different instruments to
258 Marc Weller
which reference has been made, in practice, a fairly smooth development has
occurred. During the first phase of governing under UNSRSG Bernard Kouch-
ner, participation of local actors was limited to a consultative role. In a second
phase, provision was made for limited co-decision, when, early in 2000, the UN
administration established a Joint Interim Administrative Structure (JIAS). This
consisted of the establishment of some twenty departments with responsibility
for civil administration. These departments were jointly headed by one local and
one international official. This was matched by the holding of municipal elections
and the establishment of local structures of self-government at the end of 2000.
A third phase provided for conditional self-government, to be realized under the
terms of the constitutional framework that was put forward in 2001.
C Constitutional Framework
1 Origin and Establishment
The Constitutional Framework for Provisional Self-Government was issued on
15 May as UNMIK regulation 2001/9. That is to say, its legal force is rooted in
Regulation 1999/1 (subsequently amended), which vests supreme executive pow-
ers and also quasi-legislative powers in the SRSG. This power is, in turn, derived
from the mandate contained in Security Council Resolution 1244 (1999). The
drafting of the document had caused some friction among the Western members
of the Contact Group (the quint) throughout 2000. The United States attempted
to present a draft instrument that was very much based on Rambouillet, although
the powers for Serbia/FRY had been significantly reduced in view of the recent
conflict and its outcome. Others argued that it would be premature to offer any
sort of consolidated interim constitution, which would accelerate the process of
claims for statehood on the part of the Kosovo majority parties. Instead, indi-
vidual regulations that would, over time and in stages, address issues normally
contained in a constitutional document should be issued.
The debate took a different turn with the appointment of a new SRSG, former
Danish Minister of Defence Hans Haekkerup. With amazing expedition and the
encouragement of a Presidential Statement of the Security Council, a consulta-
tive body of experts was established to draft a framework document (at the time,
the use of the word ‘constitution’ was not accepted). This body was intended to
be fully representative, but was generally boycotted by its Serb representatives.
Some of the ethnic Albanian experts also resigned towards the end of the draft-
ing process, when it emerged that the draft texts they had submitted would not
be taken as the basis for the venture. Overall, therefore, the text that emerged
showed a strong hand of internationalized drafting, through the UN’s own le-
gal advisor’s office and other international experts, especially including an expert
of the Venice Commission. Nevertheless, the ethnic Albanian experts did have
influence on the shaping of the document, which eased the process that led all
major ethnic Albanian political parties to embrace the document (although with
hesitations) as an acceptable interim solution. At the very end of the drafting, an
effort was also made to take account of the positions of the Serb leadership and
to incorporate at least some of them, despite the continued failure to participate

formally in the drafting process.
Overall, therefore, an instrument emerged that fulfilled the functions of a con-
stitution for a considerable period to the end of 2007. There was no democratically
legitimized constitutional drafting process, say, through a constitutional conven-
tion. Instead, the document was internationally established with the involvement
(or, in the case of the Serb community, with the possible involvement) of experts.
Some of these experts were in fact affiliated with the principal political parties in
Kosovo. Hence, local ‘ownership’ of the drafting process was fairly limited.
On the part of the ethnic Albanian parties, it was possible, at least for the mo-
ment, to avoid the appearance of ‘spoilers’. Instead of attacking the document,
they declared it flawed but acceptable and focused on the upcoming election
campaign in order to attempt to fill some of the positions of political power
foreseen in the document. The Serb National Council had rejected the instru-
ment and remained a spoiler in the establishment of the intended power-sharing
arrangement. However, a large number of ethnic Serbs were registered for the
Kosovo-wide elections (in stark contrast to the local elections of the preceding
year). Hence, the Serb political leadership retained the option to participate in
the system or to attempt to opt out of it and instead to exclude elections from
areas mainly inhabited by ethnic Serbs.
2 Status, Structure, and the Layering of Public Authority

Kosovo is defined in Article 1.1. as an entity under interim international admin-
istration that, because of its people, has unique historical, legal, cultural, and lin-
guistic attributes. This clause appears to acknowledge that Kosovo is an entity
that has ‘a people’. If so, then the preambular paragraph – which refers to provi-
sional self-government as a step for a peaceful and normal life for all inhabitants
of Kosovo, with a view to facilitating the determination of Kosovo’s future status
through a process at an appropriate future stage that shall, in accordance with
UNSCR 1244 (1999), take full account of all relevant factors “including the will of
the people” – takes on an interesting meaning. The two combined could give rise
to the view that Kosovo is a self-determination unit. On the other hand, the entire
text is of course subjected to the superior authority of Resolution 1244 (1999).
There exists an institutional and a functional layering of authority. The insti-
tutional layers of authority are arranged in a comparatively simple way. There
are municipalities, the Kosovo-wide layers of provisional self-government, and
the SRSG. Municipalities are the basic territorial units of local self-government,
exercising responsibilities as set forth in UNMIK legislation in force. Hence, as
opposed to the Hill draft, the municipalities are no longer the residual source of
all public authority not expressly assigned elsewhere.
The Framework also refers to communities and their members. However, as
opposed to the Rambouillet draft, there is emphatically no provision for parallel
institutions for communities. Instead, these are now defined in terms of tradi-
tional minority rights. That is to say “communities of inhabitants belonging to the
same ethnic or religious or linguistic group (communities) shall have the rights
260 Marc Weller
set forth in this Chapter in order to preserve, protect and express their ethnic,
cultural and religious identities.” To this end, they can use their own language,
receive education and have access to information in that language, as well as en-
joy equal opportunities, etc. They can also establish associations to promote their
community interests and culture and to operate educational institutions. Howev-
er, the communities are not equipped with institutions that exercise original state
powers almost by way of full functional autonomy. Hence, this layer of public
authority that was a crucial element of Rambouillet has simply disappeared.
The Kosovo-wide institutions enjoy powers that are specifically enumerated in
the Framework. These functions extend to most aspects of public policy. These
institutions also have the power to coordinate and supervise the exercise of pub-
lic powers by the municipalities. This fact, together with the finding in Article
1.1. that “Kosovo is an undivided territory throughout which the Provisional In-
stitutions of Self-Government established by this Constitutional Framework …
shall exercise their responsibilities,” clearly subordinates the municipalities to the
Kosovo-wide institutions. Hence, there is no original autonomy located within
the municipalities, other than specific power of municipal governance assigned
in legislation.
A third institutional layer consists of the SRSG, who exercises certain ‘reserved
powers and responsibilities’ that remain exclusively in his hands:
• Full authority to ensure that the rights and interests of Communities are
fully protected;
• Final authority to set financial and policy parameters and approval of the
Kosovo Consolidated Budget and its auditing;
• Dissolving the Assembly and calling for elections, especially when other in-
stitutions have acted in violation of Resolution 1244 (1999);
• Monetary policy;
• Control of the customs service;
• Appointment, removal, and disciplining of judges and prosecutors;
• Assignments of international judges and prosecutors and changes of venue
of trials;
• International legal cooperation, conclusion, and implementation of interna-
tional agreements, and exercise of foreign affairs powers;
• Authority over law enforcement institutions and correctional facilities;
• Control over the Kosovo Protection Corps;
• Etc.
While the chapeau to this provision does suggest that these are exclusive powers,
several provisions do in fact provide for the exercise of administrative authority
in relation to them ‘in cooperation’ with the Kosovo institutions. Hence, it was
expected that some of this residual authority would be exercised by the Kosovo
institutions, up to the point when the SRSG saw a need to intervene. In relation to
other functions that are indeed formally exclusive, there is room for consultation
and advice before decisions are made. However, the principle of unchallengeable
supreme powers in relation to these issue areas must be noted. Moreover, accord-
ing to a concluding provision of the Regulation, the SRSG retained a general right
to oversee the provisional institutions and its officers and to take appropriate
measures whenever their actions are inconsistent with Resolution 1244 (1999) or
the Framework.
The unique character of the SRSG’s role was also evident in the fact that he
could ‘effect’ changes to the constitutional framework either at the initiative of
a two-thirds majority of the Assembly, or at his own discretion,. This provision
does, of course, cast some doubt on the character of the instrument as a ‘constitu-
tion’ and instead confirms its sui generis character.
3 Veto-mechanisms and Guaranteed Representation

The Kosovo Assembly, the principal legislative organ of provisional self-govern-
ment, consists of 120 members elected by secret ballot in a nationwide, single-
district vote. 100 of the seats are assigned proportionally to the votes cast for par-
ties, citizens initiatives (according to a fixed list system), or individual candidates.
Additionally, twenty seats are reserved for ‘the additional representation of non-
Albanian Kosovo communities.’ Ten of these fall to those declaring themselves
respresentatives of the Kosovo Serb Community; the others are divided amongst
the Roma, Ashkali, Egyptian, Bosniak, Turkish, and Gorani communities.
The seven-member Presidency of the Assembly must contain one member
representing the Serb community and one member representing another minor-
ity community. However, there is no rotating presidency. Instead the president is
nominated by the party having achieved the highest number of votes. There is no
provision for representation of other groups through, for instance, offices of vice
presidents. The presidency, which manages parliamentary business, is invited to
attempt to agree by consensus, but failing that decides by majority vote.
There is also no specific quota for representation in function committees, other
than the general exhortation that the membership shall reflect the diversity of the
membership of the Assembly. Chairs and vice chairs of committees are appointed
according to strength of parties in the Assembly.
A Committee of the Rights and Interests of Communities does exist, however,
consisting of two members from each of the communities represented in the As-
sembly (but only one Gorani representative). The Committee can adopt a recom-
mendation in relation to draft legislation or other matters by simple majority.
Decisions of the Assembly are adopted by simple majority of members present
and voting (the quorum being 50% for decisions and 33% for debates). Within
forty-eight hours of approval of a law, any six members may submit a motion to
the presidency alleging a violation of vital interests of the community to which
at least one of them belongs. A motion may be made on grounds that the law or
provisions discriminate against a community, adversely affect the rights of the
community or otherwise seriously interfere with the ability of the community to
preserve, protect, or express its identity.
An attempt at conciliation will be made by the presidency, which is to attempt
to put forward a consensus proposal. If that is not possible, a panel consisting
of one member of the majority, one representative of the complainants and one
262 Marc Weller
presiding individual appointed by the SRSG are to issue a recommended decision

by simple majority. However, the Assembly remains at liberty to decide to accept
or reject the consensus proposal of the presidency or the recommendation of the
panel. An Assembly decision can be reviewed for compatibility with the Consti-
tutional Framework by a special chamber of the Supreme Court.
Overall, therefore, there exists no effective blocking mechanism in the Con-
stitutional Framework. There is also only limited provision for representation of
minority groups. While it is confirmed that judges, the police, and other public
officials shall be fully reflective of the diversity of the population, there is no rigid
quota system as part of the Constitutional Framework (implementing legislation
may be different). The only concession to quota rules and minority representa-
tion is the determination that out of nine ministries that are to be established
after the election, two shall be headed by minority representatives, one of them
composed of ethnic Serbs.
IV Preparations for a Final Status

UNMIK remained very much involved in the international administration of
Kosovo during the period that followed the adoption of the constitutional frame-
work. As was foreseen, the JIAS was replaced by the Provisional Institutions of
Self-government (PISG). These were essentially ministries. The assignment of the
ministries to the Kosovo parties proved difficult. In the end, the SRSG mediated
a ministerial power-sharing agreement between all parties, including minority
communities. This meant in practice that there was no parliamentary opposition,
as all ethnic majority parties were represented in the government. This practice
was only terminated after a further round of elections, resulting in a coalition
government formed between the AAK and the LDK parties, leaving the PDK
and the smaller Ora party in opposition. As before, though, minority commu-
nities were represented in government according to a power-sharing formula.
However, after a series of destructive anti-Serb riots in March 2004, the official
Serb parties in parliament withdrew cooperation from the political processes,
although ministerial posts remained occupied.
As a result of the 2004 riots, a greater impetus towards a final settlement was
generated. Nominally, a policy of ‘standards before status’ was maintained, which
would have conditioned the commencement of status talks on the fulfillment by
the Kosovo authorities of standards of good governance and community rela-
tions. In reality, the dynamics of the situation made a further delay of status talks
impossible, even if achievements in governance were somewhat uneven. In 2006,
a process of direct negotiations mediated by former President Ahtisaari of Fin-
land acting under a UN mandate commenced. A Framework for a Comprehen-
sive Settlement for Kosovo emerged from these discussions. However, Belgrade
was unwilling to accept this design, which essentially foresaw an independent
Kosovo, remaining for some time under partial international administration.
New negotiations were demanded by Russia and Serbia.
V Conclusion
Kosovo has gone through an interesting permutation of constitutional designs,
from a (barely) autonomous territory within Serbia to a quasi-federal entity in
the 1974 SFRY system. Indeed, it gave itself an independent state constitution
at one stage in 1991. Afterwards, the Hill drafts, but in the most pronounced
way the Rambouillet agreement, provided for an ever-more complex fractionat-
ing of public power in Kosovo. This was overcome to a considerable extent in
the Constitutional Framework that was established for an interim period. This
instrument was principally generated by United Nations and other international
drafters although with input from experts representing the majority population.
Its authority is based in powers derived from a Chapter VII Security Council
resolution. Residual authority remained with the SRSG to a considerable extent,
although much governmental authority, including towards the end so-called re-
serve powers, were progressively transferred to local authorities. Power-sharing
between the ethnic communities was not supported by a very strong institutional
framework, probably as a result of the Bosnian experience. This resulted in a
withdrawal of ethnic Serb populations to virtual autonomy exercised through the
municipalities. While in principle local powers were subject to control and coor-
dination by the Kosovo-wide institutions, this power proved difficult to exercise
in practice, given the lack of effective control of the central government in com-
munity enclaves, especially Northern Mitrovica.
Overall, the case of Kosovo is a difficult one to assess in terms of the perfor-
mance of power-sharing arrangements. With hindsight, it is probably clear that
the Rambouillet accords would have imposed an essentially unworkable con-
stitutional setup upon Kosovo, given the excessive allocation of powers at the
local level, blocking powers, and other consociationalist mechanisms operating
in parallel. The constitutional framework avoided these pitfalls. However, it did
not engender governance in a conventional sense. The Serb-held ministry tended
to be isolated and politically irrelevant. The Serb community and especially the
smaller minorities could not assert themselves through their representation in
parliament. The ethnic Albanian parties found it difficult, at least in the early
phases, to forge a consensus on legislative action in the Assembly, leaving much
of the legislative work to EU and UNMIK experts. However, towards the end of
the interim period, performance improved markedly. Similarly, the transfer of
powers from UNMIK to local ministries removed the tendency to blame only the
international administration for all ills befalling Kosovo and instead impelled the
local political parties to accept more responsibility for their own conduct.
The security situation of ethnic Serb populations improved somewhat. How-
ever, there was no sign of a beginning of any genuine integration at that time. In-
stead, the prospect of the forthcoming status discussion reinforced the tendency
of most political actors to retain the status quo in terms of ethnic relations and
governance in general.
Chapter 9
Power-sharing in Macedonia?
Farimah Daftary and Eben Friedman
Since its independence, representatives of Albanian parties have shared power with
Macedonians. That is part of the general political consensus in the country … Certainly
it would be hard to find a serious analyst who maintains there is a minority in the region
that enjoys greater rights and more prosperity than Albanians in Macedonia.
Branko Geroski, Editor-in-Chief, Dnevnik, Skopje (IWPR 2001a).
[T]he fact remains the [NLA] militants have exposed the weaknesses in Macedonia’s
concept of multi-ethnic democracy – a concept built for at least a decade on the prin-
ciple of one community’s, the Macedonian majority’s, dominance over another, the Al-
banians.
Kim Mehmeti, Director, Centre for Multi-cultural Understanding and Cooperation (CMUC),
Skopje (IWPR 2001b).
I Introduction
Until early in 2001, the ‘former Yugoslav Republic of Macedonia’ (hereafter
‘Macedonia’) represented an apparently successful model of preventive diplo-
macy, and had been commended by the Organization for Security and Co-opera-
tion in Europe (OSCE) and other international organizations for its commitment
to reforms designed to enhance political stability and economic prosperity. In-
deed, despite predictions that conflict would spill over into this multiethnic state
surrounded by hostile neighbours when it declared independence in September
1991, Macedonia has fared relatively well. From the drafting of its Constitution
onward, the young republic sought to develop a legal framework for the protec-
Although it seceded from Yugoslavia in 1991, Macedonia was not admitted to the
United Nations until 1993, and then only under the provisional name ‘former Yugo-
slav Republic of Macedonia.’ The country’s constitutional name is ‘Republic of Mac-
edonia.’
See, for example, OSCE 1999b.
266 Farimah Daftary and Eben Friedman
tion of its minorities (which represent roughly a third of its population) while em-
barking upon a difficult period of democratic transformation and economic tran-
sition. The preventive activities of numerous international and domestic actors
(including the United Nations (UN), the OSCE and numerous nongovernmental
organizations (Anderson 1999)) combined with the moderate approach of do-
mestic political leaders to make Macedonia a beacon of stability in a region beset
by conflict. Macedonia even withstood the immediate impact of the Kosovo con-
flict in 1999, which resulted in a huge influx of ethnic Albanian (and later Roma)
refugees, who threatened to disrupt Macedonia’s fragile interethnic equilibrium.
Nevertheless, interethnic relations had been deteriorating since Macedonia’s in-
dependence from Yugoslavia, particularly affecting relations between the ethnic
Macedonian majority and the ethnic Albanian population (the country’s second-
largest group). With this in mind, both international and domestic experts em-
phasized the importance of continued attention to the question of interethnic re-
lations, warning of possible conflict if the Macedonian government did not make
a special effort to address the issue (Najčevska and Beška 1999).
In February 2001, shooting broke out between Macedonian border guards and
ethnic Albanians belonging to an organization calling itself the National Libera-
tion Army (NLA) in Tanuševci (a village on the Kosovo border that had served
as a KLA base in 1999). When in March fighting between the NLA and Macedo-
nian security forces spread to Tetovo – the second largest city in Macedonia with
an ethnic Albanian majority – the worst fears of the international community
concerning a new Balkan war seemed to have materialized. As the NLA opened
new fronts, it became increasingly clear that this was not a small terrorist group
(as the government claimed), but rather an organized movement with a mili-
tary strategy to gain control of portions of the territory of the Macedonian state.
While the NLA claimed to be fighting for the improvement of the rights of ethnic
Albanians in Macedonia, the Macedonian government maintained that the NLA
was working towards secession or a ‘Greater Albania’. The foreign media was
quick to seize on a new story, using terms such as ‘Balkan Ulster’ (Walker 2001)
to the frustration of Macedonia’s population, which had become used to more
agreeable terms such as ‘Oasis of Peace’.
The low-level military conflict between the NLA and Macedonian government
forces continued throughout the spring of 2001 in the regions around Tetovo (to
the west) and Kumanovo (to the north), even spreading to the immediate suburbs
of Skopje in early June when the NLA occupied the village of Aračinovo. There
were also several incidents of interethnic violence (mainly in Bitola and in Sko-
pje) as crowds took to the streets in several cities and destroyed stores belonging
The NLA and the Kosovo Liberation Army (KLA) have the same abbreviation in Al-
banian: UÇK (which stands for Ushtria Çlirimtare Kombëtare, in the first case, and
Ushtria Çlirimtare e Kosovës in the second).
‘Greater Albania’ refers to the territories inhabited by ethnic Albanians: Albania it-
self, Kosovo, parts of southern Serbia (Preševo valley), and portions of western Mac-
edonia.
9 Power-sharing in Macedonia? 267
to ethnic Albanians and other minorities in response to killings of Macedonian

soldiers and police officers that were attributed to the NLA. The fears of civil war
raised by the interethnic violence were further fed by several large, ethnic Mace-
donian-led anti-government demonstrations in front of the parliament building
in Skopje. On 13 August 2001, however, after more than a month of intense ne-
gotiations mediated by European Union (EU) and United States Special Envoys
François Léotard and James Pardew, as well as by Max van der Stoel of the OSCE
and NATO’s Peter Feith, Macedonia’s four largest political parties – two ethnic
Macedonian and two ethnic Albanian – signed an agreement on a set of legal
and other measures designed to improve the situation of the country’s ethnic
Albanian community.
Beginning with a brief analysis of the roots of the Macedonian conflict, this
chapter provides an overview of events in Macedonia from February to August
2001. Domestic and international attempts to end the fighting receive attention,
as do the negotiations leading to the Framework Agreement of 13 August 2001.
The chapter also offers an analysis of the main measures proposed in the Frame-
work Agreement and an assessment of the formal grounds the document pro-
vides for power-sharing. Next, we examine domestic and international reactions
to the Framework Agreement, as well as the complex and to-date incomplete
process of implementation. Drawing on the preceding sections, in Section VI we
assess the Framework Agreement’s prospects for bringing a durable resolution to
the conflict which prompted its drafting.
From our analysis, we conclude that Macedonia constitutes neither a success-
ful case of power-sharing nor a failed one, but rather a potential case of power-
sharing at a crucial juncture. On the one hand, it is still too early to assert that
the Framework Agreement has attained its goals, and future implementation of
the accord may be compromised by the perception, widespread among ethnic
Albanians and ethnic Macedonians alike, that the terms of the Framework Agree-
ment stem from the international community’s sympathy for the ethnic Albanian
side. Moreover, even strict adherence to the terms of the Framework Agreement
would not lead to consistent and comprehensive power-sharing. On the other
hand, the fact that the Framework Agreement continues to play a central role in
setting the agenda of the Assembly of the Republic of Macedonia (Macedonia’s
parliament) in a highly sensitive, but largely non-violent, political atmosphere
provides evidence that the document may yet provide the basis for a functional
consociational arrangement.
For the text of the agreement, see http://www.president.gov.mk/eng/info/dogovor.

htm.
II Roots of the Conflict
A Delayed Spillover from Kosovo or Domestic Tensions?

The Macedonian conflict was prompted by the deliberate decision of an ethnic
Albanian paramilitary organization with roots in both Kosovo and Macedonia to
provoke clashes with the Macedonian security forces in the hope that this would
spark a conflict between the Macedonian authorities and the local Albanian
minority. While the official appearance of the NLA is linked to the attack on a
police station in January 2001 in Tearce, Macedonia, there are indications that
preparations for the NLA offensives might have already begun in mid-2000. The
explanation for the timing of events must be sought in a combination of factors:
the influence of ethnic Albanian decision-makers in Kosovo; an attempt in 2000
by the Macedonian border patrol to crack down on cross-border smuggling and
weapons trafficking by ethnic Albanians; the signing of the border delineation
agreement between Macedonia and the Federal Republic of Yugoslavia in Febru-
ary 2001; and, finally, the decision by NATO to allow Yugoslav forces back into
southern Serbia in May 2001, which in consequence drove those NLA fighters
who were there into Macedonia.
What was the overall aim of the NLA? Notwithstanding ethnic Macedonian
suspicions that the NLA was fighting for a Greater Albania and that the ethnic
Albanian population of Macedonia supported this aim, polls taken in the spring
of 2001 showed that, in fact, support among the ethnic Albanian population for a
Greater Albania was quite low (although it might have been higher for uniting the
western, Albanian-majority areas of the country with Kosovo to form a ‘Greater
Kosovo’). Further, even those who supported this idea were not necessarily ready
to take up arms. According to a poll commissioned by the US State Department’s
Office of Research, while 69% of ethnic Albanians were sympathetic to the NLA
guerrillas, 87% stated that it was personally important to them that the country
remain united. A majority of ethnic Albanians (71%) stated that they would prefer
to live in an ethnically mixed Macedonia rather than a greater Albanian state and,
of those who supported a Greater Albania, only 13% were prepared to fight for
it. It therefore cannot be maintained that the conflict in Macedonia stems from
claims to external self-determination by the ethnic Albanian population.
These poll results seem to demonstrate that the strong sympathies for the NLA
among ethnic Albanians reflected support for the NLA’s claim to be fighting for
an improvement of ethnic Albanian rights as well as a greater say in the affairs of
the state rather than any territorial aim. The Macedonian government nonethe-
less held that the NLA was a terrorist organization consisting primarily of KLA
fighters and lacking domestic legitimacy. The government accordingly blamed
the international community for having failed to effectively seal the border be-
tween Kosovo and Macedonia. The international community, while also refusing
Poll of 1,091 nationally representative adults conducted between 16 April and 3 May
2001 (cited in Judah 2001a).
to accept the NLA’s legitimacy to speak for the ethnic Albanian community of
Macedonia, nevertheless took the stance that the actual grievances were legiti-
mate and that the solution to the crisis lay in addressing these concerns through
political dialogue.
The Macedonian government’s position was consistent with various theories
predicting that interethnic conflict in Kosovo would eventually spill over into
Macedonia. However, while it was true that the NLA had clear links to the KLA
and a splinter group, the Liberation Army of Preševo, Medvedja, and Bujanovac
(UÇPMB), which was active in Albanian-majority enclaves of southern Serbia,
there was no doubt that it also had important domestic roots. Not only did the
NLA include ethnic Albanians from Macedonia who had previously fought with
the KLA (a fact that the Macedonian government recognized), but its ranks also
included significant numbers of ethnic Albanians from Macedonia who had taken
up arms more recently, with the NLA claiming that about 80% of its fighters were
from Macedonia (Naegele 2001). Thus, while the unfolding of the Macedonian
conflict seems to fit with expectations of a spillover from Kosovo, this explana-
tion must be supplemented with the domestic roots theory, which saw poten-
tial for conflict in the growing tensions between ethnic Macedonians and ethnic
Albanians. Indeed, Macedonia’s ethnic Albanian population seems to have be-
come increasingly frustrated with the ineffectiveness of its political leaders in
representing its concerns. Sidelined from politics in Kosovo, former KLA leaders
might have calculated the opportune moment to capitalize on ethnic Albanian
dissatisfaction in Macedonia, gathering support for armed action there. At the
There were two main variants of the spillover scenario. The first one assumed that the
western and northern portions of Macedonia would become the theatre of military
operations for the Albanian army or for paramilitary formations of ethnic Albanians
from Macedonia (or for both) acting in support of Albanian insurgents in Kosovo
and possibly resulting in an armed incursion by Serbian troops. The second variant
predicted that a wave of up to one million refugees from Kosovo could flood into
Macedonia and destabilize the country. For a critique of the various scenarios, see
Troebst 1997.
Although different variants of the domestic roots theory provide different genealo-
gies of the tensions that eventually erupted into the armed conflict of 2001, a com-
monly cited point of origin for a gradual but steady deterioration of relations between
ethnic Macedonians and ethnic Albanians are the Kosovo uprisings of 1981. The en-
suing government crackdown in Kosovo led to a large influx of ethnic Albanians into
Macedonia, prompting Macedonian authorities to take various measures to combat
the perceived threat from Albanian nationalism. For a discussion of these measures,
see Poulton 1989.
According to Tim Judah, many members of the Popular Movement for Kosovo
(LPK), a small party instrumental in setting up the KLA, were ethnic Albanians from
Macedonia. While some opted for politics at the end of the Kosovo war, others did
not. According to Judah, these people whom he calls “the losers of Kosovo politics”
had been agitating for a conflict in Macedonia (Judah 2001b). Stefan Troebst has
published a similar analysis (2000).
same time, in an attempt to gain the support of both ethnic Albanians and the
international community, the NLA “effectively hijacked the political programme
of the Albanian parties in Macedonia” (Engström 2002), advocating recognition
of Albanians as an equal constituent people in the preamble of the Constitution,
making Albanian an official language, and equal job opportunities in state institu-
tions and administration.
B Interethnic Relations in Macedonia: Fertile Ground for an

Albanian Insurgency?
According to official statistics based on the 1994 census,10 ethnic Macedonians
account for approximately 67% of Macedonia’s 1.95 million inhabitants, with eth-
nic Albanians constituting 23% of the total population, ethnic Turks 4%, Roma
2.25%, ethnic Serbs 2%, Vlachs 0.44%, and other ethnic groups another 2%.11 The
high birth rate of the ethnic Albanian population is often cited as a reason for the
ethnic Macedonians’ fear that they will lose control of ‘their state’, with ethnic
Macedonian insecurities magnified by Macedonia’s neighbours’ refusal to recog-
nize the existence of a separate Macedonian nation, a distinct language, or even
the name of the country. In this context, many ethnic Macedonians seek to justify
their domination of the state by the fact that this is the only state that they have
(implying that ethnic Albanians can go to Albania if they wish). Additionally, the
Macedonian government may well be correct in claiming that no other minor-
ity in the region has such extensive rights as the Albanian minority in Macedo-
nia, and the status of ethnic Albanians in Macedonia is considerably higher than
the status of ethnic Macedonians living in neighbouring Albania, Bulgaria, and
Greece.
Ethnic Albanian parties have pressed for changes of a largely symbolic nature,
arguing that these are essential for improving the rights of ethnic Albanians in
practice. They reject the 1994 census figures, claiming that ethnic Albanians rep-
resent more than a third of the population12 and that they should therefore enjoy
10 A new census had been planned for May 2001 but was repeatedly delayed, such that
it was only eventually conducted at the end of 2002.
11 For official statistics, see http://www.stat.gov.mk. As is the case throughout post-
communist Eastern Europe, Macedonian census data undercount Roma relative to
the number of persons who identify themselves as such in daily life. Estimates by
Romani nongovernmental organizations in Macedonia place the country’s Romani
population between 100,000 and 150,000.
12 Discrepancies are due to the large number of ethnic Albanians living in the country
without a legal status, with some representatives of Macedonia’s ethnic Albanian
population claiming that over 100,000 ethnic Albanians were not counted in the
1994 census as a result of application of the Law on Citizenship of 1992 (“Zakon za
državjanstvoto na Republika Makedonija,” Služben vesnik na Republika Makedonija
67/1992). Among the most problematic of the provisions from the standpoint of
many ethnic Albanian applicants for Macedonian citizenship were the citizenship
law’s requirements of fifteen years of legal residence in Macedonia and fluency in the
equal status with ethnic Macedonians in the preamble of the Constitution. They
have also long demanded that Albanian be made a second official language along-
side Macedonian. At present, it may be used at the local level in municipal bodies
if ethnic Albanians make up a majority or a ‘considerable number’ (set at 20% in
the 1995 Law on Local Self-Government) of the local inhabitants (this applies to
any minority language which satisfies this requirement).13 They also point to the
fact that they are under-represented in state administration (despite a doubling
of numbers since 1991, the percentage of ethnic Albanians in civil service is only
about 7%), in the police, and in the army, and that they generally feel like second-
class citizens. As already stated, the majority of ethnic Albanians want to live
in a shared Macedonian state, as the political, economic, and other advantages
of being governed by Skopje have until now outweighed the numerous risks of
changing the territorial status quo (Troebst 1997: 91).
Successive Macedonian governments have made efforts to build a satisfac-
tory framework for the protection of rights of the ‘nationalities’ (i.e., minorities),
granting them special rights in the fields of education, culture, and the media.14
Particularly notable are the Macedonian government’s efforts to work with the
OSCE High Commissioner on National Minorities to settle the question of higher
education in Albanian, with the private South East European University at Tetovo
admitting its first students in October of 2001.15 Still, many issues have been left
unresolved, exacerbating ethnic tensions between the ethnic Macedonian and
ethnic Albanian populations.
Ethnic tensions are directly related to ethnic imbalances that can be traced
back to the socioeconomic structure left over from Yugoslavia, of which Mace-
donia was a constituent republic from 1943 to 1991. During this period, despite
Macedonian language, with both provisions apparently targeting ethnic Albanians

who had fled Kosovo following the uprisings of 1981 in that province. There have also
been calls to count the ethnic Albanian diaspora.
13 See “Zakon za lokalnata samouprava,” Služben vesnik na Republika Makedonija
52/1995. A 1999 resolution of the Macedonian Constitutional Court repealed the
right to use minority languages in local courts, restricting its use to municipal coun-
cils. The resolution provided that a person may still address the court in a minority
language, but only through an interpreter whose costs must be borne by the state.
The 1995 Law on Local Self-Government does not contain provisions for the use of
minority languages in communication with local government offices.
14 See, for example, Articles 7, 8, and 48 of the Macedonian Constitution of 1991, an of-
ficial English version of which is available at http://www.ok.mk. These constitutional
provisions are further elaborated in separate laws in the areas of education, local
self-government, and the media. Macedonia is also a party, inter alia, to the Council
of Europe’s Framework Convention for the Protection of National Minorities. The
European Charter for Regional or Minority Languages was signed on 25 July 1996,
but has not yet been ratified.
15 The recommendations of the HCNM on Macedonia are at http://www.osce.org/
hcnm/recommendations/fyrom/index.htm. For the homepage of the South East Eu-
ropean University at Tetovo, see http://www.see-university.com/.
many human rights abuses, an extensive system of collective rights for minorities
was established. In reality, however, discrimination against certain ethnic groups,
especially against ethnic Albanians, was common practice. The socioeconomic,
cultural, and educational gap between the ethnic groups widened due to the fact
that a higher proportion of ethnic Albanians and ethnic Turks lived in predomi-
nantly rural environments.
As a result of disproportionately high dropout rates, minorities are under-rep-
resented in secondary and post-secondary education. Despite a drastic increase
in minority representation following the raising of university quotas, members
of Macedonia’s ethnic minorities still constituted only 16% of the total student
population of Skopje and Bitola in the 1998–1999 academic year (US Department
of State 2000).16 Lower levels of education compound problems of unemploy-
ment among ethnic Albanians and Roma, although all groups, including ethnic
Macedonians, are affected by high rates of unemployment.17 The problems of a
struggling economy (exacerbated by the embargo on the Federal Republic of Yu-
goslavia) have also been a consistent negative factor.
Ethnic and socioeconomic differences in Macedonia are reinforced by reli-
gious and linguistic divides. Although nearly all ethnic Albanians speak Macedo-
nian, very few ethnic Macedonians can communicate in Albanian. Additionally,
the barriers between ethnic Albanian and ethnic Macedonian communities are
sustained by a political and state system founded on the ethnic rather than the
civic principle, with ethnically-based political parties. The absence of truly mul-
ticultural and multilingual media reinforces the isolation of the ethnic communi-
ties. Even the many domestic nongovernmental organizations tend to represent
ethnic interests. Approximately a decade after Macedonia’s secession from Yu-
goslavia, the country’s social and political landscape was marked throughout by
ethnic divisions, with ethnic Macedonians and ethnic Albanians living in paral-
lel societies, such that there was little communication and interaction between
ordinary people from different ethnic backgrounds. This state of affairs had re-
sulted in growing mistrust between the ethnic Macedonian and ethnic Albanian
populations, leading in turn to occasional incidents, often over issues related to
education and use of language. Relations between ethnic Macedonians and eth-
nic Albanians are thus not a function of ancient hatreds, but rather the prod-
uct of indifference and mistrust stemming from Macedonians’ and Albanians’
lack of first-hand knowledge about one another. The situation deteriorated as
Macedonia’s independence from Yugoslavia increased the political relevance of
the country’s sizeable Albanian population.
16 The quota system was introduced at 10% in 1991, and in 1996 was adjusted to corre-
spond to each minority’s share of the total population according to the results of the
1994 census.
17 In late 2001, for example, estimates from the State Employment Bureau and the Insti-
tute of Economics indicated an unemployment rate of over 40% (Nanevska 2002).
C Main Political Actors before the Conflict

The government formed after the 1998 parliamentary elections and which ruled
until May 2001 was a coalition of the Internal Macedonian Revolutionary Orga-
nization – Democratic Party for Macedonian National Unity (VMRO-DPMNE),
the small and nominally civic Democratic Alternative (DA), and the Party for
Democratic Prosperity of the Albanians (PDP-A, or PPD-Sh in Albanian).18 From
independence until the 1998 elections, Macedonia had been ruled by changing
coalitions headed by the Social Democratic Union of Macedonia (SDSM), a party
presenting itself and viewed externally as a successor to the League of Commu-
nists of Macedonia.
If the programmatic differences between the two largest parties of the Mace-
donian majority have often been obscured by convergences in political practice
(Hristova 2000; Sasajkovski 1999),19 even programmatic differences are largely
absent between the two ethnic Albanian parties represented in the Assembly of
the Republic of Macedonia. Instead, the main divergence between PDP-A and
the Party for Democratic Prosperity (PDP, or PPD in Albanian) has been their
degree of radicalism, with PDP-A forming when part of PDP’s youth wing broke
off out of dissatisfaction with PDP’s relative complacency as a member of the
governing coalition until 1998. Moreover, the position of these two parties rela-
tive to one another has seemed to depend on which of the two is in government.
Thus, PDP became the more radical of the pair when it joined the parliamentary
opposition in 1998.
As already mentioned, from independence until the parliamentary elections of
1998, the political scene in Macedonia was dominated by the Social Democratic
Union of Macedonia, with President Kiro Gligorov playing an important role in
arbitrating conflicts among parliamentary parties. The formation of the new gov-
ernment led by Prime Minister Ljubčo Georgievski (VMRO-DPMNE) and the
subsequent election of Boris Trajkovski (also VMRO-DPMNE) therefore marked
a major change in Macedonian politics. Notwithstanding, on the one hand, the
party’s pedigree as the descendant of the original Internal Macedonian Revolu-
tionary Organization founded to fight the Ottomans and, on the other hand, the
18 Following the parliamentary elections of 1998, PDP-A merged with the smaller Peo-
ple’s Democratic Party (abbreviated as NDP in Macedonian, PDP in Albanian) to
form the Democratic Party of Albanians (DPA in Macedonian, PDSH in Albanian).
The NDP was founded in early 1990 by Iljaz Halimi, who in 1992 called for the feder-
alization of Macedonia with an autonomous Albanian region and the reconstitution
of the Republic of Macedonia as a bi-national state (Bugajski 1994).
19 The practice of smaller parties of the Macedonian majority tends to be similar. In this
context, it is worth mentioning the Liberal Party, which participated in the Govern-
ment of National Unity formed in May 2001 (discussed below); the party’s presenta-
tion of itself as civic notwithstanding (see http://www.liberalna.org.mk), its leader,
Stojan Andov, often made use of his role as parliamentary speaker in obstructing the
implementation of the Framework Agreement on behalf of the ethnic Macedonian
population.
thinly veiled anti-Albanian rhetoric of its 1998 electoral programme20 (as well as
the explicitly anti-Albanian outbursts of the party’s leader, Ljubčo Georgievski),
VMRO-DPMNE turned out to be surprisingly moderate and open to resolving
outstanding issues in interethnic relations once in government. By bringing PDP-
A into the governing coalition, VMRO-DPMNE pursued its predecessor’s infor-
mal policy of power-sharing with an ethnic Albanian party. This policy of mod-
eration towards the ethnic Albanian community paid off as the VMRO-DPMNE
candidate Boris Trajkovski won the November 1999 presidential elections over
his rival Vasil Tupurkovski (DA), thanks in large part to the votes of the ethnic
Albanian population (Krause 2000).
III The Macedonian Conflict (February–August 2001)
A The Domestic Response: Pursuit of a Military Solution

Consistent with its position that the activity of the NLA was a purely imported
crisis brought on by terrorists seeking to gain control of portions of Macedo-
nian territory and to guarantee their revenues from cross-border smuggling, the
Macedonian government first attempted to destroy the NLA by military means.
On 21 March 2001, the NLA was given a twenty-four hour deadline to lay down
arms and/or leave the country. The NLA declared a unilateral ceasefire, stated
that it did not wish to threaten the territorial integrity of the country and called
for dialogue on the rights of ethnic Albanians in Macedonia. However, the gov-
ernment refused to negotiate and ordered a full-scale military offensive on 25
March.
As the conflict continued into April, it became apparent that there would be
no easy victory for the small and ill-equipped Macedonian army; not only were
there casualties among Macedonian and Albanian combatants as well as civil-
ians, but the NLA was also successful in opening a new front in the northern city
of Kumanovo. Furthermore, the Macedonian government’s strong-arm tactics
(e.g., shelling Albanian-inhabited villages) contributed to radicalizing the eth-
nic Albanian population. For its part, the NLA perpetrated acts of interethnic
violence with the killing of eight Macedonian soldiers and police officers who
were patrolling the border near Tetovo on 28 April 2001, provoking a wave of re-
venge attacks on 30 April and 1 May (and again on 6 June following more killings)
against stores owned by ethnic Albanians and other Muslim minorities in the city
of Bitola, from which several of the victims originated. The government reacted
to these incidents by imposing curfews in Kumanovo, Bitola, and Tetovo, making
known on repeated occasions that it was considering declaring a state of war in
order to have greater means of fighting the NLA.
20 Vnatrešna makedonska revolucionerna organizacija-Demokratska partija za make-

donsko nacionalno edinstvo, Promeni za dostoinstven život: Izborna programa na
parlamentarnite izbori ‘98 (Skopje, 1998).
B Formation of a Government of National Unity

Under intense international pressure, a ‘Government of National Unity’ was
formed among the main parliamentary parties on 13 May 2001 with the aim of
finding a political solution to the crisis. Joining the previous coalition (consisting
of VMRO-DPMNE, DA, and DPA) were the SDSM, the PDP, the Liberal Demo-
cratic Party (LDP), the Liberal Party (LP), and the Internal Macedonian Revolu-
tionary Organization – Real Macedonian Reform Option (VMRO-VMRO).21
No progress was made in the talks, however, for this new government was
anything but unified. Opinions on how to deal with the crisis diverged between
the two ethnic Albanian parties, although both advocated amnesty and talks with
the NLA. Divisions within the ‘Macedonian bloc’ were even stronger, especially
between the hardline faction keen on pursuing a strong military offensive, on the
one hand – represented by Prime Minister Ljubčo Georgievski and Interior Min-
ister Ljube Boškovski (both VMRO-DPMNE) – and the more moderate SDSM
leader Branko Crvenkovski, on the other. Georgievski’s frequent inflammatory
statements against the ethnic Albanian parties, as well as the West, further po-
larized the domestic political environment. Just as detrimental was the fact that
neither of the ethnic Albanian parties clearly condemned the NLA. Apart from
efforts by President Boris Trajkovski to stabilize the situation and to come up
with a peace plan, most politicians seemed more concerned with the prospect of
upcoming parliamentary elections than with preventing a civil war.
C Reactions of the Macedonian Population

The failure of the government to explain its actions to the public and its tendency
to blame any unpopular decision on Western pressure was a public relations di-
saster and partly to blame for bringing the country to the brink of civil war. A case
in point is the handling of the Macedonian army’s withdrawal from the village
of Aračinovo, with its strategic proximity to Skopje’s airport (Macedonia’s main
airport). Following a prolonged offensive in late June 2001 to dislodge the NLA
from the village, which it had occupied since 9 June 2001, the Macedonian army
was forced to withdraw under Western pressure. The state television’s broadcast
of the evacuation of armed NLA fighters by US soldiers serving under KFOR on
25 June cemented in many ethnic Macedonians a view of the event as a major af-
front against the Macedonian people on the part of the West. That same evening,
crowds gathered in front of the parliament building to protest against the govern-
21 The ministerial positions were distributed as follows: VMRO-DPMNE received the

largest share with six out of a total of eighteen ministers and four out of thirteen
deputy ministers; SDSM, despite its large size in parliament, received the same
number of positions as the two ethnic Albanian parties: three ministers and three
deputy ministers; finally, the three smaller coalition members, LDP, LP, and VMRO-
VMRO each received one ministerial position (see RFE/RL Balkan Report, 5.36, 18
May 2001).
ment’s decision to allow the NLA to escape. Shots were fired and mobs marched
through Skopje, destroying shops owned by ethnic Albanians and demanding the
resignation of the president. An address by the president the following day barely
succeeded in calming tempers.
The impact of the conflict on interethnic relations throughout the country
was extremely serious. Already divided along ethnic and linguistic lines, domes-
tic media outlets of both ethnic Macedonian and ethnic Albanian populations
broadcast ‘hate speech’ by politicians, contributing to a radicalization of opinion
within each ethnic community. At the same time, all citizens suffered from a gen-
eral atmosphere of insecurity and the deterioration of interethnic relations. The
conflict also resulted in over 100,000 refugees and internally displaced persons,
as well as in human rights violations ranging from arbitrary arrest to the killing
of unarmed civilians by both Macedonian security and NLA forces (Amnesty
International 2001).
D Unified International Intervention

The international community demonstrated that it had learned the lessons from
previous Balkan conflicts by acting quickly and in a unified fashion, condemning
the NLA’s violent tactics. At the same time, it acknowledged that action on behalf
of the Macedonian government to improve the situation of ethnic Albanians was
warranted. The international community therefore called on the Macedonian
government to find a political solution to the crisis rather than pursue a military
path, which risked leading to full-blown war. The EU High Representative for
Common Foreign and Security Policy Javier Solana and the NATO Secretary-
General Lord Robertson made repeated trips to Skopje from late March 2001
to encourage talks among the political parties represented in parliament. While
Solana’s and Robertson’s efforts prevented the declaration of a state of war and
pressured the main parties to join in a Government of National Unity, this shuttle
diplomacy achieved little in the way of positive outcomes. The international com-
munity therefore decided to invest more resources in resolving the crisis, with
the EU and the USA each nominating (in late June 2001) a special representative
to facilitate talks on the ground (former French Defence Minister François Léo-
tard and Ambassador James Pardew respectively).
The efforts of the EU and the USA to bring about a political solution to the
crisis were supplemented by those of the OSCE. The OSCE has a long history of
conflict prevention in Macedonia: in addition to the active engagement of OSCE
High Commissioner on National Minorities Max van der Stoel from 1993 to 2001,
the OSCE Spillover Monitor Mission to Skopje had been in place since Septem-
ber 1992, making it the longest-serving OSCE mission. When the conflict broke
out, the capacity of the OSCE Mission in Skopje was doubled from a staff of eight,
typical of OSCE missions, to 16 international personnel who were sent to moni-
tor crisis areas. The OSCE Mission also consistently condemned acts of violence
by both sides in the conflict, investigated claims of human rights abuses and con-
demned the misuse of religious and cultural sites.22 The task of the OSCE Mission
was made particularly difficult by the strong anti-Western sentiment of parts of
the population reinforced by the xenophobic statements of some politicians. In
this atmosphere, there were also several attacks on Western embassies as well as
against the office of the OSCE Mission.
On 21 March 2001, the OSCE Chairman-in-Office, Romanian Foreign Minister
Mircea Geoana, announced his intention to nominate US diplomat (and former
head of the OSCE Mission in Skopje) Robert Frowick as his Personal Representa-
tive. In May of the same year, Frowick initiated the first serious attempt to bring
about a cessation of hostilities by brokering direct negotiations between the DPA,
the PDP, and the NLA on a ceasefire in exchange for an amnesty (International
Crisis Group 2001a; IWPR 2001c). In the ‘Prizren Declaration’ of 22 May 2001
(named after the town in Kosovo where the parties met), the NLA reaffirmed
recognition of Macedonian territorial integrity and demanded as a condition of
a ceasefire that ethnic Albanian representatives be given a veto over policies re-
lated to ethnic Albanians’ rights, as well as an amnesty and the integration of
demobilized NLA fighters into the police and army in Albanian-majority areas of
Macedonia. These were unacceptable demands for the ethnic Macedonian side,
and Frowick’s alleged role in arranging the meeting was condemned not only by
the Macedonian-language media but also by the international community. Even
the OSCE Mission to Skopje claimed that Frowick had been acting on his own
and that the NLA could not be considered as a partner in a political dialogue.23
Ambassador Frowick left Macedonia at the end of May and was replaced on 1 July
2001 by the former OSCE High Commissioner on National Minorities, Max van
der Stoel. In the end, the Prizren Declaration nonetheless proved useful for creat-
ing a common platform between the NLA and the two ethnic Albanian parties,
likely contributing to the NLA’s assent to the Framework Agreement despite not
being a signatory to it.
E The Internationally Sponsored Talks

Following a request for assistance – apparently motivated by the absence of other
viable options – from President Trajkovski, on 29 June NATO agreed to send
about 3,000 troops to supervise the disarming of the NLA provided that four
conditions were fulfilled: (1) a peace agreement signed by the main political lead-
ers; (2) a status of forces agreement (SOFA) with Macedonia on the conditions
of NATO troop deployment; (3) an agreed plan for weapons collection, includ-
ing an explicit agreement by the NLA to disarm; and (4) an enduring ceasefire.35
The final condition for starting talks was achieved with a new Western-brokered
ceasefire on 5 July 2001.
22 For the press releases of the OSCE Mission, see http://www.osce.org/news_scripts/.

23 RFE/RL Balkan Report, 5.37, 29 May 2001.
1 Skopje, 9 July 2001: Round 1

The round of talks that began in Skopje on 9 July 2001 involved only the four
largest parties in the Government of National Unity (VMRO-DPMNE, SDSM,
DPA, and PDP). EU and US representatives Léotard and Pardew were joined on
10 July by Ambassador Van der Stoel. The chief NATO representative in Skopje,
Ambassador Peter Feith, was responsible for liaising with the NLA. An extensive
proposal drafted by domestic legal experts, with the help of the French constitu-
tional lawyer Robert Badinter, served as a basis for the talks.24
The ethnic Albanian parties strongly criticized the initial draft and made a
counter-proposal demanding that ethnic Albanians be made a ‘constituent peo-
ple’ and that the Albanian language be placed on an equal footing with the Mace-
donian language.25 They also proposed to establish the post of vice-president, to
be filled by an ethnic Albanian, with veto power over all important government
decisions. The Albanian parties also proposed transferring control of the local
police to the municipal authorities. These demands came as no surprise as most
were part of the PDP-A’s platform in the October 1998 elections (Poulton 2000:
198);26 some had also allegedly been discussed in Prizren.
Responding to the Albanian parties’ concerns, the international mediators
met separately with those parties to prepare a revised document. Two important
concessions were obtained from the ethnic Albanian parties: the request for the
post of vice-president with veto power and demands to make ethnic Albanians
a constituent people were both dropped. The second draft was presented to the
ethnic Macedonian parties on 18 July 2001.27 Already highly suspicious because
of the separate meeting with the Albanian parties, the ethnic Macedonian parties
strongly rejected the new draft in which it was proposed that any language spo-
ken by at least 20% of the population (i.e., Albanian) would be official state-wide.
Prime Minister Georgievski stated that this would lead to “language federaliza-
tion” and condemned the “cowboy-like methods” of the international mediators
(IWPR 2001d). In protest, Lord Robertson and Javier Solana cancelled a trip to
Skopje scheduled for 19 July and the two ethnic Albanian leaders pulled out of the
talks. Talks resumed the next day, but at the expert level only, with separate meet-
ings between the international mediators and each side. Meanwhile, the situa-
tion on the ground deteriorated, with ceasefire violations and reports of ‘ethnic
cleansing’ in the form of expulsions of ethnic Macedonians from the Tetovo area
confirmed by OSCE and other monitors.
24 This first draft was published in the Macedonian-language daily Dnevnik on 10 July
2001. For an overview of its contents, see RFE/RL Balkan Report, 5.48, 12 July 2001.
25 Published in the Macedonian-language daily Večer on 12 July 2001.
26 Also see Kongresi i Partisë Demokratike Shqiptare, Programi i Partisë Demokratike
Shqiptare (Tetovo, 1997); Sobranie na Partijata za demokratski prosperitet, Program-
ski opredelbi (Tetovo, 1999).
27 The second draft was published in Dnevnik on 19 July 2001.
2 Ohrid, 28 July 2001: Round 2

A government ultimatum, together with an agreement reached by Ambassador
Feith with the NLA in Prizren on 25 July 2001, prompted the NLA to begin with-
drawing from its positions near Tetovo and to respect the ceasefire. Following a
meeting in Skopje on 26 and 27 July between the four parties in the talks as well
as Lord Robertson and Javier Solana, it was announced that negotiations would
resume. While Tetovo was originally chosen as the site of the talks, as a result of
security concerns the talks began on 28 July at the Macedonian president’s official
residence on Lake Ohrid. A cautious breakthrough on the status of the Albanian
language was reported on 1 August, followed on 5 August (after another visit by
Solana) by a compromise agreement on the administration of the police.28
Despite reported progress on the political front, the armed conflict continued
even after an announcement was made that a final agreement had been initialled
on 8 August. In apparent retaliation for the extra-judicial killing by the police of
five alleged NLA members in Skopje on 7 August, five ethnic Macedonian road
workers were kidnapped and abused, allegedly by the NLA.29 The killing of at
least seventeen soldiers on 8 August in two separate attacks was also attributed
to the NLA. Editorials in the Macedonian-language press put pressure on the
government to resume military action. Over the weekend of 11–12 August, the
country saw some of the worst fighting as both sides attempted to gain ground
before the official signing of the agreement. Under heavy Western pressure, a
unilateral ceasefire was declared by the government in the evening of 12 August,
in time for a low-key signing ceremony the following day in Skopje.
IV The Framework Agreement of 13 August 2001: Towards Power-sharing

in Macedonia?
A Main Measures Proposed

The Agreement of 13 August 2001 aims to provide a “framework for securing the
future of Macedonia’s democracy and permitting the development of closer and
more integrated relations between the Republic of Macedonia and the Euro-At-
lantic Community” (Preamble). In fact, the document is an attempt to address
many of the grievances of ethnic Albanians and to give them a greater say in
managing the affairs of the country, which a system of majoritarian democracy
alone cannot guarantee. It consists of ten main sections and three annexes. The
Agreement begins by rejecting the use of violence in the pursuit of political aims
and reaffirms Macedonia’s sovereignty and territorial integrity. Next, it outlines a
series of proposed measures to improve the status of ethnic Albanians in Mace-
donia, with the measures followed in turn by a discussion of constitutional and
28 “Macedonian Parties Resolve Police Issue,” RFE/RL Newsline, 6 August 2001, at

http://www.rferl.org/newsline.
29 See the Amnesty International news release of 10 August 2001 (EUR 65/004/2001).
Also see the press release by Human Rights Watch (2001).
other legal reforms necessary for their implementation. Before evaluating the
potential of the Agreement to provide a lasting resolution of the Macedonian
conflict, we will outline the measures proposed in the document.
1 Promotion of the Civic Approach

The Framework Agreement endorses a civic rather than ethnic concept of the
state and stresses the need to “continually ensure that the Constitution fully
meets the needs of all its citizens,” while also emphasizing that the multieth-
nic character of Macedonian society must be preserved (Section 1). A series of
constitutional amendments are thus proposed in Annex A of the Agreement to
replace all reference to “nationalities” (minorities) by the term “communities.”
This concerns, first and foremost, the infamous preamble of the 1991 Constitu-
tion, which defined the republic as “a national state of the Macedonian people, in
which full equality as citizens and permanent coexistence with the Macedonian
people is provided for Albanians, Turks, Vlachs, Roma and other nationalities
living in the Republic of Macedonia.” According to the Framework Agreement,
the revised preamble is to refer only to the “citizens of the Republic of Macedo-
nia,” thereby transforming the Macedonian nation-state into a state of all of its
citizens, regardless of ethnic, religious, or linguistic background. Favoured by the
EU and US Special Envoys, this civic solution was already proposed in the first
draft by the ethnic Macedonian side as an alternative to making the country a bi-
national state of ethnic Macedonians and ethnic Albanians (Pearson 2002: 6).
2 Language-related Measures
The Framework Agreement contains several major changes concerning linguistic
rights, which in practice will have the greatest impact for ethnic Albanians. Con-
cerning the state level, the Framework Agreement states that any language spo-
ken by at least 20% of the total population is also an official language throughout
the republic (Section 6.5). Any official language may also be used in state organs,
and separate legislation is to be adopted to codify the use of official languages
other than Macedonian. By the terms of the Agreement, Albanian is the only
language to qualify as an official language state-wide and may thus be spoken in
parliament in addition to being used for the publication of state laws.
With respect to the local level, the Framework Agreement restates the right
of members of any (minority) community to use their language in units of local
self-government if that language is spoken by at least 20% of the local population
(Section 6.6). It also grants a new right to speakers of any language that qualifies
as official at the state level (i.e., Albanian) to use in official communication at the
local level with regional representatives of the central government, who must re-
spond in Macedonian as well as in that language (Section 6.5). This provision dif-
fers from the one contained in the first draft, by which regional authorities were
not obliged to respond in the minority language. At the same time, the protection
of the rights of Macedonian speakers is ensured by making Macedonian the offi-
cial language throughout the country, regardless of the number of its speakers in
any municipality (Section 6.4). Moreover, Macedonian remains the sole language
of the republic’s external relations.
The Framework Agreement’s potential impact on other minority languages is
minimal insofar as only Albanian qualifies for status as an official language for use
at the state level as well as in communication with organs of the central govern-
ment at lower levels.30 Thus, while the Agreement allows speakers of Albanian to
make use of their native tongue in communication with regional representatives
of the central government, it contains no similar provision for the use of other
languages spoken by 20% of the population of a given municipality. The situation
is different, however, when it comes to minority language use in organs of self-
government. The use of such languages in organs of municipal government was
already allowed by the 1995 Law on Local Self-Government and is not affected by
the Framework Agreement. By way of contrast, one local-level innovation of the
Framework Agreement for speakers of minority languages other than Albanian
is the option for municipalities to allow the use of languages spoken by less than
20% of the local population in communication with municipal administrative or-
gans (Section 6.6).
3 Measures Concerning Local Police

A particularly controversial issue of the talks that produced the Framework
Agreement was the control of local police. Accordingly, the final proposal regard-
ing responsibility for local police is a compromise, with local police commanders
to be appointed by local authorities from a list drawn up by the Interior Ministry
(Section 3.3). At the same time, the police will remain under the control of the
central government. In line with the commitment of the signatories to ensure
that the police force reflects the ethnic composition of the general population,
the Framework Agreement calls for a total of 1,000 new police officers “from
communities not in the majority” to be hired and trained by 2004 (500 by July
2002 and a further 500 by July 2003) (Annex C, Section 5.2). The new recruits
are to be sent to areas inhabited by members of their own minorities. In practice,
most of the new recruits will be ethnic Albanians, with the numerical goal of
the planned recruitment to increase the proportion of Albanians in the police
force from approximately 3% in 2001 to approximately 23% (i.e., ethnic Albanians’
share of Macedonia’s general population according to the 1994 census).
4 Special Majorities
Section 5 of the Framework Agreement proposes special parliamentary proce-
dures in the form of a system of “double majorities” that are designed to prevent
representatives of minority communities from being outvoted when adopting
30 While the 20% threshold is not uncommon at the local level (see, for example, the
1999 Slovak Law on the Use of Minority Languages (Zákon č. 184 o používaní jazykov
národnostných menšín, at http://www.zbierka.sk)), Macedonia’s smaller minorities
have called for a national threshold as low as 3% on the grounds that none of them
can even approach the 20% requirement.
laws that directly affect those communities (i.e., laws concerning culture, edu-
cation, personal documents, and the use of languages and symbols). The same
section of the Agreement provides for similar procedures to be applied when
adopting laws on local finances, local elections, the city of Skopje, and the bound-
aries of municipalities. In each of these cases, laws can only be approved by a
parliamentary majority “that includes a majority of the votes of representatives
claiming to belong to the communities not in the majority in the population of
Macedonia” (i.e., minority votes). This mechanism is similar to what was pro-
posed for Bosnia-Herzegovina, although much simpler, as it will require depu-
ties in parliament to identify themselves either as members of the majority or of
“communities not in the majority in the population of Macedonia.” Not specified
in the Framework Agreement, however, is whether membership in a minority is
automatic (e.g., based on whether the deputy in question is a member of a politi-
cal party associated with a particular ethnic minority) or instead is based on a
more flexible system of self-identification, which could vary from vote to vote.
The practical relevance of these considerations of group membership is demon-
strated by the fact that although the Macedonian parliament in place when the
Framework Agreement was drafted and approved included members of several
of the country’s smaller (i.e., non-Albanian) minorities,31 only one member of a
smaller minority (Amdi Bajram of the Union of Roma of Macedonia (SRM)) had
been elected as a representative of a minority party.
The Framework Agreement foresees different types of majorities (simple and
qualified or two thirds) for different types of decisions. In most cases of laws
directly affecting minority communities, the Framework Agreement calls for
a simple majority of votes that also includes a simple majority of the minority
votes. In the case of the constitutional amendments proposed in Annex A of the
Framework Agreement and a revised Law on Local Self-Government, however,
the Agreement calls for a special qualified majority of two thirds of the votes,
including two thirds of the minority votes.
5 Equitable Representation
Section 4 of the Framework Agreement (Non-Discrimination and Equitable
Representation) reaffirms the general principle of non-discrimination and equal
treatment before the law. In particular, Section 4 focuses on employment in pub-
lic administration and public enterprises, as well as access to public financing for
businesses. In addition to the measures to ensure fair representation of ethnic
Albanians in the police force (discussed above), laws regulating employment in
public administration are also to be amended in order to ensure equitable repre-
sentation of the “communities” in all central and local public bodies as well as at
all levels of employment. These proposed measures respond to observations that
31 According to the Macedonian Ministry of Information, the Assembly resulting from

the parliamentary elections of 1998 consisted of ninety-two Macedonians, twenty-
five Albanians, two Roma and one Bosniak. Ministerstvo za informacii na Republika
Makedonija, Fakti (Skopje, 1999), 44.
ethnic Albanians are under-represented in the workforce in general, and at the

top levels of administration and decision-making in particular. Not resolved by
the Framework Agreement are the financial means necessary to recruit and train
persons from the under-represented communities.
While the post of Public Attorney (i.e., ombudsman) was established in 1997,
its lack of influence during its first four years of existence led to an expansion of
the Public Attorney’s area of competence in the Framework Agreement. In par-
ticular, the ombudsman is to ensure observation of the principles of non-discrim-
ination and equitable representation of (minority) communities in public bodies
and in other areas of public life. Additionally, the ombudsman may contest the
conformity of laws with the Constitution before the Constitutional Court, one-
third of the members of which must now be chosen in parliament by a simple
majority of votes, including a simple majority of minority votes. This same proce-
dure applies also to the election of three members of the Judicial Council, as well
as to the election of the ombudsman himself.
6 Decentralization of the Unitary Macedonian State

The Framework Agreement states that the unitary character of the Macedonian
state is inviolable, thereby excluding federalism as well as all forms of territorial
autonomy. At the same time, however, the Agreement notes the importance of
developing local self-government as an essential step in encouraging the partici-
pation of citizens in democratic life and promoting respect for the identity of the
ethnic communities. A cornerstone of the Framework Agreement is thus the ex-
pansion of the role of local self-government and municipalities in what had been
an extremely centralized country (Section 3). Indeed, when the Agreement was
approved, there was no intermediate level of administration between the central
government and the country’s 124 municipalities.32 The Framework Agreement
accordingly states that a revised law on local self-government must be adopted to
reinforce the powers of elected local officials and to enlarge their competencies in
the areas of public services, urban and rural planning, environmental protection,
local economic development, culture, education, social welfare, and health care.
7 Other Measures
A symbolic concession of great importance to ethnic Albanians and other mi-
norities in Macedonia is the provision of the Framework Agreement that official
personal documents may be issued in a minority language and its alphabet, in ad-
dition to Macedonian, which uses the Cyrillic alphabet (Section 6.8). The Agree-
ment also stipulates that local authorities will be allowed to display emblems of
community identity on local public buildings (an issue which in 1997 led to riots
32 The administrative changes of 1996 more than tripled the previous number of mu-
nicipalities in the country (i.e., 35), resulting in an extremely large number for a coun-
try as small as Macedonia. See “Zakon za teritorijalnata podelba na Republika Make-
donija i opredeluvanje na podračjata na edinicite na lokalnata samouprava,” Služben
vesnik na Republika Makedonija 49/1996.
in the west Macedonian towns of Gostivar and Tetovo in response to the removal
of Albanian and Turkish flags from municipal buildings by the police). Addition-
ally, the Agreement mentions the possibility of state funding for the new, primar-
ily Albanian university in Tetovo (Section 6.2).
A provision of the Framework Agreement inconsistent with the principle of
non-discrimination is the document’s approach to Article 19 of the Constitution,
which addresses the status of religious communities. While Article 19 was due to
be amended through the elimination of its reference to the Macedonian Ortho-
dox Church, thus granting all religious communities the same status, the Agree-
ment calls for the amendment to state that “the Macedonian Orthodox Church,
the Islamic Religious Community in Macedonia, the Catholic Church, and other
Religious [sic] communities and groups are separate from the state and equal be-
fore the law.” By extending privileged status to Macedonia’s Muslim and Catholic
populations in addition to the Orthodox population, the phrasing proposed in
the Framework Agreement risks discriminating against smaller religious com-
munities. Representatives of the Macedonian Orthodox Church, on the other
hand, complained about the placement of Islam and Catholicism on the same
level with Orthodoxy.
B Provisions Concerning Implementation

The Framework Agreement sets strict deadlines for effecting the changes outlined
in the document (Section 8). In addition to the proposed constitutional amend-
ments contained in Annex A, a revised Law on Local Self-Government was to be
debated and adopted within forty-five days of the date of the official signing of
the Framework Agreement (i.e., by 27 September 2001). Additionally, a long list
of legislative changes was to be effected within a rigid timeframe (Annex B).
Changes to be made by the end of the term of the parliament in power when the
Framework Agreement was signed (thought at the time to be the end of 2001)33
included: a law on local self-government finances; provisions pertaining to police
located in the municipalities; amendments to the laws on the civil service and
public administration to ensure equitable representation of the communities;
amendments to the Rules of Procedure of the Assembly to enable use of the Al-
banian language; new legislation regulating the use of languages in state organs;
and new legislation on the issuance of personal documents. All other laws were
also to be checked to ensure that their provisions regarding the use of languages
were fully compatible with those of the Framework Agreement.
Changes to be made by the end of 2002 were: a revised law on municipal
boundaries; a revised law on electoral districts; and amendments to the law on
the Public Attorney and other relevant laws to enact the change concerning the
new responsibilities and financing of this position.
33 In fact, however, parliamentary elections were subsequently postponed until 15 Sep-

tember 2002.
C A Framework for Complex Power-sharing?

The five main elements of a full-fledged power-sharing arrangement are: (a) a
power-sharing executive; (b) veto rights; (c) proportionality; (d) arbitration
mechanisms; and (e) segmental autonomy (Schneckener 2000; cf. Lijphart 1977).
Below, we juxtapose these elements with the provisions of the Framework Agree-
ment. In so doing, we aim to demonstrate that the Framework Agreement is an
awkward attempt at combining an integrative, civic approach emphasizing equal
rights for all citizens with elements of consociational democracy and a regime
of minority rights. Attempting, on the one hand, to establish a civic state, the
document perpetuates divisions between the majority and non-majority commu-
nities, on the other. An example of the latter tendency is the special parliamen-
tary mechanism proposed for certain situations, which forces deputies to make
a choice about whether they belong to the majority or the non-majority commu-
nity. At the same time, the Agreement reaffirms the unitary character of the state
while introducing many important elements of power-sharing at the central level
between the two largest communities. Thus, the Framework Agreement simulta-
neously rejects territorial autonomy and adopts measures which usually call for
complementary territorial arrangements.
1 A Power-sharing Executive
As previously mentioned, Macedonia has had an informal power-sharing execu-
tive due to the participation of an ethnic Albanian party in government since
1992. From 1994 to 1998, the PDP had nineteen representatives in parliament,
five ministers (including one Deputy Prime Minister) and four deputy ministers.
Following the October 1998 elections, the DPA held fourteen of the 120 seats in
parliament, five of twenty-seven government ministers (including one Deputy
Prime Minister), and five deputy ministers (out of twenty-one).34 Although this
informal power-sharing arrangement at the executive level was not formalized
in the Framework Agreement, the results of the September 2002 parliamentary
elections provide additional evidence that a power-sharing executive constitutes
an unwritten rule of Macedonian politics, with former NLA leader Ali Ahmeti’s
Democratic Union for Integration (DUI in Macedonian, BDI in Albanian) in
charge of four ministries (education, health, justice, and transport), as well as a
minister without portfolio in the new SDSM-led government, which consists of a
total of seventeen ministries.35
34 Given that the total number of ministries had increased, ethnic Albanians claimed
that they were less represented than in the previous government. However, the situ-
ation had improved in terms of providing a more significant presence of DPA repre-
sentatives among high-level state administrators.
35 For the results of the 2002 parliamentary elections in Macedonia, see http://www.
ecmi.de/emap/mk_B04.html.
2 Veto Rights
Another element of a power-sharing arrangement is a veto power designed to
encourage parties to reach a consensus by creating an atmosphere of security in
which each group can, if necessary, block decisions in certain policy areas. Veto
rights can be divided into three types: delaying, indirect, and direct. While the
demand of the Albanian side for the creation of a post of vice-president with a
veto right was not granted, the special system of double majorities to be used in
specific situations constitutes an indirect veto power for minority representa-
tives. In particular, the system of double majorities is designed to ensure that no
law directly affecting the ethnic Albanian community can be adopted without the
consent of ethnic Albanian deputies. According to the terms of the Agreement,
the influence of ethnic Albanian politicians is not strictly limited to laws related
to language, education and culture, as Albanian deputies’ consent is also neces-
sary for the adoption of revised laws on local self-government, local finances, and
local elections. In these cases, a simple majority of votes, including a majority of
the votes of “representatives claiming to belong to communities not in the major-
ity in the population of Macedonia”, is necessary (Section 5.1).
3 Proportionality
The extensive measures to ensure proportional representation of the minority
communities in state institutions and in public administration have already been
outlined. More important than overall statistics, however, is the degree to which
ethnic Albanians and other minorities will be represented at all levels, and espe-
cially in positions of responsibility.
4 Arbitration Mechanisms
A further indication of a move from a state system based on general non-dis-
crimination and minority protection towards power-sharing is the creation of a
Committee on Inter-Community Relations to replace the weak and ineffective
Council for Inter-Ethnic Relations. The purpose of the Committee is to settle
disputes on issues affecting inter-community relations and regarding laws di-
rectly affecting the minority communities. The members of the Committee, to
be elected by parliament from among its members rather than proposed by the
president, are to consist of equal numbers of ethnic Macedonians and ethnic Al-
banians (seven each), and one member each from among ethnic Turks, Vlachs,
Roma, and two other communities.36 Although one purpose of the Committee on
Inter-Community Relations is to ensure that the concerns of the smaller minori-
ties are heard, the influence of such minorities has been reduced relative to their
previous position in the Council for Inter-Ethnic Relations, which had included
two members each from the Macedonian, Albanian, Turkish, Vlach, Romani, and
36 If there are not enough members of parliament representing these communities, the
Public Attorney/ombudsman may appoint persons from outside following consulta-
tion with the relevant community leaders.
‘other’ nationalities.37 In this sense, one of the major mechanisms of power-shar-

ing in Macedonia effectively increases ethnic Albanian representation at the ex-
pense of the representation of the smaller minorities. The Public Attorney (i.e.,
ombudsman), on the other hand, can be viewed as a more equitable arbitrator
among ethnic communities insofar as election to the position follows from a
majority vote in parliament, also including “a majority of the votes of the total
number of Representatives claiming to belong to the communities not in the ma-
jority in the population of Macedonia.”38 On the whole, however, the (incomplete)
power-sharing arrangement established by the Framework Agreement reduces
the influence of Macedonia’s smaller minorities in favour of the ethnic Albanian
community.
5 Segmental Autonomy
Absent from the provisions of the Framework Agreement is any form of seg-
mental autonomy, either territorial or cultural/personal. A crucial element of
power-sharing arrangements, segmental autonomy gives the parties involved in
power-sharing rights to manage their own affairs in certain policy areas (e.g., cul-
ture, education, economy, health services, security). While territorial autonomy
would have been theoretically possible for the western portions of the country in
which ethnic Albanians are concentrated, the Macedonian government viewed
such an arrangement as unacceptable on the grounds that it could serve as a first
step towards secession. Moreover, a simple territorial arrangement would not
have satisfied the Albanian side concerned primarily with greater participation
in the affairs of the state. However, while the planned decentralization does not
give ethnic Albanians group autonomy, it is intended to increase their room for
manoeuvre by way of ethnically neutral mechanisms; by calling for the transfer of
competencies from the central government to the local governments, the Frame-
work Agreement facilitates ethnic Albanian self-government in municipalities
where ethnic Albanians constitute the local majority.
A summary review of the five elements of power-sharing as they apply to Mace-
donia provides a neat demonstration of the incompleteness of the power-sharing
arrangement promulgated in the Framework Agreement. First, while the docu-
ment does nothing to prevent continuation of the informal power-sharing at the
executive level that has characterized Macedonian politics since independence
and may indirectly encourage such continuation through a proportional electoral
system that increases ethnic Albanian representation in parliament, this central
element of consociational democracy has not been embodied in legislation. By
way of contrast, veto rights, another central element of power-sharing, have been
encoded in the various requirements for double majorities in parliament, while
both the Committee on Inter-Community Relations and the institution of the
37 Also included on the Council for Inter-ethnic Relations was the President of the As-
sembly of the Republic of Macedonia. See Article 78 of the 1991 Constitution.
38 Constitution of the Republic of Macedonia (2001), Article 77, at http://www.sobranie.
mk/Eng/rule.htm.
ombudsman constitute formal mechanisms of arbitration. On the other hand,

while the Framework Agreement also calls for proportional representation in
state institutions and public administration, implementation in this area has thus
far been limited mostly to formal party politics. Possibly the most conspicuous
omission in the Framework Agreement, however, is that of any provision for seg-
mental autonomy, territorial or otherwise.
In light of the above, even if it were implemented consistently, the Framework
Agreement could not produce a full-fledged power-sharing arrangement; not
only are some essential elements of power-sharing missing, but there are major
contradictions that prevent this document from being a formal power-sharing
arrangement. Perhaps most problematic is that the groups concerned by the ar-
rangement are not specifically named. While the measures outlined in the Agree-
ment clearly relate to the country’s two largest groups (i.e., ethnic Macedonians
who constitute the majority, and ethnic Albanians, who constitute the only mi-
nority that comprises more than 20% of the general population of the country),
the constitutional amendments proposed in the Framework Agreement seek to
do away with distinctions based on ethnic differences by eliminating explicit ref-
erences to concrete ethnic groups. Given the various ethnic communities’ re-
spective numbers and the fact that Macedonia’s political system is based on eth-
nic parties, the Framework Agreement can at most enforce what was previously
an informal power-sharing arrangement limited to ethnic Albanians and ethnic
Macedonians, and only at the executive level.
D Vertical and Horizontal Aspects of the Arrangement

Not outlining all of the legislation for which it calls, the Framework Agreement
provides little information as to how power will be divided vertically other than
regarding issues such as control of the local police. Nonetheless, the Agreement
suggests that local governments will enjoy greater influence in the areas of cul-
ture, education, environmental protection, health care, local economic develop-
ment, planning (both urban and rural), public services, and social welfare. The
Framework Agreement also foresees decentralization in the area of human and
minority rights, as the Public Attorney is expected to establish decentralized of-
fices. Language matters are also to be decentralized to the extent that municipali-
ties may decide on the languages to be used in their administrative organs.
If the Framework Agreement provides some indication as to how power will be
divided vertically, it also suggests by omission that there will be no correspond-
ing horizontal division, for the document makes no mention of dividing jurisdic-
tions between ethnic Macedonians and ethnic Albanians. While the Framework
Agreement leaves room for speculation as to which ministries will be headed
by Albanians, such speculation was equally possible in the period preceding the
conflict.
E Conditions Affecting Successful Implementation of Power-sharing

The conditions for implementing any form of power-sharing in Macedonia have
generally been unfavourable (Schneckener 2000: 10-18). Reached through ex-
ternal pressure rather than internal compromise, the Framework Agreement is
viewed by considerable segments of the ethnic Macedonian population as a bit-
ter pill Macedonia was forced to swallow by pro-Albanian Western actors. Im-
plementation of the Framework Agreement is also likely to be hindered by the
presence of mutually reinforcing cleavages in Macedonia’s general population, as
ethnic differences are generally reinforced by linguistic, religious, and socioeco-
nomic differences. Moreover, neither ethnic Macedonian nor ethnic Albanian
political elites have had the unilateral support of their respective ethnic commu-
nities. Also problematic for the implementation of a power-sharing agreement is
the exclusion of one of the parties to the conflict – the NLA – from the negotia-
tions to end the conflict. Thus, the Framework Agreement can be seen as an ex-
ternally imposed solution on two levels: first by the international community on
Macedonia as a whole, and second by the parliamentary political parties on the
NLA. Moreover, the success of the parliamentary parties in bringing the fighting
to a conclusion despite the NLA’s absence from the negotiations combined with
various features of the conflict (e.g., the relatively low number of casualties and
the fact that the fighting began in the wake of a major government corruption
scandal) tend to support the position of some observers that the armed conflict
had in fact been staged (e.g., Vankovska 2002: 7-8; International Crisis Group
2002a).
V Reactions to the Framework Agreement
A Reactions of Domestic Politicians

Prime Minister Ljubčo Georgievski repeatedly sought to distance himself from
the terms of the Framework Agreement, which he claimed to have signed under
international pressure, and then only as leader of VMRO-DPMNE. Georgievski
also displayed outrage when DPA leader Arben Xhaferi addressed the press in Al-
banian immediately following the signing ceremony of 13 August 2001. On 3 Sep-
tember 2001, Georgievski declared in parliament that to approve the Framework
Agreement would send a message that terrorism pays. Unlike Georgievski, Presi-
dent Trajkovski made some effort to defend the Framework Agreement itself,
describing it in his 31 August 2001 address to the Macedonian parliament as a
path chosen by the legitimately elected leaders of a large number of citizens, both
ethnic Macedonian and ethnic Albanian. Trajkovski’s main argument, however,
was that the Framework Agreement was the only alternative to civil war. SDSM,
on the other hand, generally reacted positively and refrained from criticizing the
Framework Agreement. On the ethnic Albanian side, the defence of the Agree-
ment by DPA and PDP was lukewarm. Finally, representatives of some of the
country’s smaller minorities – particularly ethnic Turks and Roma – expressed
concern that the Framework Agreement would lead to the neglect of their needs
in a new bi-national order.
It is extremely likely that Georgievski’s efforts to undermine the legitimacy of
the Framework Agreement hurt the document in the eyes of many segments of
Macedonia’s general population. Moreover, even if Georgievski and other domes-
tic politicians had wished to sell the Agreement to their constituents, their task
would be made more difficult by the fact that most politicians have lost legiti-
macy in the eyes of the people. A poll conducted in 2001 confirms this trend, also
showing that neither belligerent rhetoric nor a more compromising approach
bore fruit as all four political parties involved in the talks lost popularity.39 Even
more significant is that one third of all people planned not to vote at all (31.6% of
ethnic Macedonians, 14.3% of ethnic Albanians) in the next parliamentary elec-
tions. The disposition to boycott was strongest in the Turkish ethnic community
(50%), a reflection of ethnic Turks’ disaffection with the Macedonian-Albanian
political debate. However, voter turnout of nearly 75% in September 2002 con-
founded these earlier, more pessimistic predictions about the elections.
B Reaction of the NLA

Although the non-participation in the negotiations of one party to the conflict
– the NLA – would normally be considered a negative factor, the NLA’s indirect
involvement through a NATO liaison seems to have secured its support for the
Framework Agreement. Thus, despite the fact that one of his main demands (i.e.,
a transfer of control over the police to the municipalities) was not satisfied, NLA
leader Ali Ahmeti indicated on 20 August 2001 that he was satisfied with the
Agreement and that the NLA would surrender its weapons.
C Legitimacy of the Framework Agreement in the Eyes of the Population

Most problematic is that the Agreement was reached through strong external
pressure on seemingly unwilling political actors rather than through internal
compromise. Additionally, the Framework Agreement contains many measures
that are unpopular with ethnic Macedonians, and there has been little support
for it in the Macedonian-language press (which has termed it the ‘Macedonian
Dayton’ and even compared it to the 1938 Munich Agreement). Moreover, while
the change to the Preamble proposed in the Agreement is offensive to many eth-
nic Macedonians as well as to the smaller minorities, it also does not satisfy the
demands of the ethnic Albanians for a bi-national state.40 The low degree of sup-
39 Poll published on 20 August 2001 by the Skopje daily Utrinski Vesnik, cited in RFE/
RL Balkan Report, 5.60, 24 August 2001.
40 According to a poll from early May 2001, 83% of ethnic Macedonians said they were
against a revision of the Constitution while 77% of ethnic Albanians believed imme-
diate changes were necessary (poll conducted by the Center for Ethnic Relations in
Skopje, published in Utrinski Vesnik on 30 May 2001, cited in “Macedonians Against,
port among the ethnic Macedonian population may also be related to the failure
of the government to explain the implications of the measures proposed; there
has been little debate in the media (even in the Albanian-language media) on
the actual implications of the measures proposed, leading to misunderstanding
about the actual implications for the Macedonian state (Rusi 2001). Thus, the
poll of August 2001 indicated that two thirds of ethnic Macedonians believed
that decentralization would lead to federalization, which they equated with an
eventual dissolution of the state. Nonetheless, it should be noted that the Frame-
work Agreement is supported by a significant segment of the ethnic Macedonian
population. A poll taken in September 2001 indicated that while 50.7% of eth-
nic Macedonians opposed the peace plan, 43.7% approved of it.41 Also providing
some indication of ethnic Macedonian support for the Framework Agreement
is SDSM’s clear victory in the 2002 parliamentary elections. The September poll
indicated considerably stronger support for the accord among ethnic Albanians
(78% in favour, 12.9% against).
D Reaction of the International Community

The international community was unanimous in welcoming the Framework
Agreement and in insisting that it be implemented as soon as possible. In par-
ticular, the EU, USA, and NATO responded quickly and favourably to the signing
of the document, with the UN issuing a statement on the same day the Frame-
work Agreement was signed. International human rights organizations, such as
Amnesty International (AI), also welcomed the Framework Agreement out of a
belief that the document would redress past human rights abuses and institu-
tionalized discrimination. At the same time, AI and others expressed concern
over the implications of some of the constitutional amendments for the rights of
Macedonia’s smaller ethnic minorities.42
VI Implementation of the Framework Agreement

Implementation of the Framework Agreement involves a series of complex pro-
cesses, each of which represents a potential obstacle to successful conflict reso-
lution. Additionally, not all of the legislation adopted since the signing of the
Agreement is consistent with the provisions of the Agreement. Thus, not only do
the elements of power-sharing outlined in the Framework Agreement fall short
of a full-fledged power-sharing arrangement, but the measures enacted to imple-
Albanians for Changes of Constitution,” MILS News, 6 June 2001, at http://groups.

yahoo.com/group/balkanhr/message/2306.
41 Poll published in Utrinski Vesnik and cited in “Macedonian Poll Shows Depth of
Ethnic Divide,” RFE/RL Newsline, 4 September 2001.
42 See Amnesty International news releases of 15 August, 7 September, and 10 Septem-
ber 2001, at http://web.amnesty.org. Also, see the statements by the Budapest-based
European Roma Rights Center, at http://www.errc.org.
ment the Agreement also reflect additional compromises among parliamentary

actors in Macedonia concerning the terms of the document.
A NATO’s Operation “Essential Harvest”

The first step taken by the international community to assist with implementation
of the Framework Agreement was to disarm the NLA. After NATO had secured
an agreement to disarm from the NLA and President Trajkovski had secured par-
liamentary approval for an amnesty, the North Atlantic Council authorized the
deployment of 3,500 NATO troops to Macedonia. Operation ‘Essential Harvest’
(27 August–26 September 2001) was a ‘soft’ mission to collect and destroy 3,300
weapons voluntarily surrendered by the NLA. By 26 September, NATO had actu-
ally exceeded its mission, collecting and destroying 3,875 weapons.
From its inception, Operational Essential Harvest was criticized by Macedo-
nian politicians and the Macedonian-language press as an essentially symbolic
mission rather than a genuine effort to disarm the NLA. Presumably because
of Macedonian authorities’ perception that the international community would
dictate the terms under which it would move toward implementation of the
Framework Agreement, the Macedonian government apparently never prepared
its own plan for disarming the NLA. Within the broader ethnic Macedonian
population, on the other hand, there were widespread fears that NATO interven-
tion would lead to the creation of an international protectorate (as in Kosovo)
or to territorial division along ethnic lines (as in Bosnia) by helping the NLA
to secure control over portions of Macedonian territory. Thus, on 11 September
2001, major Macedonian dailies ran editorials expressing concern that the coun-
try might soon be forced to accept a new international force, possibly leading to
the partition of Macedonia along ethnic lines.43 The number of weapons to be
collected also sparked controversy, for the Macedonian government estimated
the NLA stockpile to be as high as 60,000. It was argued that even if the majority
of weapons was collected that the NLA could easily procure more from Kosovo
and other sources. Further complicating matters was the need for NATO to syn-
chronize the collection of weapons with the debates over the Framework Agree-
ment in parliament; whereas the Macedonian government had demanded that
all weapons be collected before parliament would start debating the Agreement,
the NLA wanted the constitutional amendments to be adopted as weapons were
surrendered.
Among the ethnic Macedonian population, strong negative perceptions of
NATO date back to the NATO bombing of Serbian forces in Yugoslavia during
the 1999 Kosovo crisis. As with many other issues, the position of the ethnic Al-
banian population was radically different: it strongly supported NATO’s involve-
ment and hoped that it would protect ethnic Albanians against harassment by
Macedonian security forces. According to the May 2001 poll cited above, 57.9%
of ethnic Macedonians said they did not trust NATO while 76.3% of ethnic Al-
43 See “Macedonia ‘On Horns of Dilemma’, ” RFE/RL Newsline, 12 September 2001.

banians did. Thus, whereas many ethnic Albanians see the Framework Agree-
ment and NATO’s role in promoting it as important victories in the (continuing)
battle for improved status in Macedonia, a large number of ethnic Macedonians
view the Framework Agreement as a defeat at the hands of not only the ethnic
Albanian population, but also the international community. In this manner, the
mutual reinforcement of common ethnic Albanian and ethnic Macedonian per-
ceptions concerning the partiality of the international actors responsible for the
Framework Agreement poses a significant threat to the implementation of the
power-sharing arrangement outlined in the document.
B Preliminary Steps towards Implementation: August–November 2001

From the outset, the adoption of the constitutional amendments and other leg-
islative modifications promised to be a long and difficult process complicated
by the strong opposition of many ethnic Macedonian deputies, particularly rep-
resentatives of VMRO-DPMNE and parliamentary speaker Stojan Andov (LP).
Further, although DPA, PDP, and SDSM were generally in favour of the proposed
amendments, many deputies were afraid of supporting the Framework Agree-
ment for fear of appearing insufficiently patriotic.44 Another complication was
that the amendments, originally proposed in English, had to be translated into an
acceptable Macedonian version.
The day after NATO completed the first phase of Operation Essential Harvest,
parliament began debating the Framework Agreement. The first day of parlia-
mentary debate, 31 August 2001, did not bode well, with Stojan Andov delaying
discussion in response to a demonstration outside the parliament building, then
setting new conditions to be met by the NLA to allow for the return of displaced
civilians before deliberation would continue. Despite predictions to the contrary,
however, parliament approved the Framework Agreement on 6 September. Of
the 112 deputies voting, 91 voted in favour, 19 voted against, and 2 abstained.
On 7 September 2001, NATO’s Task Force Harvest embarked on Phase 2 of the
weapons collection – to gather an additional third of the weapons while parlia-
ment began debating the constitutional amendments. Six days later, NATO an-
nounced that Phase 2 was complete, with the weapons tally at 2,200. Instead of
proceeding with adopting the amendments, however, parliament began debating
a referendum on the Framework Agreement, a move which threatened to sig-
nificantly delay the legislative process (as a simple majority is sufficient to call a
referendum). This debate, too, was suspended after DPA leader Xhaferi raised the
possibility of holding a separate referendum to decide on which constitutional
changes the ethnic Albanian community wanted. Xhaferi was aided in his effort
to stall the debate over a referendum by the SDSM leader Branko Crvenkovski,
44 At the time of the debates in parliament over the Framework Agreement, VMRO-
DPMNE occupied 47 of the 120 seats in the Macedonian Assembly, while the two
ethnic Albanian parties occupied an additional 25. SDSM, on the other hand, held 27
parliamentary mandates.
who threatened that his party would leave the governing coalition if parliament
decided to hold a referendum. NATO Secretary-General Robertson travelled to
Skopje on 14 September to ensure that the political aspect of the implementation
process would move on. According to Andov, however, the second phase of the
parliamentary debate would end on 20 September without the adoption of the
constitutional measures, which would have to wait until all weapons were col-
lected and Macedonian forces had moved back into the NLA-controlled areas.
The issue of a referendum was finally set aside on 18 September and parliament
returned to the constitutional amendments.
As an illustration of the difficulties encountered in approving the amendments,
one can cite the proceedings of 21 September 2001. On that day, parliament ap-
proved amendments on the preamble of the Constitution, on provisions regulat-
ing the use and status of the Albanian language (Article 7), and on provisions
regarding the equitable representation of minority communities in state institu-
tions (Article 8). Each of these measures was adopted with a majority of sixty-two
votes. By way of contrast, the proposed changes to Article 19 on the status of the
Macedonian Orthodox Church failed to pass by one vote. While the amendments
were supported by the SDSM, PDP, DPA, Liberal Democrats (LDP), and LP, ten
VMRO-DPMNE deputies also voted in favour. About thirty VMRO-DPMNE
deputies voted against, along with VMRO-VMRO, DA, New Democracy (ND),
and Amdi Bajram (SRM), who apparently objected to the removal of a reference
to Roma as a “nationality” in the preamble of the Constitution.
Despite the delays, NATO announced that it would proceed with the third
and final phase of weapons collection. For its part, the NLA stated that it would
surrender the remainder of its weapons. As news came on 24 September that
the 3,300 weapons had already been collected, parliament adopted the package
of fifteen constitutional amendments as well as the new preamble by a narrow
majority (sixty-eight in favour, twenty-four against and four abstentions).
In October 2001, the third phase of the parliamentary debate (i.e., the pro-
cess of ratification) was slowed as the media (particularly television) launched an
extensive public discussion on the provisions of the Framework Agreement. At
roughly the same time, Andov raised the issue of a referendum again, while Presi-
dent Trajkovksi and SDSM proposed changes to the proposed amendments. Eth-
nic Albanian deputies, on the other hand, boycotted some sessions of parliament
in protest against piecemeal discussion of the amendments. Following visits by
the EU Commissioner for External Relations Chris Patten and by Javier Solana,
the parties reached a new compromise (albeit over PDP objections).
Even amid the ongoing wrangling in parliament, there were some positive
developments on the security front. On 9 October 2001, President Trajkovski
– with the backing of the Macedonian government – declared an amnesty for
NLA fighters not guilty of war crimes. In light of the high degree of ethnic po-
larization surrounding the issue of amnesty, Trajkovski’s declaration constituted
a significant event.45 Still, the ethnic Albanian parties in parliament called for
stronger guarantees in the form of a law. Another positive development was the
deployment (beginning on 22 October 2001) of ethnically mixed police patrols
(accompanied by NATO troops and OSCE monitors) in some villages in the crisis
areas. Additionally, the Ministry of Defence announced the demobilization of ap-
proximately 1,200 reservists.
On 1 November 2001, parliament resumed discussion of the package of fifteen
constitutional amendments. Two weeks later, in a surprise vote held shortly af-
ter midnight to avoid demonstrations, the Macedonian Assembly ratified a new
Constitution containing the fifteen amendments with a comfortable two-thirds
majority (ninety-four deputies in favour, fourteen against). The ratification came
just three days after some of the worst incidents since the signing of the Frame-
work Agreement, involving ethnic Macedonian civilian hostages as well as both
Macedonian police (including the special units called ‘Lions’ and ‘Tigers’) and the
NLA. In combination with discussions with the new EU representative in Skopje,
Alain Le Roy, the violence of 11–12 November played an important role in con-
vincing PDP to abandon its opposition to the new preamble.
C Legal Measures Adopted in Implementing the Framework Agreement

1 The 2001 Constitution
The Macedonian Assembly adopted an amended Constitution on 15 November
2001. Contrary to the provisions of the Framework Agreement, which had called
for an ethnically neutral preamble, the version adopted as Amendment IV refers
to “[t]he citizens of the Republic of Macedonia, the Macedonian people, as well as
citizens living within its borders who are part of the Albanian people, the Turkish
people, the Vlach people, the Serbian people, the Romani people, the Bosniak
people and others.” (Constitution of the Republic of Macedonia 2001). While the
new preamble does away with the term “national state” used in the preamble of
1991, it also gives pride of place to “the Macedonian people” in the list of peoples.
In tension with this apparent priority, however, is the use of the term “people” in
reference to the other groups mentioned in the Preamble. While this would seem
to indicate a promotion from these groups’ prior status as ‘nationalities’, reference
to these same groups as “communities” elsewhere in the Constitution (see, for
example, Amendment VIII, which replaces Article 48) further confuses the issue.
Overall, the preamble of 2001 suggests that Macedonia is neither a civic state, nor
a nation-state, nor yet a bi-national state.
Whereas the Framework Agreement proposed to regulate relations among re-
ligious communities by extending privileged status to Macedonia’s Muslim and
Catholic populations as well as to the Macedonian Orthodox Church, Amend-
ment VII to the Constitution takes this approach further by incorporating addi-
tional groups. Thus, rather than employ terminology that would allow all religious
45 According to the poll of May 2001, 81.8% of ethnic Macedonians opposed granting
amnesty to NLA fighters while 98.4% of ethnic Albanians favoured amnesty.
communities to enjoy the same status, Amendment VII mentions “[t]he Macedo-
nian Orthodox Church, as well as the Islamic Religious Community in Macedo-
nia, the Catholic Church, Evangelical Methodist Church, the Jewish Community
and other Religious communities and groups.” In this manner, Amendment VII
extends privileged status to Macedonia’s Jews and Methodists as well as to the
groups enumerated in the Framework Agreement, leaving intact the possibility
for discrimination against smaller religious communities.
Other constitutional amendments adopted on 15 November 2001 are consis-
tent with provisions outlined in the Framework Agreement and discussed above.
These include measures related to the use of minority languages in an official
capacity (Amendment V), measures to prevent ethnic Albanian deputies from
being outvoted in the adoption of future legislation (Amendments X–XII and
XIV–XVI), measures to bring about proportional representation in public life in
general as well as in public administration in particular (Amendments VI, XI, and
XII), and measures concerning arbitration mechanisms in the form of the Public
Attorney and the Committee for Inter-Community Relations (Amendments XI
and XII, respectively).
2 The Law on Local Self-government of 2002

Missing the 27 September 2001 deadline by nearly four months, the new Law on
Local Self-Government was passed on 22 January 2002 by the two-thirds ma-
jority required in Amendment XVI to the Constitution.46 As stipulated in the
Framework Agreement, the 2002 Law endows the municipalities with a larger
set of competencies relative to the 1995 law of the same name, including public
services, urban and rural planning, environmental protection, local economic
development, culture, education, social welfare, and health care.47 Additionally,
the Law enacts elements of the Framework Agreement in its provisions for the
use of Macedonian and minority languages (Articles 89 and 90), in requiring
double majorities for local regulations concerning culture, languages, and alpha-
bets used by less than 20% of the citizens of a given municipality (Article 41), and
in calling for the establishment of a Committee for Inter-Community Relations
in each municipality in which more than 20% of the inhabitants “are members of
a certain community” according to the most recent census (Article 55). Although
the committees are to consist of “an equal number of representatives from each
community represented in the municipality,” the Law provides no indication as
to how to determine what constitutes a community which must then be repre-
sented. Anticipating a still forthcoming law on municipal boundaries, the Law on
Local Self-Government of 2002 also contains provisions concerning appropriate
grounds for the creation of new municipalities:
46 Official Gazette of the Republic of Macedonia 5/2002. An English translation of the

law is available at http://www.ecmi.de/emap/mk_CO4_O4b01.html.
47 For an analysis of administrative centralization in Macedonia before the 2002 Law on
Local Self-Government, see Popovski and Panov 1998.
The territory on which a municipality is established should represent a naturally, geo-

graphically and economically linked entity, with communication among populated
places and gravitation towards the common center, and it should have infrastructure
facilities as well as facilities of social standard built therein (Article 17; see also Article
16).
Reactions to the Law on Local Self-Government have been mixed, with some
ethnic Macedonians claiming that the new legislation goes too far in reducing
central control while some ethnic Albanians maintain that Skopje retains too
much power.48 Insofar as municipalities in Macedonia – regardless of the ethnic
composition of the local population – had suffered under the previous legal order
from a dearth of competencies, the 2002 Law on Local Self-Government seems
to constitute a major improvement. Nonetheless, the actual functioning of the
municipal governments with their newly acquired powers will depend in large
part on the definition of local populations in the forthcoming law on municipal
boundaries, as well as on pending legislation concerning the financing of the mu-
nicipalities.
3 The Law on Amnesty

Another major piece of legislation related to implementation of the Framework
Agreement is the Law on Amnesty of 26 February 2002.49 Although no men-
tion of an amnesty was made in the Framework Agreement, it was promised by
President Trajkovsi on 30 May 2001 as part of his peace plan. The chances of a
resumption of violent action remained high as long as an amnesty had not been
declared, for the absence of such a declaration added to the general climate of
uncertainty over the future of former NLA fighters.50 Due to fears of being ar-
rested if they returned home, many NLA fighters seemed to be seeking refuge in
Kosovo, taking their weapons with them (Wood 2001). Others were reportedly
firing on Macedonian forces seeking to regain control of areas formerly held by
the NLA. Clashes were thus reported between the NLA and government forces
in September and early October of 2001 as Macedonian police sought to move
into these areas despite NATO’s appeal to the Macedonian government to refrain
from sending its forces until an amnesty was in place.
Passed for the explicit purpose of reintegrating into normal life participants in
the conflict that led to the Framework Agreement, the Law on Amnesty exempts
from prosecution persons who committed acts not falling under the jurisdiction
of the Hague Tribunal (Article 1). Additionally, the amnesty law terminates crimi-
nal proceedings against those same persons while providing full exemption from
48 “Macedonian Government Passes Local Autonomy Law …”; “… As Controversy Re-

mains,” both RFE/RL Newsline, 25 January 2002.
49 Official Gazette of the Republic of Macedonia 18/2002. English translation at http://
www.ecmi.de/emap/mk_C04_04b01.html.
50 Whereas some KLA fighters were integrated into local police forces in Kosovo, there
are no similar provisions for these (mostly) young men in Macedonia.
serving any time in prison for their involvement in the conflict. Also covered by
the amnesty are persons who did not report for duty in the Macedonian army as
well as those who deserted it in the course of the conflict (Article 2). Finally, the
Law reduces the sentences of most of the general prison population by 25% (ex-
ceptions include persons convicted for crimes against humanity, involvement in
the narcotics trade, and persons sentenced to life terms) in an attempt to prevent
“mutiny with unwanted material and human consequences” (Article 3).
4 Electoral Legislation
On 25 and 26 June 2002, the Macedonian Assembly passed a package of electoral
laws, including the Law on Election of Members of Parliament of the Republic of
Macedonia, the Law on Voters’ List, and the Law on Election Districts for Elec-
tion of Members of Parliament of the Republic of Macedonia.51 Required by the
Framework Agreement by the end of 2002, the Law on Election Districts divides
Macedonia into six districts of approximately equal population.52 The Law on
Election of Members of Parliament stipulates that the Assembly’s 120 deputies
will be elected from the six districts on rigid party lists using the D’Hondt for-
mula of proportional representation (PR) (Articles 2, 83, 96). In so doing, the new
Law on Election of Members of Parliament departs from its predecessor, which
provided for the distribution of eighty-five parliamentary mandates by a majority
run-off system and thirty-five mandates by fixed-list PR in a single, state-wide
district.53
5 The Status of the Albanian Language

On 19 June 2002, the Macedonian parliament adopted nine laws aimed at mak-
ing Albanian an official language. Not addressed by this legislation was the use
of the Albanian language on passports.54 Following repeated disputes in parlia-
ment, on 13 March 2003 SDSM, LDP, and DUI struck a compromise by which
ethnic Albanian citizens of Macedonia could apply for passports with the Alba-
nian language on the cover in addition to the standard Macedonian and English
inscriptions.55 In return for this concession, the DUI abandoned its demand that
the use of Albanian be allowed for chairing debates in parliament and meetings
of parliamentary committees.
51 Whereas the first two laws were published in Official Gazette of the Republic of Mac-
edonia 42/2002, the Law on Election Districts was published in Official Gazette
43/2002.
52 For a map of the new districts, see http://www.dik.mk.
53 “Zakon za izbor na pratenici vo Sobranieto na Republika Makedonija,” Služben ves-
nik na Republika Makedonija 28/1998.
54 Legislationline Newsletter, 4, 4 September 2002, at http://www.legislationline.org.
55 “Compromise Reached over Language Issue in Macedonia,” RFE/RL Newsline, 14
March 2003.
D Pending Legislation Affecting Implementation of the Framework

Agreement
While the above discussion demonstrates that the Framework Agreement has not
been ignored in the Assembly of the Republic of Macedonia, much remains to be
done for the accord to be fully implemented, with a plan drafted on 26 December
2002 by Macedonian officials in the presence of EU special representative Alexis
Brouhns, as well as US Ambassador Lawrence Butler, outlining necessary amend-
ments to over ninety laws to this end (Saskova and Janeva 2003). Of the pending
legislative changes, a set of proposed amendments to the 1992 Law on Citizen-
ship stands out as particularly sensitive, as approval of these changes would allow
many ethnic Albanians originally from Kosovo, who had been excluded by the
law, to become Macedonian citizens.56 The proposed changes are likely also to
affect considerable numbers of Roma and ethnic Turks.
E The Role of International Actors

Recognizing the need for international aid in implementing the Framework
Agreement, Annex C of the Agreement calls for extensive international support.
International monitors are invited to go to the sensitive areas, and international
support is requested for facilitating the return of refugees and reconstruction.
Also requested is aid in strengthening local self-government as well as in training
local police, minority lawyers, and judges. Additionally, Annex C calls for aid to
projects aimed at strengthening the media.
Even before the drafting of the Framework Agreement, the OSCE had been
actively involved in the crisis areas. Following an official invitation by the Mace-
donian foreign minister to assist with implementing the Agreement,57 the OSCE
announced on 1 October 2001 that the size of the Mission in Skopje would be
increased from 51 to over 200 international monitors. These monitors have been
involved in confidence-building and police training as well as in reporting on
humanitarian issues. Ambassador Van der Stoel, whose discrete role of facilitator
and source of advice in the internationally sponsored talks was recognized by the
US and EU representatives, also continued to be involved, along with the OSCE
Chairman-in-Office.
As past experience in Macedonia has shown, international monitors have an
important role to play in parliamentary elections and the population census. Held
on 15 September 2002, the most recent parliamentary elections in Macedonia
have been evaluated considerably more favourably by domestic as well as inter-
56 Some ethnic Albanians claim that the requirements of the 1992 citizenship law re-
sulted in the exclusion of over 100,000 ethnic Albanians.
57 Speaking in Vienna on 30 August 2001, the Macedonian Minister of Foreign Af-
fairs, Ilinka Mitreva, commended the OSCE’s efforts in Macedonia and called for the
OSCE’s further involvement in monitoring through the OSCE Mission as well as for
assistance in implementing Annex C of the Agreement.
national observers than had any of the elections previously held there. Running
in an electoral coalition with several smaller parties, including parties represent-
ing Bosniaks, Roma, Serbs, Turks, and Vlachs (among others), the SDSM under
Branko Crvenkovski won a narrow majority in parliament. Although Macedonia
lacks special provisions to facilitate the entry of minority parties into parliament,
cooperation with SDSM resulted in the election to parliament of several candi-
dates representing minority political parties, including Neždet Mustafa of the
United Party of Roma of Macedonia (OPRM in Macedonian, PPRM in Romani).58
On the other hand, the change to a pure proportional electoral system is likely
to have been responsible for another Rom, former MP Amdi Bajram (SRM), not
being re-elected.
Originally scheduled for May 2001, the population census was initially post-
poned to October 2001 due to the crisis. Although this date is reiterated in the
Framework Agreement, additional postponements followed, with the census
conducted only in the last quarter of 2002.59 Insofar as it has the potential to
settle the sensitive issue of the size of Macedonia’s various ethnic communities,
the census constitutes an essential element of the current reforms. While the re-
sults of the census have not been released as of March 2003 (and no information
is available as to when the results will be released), preliminary indications are
that radical departures from the results of the last (1994) census are few and not
significant for the overall implementation of the Framework Agreement.
Another sensitive task yet to be undertaken will be the redrawing of munici-
pal boundaries, due to be conducted under international supervision within one
year of completion of the census. Some experts have expressed concern that the
new boundaries might be drawn with the aim of obtaining ethnically pure mu-
nicipalities, thereby reinforcing the divide between ethnic groups. Apart from
ethnic Albanians, it is likely that only ethnic Turks and Roma will qualify for the
20% threshold, as other minorities are either too small or too highly dispersed
throughout the country. Nonetheless, a municipality dominated by one ethnic
group could decide to allow the use of another group’s language, and in so doing
provide an indication of interethnic accommodation and allay concerns regard-
ing the rights of other ethnic communities under a given local ethnic majority.
F Financial Assistance from the International Community

Implementation of the many measures contained in the Framework Agreement
will be expensive. With this in mind, the EU promised large amounts of aid to
Macedonia, including USD 27 million towards damages and infrastructure re-
pair, USD 44 million in state budget support and USD 37.7 million as part of a
58 Prior to his election to parliament, Mustafa served six years as mayor of the Romani-
majority municipality of Šuto Orizari on the outskirts of Skopje.
59 The reason for the delay was that the census was to be held after the parliamentary
elections, which were in turn dependent on the timely passage of relevant legislation
(discussed above).
broader package. The direct link between the process of implementing the con-
stitutional and other legal changes proposed in the Framework Agreement to
financial support from the international community for Macedonia was made
clear by François Léotard, who stated on 2 October 2001 that the international
donor conference foreseen in the Framework Agreement and scheduled for later
that month was in jeopardy due to the delays by parliament (“no vote, no mon-
ey”). That this was no idle threat was confirmed three days later when the EU
called off the conference. As the EU Commissioner for External Relations, Chris
Patten, said in Skopje on 4 October 2001, it was inconceivable for the EU to invite
donors to write cheques to support a political agreement that has not yet been
endorsed and implemented. Following re-evaluation of the situation in Macedo-
nia, the postponed donor meeting was held in Brussels on 12 March 2002. The
total contribution of the European Commission to Macedonia in 2001-2002 was
EUR 188 million.60
VII Prospects for Durable Peace in Macedonia?
A The Possibility of Renewed Conflict

Although the level of social disorder has generally been low since the signing of
the Framework Agreement, sporadic incidents – some of them involving special
units of the police – have contributed to maintaining tensions. While most of the
conflict’s refugees and internally displaced persons have returned to their homes,
the process has been slow for fear that returnees’ security will not be guaranteed.
It is therefore crucial that ethnic Albanians be incorporated into the local police
force as quickly as possible to boost confidence among ethnic Albanian civil-
ians.61 The continued presence of NATO forces, although they are not mandated
to protect the civilian population, may also provide some reassurance.
While the NLA’s cooperation in Operation Essential Harvest and its disband-
ment announcement of 27 September 2001 were favourable developments,
doubts have persisted that former NLA fighters will integrate into normal life
as peaceful citizens as per the instructions of their leader, Ali Ahmeti. Particu-
larly troubling is the mysterious genesis of the ‘Albanian National Army’ (ANA or
AKSH in Albanian) (Matoshi 2001),62 which rejected the Framework Agreement
and claimed to be fighting for a Greater Albania, and also claimed responsibility
for the seventeen Macedonian soldiers killed on 8 August 2001. Whatever the
relationship between the ANA and the NLA, delays in parliament surrounding
the implementation of the Framework Agreement could lead to frustration and
renewed armed action.
60 See http://europa.eu.int/comm/external_relations/see/news/ip02_399.htm.
61 100 ethnic Albanians began police training on 17 September 2001. Within the first
month, there were complaints of harassment of the new recruits by ethnic Macedo-
nians in the force.
62 See also Rüb 2001; IWPR 2003.
One of the principal aims of Ahmeti and others within the NLA was to carve
for themselves a role in Macedonian political life in a context of decreasing le-
gitimacy of the two main ethnic Albanian parties. Despite the fact that the NLA’s
attempt to enter the Macedonian political scene by force was initially thwarted
by a refusal to include the NLA in the negotiations that produced the Framework
Agreement, Ahmeti has nevertheless succeeded not only in endearing himself to
the international media, but also in founding a political party considerably more
powerful than DPA or PDP after the 2002 parliamentary elections. It remains to
be seen, however, what kind of role Ahmeti’s DUI will play in parliament and in
the current SDSM-led government.
On the Macedonian side, there may still be support for a renewed military of-
fensive among hardliners seeking to improve their political ratings in the context
of a serious erosion of political legitimacy, particularly among the losers in the
parliamentary elections of 15 September 2002. Those in favour of a renewed con-
flict might be counting on the support of the ethnic Macedonian population and
on the feeling that it has been forced to make unacceptable concessions to the
ethnic Albanians under the threat of violence. Further, NATO envoy Peter Feith
has relayed concerns about the presence of ethnic Macedonian paramilitaries
near Tetovo and in other parts of the country. While the Macedonian govern-
ment has consistently denied the existence of paramilitaries, it has also not col-
lected all of the weapons distributed to civilians during the conflict.
B Fears of a Security Vacuum after the Departure of NATO

Concerns about a security vacuum following the end of Operation Essential Har-
vest were temporarily allayed when an agreement was reached on 27 September
2001 over a new NATO mission of 1,000 lightly-armed soldiers under the lead-
ership of Germany. Operation ‘Amber Fox’, which took over immediately upon
completion of Essential Harvest, was initially due to be active for three months,
subject to renewal upon agreement by both sides. Amber Fox’s mandate was ini-
tially limited to guaranteeing the security of the EU and OSCE monitors who had
gone to Macedonia to assist with the implementation of the Framework Agree-
ment until the end of 2001.63 While the size of the new NATO force was more
than the Macedonian government had wanted, it was also far from the large,
longer-term NATO presence advocated by some international experts.64 Indeed,
NATO, and particularly Germany, had insisted on 1,200 troops with an initial
mandate of nine months. President Trajkovski, on the other hand, had asked for
only 700 troops to stay for a period of three months. Also subject to dispute were
the terms of the mandate, with an initial formulation envisaging NATO involve-
63 According to the head of Task Force Fox, its mission was: “Liaison and exchange of
information with international organizations and Macedonian authorities; in emer-
gency situations, after consultation with the Government as appropriate, to provide
medical evacuation and extraction for monitors.” (NATO 2001).
64 See, for example, International Crisis Group 2001b; 2001c.
ment in the implementation of the Framework Agreement and of the proposed

constitutional changes. The formulation was dropped when some ethnic Mace-
donian politicians objected to giving the mission a political mandate.65 Addition-
ally, the UN option preferred by President Trajkovski was not realized, as the UN
Security Council seemed reluctant to take on a lead role.66 Nonetheless, a state-
ment from the Security Council granted badly needed support to multinational
efforts to establish a security presence in Macedonia. After staying through the
September 2002 parliamentary elections as well as the census, Amber Fox was
succeeded on 16 December 2002 by a smaller mission, called ‘Allied Harmony’,
followed in turn by a handover of the military mission to the EU is pending, with
the mandate of the EU mission tentatively set at six months.67
C Impact of the Framework Agreement on Interethnic Relations in

Macedonia
Insofar as the debate over Macedonia’s future involved only representatives of the
country’s two largest ethnic groups, members of some of the smaller minorities
– particularly Roma and ethnic Turks – were alienated by the negotiations. While
the terms of the Framework Agreement provided some reassurance that Mace-
donia will not become a bi-national state, some representatives of the smaller
minorities see the position accorded to ethnic Albanians in the new legal order
as problematic. Still, it is to be hoped that the smaller minorities will reap some
of the benefits negotiated by the ethnic Albanian parties, such as the measures to
ensure adequate representation of non-majority communities in public bodies.
As a sign of goodwill and commitment to the concept of a civic state, the ethnic
Albanian representatives in government and in parliament would do well to seek
also to take the interests of other minorities into account.
Since the signing of the Framework Agreement, the position of the interna-
tional community in general has been that implementation of the Agreement is
the only current option for resolving the Macedonian conflict. However, consid-
erably more explanation will be necessary if the provisions of the document are to
be accepted by the general population in Macedonia. At the same time, although
the population has exhibited a high degree of scepticism that the Framework
Agreement will provide a durable resolution to the crisis, a poll conducted short-
ly before the signing of the Agreement indicated that most people believed that
65 “Macedonia – One Down,” Transitions Online, 25 September–1 October 2000, at

http://www.tol.cz.
66 There were calls for reviving the UN Preventive Deployment Force (UNPREDEP),
which had been successful in preventing spillover of conflict in the country from
1995 until February 1999 when a Chinese veto at the UN Security Council blocked
the renewal of its mandate. UNPREDEP was preceded by the UN Protection Force
(UNPROFOR), first deployed in January 1993.
67 “Skopje, Paris Agree to Limit EU Military Mandate in Macedonia,” RFE/RL Newsline,
3 March 2003.
ethnic Macedonians and ethnic Albanians can still live together.68 Efforts must,
therefore, focus on rebuilding trust between ordinary citizens, as dialogue at the
local level will be indispensable for repairing the huge damage to interethnic rela-
tions caused by the conflict. A related step in conflict resolution will be to build
on the existing (albeit tenuous) loyalties of both groups to the common state.
Here, the positive example of restraint shown by local leaders in some multiethnic
towns (e.g., Kumanovo) during the conflict offers grounds for cautious optimism,
as does the orderly conduct of the September 2002 parliamentary elections.
VIII Conclusion: An Arrangement Highly Dependent on International

Support
Well over a year after the signing of the Framework Agreement, Macedonia is
at best in the very early stages of “the peaceful and harmonious development of
civil society” that the document seeks by its own account to promote (Preamble).
This being the case, it would be premature to call Macedonia – based on the
provisions of the Framework Agreement – a successful complex power-sharing
arrangement. Moreover, the provisions of the Framework Agreement are such
that even their full and consistent implementation would not result in such an
arrangement. These facts notwithstanding, however, the Framework Agreement
remains at the centre of political life in Macedonia, and its continued implemen-
tation seems to provide the most favourable prospects for the country’s develop-
ment as such. Here, the international community has a central role to play in
maintaining a regional environment conducive to the Framework Agreement’s
implementation.
The conflict of 2001 has demonstrated that Macedonia exists at the mercy of
destabilizing regional factors, with the situation of regional interdependence
heightened by political instability and tense interethnic relations within Mace-
donia itself. Outstanding issues, such as the status of Kosovo, should therefore
be resolved as soon as possible in order to enhance regional stability. In light
of Macedonia’s precarious regional position, the continued implementation of
the Framework Agreement depends in large part on international assistance. In
particular, a NATO presence will likely be necessary to help Macedonian and
Yugoslav forces to control the border regions and to prevent a build-up of forces
by the NLA and other groups, as well as to stem arms trafficking, the drug trade,
and other forms of illegal business. This task could also be carried out through a
revival of UNPREDEP or another multinational force.
Given popular views of NATO in Macedonia, the need for an international
security presence must be explained to the population by political leaders on all
68 60% of both communities think that the two groups can live together in the future.
22% of ethnic Macedonian respondents do not believe in peaceful coexistence; 40%
of ethnic Albanians did not know or did not respond at all (poll cited in “Macedonia:
Peace Talks Successful, Civil War Inevitable?,” RFE/RL Balkan Report, 5.57, 7 August
2001).
sides. At a minimum, this means presenting the positions of all parties in both the
Macedonian- and Albanian-language media. Another way in which the interna-
tional community can assist in enhancing Macedonia’s internal as well as external
stability is by recognizing the country under its constitutional name, “the Repub-
lic of Macedonia” (particularly given repeated indications from Greece that the
name dispute is near resolution). Settlement of this longstanding issue would
greatly contribute to reducing the insecurity of the Macedonian people, thus im-
proving the chances for success of any power-sharing arrangement. Additionally,
Macedonia needs international financial assistance and expertise to enact reforms
and to fight corruption, with financial and economic assistance packages needed
to ensure resources adequate for implementing the provisions of the Framework
Agreement as well as to bring about an improvement of the economic situation
of all citizens. The end goal of this ensemble of measures should be Macedonia’s
integration into European structures, with continued international aid part and
parcel of the integration process.69
69 A significant step was the signing of a Stabilization and Association Agreement on 9

April 2001 between the EU and Macedonia. The Stabilization and Association Proc-
ess was initiated by the EU during the NATO campaign in Kosovo as an upgrading of
the EU’s Regional Approach to Southeastern Europe.
Chapter 10
Gagauzia and Moldova:
Experiences in Power-sharing
Priit Järve
I Introduction
Several syndromes are haunting the post-communist countries. One of the most
paralysing is the ‘weak state syndrome’. As George Schöpflin points out, with the
collapse of communism not only the ideology but also the communist state dis-
integrated. Despite their questionable legitimacy, the institutions that the com-
munist state had established had provided a degree of order and coherence seen
as ‘normal’ by many. Attempts to establish the authority of post-communist po-
litical formations could neither rely on the state, which had disintegrated, nor on
civil society, which did not exist under communism (Schöpflin 2001: 8-9). Hence,
post-communist politics and social-cultural coherence came to depend heavily
on ethnicity, the third classic raison d’etre of traditional European polities.
This point of departure for post-communist politics could not but give rise to
feeble state institutions fortified by ethnonationalist rhetoric. This combination
did not produce the order and coherence required for the normal and peaceful
development of these societies. Instead, the ethnocultural preferences of ethnic
majorities, imposed on the whole population in certain multiethnic countries,
induced claims of smaller ethnic groups to share power through autonomy or to
become independent altogether.
The weak state institutions have been unable to defuse the resulting ethnopo-
litical tension and conflicts. In addition to initial ethnocultural disagreements,
these conflicts have been fuelled by the opposing economic interests of emerging
ethnic elites in the distribution of state property through privatization or other-
wise. As weak state institutions seek to establish themselves by claiming as much
power as they can, the atmosphere remains unfavourable for resorting to power-
sharing. For example, the aversion to territorial autonomies in Central and East-
ern European countries is overwhelming.
The Republic of Moldova, a post-Soviet country, reflects this general pattern.
The ‘weak state syndrome’ has plagued Moldova ever since its independence in
1991. At the same time, Moldova is exceptional because in 1994 it legislated au-
308 Priit Järve
tonomy for the territorial unit of Gagauzia (Gagauz-Yeri) implying codified pow-
er-sharing between the centre and the autonomous region. However, the con-
ception of the Gagauz autonomy was closely related to the armed ethnopolitical
conflict in the Moldovan region of Transdniestria in 1992, which deeply shocked
the whole country. Thus, the creation of this autonomy was not a manifestation
of a noble nationwide wish to be democratic and to share power, but rather one
that sought to avoid the worst, a recurrence of the Transdniestrian scenario in
Gagauzia.
This aim has been achieved. However, the six years of Gagauz autonomy have
shown, as some predicted, that the relations between the national government
and the authorities of the autonomous region could not escape fundamental dis-
agreements over vital aspects of the power-sharing. A weak state seems to be
inherently unable to provide for a strong, well-functioning autonomy that enjoys
clear-cut power-sharing arrangements with the centre. Logically, a weak state
can only support a weak autonomy, as will be shown is the case for Gagauzia and
Moldova.
Another lesson of the last six years is that the interaction between Gagauzia
and Transdniestria has remained an important factor in the domestic politics of
Moldova. In 1994, it was the painful experience of the bloody conflict in Transdni-
estria that inspired the Moldovan authorities to accept the Gagauzian autonomy.
Later, it was the ineffective implementation of the autonomy that gave Transdni-
estria reason for its refusal to negotiate any similar status within Moldova with
the central government.
II Brief History
A Outline of the Self-determination Dispute

There are two interrelated contexts that are vital for understanding this conflict
and its settlement in Gagauzia. The broader context was an upsurge of national-
ism in the so-called national republics of the Soviet Union before and after its
Meaning the ‘land of Gagauz’ in Gagauzian and is locally used as the synonym for
‘Gagauzia’.
In the summer of 1992, a full-fledged armed conflict broke out in the Dniester river
valley. The army of the newly independent Republic of Moldova tried to clear the city
of Bendery of forces from the breakaway Transdniestrian authorities. After a week
of fierce fighting in and over this town in June 1992, the 14th Army of the Russian
Federation, stationed on the eastern bank of the Dniester, intervened and ended by
force the fratricidal operations of Moldovans and Transdniestrians. To the present
day, no exact casualty figures exist. It is estimated that some 1,000 people were killed
and some 100,000 were forced to flee their homes. In July 1992, a peacekeeping con-
tingent of the Russian Federation was stationed along the Dniester Valley to con-
trol a Security Zone 225 kilometres long and 4 to 15 kilometres wide together with
Moldovan and Transdniestrian army units. Since then, the conflict has remained
frozen, though by no means resolved (Järve 1998: 6).
10 Gagauzia and Moldova: Experiences in Power-sharing 309
dissolution, which did not leave Moldova untouched. The specific context was
provided by an armed conflict between the central government of Moldova and
the breakaway region of Transdniestria, which was to encourage a more con-
structive approach into the government’s policy towards its ethnic minorities and
ethnopolitical conflicts.
Allen H. Kassof, President of the Project on Ethnic Relations, notes that: “In
every two-sided interethnic conflict there are at least four sides” (Kassof 2001: 2).
He refers to the fact that both sides are usually divided, typically into ‘hawks’ and
‘doves’. The hawks from either side will thus attack any agreement between the
doves. Usually, the hawks initiate the conflict and moderates or doves are needed
to reach an agreement to regulate it. This is exactly what happened during the
conflict in Moldova between Chisinau, the capital of Moldova, and Comrat, the
largest town and administrative centre of Gagauzia.
In 1989, the Popular Front of Moldova (PFM) started demanding that the Mol-
dovan language (practically the same as Romanian) be constituted as the only
state language of Moldova. At that time, similar demands were voiced in all the
republics of the USSR by the respective popular fronts. Estonia had been the first
to adopt a law on its state language in January 1989 (Grin 1991). Moldova fol-
lowed suit on 31 August 1989. However, the PFM was not only anti-Moscow and
anti-Russian, but also pro-Romanian, openly advocating Moldova’s merger with
its neighbour. The non-Moldovan nationalities, who comprised more than one
third of the country’s population, including the Gagauz, perceived this language
policy as a threat to their cultural and political status quo.
As a result, the non-Moldovans joined forces and started to pursue separatist
policies. The two regions in Moldova with a high concentration of non-Moldo-
vans – Transdniestria in the east and Gagauzia in the south of the country – op-
posed the policies of the PFM and demanded territorial autonomy. These claims
produced two different, but interrelated, sequences of events, which have had a
major impact on the development of Moldova ever since, turning it into an area
of lingering ethnopolitical conflict.
The Gagauz are a small Turkic-speaking (Gagauz) Orthodox Christian people,
settled in what is now the southern part of Moldova, a territory of 1,831 square
km (5.5% of the territory of Moldova). In 2000, there were 161,100 inhabitants in
Several years later, the Popular Front of Moldova was characterized in the following
way: “The Popular Front of Moldova (PFM), a political force associated with dra-
matic changes in the society, focused on ethnic problems. It promoted attention to
these specific problems without contributing to the consolidation of the society. The
achievement of a political goal in that period was associated, as a rule, with the no-
tion ‘defeat your enemy’. Any hint of the need for dialogue or reasonable compromise
was interpreted as treason and rejected from the start. (UNDP 2001a: 29).
According to the census of 1989, Moldovans (Romanians) made up 64.5% of the pop-
ulation of Moldova, followed by Ukrainians (13.8%), Russians (13.0%), Gagauz (3.5%),
Bulgarians (2.0%), Belarussians (0.5%), Roma (Gypsies) (0.3%), Germans (0.2%), Poles
(0.1%), and others.
310 Priit Järve
Gagauzia, of which over 81.4% were ethnic Gagauz (3.8% of the country’s total
population). Most (92%) consider Gagauz as their native language, but 73% also
use Russian as a second language. Only 4%, however, speak Moldovan (Roma-
nian). Having no kin-state, the Gagauz started to claim autonomy in 1989, but the
moves to establish a separate administration were suppressed by Moldova. The
confrontation deepened after the Republic of Moldova declared its independence
on 27August 1991.
According to the assessment of the OSCE (OSCE 1994a), shared by many ex-
perts, the language issue lies at the very origin of the conflict in Moldova. Simi-
larly, in Transdniestria, the language legislation of 1989 is widely regarded as the
cause of the subsequent political troubles and the armed conflict. Long before
the declaration of independence and months before the possibility of unification
with Romania was publicly discussed, the Moldovan language legislation became
the clear signal for a process of emancipation from the Soviet legacy.
On 30 August 1989, the Constitution of the Moldavian SSR was amended by
Article 70, which decreed Moldovan (Romanian) written in the Latin alphabet as
“the state language.” Russian was described as the interethnic “language of com-
munication,” and the language of the Gagauz population was to be protected and
developed. On the following day, the Law on the Use of Languages on the Territo-
ry of the Moldavian SSR was passed, stating that Russian would be used through-
out Moldova on the same footing as Romanian, and that Romanian, Gagauz, and
Russian would be the “official languages” in areas with a predominantly Gagauz
population. The use of various minority languages (such as Ukrainian, Bulgarian,
Yiddish) was also guaranteed. Article 7 of the law prescribes that persons holding
positions in state administration and public organizations, which bring them into
contact with citizens (public health, education, culture, mass media, transport,
trade, services, etc.) must know Romanian, Russian, and, in areas with a Gagauz
population, also Gagauz at a “level necessary for fulfilling their professional ob-
ligations.” Article 7, although seemingly balanced, had an asymmetric impact, as
practically all Romanian speakers knew Russian, but not all Russophones were
able to speak Romanian. Therefore, this article was perceived by Russian speakers
on both sides of the Dniester river as an existential threat to their culture.
By early 1992, Chisinau had practically lost its sovereignty over the territories
of Transdniestria and Gagauzia. In March 1992, an armed conflict in Transdni-
estria broke out when the Dniester National Guard attacked a Moldovan police
station in the town of Dubossary. The Moldovan president ordered counter at-
tacks against the Transdniestrian separatists. Fighting spread quickly and also es-
calated in intensity. After weeks of fighting and many hundreds of casualties, the
Russian 14th Army, which had remained stationed in Moldova from the Soviet
period and had been officially neutral, under the leadership of the late General
According to the data of the Gagauz local administration, in 2000, 81.4% of the popu-
lation of Gagauzia were Gagauzians, 5.2% were Bulgarians, 4.6% were Russians, 4.4%
were Moldovans, 3.3% were Ukrainians, and 1.1% were of other ethnic origin (UNDP
2001b: 20).
Lebed intervened and stopped the armed conflict. In July 1992, the presidents of
Moldova and Russia signed a ceasefire agreement in Moscow. The same month, a
peacekeeping contingent of the Russian Federation was stationed along the Dni-
ester valley, on both sides of the river.
In the southern territories of Moldova where the Gagauz proclaimed and set
up their own administrative institutions, the tension had already reached its peak
in 1990. Nevertheless, the chances of resolving the conflict in Gagauzia by peace-
ful means were greater than in the case of Transdniestria, as armed clashes only
occurred occasionally and did not escalate. Consequently, the death toll remained
very low.
In February 1994, the pro-Romanian PRM suffered a major defeat at the par-
liamentary elections. Then, in March 1994 over 90% of the population voted in
favour of an independent Moldova in a consultative referendum. This meant that
the voters were against a union with Romania. The newly formed coalition gov-
ernment, consisting of anti-unionist forces, was much more receptive to the con-
cerns of the two breakaway republics than the previous one. It was also keen on
resolving at least one of the ongoing disputes as soon as possible. A new Consti-
tution of the Republic of Moldova was adopted in July 1994; Article 111 provided
for the possibility of autonomy for the two regions of Moldova without naming
them.
On 23 December 1994, the parliament of Moldova adopted an Organic Law on
the Special Legal Status of Gagauzia (Gagauz-Yeri) (hereafter referred to as the
1994 Law), which constituted Gagauzia as an autonomous territorial unit, grant-
ing the Gagauz people a form of self-determination within Moldova. In the legal
system of Moldova, the organic law occupies a status between the Constitution
and the ordinary laws. The 1994 Law is the most important legal document that
defines, although in part ambiguously, the power-sharing between the central
government of Moldova and the autonomous region. Since 1994, the autono-
mous region has been busy constituting itself. It elected its governing bodies and
officials. By 1998, the People’s Assembly of Gagauzia, the highest legislative body
of the autonomous region, had already passed more than fifty local laws.
On 14 May 1998, the Gagauz People’s Assembly passed a Basic Law for Gagau-
zia. This Code of Gagauzia (Gagauz Yerin Temel Kanonu in Gagauz, Regulamen-
tul Gagauz Yeri in Moldovan) specifies the rules laid down in the 1994 Law. So
far, the Gagauzia power-sharing arrangement promulgated by the 1994 Law is the
only case in Central-Eastern Europe and the former Soviet Union where territo-
rial autonomy has been established for an ethnic group.
B Main Protagonists
The main protagonists are the different political movements (such as the Popular
Font of Moldova (PFM) and the movement of Gagauz People) and elite groups
representing the perceived interests of Moldova as a whole, on the one hand,
Gagauz People (Gagauz Khalki) was the first Gagauz national movement.
312 Priit Järve
and those of Gagauzia, on the other hand. At the initial stage of the conflict, the
respective movements played a very important role in mobilizing people on a
large scale, creating paramilitary units, shaping the general political atmosphere
and legitimizing the leaders of these movements. At a later stage, the opposition
to these mass-led movements was transformed into an elite-led debate over the
conception, articulation, and implementation of Gagauz autonomy.
These protagonists, movements, and elites alike, were, in fact, the agents as
well as the results of the disintegration of the Soviet Union. Both sides were rep-
resented by individuals who had been willing and able to become or stay politi-
cally active under Gorbachev’s policies of perestroika and glasnost. The Gagauz-
ian side mostly included ethnic Gagauz, whose everyday language tended to be
Russian. Many had held high positions in local Soviets and Communist Party
organizations. People with similar Soviet political experience and affiliations
took over the leading positions on the Moldovan side, though, in contrast to the
Gagauzians, they knew Moldovan. Moldova was one of the cases that showed
how the communist cadre policy of promoting the representatives of titular eth-
nic groups and/or nations (the policy of korenizatsiya, or ‘indigenization’) had
grave unintended consequences for the Soviet Union when it found itself in a
deepening crisis by the end of the 1980s. As long as the Soviet system was able
to deliver economically, the ethnic communist leaders were contributing to the
functioning of this system – which could itself be seen as a sort of power-sharing
arrangement between Moscow and the so-called national republics. However, in
the serious crisis of 1991, these local leaders were able to switch to an ethnic al-
ternative and replace ‘Soviet values’ with particular local interests. The existence
of this ethnic alternative largely explains not only the stunningly swift and almost
non-violent disintegration of the Soviet Union but also the subsequent emer-
gence of ethnopolitical conflicts on its former territory.
Over time, the composition of the opposing groups has changed, some of this
being due to intra-elite competition, and moderate politicians have been included
(Crowther 1998). The conflict between Chisinau and Gagauzia transformed into
a political and economic debate within and over the legal framework provided by
the 1994 Law.
Today, the main actors on the Gagauz side are the Governor of Gagauzia
(Bashkan) and the other leading members of the regional administration, as well
as the Chairman of the People’s Assembly of Gagauzia and other members of the
Assembly. On the Moldovan side, all the main political institutions, such as the
president, the parliament, political parties, and the members of government can
be seen as regular participants in the debate, depending on the nature of the is-
sues being discussed.
On the initial stage of the korenizatsiya policy in Moldova, see King 1998.
C The International Community

Moldova became a member of the Conference for Security and Cooperation in
Europe (CSCE) on 30 January 1992, a member of the UN on 2 March 1992 and a
member of the Council of Europe (CoE) on 13 July 1995. Since 1992, the Republic
of Moldova has been monitored and assisted by the international community
(the CSCE Mission to Moldova was started in 1993) largely because of its eth-
nopolitical conflict with the breakaway region of Transdniestria. Since the end of
the armed conflict there, political relations between the government of Moldova
and the Transdniestrian authorities have remained deadlocked. At the same time,
attempts were undertaken to defuse the simmering ethnopolitical conflict with
Gagauzia. An ad hoc Parliamentary Commission was created to work out a ‘Sta-
tus for Settlements Inhabited Predominantly by Gagauzes’. Given the novelty and
sensitivity of the issue, Moldova asked the CoE and the CSCE experts to evaluate
the draft law on the special status of Gagauzia, after its first reading in the parlia-
ment of Moldova in July 1994.
The response of the international experts was predominantly negative. In their
opinion, the rights that provided for Gagauz autonomy were too extensive and
liberal. For example, while the draft law described Gagauzia as a subject of in-
ternational law, the legal status of the non-Gagauz in an autonomous Gagauzia
remained unclear, which was likely to lead to problems with non-Gagauz minori-
ties. The experts thus proposed that to help resolve its ethnopolitical conflicts
Moldova should adopt a general law on national minorities.
The Gagauz resolutely objected to such an approach. They stressed that the
Gagauz, having no kin-state, are not a national minority, but a people who enjoy
the right to state independence. However, they emphasized that they would not
insist on Gagauz independence so as not to undermine the territorial integrity of
Moldova. Instead, they would be satisfied with an autonomy arrangement.
In 1994, the government of Moldova decided in favour of territorial autonomy
for the Gagauz. Therefore, the efforts of international experts focused on suggest-
ing changes that would limit the autonomy rights, on the one hand, and enhance
the rights of other ethnic groups in Gagauzia, on the other. These suggestions
were forwarded to the ad hoc Parliamentary Commission, which then sought to
incorporate them before the second and final reading of the basic law in Decem-
ber 1994. At first, the hardliners among the Gagauzian leaders objected to these
changes, but later they agreed, obviously considering the autonomy arrangement
as far more important than the details of it.
The Gagauzian side presented these views to the OSCE High Commissioner on Na-
tional Minorities during his visit to Gagauzia in December 1994. I am indebted to
Professor Stefan Troebst of the University of Leipzig for kindly providing me with his
manuscript (Troebst 1994). In December 1994, Professor Troebst was a member of
the CSCE Mission to Moldova and accompanied the HCNM to Gagauzia.
314 Priit Järve
III The Origins of the Power-sharing Agreement
A The Route to the Current Agreement

Chinn and Roper date the beginning of the journey towards autonomy to the
early 1980s, when a small number of Gagauz intellectuals created a cultural club
to discuss how to preserve the language and culture of their people. During these
years, the Soviet authorities prosecuted such activities (Chinn and Roper 1998:
91). The political atmosphere changed under Gorbachev’s perestroika policy and
in May 1988 the first Gagauz nationalist movement, the Gagauz Khalki (Gagauz
People) was founded. Initially, it supported the Popular Front of Moldova but
moved to oppose it after the adoption of the language law in August 1989. The
Gagauz felt threatened by this legislation and the ensuing politics of ‘Romaniza-
tion’ and ‘de-Russification’ pursued by the Moldovan authorities.
In this situation, the Gagauz leadership sought closer relationships with pro-So-
viet forces. On 12 November 1989, delegates and local Gagauz assembled in Com-
rat to proclaim a Gagauz Autonomous Soviet Socialist Republic (ASSR) within
Moldova. The Moldovan Supreme Soviet immediately rejected this proclamation.
The initial months of 1990 saw Chisinau providing little economic or cultural aid to
the Gagauz-populated regions, so the Gagauz delegates and officials assembled in
Comrat on 22 June 1990 and once again declared the creation of the Gagauz ASSR
within Moldova. Again, no substantial practical moves were immediately initiated
to validate this claim and no violence occurred (Thompson 1997: 91–92).
On 19 August 1990, the Gagauz leadership proclaimed a Gagauz Soviet Social-
ist Republic, which would be independent from Moldova, but part of the Soviet
Union. They also decided to hold elections for the Supreme Soviet of the newly
self-proclaimed republic on 28 October 1990. In turn, on 26 October 1990, the
Supreme Soviet of Moldova decreed an emergency situation in the southern part
of Moldova (where the Gagauz live). To prevent the elections taking place, the
Moldovan government mobilized approximately 40,000 armed volunteers and
moved them to the South.
In response, the Gagauz organized their own voluntary paramilitary units. An-
other column of volunteers arrived from Transdniestria, through the territory of
the Ukraine, to support Gagauzia. Fortunately, this potentially dangerous confron-
tation between the Gagauz and Transdniestrian forces, on one side, and the Mol-
dovan volunteers, on the, other did not develop into an armed conflict. The Gagauz
leaders appealed to Moscow to send forces to the region. Moldovan police and So-
viet Interior Ministry troops managed to stop the Moldovan volunteers, and thus
prevented bloodshed. At the negotiations, in which the Commander of the Soviet
interior forces participated, an agreement was reached to withdraw both the Mol-
dovan and Gagauzian volunteers. Shortly afterwards, Moldova dissolved its volun-
tary units. These developments, however, alienated the Gagauz even further from
the Moldovans. The central authorities effectively lost control over the Gagauz area
around Comrat and some Moldovan policemen were killed in 1991 and 1992 dur-
ing Gagauz attacks on Moldovan police stations. It was only in the autumn of 1992,
after the armed conflict in Transdniestria had ceased, that political negotiations on
autonomy status for Gagauzia were initiated (Neukirch 2000).
Until then, all declarations and decisions made by the Gagauz between 1989
and 1991 had been systematically overruled or banned by the Moldovan Supreme
Soviet. The latter’s proposal in 1991 to create a special Gagauz ‘county’ was, in
turn, rejected by the Gagauz, as this would have deprived them of the right to
make decisions on political matters and legislative issues.
The situation changed after Moldova suffered the 1992 shock of armed conflict
with Transdniestria. In 1993, the Moldovan parliament discussed several draft
laws on a special status for Gagauzia. However, the pro-Romanian members of
the parliament rejected these drafts as too liberal. Only after a new, and less pro-
Romanian, government and parliament was elected in 1994, was a more con-
structive discussion of Gagauz autonomy started, which led to the adoption of
the 1994 Law.
B Role of International Actors

There has been no extensive international involvement in the management of the
Gagauzia conflict. Some authors have claimed that the success of the arrange-
ment was indeed partially due to the fact that it remained bilateral (Chinn and
Roper 1998: 97). However, sporadic international involvement can be identified.
For example, the CoE initially stated that Gagauzia was not entitled to the right
of self-determination. The CoE and the CSCE repeatedly stated that Chisinau
should solve the Gagauz problem by means of introducing general minority leg-
islation. Most conspicuously, the mandate of the CSCE Mission to Moldova, es-
tablished in April 1993 in the wake of the Transdniestrian conflict, did not even
mention Gagauzia.
However, on 13 November 1993, soon after its establishment, the CSCE Mis-
sion to Moldova issued its Report No. 13 (CSCE 1993). This report mostly pro-
posed ways for resolving the conflict between Chisinau and Transdniestria, but it
also made references to Gagauzia, which demonstrate that despite Transdniestria
being the main focus, the Mission had a holistic approach to Moldova.
In the Transdniestrian conflict, the CSCE tried, naturally, to encourage both
sides towards a compromise. Pushing aside the extreme positions, the report re-
jected the Transdniestrian proposal for a ‘confederation’, which presupposed the
existence of separate states and aimed at distributing all power among them at
the expense of the centre. Neither did the report attempt to please those circles in
Chisinau which considered Transdniestria as simply another part of the country
and denounced any attempt to let it benefit from a special status. Instead, the
report developed compromise positions on how to re-establish the territorial in-
tegrity of Moldova by offering Transdniestrians a measurable degree of self-rule.
The report preferred the term ‘autonomy’ to ‘special status’. It underlined that
under the Soviet regime ‘autonomy’ had had little real significance and was thus
compromised. This explains why ‘autonomy’ as a term is generally unpopular with
many Moldovans and why ‘special status’ was preferred by Moldova. However, as
316 Priit Järve
the report pointed out, in the context of a centralized state like Moldova, a ‘spe-
cial status’ for a region, by definition, had to entail autonomy rights because the
term implies that this region is governed less directly, thus leaving more room for
self-rule (autonomy). Moreover, the report claimed that experience showed that
autonomy solutions such as those in South Tyrol, the Spanish Basque country, or
the Finnish Åland Islands could go as far as, or even further than, devolution in
some federal states.
Further, the report referred to a CSCE Expert Meeting on National Minorities
in Geneva in July 1991, which had proposed local and autonomous administration
and territorial autonomy, including the establishment of legislative and executive
organs, as a mechanism for multiethnic cohabitation. It went on to discuss the
possible future structures of the Moldovan state as a whole and assessed that
three federal states, as suggested by some (Transdniestria, Gagauzia, and the re-
maining, but by far largest, part of the country), would create an unbalanced state
structure. Instead, the report suggested to subdivide the country into eight to ten
regions (one or two of them being Transdniestria, one the area around Bendery,
another one the preponderantly Gagauz-inhabited area around Comrat) and to
transfer power from the capital to the new regional structures. In the opinion of
the Mission, a balanced distribution of power between the capital and a region
had to be the essence of any special status.
The report contains a passage which shows that a draft law on the special status
of the territory largely populated by the Gagauz contributed some ideas to the
proposal of the Mission to set up a Special Region of Transdniestria within the
Republic of Moldova. This Special Region was envisaged to be an integral part of
the Republic of Moldova but also enjoying considerable self-rule, having its own
executive, elective assembly, and court.
Thus, while the conflict in Transdniestria was the Mission’s main concern, it
viewed Transdniestria and Gagauzia as similar challenges to the unitary state
structure of Moldova. A comparison of its Report No. 13 with the 1994 Law shows
that some of the ideas and solutions proposed in the report for Transdniestria
were eventually applied to Gagauzia in the 1994 Law. However, the special auton-
omous status to be granted to Gagauzia initially caused objections from the CoE.
It criticized the extensive downward delegation of powers to the autonomous
region (REF/RL Daily Report 1994a).
Although Gagauzia was not part of the CSCE/OSCE’s original mandate, the
organization later expanded its activities to include the autonomous region. In
1994, it was the Moldovan government that decided to invite both the CoE and
the OSCE High Commissioner on National Minorities (HCNM) to comment on
the draft law on the special status of Gagauzia. Both were concerned that grant-
ing territorial autonomy to the Gagauz would create new ethnic minorities in-
side Gagauzia, where substantial numbers of Moldovans, Bulgarians, Ukrainians,
and Russians were living. As discussed previously, a general minority legislation
was considered as an alternative to the autonomy arrangement. There was also a
broader concern that the law would create a precedent for the whole of Europe
where governments would oppose Gagauz-style solutions for providing territo-
rial autonomy to national minorities. Additionally, experts and other international

actors noted that the formulations of the law were too ambiguous on taxation
and other financial matters. Indeed, while the Gagauz example did not lead to a
Europe-wide upsurge of demands for autonomy, Article 18 of the 1994 Law, which
establishes budgetary relations between Moldova and Gagauzia, did cause most of
the later friction between the autonomous region and the central government.
In June 1994, the Moldovan government invited the HCNM to visit Moldova
and evaluate interethnic relations in the country. The visit of the HCNM took
place at the beginning of December 1994, when the second reading of the law
on Gagauzian autonomy in the Moldovan parliament was pending. The HCNM
met with practically all those in Chisinau and Gagauzia who were engaged in the
negotiation process. In Gagauzia, he explained to the local leaders that certain
concessions were necessary to achieve a settlement with the central government
on the special status for Gagauzia. The most important concessions in this regard
were that Gagauzia was no longer described as a subject of international law and
that the legal status of the non-Gagauzes in the future Gagauzia was enhanced by
inserting a new article into the draft law (Troebst 1994: 4). It is believed that these
meetings with a high-ranking representative of the international community con-
tributed to the adoption of the law at the second reading on 23 December 1994.
Apart from the visit of the HCNM, the OSCE monitored the elections and
evaluated the efficiency and functional validity of the electoral system in the au-
tonomy agreement. The OSCE, for instance, assessed the elections for the Gov-
ernor of Gagauzia and the elections for the People’s Assembly of Gagauzia in
August–September 1999.
In 1999, the mandate of the OSCE Mission to Moldova included such responsi-
bilities as: the monitoring of interethnic relations (e.g., post-conflict rehabilitation
in Gagauzia), the provision of advice and expertise, as well as other contributions
including a framework for a political settlement that complied with international
obligations and commitments regarding human and minority rights, or the me-
diation of the disputes or complaints in the human dimension areas in coopera-
tion with local authorities.
It may be argued that there were important inputs from other external actors,
such as the Ukraine and Turkey. The latter played a significant role in the even-
tual arrangement of Gagauzian autonomy by promising investments and devel-
opment assistance for the southern regions. The visit of the president of Turkey,
Suleyman Demirel, to Moldova in 1994, prior to the adoption of the law on the
special status of Gagauzia, has been described as crucial to the resolution of the
Gagauzia issue (RFE/RL Daily Report 1994b).
C Final Negotiations
The negotiations of a special Parliamentarian Commission with the participation of
the Gagauz representatives were initiated in 1993, but the parliament was not able
to decide on the status of Gagauzia as drafted and proposed by the Commission.
After a new parliament was elected in 1994, the work was resumed. On 1 April 1994,
318 Priit Järve
another special Parliamentary Commission (with the participation of the Gagauz)

was formed, which worked on the draft law on the special status of Gagauzia for
several months, using recommendations from the CoE and the CSCE.
In July 1994, the Moldovan parliament adopted the draft law at its first reading
and on 23 December 1994, after it was amended following the recommendations
of international experts, it was adopted at the second reading. Finally, Gagauz
autonomy had been formally recognized.
IV Initial Reaction to the Agreement

The agreement, once reached by the adoption of the 1994 Law, was met with
positive evaluations both in Moldova and abroad. In 1997, one senior Gagauz of-
ficial characterized the agreement on Gagauz autonomy as a breakthrough in the
domestic affairs of Moldova, and as “an act of wisdom of the peoples of Moldova
and Gagauzia” (Järve 1998: 21).
Pro-Romanian politicians in Moldova, on the other hand, strongly objected to
the right of the Gagauz to external self-determination in the event of Moldova
changing its status (1994 Law Art 1). In 1994, the representatives of the oppo-
sition ‘Gagauz People’s Party’ told the HCNM in Comrat that their party was
against territorial autonomy for the Gagauz, accusing the then Gagauz leadership
of being in close contact to red-brown forces (Zhirinovskii’s Liberal-Democratic
Party and Zyuganov’s Russian Communist Party) in Moscow. They described the
Gagauz as a national minority (Troebst 1994: 9). However, the influence of these
groups remained marginal.
International involvement continued in connection with the 1998 adoption
of the Legal Code of Gagauzia (regarded by the Gagauz as their constitution),
which was written by the Gagauz and reviewed by a number of experts at home
and abroad. On 17 March 1998, the Moldovan Supreme Court nullified the deci-
sion of the People’s Assembly of the autonomous region to hold a referendum
on a constitution for the region at the same time as the 22 March elections to
the Moldovan parliament. Following that decision, the Central Electoral Com-
mission revoked its earlier decision to allow the referendum. A spokesman for
the commission said the People’s Assembly had failed to bring the draft regional
constitution into line with Moldova’s Basic Law. However, after receiving opin-
ions from the CoE,10 the Moldovan Ministry of Justice, and independent lawyers
The legal argument for stopping the referendum was that the draft of the Code of
Gagauzia had been distributed only in Russian, thus violating the law on languages,
and also only in a small number of copies, so that the electorate was not sufficiently in-
formed about its substance. Thus, a translation of the Russian version into Moldovan
and Gagauzian was done and as a result the Code of Gagauzia happens to be the only
legal act of Gagauzia also available in Gagauzian and Moldovan, as all other legislation
has been published in Russian only. (I owe this clarification to Claus Neukirch.)
10 On 7 May 1998, the Venice Commission concluded in its “Opinion on the Legal Code
of Gagauzia” that the draft Legal Code constituted a good basis for defining the rules
from/in the USA – all of whom agreed that the document did not contradict the
Constitution and was in line with international provisions on local autonomy
– the People’s Assembly of Gagauzia unanimously voted in favour of the region’s
Basic Law on 14 May 1998.
The active cultural and economic involvement of Turkey continued after the
autonomy agreement was adopted. In June 1998, the Turkish President Suleyman
Demirel came to Moldova to discuss bilateral relations and economic cooperation
with his host, President Petru Lucinschi, and to visit Gagauzia. Demirel thanked
his hosts for the way Moldova had solved the problem of Gagauz autonomy, saying
that the Gagauz population is “a bridge” in the two countries’ friendship. Demirel
announced that Turkey had agreed to grant Moldova a usd 35 million credit, of
which usd 15 million was to be used for improving the water supply system in
Gagauzia (RFE/RL 1998; Turkish Press Review 1998). In fact, Turkey has played
the role of surrogate kin-state – due to cultural and language proximity with the
Gagauz – both before and after the power-sharing agreement was concluded.
At the initiation of the autonomy agreement, experts pointed to the problem of
borders of the autonomous region (the settlement pattern of the Gagauz is geo-
graphically dispersed) and to the vaguely-defined taxation and other economic
matters. The border issues, somewhat surprisingly, have not caused any major
trouble. At the same time, economic matters, including taxation, have proved
to be major stumbling blocks on the way to harmonious relations between the
national centre and the autonomous region, causing heated debates and disagree-
ments on how Article 18 of the 1994 Law should be interpreted.
V The Power-sharing Arrangement and Its Implementation

The power-sharing arrangement between Chisinau and Gagauzia (Comrat) is
based on the Moldovan Constitution, Article 111 of which stipulates the possibil-
ity of territories with a special status in the country.11 An explicit and detailed ar-
governing Gagauz autonomy. Yet, the Commission pointed out that the draft Code
regrettably incorporated numerous provisions derived from other legal texts that
had simply been copied. The result is that the draft Code covers matters which in
actual fact fall outside the jurisdiction of Gagauzia, raising doubts in the reader’s
mind as to which is the right text (the Legal Code or the text copied). The Commis-
sion recommended that all the provisions copied from other texts be systematically
expunged from the draft. See http://www.venice.coe.int/site/interface/english.htm).
11 Article 111 “Special Autonomy Statutes” of the Moldovan Constitution reads:
(1) The places on the left bank of the Nistru river, as well as certain other places in
the south of the Republic of Moldova may be granted special forms of autonomy
according to special statutory provisions of organic law.
(2) The organic laws establishing special statutes for the places mentioned under
paragraph (1) above may be amended if three fifths of the parliament members
support such amendments.
Available at http://oncampus.richmond.edu/~jjones//confinder/moldova3.htm#C8”.
320 Priit Järve
rangement of the autonomy of Gagauzia is laid down in the 1994 Law, which can
only be amended by a three-fifths majority of the Moldovan parliament.
The preamble of the 1994 Law says that the law is adopted “with the aim of
satisfying the national needs and preserving the national identity of the Gaga-
uzes, bringing about their full and comprehensive development, fostering their
language and national culture, and assuring their political and economic self-reli-
ance.”12 It also confirms “the equality of rights of all citizens who live on the terri-
tory of the autonomous territorial unit, which is being created without regard to
nationality or other distinctions.”13
Article 1 of the organic law stipulates:
(1) Gagauzia (Gagauz-Yeri) is an autonomous unit, with a special status as a form

of self-determination of the Gagauzes, which constitutes an integral part of the
Republic of Moldova.
(2) Gagauzia shall, within the limits of its competence, resolve by itself questions of
political, economic and cultural development in the interests of all its popula-
tion.
(3) All rights and liberties defined in the Constitution and legislation of the Republic
of Moldova shall be guaranteed on the territory of Gagauzia.
(4) In the case of a change of the status of the Republic of Moldova as an independent
state, the people of Gagauzia shall have the right of external self-determination.
A change of the status of Moldova as an independent state refers to Moldova’s

mooted unification with Romania, one of the main aims of the Popular Front – an
idea which is still alive (Prihodko 2001). For the Gagauz, who were and remain
against such a union, it was important to have the right of external self-determi-
nation reserved for this eventuality. The radical elements on the Moldovan side,
particularly the Popular Front, have always objected to this right of the Gagauz.
A ‘Vertical’ Aspects of the Agreement

1 Division of Power
According to its Constitution, the Republic of Moldova is still a unitary state,
because Article 111 of the Constitution stipulates only the possibility of a special
status for certain regions on the left bank of the Dniester river and in the south
of the country. The Gagauz autonomy is not mentioned in the Constitution, as it
was adopted before the special status was created for Gagauzia. There have been
no amendments to the Constitution, which remains a source of serious concern
12 References to the text of the 1994 Law follow the office translation by the OSCE
Mission to Moldova. A Russian version of the 1994 Law is available at http://www.
minelres.lv/NationalLegislation/Moldova/Moldova_Gagauzia_Russian.htm
13 The term ‘nationality’ in the Eastern European context in general, and in the post-
Soviet context in particular, means ‘ethnic origin’, not ‘citizenship’, as is common in
the West.
for the Gagauz. Vertically, the Republic of Moldova has three levels of govern-
ment: the state level (the capital Chisinau), the county level (since 1998 this has
been nine counties, Chisinau municipality, Gagauzia, and Transdniestria), and
the local level. On 27 December 2001, the parliament of Moldova adopted a new
law on the administrative-territorial structure of the country, abolishing the nine
counties created by the 1998 law and restoring the thirty-two rayons of the Soviet
period.
According to Article 7 of the 1944 Law, “the representative authority of Gagau-
zia shall be the People’s Assembly (Halk Toplushu), which shall have the power,
within the limits of its competence, to enact laws.” It is elected every four years
by a universal, equal, direct, secret, and free vote on the principle of one deputy
for each 5,000 voters, with at least one deputy from each locality that belongs to
Gagauzia. Citizens of Moldova who have attained the age of 21 and who live in
the electoral district (locality) they wish to represent on the twenty-one of the
elections may be deputies of the People’s Assembly (Art 8).
According to Article 12(1), the People’s Assembly shall adopt, within the limits
of its competence, local laws by a majority of votes of the elected deputies. Article
12(2) explains that this competence includes the fields of:
a) science, culture, education;
b) housing management and urban planning;
c) health services, physical culture, and sports;
d) local budget, financial, and taxation activities;
e) economy and ecology; and
f ) labour relations and social security.
Further, Article 12(3) stipulates that the competence of the People’s Assembly
shall also include:
a) solving in a legal manner of questions of the territorial organization of
Gagauzia, the establishment and alteration of the categories of localities, the
borders of the regions, towns and villages, and their naming;
b) participation in the implementation of internal and external policy of the
Republic of Moldova connected with the interests of Gagauzia;
c) defining the structure of the organization and activity of local public ad-
ministrative authorities of Gagauzia and of citizens’ associations, with the
exception of political parties and other voluntary political organizations;
d) setting, organizing, and conducting elections of deputies for the People’s As-
sembly and approving the composition of the Central Election Committee
for carrying out the elections; setting elections for local public administra-
tive authorities of Gagauzia;
e) conducting local referendums concerning questions that are within the
competence of Gagauzia;
f ) adoption of regulations and symbols of Gagauzia;
g) creation of titles of honour and the institution of decorations;
h) examining the question and the initiation of a proposal to the parliament of
the Republic of Moldova concerning the declaration of a state of emergency
322 Priit Järve
on the territory of Gagauzia and the introduction under such circumstances

of a special form of administration in order to ensure the protection and
security of the inhabitants of Gagauzia; and
i) the right to appeal in a manner fixed by law to the Constitutional Court of
the Republic of Moldova with a case concerning the voiding of enactments
by the legislative and administrative authorities of the Republic of Moldova
if they infringe on the authority of Gagauzia.
Thus, division of powers at the international level is defined in Article 12(3b) as

a competence of the People’s Assembly to participate “in the implementation of
the internal and external policy of the Republic of Moldova connected with the
interests of Gagauzia.”
The participation of the Gagauz in the government of Moldova has two main
forms. First, the Governor of Gagauzia (Bashkan), who is elected for a four-year
term, becomes a member of the government of the Republic of Moldova after a
decree of the president of Moldova (Art 14). Second, on the recommendation of
the Governor of Gagauzia, the directors of departments of the Executive Com-
mittee (the Gagauzian administration), appointed by the People’s Assembly, shall
become members of the boards of the corresponding ministries and of depart-
ments of the Republic of Moldova (Art 19). In addition, decrees and orders of the
Governor of Gagauzia and the Executive Committee shall be presented to the
government of the Republic of Moldova with a ten-day period from the date of
adoption (Art 17(3)).
In the opinion of the Venice Commission, the extent of the powers conferred
on the Gagauzian autonomous institutions is striking. The Commission stresses
that the People’s Assembly can legislate on almost any important area with the
exception of defence and foreign policy. It is difficult, in the opinion of the Com-
mission, to see that the creation of a legislature in Gagauzia whose laws are ca-
pable of ousting the national laws is consistent with Article 60 of the Constitution
of Moldova in its conferring of sole legislative competence on the national parlia-
ment. The problem also arises with Article 66 of the Constitution, which empow-
ers the parliament to ensure legislative unity of regulations throughout Moldova
(Council of Europe 2002a: Items 25 and 26).
As a remedy, a change from the right to legislative power to the right to legisla-
tive initiative is proposed for the People’s Assembly in the draft Law on Modifi-
cation and Addition in the Constitution of the Republic of Moldova. The Venice
Commission supports this change in its Opinion of 12 March 2002, as this would
bring Moldova into line with most countries with territorial autonomies where
the final decision on legislative initiatives belongs to the national parliament
(Council of Europe 2002a: Items 31).
In other observations of its 12 March Opinion, the Venice Commission notes
that while Article 12 of the 1994 Law gives the People’s Assembly the right to sub-
mit to the Constitutional Court legal acts adopted by the legislative or executive
authorities of the Republic of Moldova that it considers to infringe on the autono-
mous powers of Gagauzia, the authorities of Moldova have no power to submit
any legal act of the People’s Assembly to the Court that these authorities consider
to exceed the powers of the Assembly. The Commission recommends that the
authorities of Moldova should have the power to challenge the constitutionality
of the normative acts of the autonomy.
2 Constitution and Design of the Polity

Starting from 1995, when the 1994 Law entered into force, the Gagauz region
constituted itself as ‘an autonomous territorial unit’, which is supposed to have
a higher status than the counties and be comparable with the municipality of
Chisinau, the capital city.
According to Article 5 of the 1994 Law:
1) Gagauzia shall include localities in which Gagauzes constitute over 50% of
the population.
2) Localities in which Gagauzes constitute less than 50% of the population may
be included in Gagauzia on the basis of the freely expressed will of a majority
of the electorate revealed during a local referendum conducted on the initia-
tive of no less than one third of the electorate of the corresponding locality.
3) Localities referred to in parts (1) and (2) of the present Article shall be in-
cluded in Gagauzia on the basis of the result of the local referendum con-
ducted by the Government of the Republic of Moldova in each community.
4) Localities included in Gagauzia shall reserve the right to secede from
Gagauzia as a result of a local referendum conducted at the initiative of at
least one third of the electorate, but not before one year after the date it was
included in Gagauzia.14
Article 14 of the 1994 Law legislates that:

1) The highest official of Gagauzia shall be the Governor (Bashkan). All public
administrative authorities in Gagauzia are subordinate to him.
2) The Governor of Gagauzia shall be elected on a contested basis for four years
by universal, equal, direct, secret, and free vote.
3) The Governor of Gagauzia must be a citizen of the Republic of Moldova who
has reached thirty-five years of age and who has a command of the Gagauz
language.
4) The Governor of Gagauzia shall be appointed as a member of the Govern-
ment of the Republic of Moldova after a decree by the President of the Re-
public of Moldova.
5) No person shall be elected to the office of the Governor of Gagauzia more
than twice in a row.
6) The Governor of Gagauzia shall direct the activity of the public authorities
and shall be responsible for carrying out the powers vested in him by law.
14 Article 8(6) of the Code of Gagauzia restricts this right of secession only to those
localities where the Gagauzes constitute less than 50%.
324 Priit Järve
7) The Governor of Gagauzia shall, in conformity with the present Law, is-
sue decrees and orders that shall be carried out on the whole territory of
Gagauzia and shall come into force on the day of their publication.
8) The Governor of Gagauzia shall provide the People’s Assembly on an annual
basis with reports on the activity of the public administrative authorities of
Gagauzia.
9) The functions of the Governor of Gagauzia shall be suspended before his
term expires in the event he does not observe the Constitution of the Re-
public of Moldova, the present Law, local laws, and the resolutions of the
People’s Assembly, or if he commits an infraction.
10) A decision to remove the Governor of Gagauzia from office shall be adopted
by a two-thirds vote of the elected deputies of the People’s Assembly, but a
decision to remove other officials of public administrative authorities shall
require a majority of the votes of the deputies of the People’s Assembly.
The executive body of Gagauzia, the Executive Committee of Gagauzia, is leg-

islated by Articles 16 and 17. Article 16 says that the Executive Committee shall
be appointed by the People’s Assembly at its first session for the duration of its
mandate. The Assembly shall appoint the membership of the Committee on the
proposal of the Governor of Gagauzia by a vote of the majority of the deputies.
The Venice Commission has noted that the range of responsibilities of the Ex-
ecutive Committee is equally comprehensive, as it has to implement all acts ad-
opted by the People’s Assembly (Council of Europe 2002a: Item 25).
Article 17 stipulates the following competences of the Executive Committee:
a) the implementation and observance of the Constitution and of the laws of
the Republic of Moldova and of the enactments of the People’s Assembly;
b) participation in the functioning of specialized central administrative au-
thorities of the Republic of Moldova in matters relating to the interests of
Gagauzia;
c) the regulation in conformity with the law on the whole territory of Gagauzia
of property rights and the management of the economy, social, and cultural
development, the local budget and financial system, social security, remu-
neration, local taxation, environmental protection, and the rational use of
natural resources;
d) the definition of the framework and priorities for economic development
and scientific and technical progress;
e) the working out of programmes of economic, social, and national-cultural
development and of environmental protection, and their implementation
after approval by the People’s Assembly;
f ) the drawing up of the budget of Gagauzia, its submission to the People’s As-
sembly for approval, and its execution;
g) the resolution of questions of ecological security, the rational use, protection
and regeneration of natural resources, the setting of quarantines, and the
declaration of zones affected by natural disasters;
h) the drawing up and carrying out of programmes in the areas of education,

culture, public health, physical culture of sports, social security, as well as
protection and use of the historical and cultural monuments;
i) the maintaining of equal civil rights and liberties, national and civic har-
mony, and protection of legality and of public order;
j) the drawing up and promotion of a scientifically valid demographic policy,
and a programme of urban development and housing management;
k) the use and development of the national languages and cultures on the ter-
ritory of Gagauzia.
Article 17 also gives the Executive Committee the right to initiate legislation in
the People’s Assembly. This right has been actively used by the authorities of the
new autonomous region. From April 1999 to January 2001, for example, the Peo-
ple’s Assembly adopted 29 local laws, including laws on the budgets of 2000 and
2001, and nine decrees (Sbornik zakonodatelnych aktov Gagauzii 2001: 221). Ac-
cording to information from the Chancellery of the People’s Assembly of Gagau-
zia, from 1995 through September 2001, all in all, ninety-four local laws have been
adopted.
The government of Moldova played an important role at the beginning of the
implementation of the 1994 Law by organizing the referendums and elections
necessary for the establishment of the Gagauzian autonomous region in 1995.
Later, however, the Gagauz became more and more critical of the role of Chisinau
in this implementation.
3 Electoral Mechanisms
On the same day the 1994 Law was adopted, the Moldovan parliament also
passed the Resolution on Implementation of the Law on the Special Legal Status
of Gagauzia (Gagauz Yeri), which tasked the government with several obligations
regarding the elections to be carried out in the southern regions of Moldova to
constitute the Gagauzian autonomous region.
The government had to draw up and adopt a Temporary Regulation on the Lo-
cal Referendum based on the Law on Local Elections, a Temporary Regulation on
Elections for the Governor of Gagauzia and a Temporary Regulation on Elections
for the People’s Assembly of Gagauzia. The Resolution of 23 December 1994 also
charged the government with organizing and conducting a local referendum in
several localities in the south of Moldova, as well as organizing and conducting
elections for the People’s Assembly of Gagauzia and the Governor of Gagauzia.
The government of Moldova fulfilled these tasks. In March 1995, referendums
were held in thirty-six districts in the south of Moldova. The voters were asked
the following question: “Do you favour your community entering Gagauz Yeri,
which is an integral part of Moldova?” As a result, three cities and twenty-six
villages opted to join the autonomous region. Voter turnout was 79% and inter-
national observers from the CoE, the OSCE, and Turkey judged the referendums
free and fair (Thompson 1997: 101).
326 Priit Järve
In May 1995, elections for the People’s Assembly and for the Governor of Gaga-
uzia were held. In 1999, elections for the People’s Assembly of Gagauzia and the
Governor of Gagauzia were held for the second time, as prescribed by the 1994
Law. As international observers have not noticed major violations of laws and
regulations during any of these elections, one can say that the electoral system in
Gagauzia is functioning normally. Still, the Gagauz would like to have reserved
seats for the autonomous region in the parliament of Moldova. They do have
reserved seats in the executive structures, as discussed below.
B ‘Horizontal’ Aspects of the Agreement

Issues such as communications and media, health and social services, educa-
tion, human and minority rights have not caused significant problems during the
implementation process. However, their practical implementation suffers from a
lack of resources. At the same time, the economy, administration of justice, ex-
ternal relations, and trans-border cooperation are generating political and legal
problems.
1 Communications and Media

In the 1994 Law, communications and media were not explicitly referred to.
However, Article 1 states: “Gagauzia shall, within the limits of its competence,
resolve by itself questions of political, economic and cultural development in the
interests of all its population.” This implies communications and media.
Communications and media follow the linguistic pattern of Gagauzia. All me-
dia operating in Gagauzia tends to be predominantly in Russian. National radio
and television broadcast programmes in Gagauz only once a month. The auton-
omy agreement has not made a difference to the proportion of Gagauz language
programmes on national radio and television.
Communications and media suffer from a lack of resources that can have seri-
ous political consequences. For example, because of technical limitations, it is
not possible to cover the whole autonomous region with one single TV channel.
The Gagauzian authorities suffered from this drawback during the political crisis
of early 2002 when they were not able to communicate their own messages to
the population, while broadcasts from Chisinau could be received without prob-
lems.

Gagauzia is part of the Moldovan healthcare and pension systems and has to
keep to the established standards and costs of providing these services. However,
using its legislative power, and after overcoming some resistance from Chisinau,
it was the first region in Moldova to introduce a health insurance system instead
of a state-provided health care system.
3 Education
The Moldovan Law on the Use of Languages (Art 18) provides that “the state
shall guarantee the right to secondary education, non-specialized intermediate
studies, specialized, technical and vocational intermediate studies, and higher
education in Moldovan and Russian, and shall create the necessary conditions
for observing the rights of citizens of other nationalities who are resident in the
Republic to education and studies in their native language (Gagauz, Ukrainian,
Bulgarian, etc.).”
As Article 3 of the 1994 Law notes, Moldovan, Gagauz, and Russian will serve
as the official languages of Gagauzia. Moldovan and Russian are designated as the
languages of correspondence among the public administration authorities. The
People’s Assembly has the right to adopt local laws on education (Art 12(2a)).
In practice, besides the lack of resources, many problems of the education sys-
tem are related to the poor status of the Gagauz language. The authorities of So-
viet Moldova had introduced Cyrillic script for the Gagauz language in 1957. The
Gagauz language was then unable to establish itself during the following years as
the main language of the Gagauz and is still facing difficulties. In 1989, 87.5% of the
Gagauz claimed Gagauz as their native language, but a 1998 sociological survey
established that only 37.8% of the adult population of Gagauzia knew the written
form of the language, while 44.1% spoke Gagauz with their children at home.
With regard to the language of education for their children, 80.6% of the respon-
dents preferred Russian, 4.6% Gagauz in combination with other languages, 2.6%
Gagauz only, 2.6% English, and 1.4% Moldovan (Kristioglo 2000: 223-224).
Nevertheless, under autonomy a sound legal basis has been created for the de-
velopment of culture, which is regarded as a most significant achievement. Local
experts refer to the positive role in this of Article 27 of the International Covenant
on Civil and Political Rights, to which Moldova is a party. At the same time, they
complain about an overload in the learning of languages. The schools in Gagauzia
teach the three local official languages – Russian, Moldovan, and Gagauz – as
compulsory subjects, accompanied by English or German as foreign languages.
This has two major negative effects: many children cannot learn any of these lan-
guages properly, and the time left for other subjects is insufficient, which impairs
the general level of education that can be provided (Kristioglo 2000: 222-224).
Gagauzian authorities are in charge of their own education system. They have
to follow nationwide standards and programmes, but they are independent in se-
lecting and appointing their school headmasters and teachers. A special research
unit has been created by the Gagauz to elaborate a new concept and launch ‘na-
tional schools’, i.e., schools for the Gagauz, Bulgarians, and Ukrainians, to devel-
op curriculum items and courses in the Gagauz language and related disciplines.
The university and the teachers’ college in Comrat are preparing specialists for
local institutions of education and culture. In this effort, the assistance of Turkey
is highly appreciated by the Gagauz. Still, material conditions in schools remain
difficult. In September 2001, teachers in Gagauzia had not been paid their sala-
ries for three months; electricity in some schools was disconnected due to unpaid
bills.
328 Priit Järve
4 Economic Policies
The legal basis of the economic autonomy of Gagauzia is provided by Articles 6
and 18 of the 1994 Law. Article 6 states:
The land, mineral resources, water, flora and fauna, other natural resources, movable
and real property situated on the territory of Gagauzia shall be the property of the
people of the Republic of Moldova and at the same time shall represent the economic
basis of Gagauzia.
Article 18 of the 1994 Law establishes budgetary relations between Moldova and
Gagauzia in the following way:
2) The local budget shall include any type of payment fixed by legislation by the Re-
public of Moldova and by the People’s Assembly.
3) The mutual relationships of the budget of Gagauzia and of the state budget shall be
established in conformity with the laws of the Republic of Moldova on the budget-
ary system and on the state budget for the corresponding year in the form of fixed
payments out of all forms of taxes and payments.
To enable implementation of Articles 6 and 18, Article 12 authorizes the People’s

Assembly to adopt laws in the areas of local budget, financial and taxation activi-
ties, economy, and ecology.
In 1995, Charles King predicted “as the central government and the new Gaga-
uz administration are beginning to discover, in formulating a comprehensive mi-
norities policy, the devil is always in the details” (King 1995: 25). This has proved
to be right, especially in economic and financial issues, such as the dividing of
competencies in property relations and taxation between the autonomous region
and the centre. However, solutions that would satisfy both sides have not yet been
found and major disagreements over these issues have persisted (Neukirch 2000:
10–12).
A major attempt to find a compromise took place on 11–12 December 2000 in
Chisinau, when the OSCE Mission to Moldova organized a workshop on “Chi-
sinau-Comrat Center-Region Relations in the Fields of Fiscal-Budgetary Policy,
Public Property and Adaptation of Legislation” and secured high-level participa-
tion. This workshop once more highlighted the unresolved issues in the relations
of Gagauzia and the central government (OSCE 2000d).15
As observed in the workshop materials, one of the major issues of debate was
Article 18 of the 1994 Law. From the very beginning, the discussion in the work-
15 I am thankful to Mr Gottfried Hanne, Member of the OSCE Mission to Moldova, for

providing me with the materials of this workshop, rendering invaluable assistance
and sharing with me his observations concerning the relations between Chisinau and
Comrat during my visit to Moldova 3–10 September 2001. This sub-section draws on
the materials of the mentioned workshop.
shop concentrated on the sources from which the Gagauz regional budget should
be drawn. The Gagauz side expressed the view that the sources of the region’s
budget should be all the taxes and payments collected on the territory of Gagau-
zia, plus customs and excise taxes for goods consumed in Gagauzia. According to
the Gagauz point of view, the reasonable needs of the centre should be estimated
and all revenues left should remain in the autonomous region.
Moldovan government representatives, on the other hand, stressed that it was
technically impossible to estimate the customs and excise taxes collected for
goods consumed in Gagauzia, and that customs, excise tax, and VAT are usually
revenues exclusive to the centre and cannot be considered legitimate revenues of
second level territorial authorities.
The discussion at the workshop showed the incompatibility of the basic posi-
tions. The Moldovan side tended to neglect the special nature of the autonomy
status of Gagauzia in comparison to the status of ordinary counties and main-
tained that there is no need for any special treatment of Gagauzia in comparison
to ordinary counties in financial terms. The Moldovan Finance Ministry held the
position that, after the reform of Moldova’s financial and tax system, Article 18(2)
of the 1994 Law contradicts other, ordinary Moldovan legislation and should be
amended.
The Gagauz side stressed that the position of the 1994 Law as an organic law
was superior to the ordinary finance and tax legislation of Moldova and demand-
ed the full implementation of that law and its Article 18. The Gagauz side viewed
the autonomy status as a sufficiently legitimate basis for Gagauzia to receive more
financial resources than ordinary counties. The proposal to assign to the centre
only those financial means that can be considered reasonably necessary to cover
the centre’s needs, but to hold all other revenues in Gagauzia, gave the impres-
sion that some Gagauz representatives regarded their autonomous region more
as a state in itself than an autonomous region within a state. Such attitudes can
be considered typical given the political fighting over scarce resources when old
Soviet ‘know-how’ – to get one, ask for two – is used by the negotiators. Here it
is employed thus – to get autonomy, ask for a state.
While no agreement on common recommendations or principles was pos-
sible due to incompatible interpretations of Article 18, both the establishment of
a permanent common commission between Chisinau and Comrat – to address
disagreements at an early stage – and of a representative office of the Gagauz
People’s Assembly at the parliament of the Republic of Moldova were supported.
It was thought that these organs could also address the needs of exchange of in-
formation and regular consultations in the financial and economic spheres.
A year later, at the end of 2001, when the Communist Party, which had prom-
ised to solve the problems of Gagauzia, had been in power for almost a year, the
Gagauz were given finally the right to collect excise tax on their territory after the
President of Moldova returned the Law on the Budget of 2002 to the parliament
for amendments. While the transfers of funds from the centre to Gagauzia were
simultaneously reduced, according to the Ministry of Finance’s estimates of the
expected volume of excise tax, the Gagauz remained hopeful that the amount
330 Priit Järve
collected in excise tax would in fact be higher, thus leaving them better off than
before.16
5 Administration of Justice
Articles 20, 21, 22, 23, and 24 regulate the administration of justice and law en-
forcement in Gagauzia.
Article 20 states:
1) The Tribunal of Gagauzia shall be a court of second instance in relation to
inferior courts and shall examine, in the capacity of a court of first instance,
the most complicated civil, administrative and penal cases.
2) Judges of the judicial bodies of Gagauzia shall be appointed by a decree of
the President of the Republic of Moldova on the recommendation of the
People’s Assembly of Gagauzia, with the agreement of the Superior Council
of Magistrates.
3) The Chairman of the Tribunal of Gagauzia shall be an ex officio member of
the Supreme Court of Justice.
Article 21 legislates:
1) The Office of the Prosecutor of Gagauzia shall carry out its duties in accord-
ance with the Law on Prosecutor’s Office of the Republic of Moldova (with
the exceptions foreseen in the present law).
2) The Prosecutor of Gagauzia shall be appointed by the General Prosecutor of
the Republic of Moldova on the recommendation of the People’s Assembly,
and shall be an ex officio member of the board of the Prosecutor General of
the Republic of Moldova.
3) Subordinate prosecutors shall be appointed by the Prosecutor General of the
Republic of Moldova on the recommendation of the Prosecutor of Gagauzia,
with the agreement of the People’s Assembly.
According to Article 22, the Chief of the Department of Justice of Gagauzia is

appointed and dismissed by the Minister of Justice of the Republic of Moldova
on the recommendation of the People’s Assembly. The Chief of the Department
of Information and Security of Gagauzia (Art 23), and the Chief of the Depart-
ment of Internal Affairs of Gagauzia (Art 24) are appointed and dismissed by the
corresponding central authorities on the recommendation of the Governor of
Gagauzia with the agreement of the People’s Assembly.
The local authorities coordinate the municipal police and policy commissari-
ats while Carabineer troops (interior forces) are the responsibility of the central
authorities (Art 24). Responsibility for the appointment and dismissal of senior
police officers is shared between the central authorities and Gagauzia.
The Prosecutor of Gagauzia, the Chiefs of the Departments of Justice, Infor-
mation and Security, and of Internal Affairs are all ex officio board members of
corresponding Ministries of Moldova (Arts 22, 23, and 24).
16 I owe this clarification to Mr Gottlieb Hanne of the OSCE Mission to Moldova.

The administration of justice in Gagauzia is suffering from the fact that the au-
thorities in Chisinau have failed to set up the Tribunal of Gagauzia as part of the
legal system of Moldova. As this tribunal must also confirm the results of elec-
tions in Gagauzia, its absence could be used to declare all elections technically il-
legal. So far, no central authority has done so, but the administration of Gagauzia
is living under constant threat that one day this may happen. This again demon-
strates the reluctance and/or inability of the central government to fully imple-
ment the 1994 Law as required, which is putting the autonomy under a long-term
threat. The central government has not carried out several of its obligations as
required by the resolution of 23 December 1994 of the Moldovan parliament on
the implementation of the 1994 Law. Among the tasks to be carried out within
a month were the following: to submit to the parliament suggestions on actions
necessary for introducing conforming legislation to accord with the 1994 Law,
and to bring its own enactments into accord with the named law.
Another drawback of the court system of Gagauzia is that while Gagauzia
has adopted laws on economy, it has no specialized court for economic issues.
Whenever these cases reach the courts in Chisinau, the laws of Gagauzia can be
easily ignored.
6 Human and Minority Rights

In October 1996, Moldova ratified the Framework Convention for the Protection
of National Minorities of the Council of Europe. The report submitted by Mol-
dova pursuant to Article 25 (1) of this convention17 presents a positive evaluation
of the Moldovan government towards the settlement in Gagauzia and contains
data on different aspects of the life of the autonomous region.
The administration of Gagauzia has taken measures to ensure the participation
of non-Gagauz minorities living on the territory of Gagauzia. Article 10 of the
1994 Law prescribes that a non-Gagauz person is to be elected as a vice-chair-
person of the People’s Assembly. This prescription has actually been followed
in practice. The executive branch also employs representatives of other ethnic
groups. According to the telephone directory of the senior officials of the Gagauz
Executive Committee, approximately one third of them have non-Gagauz last
names. One of the deputy chairpersons of the People’s Assembly is a member
of the board of the Department of National Relations of the Moldovan govern-
ment.18
17 See http://www.humanrights.coe.int/minorities/Eng/FrameworkConvention/StateRe-
ports/2000/moldova/moldovan.html.
18 Theoretically, the Gagauz ex officio members of various ministerial boards in Chisi-
nau, starting with the Bashkan as an ex officio member of the central government,
have to participate in all the meetings of these bodies. Unfortunately, statistics on
their actual participation do not exist. However, given the time and other resources
needed for frequent travel between Chisinau and Comrat, and taking into considera-
tion the other tasks of those officials in Gagauzia, it is quite possible that they could
not attend all meetings in Chisinau, thus missing important information. This situ-
332 Priit Järve
On 19 July 2001, the parliament of Moldova adopted the “Law on the Rights of
Persons Belonging to National Minorities and on the Legal Status of their Orga-
nizations.” Article 6 of this law stipulates that the state shall create the conditions
necessary for the realization of the right of national minorities to be educated in
their mother tongue, with an explicit reference to Gagauz as one of the minority
languages.
7 External Relations and Trans-border Cooperation

Gagauzia is authorized to take part in the foreign relations of Moldova (Art
12(3b)). This participation is implemented by including the representatives of
Gagauzia in official delegations. In practice, Gagauzia has already sent its own
delegations to Turkey, for example, in March 2002. Due to cultural and linguis-
tic proximity, Gagauzia maintains close relations with Turkey and discusses eco-
nomic, cultural, and educational issues. Gagauzia has started representations
dealing with economic matters in Turkey and in Transdniestria. There are also
plans to have representation in Belarus. Additionally, Gagauzia has an agreement
of economic cooperation with the Odessa oblast of Ukraine; similar agreements
with some oblasts of the Russian Federation are pending.
It is difficult to speak about trans-border cooperation of Gagauzia in the strict
sense of the word because Gagauzia as a region has no fixed borders. Only the
borders of settlements that belong to the autonomous region are fixed. So far, this
has not emerged as a major problem for the Gagauzian autonomous region.
With regard to external relations and trans-border cooperation, Gagauzia has
created political concern in Chisinau by opening and maintaining representation
in Transdniestria. In March 2002, during a deepening political confrontation be-
tween Gagauzian authorities and the central government, the representative of
Gagauzia in Transdniestria was arrested by Moldovan security officials.19
C Successes and Failures of Implementation

The day the 1994 Law was adopted, the Moldovan parliament also passed the
resolution “On Implementation of the Law on the Special Legal Status of Gagau-
zia” (Gagauz Yeri), which tasked the government with urgent obligations in har-
monizing the legislation of Moldova with the 1994 Law.
The Resolution of 23 December 1994 also charged the government with tasks
of exercising control over the implementation the 1994 Law, organizing and con-
ducting a local referendum in several localities in the south of Moldova, meaning
Gagauzia, as well as organizing and conducting elections for the People’s Assem-
bly of Gagauzia and the Governor of Gagauzia.
ation might well have contributed to the disagreements between the autonomous
region and the centre.
19 See http://www.gagauz-press.narod.ru/002-07-03.htm. At first, this official faced an
old Soviet-style threat of being kept in psychiatric hospital. However, a few months
later he was released.
As far as the elections and referendums were concerned, the government ful-
filled its obligations. However, the government has not submitted to the parlia-
ment suggestions on actions necessary for introducing conforming legislation to
comply with the organic law; it has not brought its own enactments into accord
with this law, and it has not ensured reconsideration and invalidation by minis-
tries and departments of their enactments that conflict with the law.
This legislative passiveness is creating serious problems in the relations be-
tween Chisinau and Gagauzia. The Gagauz leadership has gained an impression
that Chisinau is systematically and deliberately ignoring the special status of
Gagauzia, and thereby trying to lower it back to the level of an ordinary county
of Moldova.20
Since the adoption of the 1994 Law, the legal environment has become more
complicated. In July 1994, the new Constitution was adopted, which provided for
the possibility of regions with a special status, but did not name these regions
explicitly. This constitutional construction was used to allow the organic law on
the special legal status of Gagauzia. As an organic law has a higher status than
ordinary laws, the latter had to be harmonized with the 1994 Law. However, as
noted, the government has failed to do so. In practice, the government agencies
are applying these non-harmonized ordinary laws to Gagauzia as if the 1994 law
had never been adopted. Moreover, since 1994, other organic laws have been ad-
opted. The validity of the stipulations of the 1994 Law in this context is becoming
increasingly ambiguous. To defend the special status of Gagauzia, the leader-
ship of the region started demanding that the Constitution of Moldova should be
amended to include an explicit reference to the autonomous region of Gagauzia.
Elisabeth Teague, who visited Gagauzia with the OSCE HCNM in 1994, was
back in Gagauzia in February 2001 as an election observer. Talking about her
impressions at the Association for the Study of Nationalities (ASN) Meeting in
New York two months later, she underlined the disappointment of the Gagauz,
who felt cheated by the Moldovan authorities. According to Teague, the Gagauz
particularly complained about the lack of financial support for the region from
the central government. They were also unhappy with Moldova’s electoral sys-
tem, which leaves them underrepresented in the national parliament. As a result,
they were calling for a change of the election law to guarantee a quota of seats in
the national parliament.
In the elections of February 2001, the Party of the Communists of Moldova
came to power. To gain the support of the Gagauz people, the party promised to
confirm the special status of the region in the Constitution of Moldova. On 10
May 2001, the parliament of Moldova adopted a resolution “On the Creation of
a Commission for Harmonizing of Legislation with the Constitution of Moldova
in Questions Dealing with the Special Status of the Autonomous Territorial Unit
of Gagauzia.” Of the nineteen members appointed to the commission seven were
20 Interviews with Mr D. Croitor, the Governor of Gagauzia and with Mr V. Yanioglo,

the First Vice-Chairman of the Executive Committee in Comrat on 7 September
2001.
334 Priit Järve
from Gagauzia. However, the Gagauz decided to suspend their membership after
the first meeting as they saw an intention to revise the 1994 Law in the very name
of the commission. The Gagauz stressed that it is the Constitution of Moldova
that has to be harmonized with the 1994 Law. Regardless of this act of protest, the
Gagauz continued working on their draft amendments to the Moldovan Consti-
tution.21
Of all international actors involved, the CSCE/OSCE has contributed most to
the implementation of the Gagauzian autonomy. The OSCE has monitored the
situation in Gagauzia, observed the election of Gagauzian authorities and orga-
nized discussions of topical issues (for example, the workshop on legislative and
budgetary issues in December 2000).
In January 2001, based on the decision made by the Executive Committee of
Gagauzia, the Governor of Gagauzia invited the United Nations Development
Programme (UNDP) to help prepare the Regional Development Programme
“Gagauz-Yeri,” which would include analysis of the socio-economic situation of
Gagauzia and recommendations for regional policy. By September 2001, the pro-
gramme had been assembled and published (UNDP 2001b).
A Impact on Society
The establishment of Gagauzian autonomous region has greatly inspired the local
population, which sees it as the achievement of its internal national self-determi-
nation. The Gagauz have real possibilities to enjoy the right to use and develop
their language and culture, including education in their mother tongue. They
are more in control of their economy and environment than before. They can
devise their own laws and enforce them. The autonomous region is developing
economic and cultural relations with countries including Turkey, Russia, Italy,
and others. However, the previous economic backwardness of the region has not
yet been overcome. The lack of enforcement of the 1994 Law on the part of the
Moldovan authorities is inducing negative attitudes towards Chisinau among the
Gagauz leadership.
There has been no violence in the relations between Comrat and Chisinau
since the establishment of the autonomous region. During the political confron-
tation of February-March 2002, there were threats of violence and even incidents
of actual violence but no casualties. Public order has been maintained; however,
the discontent of the population of Gagauzia with the dire economic situation is
mounting while Chisinau’s financial policy is regarded as unfair and one of the
main reasons for the economic difficulties.
21 Interview with Mr V. Yanioglo, the First Vice-Chairman of the Executive Committee

of Gagauzia, on 7 September 2001 in Comrat.
B Dependence upon Specific Actors

Post-Soviet countries are understandably quite sensitive to geopolitical pres-
sures. Moldova is located between the expanding NATO and Russia. The coun-
try is torn by opposite intentions – the ruling Communist party has declared its
orientation towards the Union of Russia and Belarus; on the other hand, a sizable
group of Moldovan citizens consider unification with Romania as a preferable
option and about 300,000 of them have also become citizens of Romania. In the
case of Moldova’s putative unification with Romania, the Gagauz have the right
to external self-determination (Art 1(4) of the 1994 Law), which means that the
power-sharing with Moldova would fall apart. However, the current probability
of this occurring is small.
Moldova’s rapprochement with Russia would leave the power-sharing system
in place and it can be argued that Russia could urge Chisinau to further imple-
ment and develop power-sharing with Gagauzia, which Russia might regard as a
Russian-speaking region. It is worth mentioning that in the first half of 2001 Rus-
sia invested into the Moldovan economy a record high usd 93.2 million, which
amounted to 77% of all Russian investments in the CIS countries during that pe-
riod.22 Russia has signalled interest in importing larger amounts of red wine from
Gagauzia, which could boost the local economy.23 If Chisinau and Moscow try
to resolve the problem of Transdniestria by providing broad autonomy to this
breakaway region, Gagauzian autonomy will very likely undergo fuller imple-
mentation and further sophistication.
The management of the crisis in the relations between Chisinau and Comrat
at the beginning of 2002 demonstrated that the international organizations, par-
ticularly the OSCE and the CoE, could play an important role in bringing the par-
ties together for peaceful and constructive negotiations, thus helping to prevent
the power-sharing system from collapse.
In December 2001, Mr V. Mishin, President of the Commission on Changes to
the Constitution of Moldova, submitted a request to the Venice Commission for
an opinion on the draft law “On Changes and Amendments to the Constitution
of the Republic of Moldova with Regard to the Status of the Gagauz Autonomy.”
A group of Rapporteurs of the Venice Commission visited Moldova in February
2002 to discuss the draft law on constitutional changes relating to autonomous
regions within the Republic of Moldova, with particular reference to the terri-
22 Nezavisimaya Moldova, 6 September 2001. At the same time, the Western coun-
tries do not seem very eager to invest in Moldova. An Austrian airline informed
the passengers it flies to Chisinau from Vienna, quite straightforwardly: “Currently
you should not do business in Byelorussia and Moldova. Starting a business in these
small markets is still a high risk and often causes enormous problems” (“Never fall
in love with the interpreter!” On top, Tyrolean Airlines in-flight magazine, No.2/01,
47).
23 7 September 2001 interview with Mr D. Croitor, the Governor of Gagauzia in Com-
rat.
336 Priit Järve
torial autonomy of Gagauzia with representatives of the Moldovan and Gagauz

authorities. In March 2002, the Commission issued its opinion on the draft Law
on Modification and Addition in the Constitution of the Republic of Moldova
(Council of Europe 2002a).
The role of international assistance became crucial once again during the polit-
ical crisis at the beginning of 2002. Then, Resolution 1280 and Recommendation
1554 of 24 April 2002 of the Parliamentary Assembly of the Council of Europe
were much anticipated in Moldova, which had to live with months of anti-gov-
ernment demonstrations in the centre of Chisinau, while at the same time it had
a deepening conflict with and within Gagauzia. All sides of the Moldovan conflict
interpreted these documents as a success for them. This helped resolve the po-
litical tension, stopped public demonstrations in Chisinau and ended the mutual
accusations between government and opposition. The government of Moldova
had been given one more chance to work constructively on the country’s many
problems (Plugatarev and Prihodko 2002).
C Difficulties of Power-sharing
During 2001, the system of power-sharing was apparently running into trouble
rather than consolidating as had been hoped, as the communists had not been
able to deliver what they had promised the Gagauz leaders before the national
elections of February 2001. In fact, the Communist Party of Moldova had signed
an agreement with the members of the legislative body of Gagauzia that, if elect-
ed, it would support moves to enshrine the autonomous status of Gagauzia in
the Constitution of Moldova; ensure Gagauz representation in state bodies at
all levels; clarify the division of powers between Gagauzia and the centre; raise
the issue of Moldova’s joining the CIS Customs Union and the Belarus-Russia
Union; protect the Gagauz language; elevate Russian to the status of second state
language alongside Moldovan; and pursue a peaceful resolution of the Transdni-
estrian conflict.24
On 25 February 2001, the Communist Party of Moldova gained 50.7% of the
votes at the parliamentary elections, which gave it an absolute majority of 71 seats
out of 101. The OSCE described the elections as free and fair. The autonomous
district of Gagauzia gave them the highest vote of any of the regions of Moldova,
with 80.57% of the votes cast.25
The victory of the communists led to high expectations in Gagauzia as a result
of their very specific pre-election promises to Gagauz officials. However, disap-
pointment with the new government started to spread among the Gagauz as the
authorities in Chisinau continued to use the law, that were not harmonized with
the organic law on Gagauzia, and tension between Chisinau and Comrat started
to mount again. Of greatest significance though was that in 2001 the transfers
24 Communication from Elizabeth Teague on 6 April 2001 at the Association for the
Study of Nationalities (ASN) Annual Convention in New York.
25 The results of the elections are available at http://www.ifes.md.
from the central budget to Gagauzia were established well below the level of
transfers to the other regions of Moldova.
Some symptomatic developments in Gagauzia deserve specific mention. On
19 August 2001, the Gagauz authorities celebrated the 11th anniversary of the
attempt to secede from Moldova with a festival in Comrat. More than 1,000 resi-
dents and officials attended the unveiling of a commemorative stone in Comrat.
The stone was dedicated to the memory of the activists who had contributed to
the secession attempt of the region from Moldova. Speaking at the ceremony,
Gagauz officials expressed regret that no one from Chisinau had accepted the
invitation to take part in the celebrations. This was seen by Gagauz lawmakers to
cause even more tension in Gagauzia’s relations with the central government of
Moldova. The Speaker of the People’s Assembly said in his speech that if the Mol-
dovan authorities failed to adjust national legislation to accommodate Gagauz
laws, the Gagauz authorities would have to re-activate the 1990 declaration of
independence and set up their own state structures (BASA-PRESS 2001). Such
rhetoric meant that Gagauzia was again considering confrontation with the cen-
tral government by re-launching the secessionist agenda of 1990. The very selec-
tion of 19 August 1990 for the celebration was itself symbolic. It recalled the sharp
disagreements between Comrat and Chisinau when the Gagauz were determined
to break away from Moldova. The message sent by these celebrations from Com-
rat to Chisinau was that a serious worsening of their relations was still possible.
The Governor of Gagauzia stated in early September 2001 that: “Today, we are a
territory which is discriminated against,” and added that “the non-attentiveness
of Moldova is pushing us into the embrace of Transdniestria. We have explained
it all to the President of Moldova but nothing has changed.”26
On 11 September 2001, the People’s Assembly adopted a Resolution “On the
Socio-political, Financial and Economic Situation in Gagauzia”, which stated that
the highest leadership of Moldova “deliberately does not implement” the resolu-
tion of the Moldovan parliament of 23 December 1994 “On the Implementation
of the Law on the Special Legal Status of Gagauzia (Gagauz Yeri)”. The Speaker
of the People’s Assembly claimed that the new leadership of Moldova had done
nothing to change the situation by which the government is attempting to revise
this law while simultaneously blocking the amendments to the Constitution on
the division of competences between Chisinau and Comrat. Moreover, Gagauz-
ia’s right to form its own budget is constantly violated. As a result, Gagauzia is
suffering considerable financial and economic losses each year. Chisinau has at-
tempted several times to privatize properties of Gagauzia without the knowledge
and consent of the Gagauzian authorities. The People’s Assembly then concluded
that the destructive methods of Chisinau “can totally undermine the confidence
of the people of Gagauzia.”27 A Moldovan newspaper reported that the Gagauz
26 7 September 2001 interview with Mr D. Croitor, the Governor of Gagauzia in Com-

rat.
27 See Nezavisimaya Gazeta, 22 September 2001, at http://www.ng.ru/cis/2001-09-
22/5_nega.html.
338 Priit Järve
have threatened to boycott the upcoming local elections because of their disap-
pointment with the government (Tkachuk 2001).
By the beginning of 2002, the general political situation in Moldova had de-
teriorated considerably. Negotiations with the breakaway region of Transdnies-
tria had broken down and been replaced by mutual accusations. The decisions
of the Ministry of Education to start compulsory Russian in the second year of
elementary school instead of the fifth year, and to replace the high school course
“History of Romanians” with a new “History of Moldova” course helped some
political factions to mobilize thousands of people, including schoolchildren, for
daily anti-government, anti-communist protest demonstrations in the centre of
Chisinau from the beginning of January. These demonstrations continued even
after the Minister of Education reversed these decisions and resigned at the end
of February.28
In addition, the President of Moldova promulgated a new law on 25 January
2002 on the administrative-territorial organization that re-established the thirty-
two districts (rayons), which had existed until 1998. For a swift implementation
of this law, extraordinary local elections were set for 7 April 2002. This decision
meant cutting the mandates of sitting local governments by one year, which natu-
rally caused protests among those elected. Gagauzia, having their own schedule
of elections, was clearly against April 2002 elections. Finally, the Constitutional
Court ruled that the terms of elected officials must be honoured and the next lo-
cal elections could only take place in May 2003 as originally scheduled.
While Gagauzia has no need to go back to the district system (rayons) as it did
not abolish its three districts in 1998 when the previous law on administrative-
territorial arrangement was enforced, the new law, nevertheless, causes a funda-
mental problem for Gagauzia. So far, the districts of Gagauzia only have heads
of administration, appointed by the Bashkan, but no elected councils. Councils
are elected at the local level, in villages and towns. In May 2003, the councils
in all districts of Moldova were to be elected and would then report directly to
Chisinau. This threatened to leave the administration of Gagauzia without any
meaningful business unless the law is changed to take the existence of the Gagauz
autonomous region into account.
But more importantly, in the beginning of 2002, the central authorities launched
accusations of corruption against the leaders of Gagauzia, including the Bashkan
and the Speaker of the People’s Assembly, by invoking the results of investiga-
tions of the State Auditing Chamber, which had been completed in August 2001.
On 31 January 2002, referring to that investigation, a group of members of the
People’s Assembly of Gagauzia, allegedly inspired and supported by the central
authorities, adopted a decision to dismiss the Bashkan and to call a referendum
in Gagauzia for approval of this decision. However, because of several violations
of Moldovan and Gagauzian laws in the process of adoption, other members of
28 These demonstrations ended only after the Parliamentary Assembly of the Coun-
cil of Europe on 24 April 2002 adopted its Resolution 1280 (2002), “Functioning of
Democratic Institutions in Moldova.”
the People’s Assembly and legal experts of Gagauzia considered these decisions
illegal. Crucially, the decision to dismiss the Bashkan did not get the two-thirds
majority as required by the 1994 Law, remaining two votes short of the necessary
twenty-three.
On 5 February 2002, the OSCE Mission to Moldova issued a press release
“Leadership of the Gagauzian Autonomy Under Attack” in which it noted “with
deep concern a number of activities carried out in recent days against the demo-
cratically elected authorities of Gagauzian Autonomy.” Apart from the decisions
taken by the group of deputies of the People’s Assembly, the press release also re-
ferred to a 2 February attempt to take over the TV station in Comrat by the same
group together with Moldovan security agents. Further, it mentioned the decree
of the Minister of Interior to temporarily dismiss the Chief of the Administration
of Interior of Gagauzia and the police chiefs of two Gagauzian districts without
prior agreement with Gagauz authorities, and thus in violation of the 1994 Law.
Meanwhile, the situation in and around Gagauzia became more complicated.
Moldovan troops and special forces were moved closer to Gagauzia. On 8 Febru-
ary, the same group of disgruntled People’s Assembly deputies, ignoring proper
procedures once again, adopted a decision to set the date of a Gagauzian referen-
dum on the dismissal of the Bashkan for 24 February. On 9 February, the presi-
dent of Moldova, speaking on the national television, supported the referendum
and urged the Gagauz to participate in it. The confrontation between Gagauzia
and central authorities was again escalating. Chisinau claimed that the leadership
of Gagauzia was corrupt and had no right to remain in office. Comrat maintained
that the communist government of Moldova wanted to dissolve the inconvenient
autonomous region and subject its territory to direct rule by the President of
Moldova. With the assistance of Transdniestria, Gagauzia launched its websites
www.sos-gagauz.narod.ru and later www.gagauz-press.narod.ru to publicize its
situation. On 12 February, the Bashkan issued a memorandum for the media in
which he revealed having received warnings from Chisinau to resign or risk his
life.29
On 13 February, opposition deputies of the People’s Assembly established the
electoral commission for the referendum. On 17 and 18 February, the Bashkan
issued decrees, prohibiting all preparations for the referendum and declaring
agitation to that end as illegal activities. Nevertheless, on 23 February, some 100
police and security workers arrived in Comrat from Chisinau to protect the poll-
ing stations of the referendum. Gagauzian police, together with local volunteers,
forced them out of the region and established additional security measures on the
border of Gagauzia. The authorities of Comrat closed and sealed the local office
of the Moldovan Information and Security Service twice – on 23 and 27 Febru-
ary. In regard to this, the Procurator General of Moldova initiated a criminal
investigation on 27 February into the allegedly illegal actions of the Bashkan, of
the Speaker of the People’s Assembly, and of the Head of the Legal and Protocol
29 See http://www.gagauz-press.narod.ru/o-08.htm; also in Moldavskie vedomosti, 16

February 2002, at http://www.vedomosti.md/index.asp?doc=1_2.
340 Priit Järve
Department of the People’s Assembly. Meanwhile, the 24 February referendum

failed when only 18% of the voters took part in it.
On 4 March, the president of Moldova, speaking in a Moldovan TV broadcast,
accused the Bashkan and the Speaker of the People’s Assembly of having ‘sold
themselves’ to Transdniestrian authorities and of having stolen large amounts
of humanitarian aid provided by Turkey. Three days later, the Baskhan and the
Speaker of the People’s Assembly sued the president for slander,30 while the Head
of the Legal and Protocol Department of the People’s Assembly, who was also
the Representative of Gagauzia in Transdniestria, was arrested in his Comrat of-
fice and taken to Chisinau by armed Moldovan security agents. On 7 March, the
Bashkan also sued the People’s Assembly for its decisions of 31 January on the
dismissal of the Bashkan, of 8 February on establishing the date of the referen-
dum, and of 13 February on creating electoral bodies for the referendum.31
On 11 March, the Gagauz leadership issued appeals to the media, to foreign
ambassadors in Moldova, and to the OSCE and other international organizations,
asking for help in arranging a meeting with the president of Moldova. The ensu-
ing contact of foreign diplomats with the president might have helped. On 20
March, the president of Moldova had long-awaited meetings with the Bashkan
and with a group of cooperative members of the People’s Assembly. In the press,
these meetings were characterized as constructive and as the beginning of a new
phase in relations between Chisinau and Comrat. At the meeting, the Bashkan
suggested that feeding the president with one-sided and distorted information on
the problems of Gagauzia had been the cause of the crisis. The Bashkan and the
President agreed to have monthly meetings in the future to discuss pressing is-
sues. The president promised to activate the harmonization of legislation, to hold
a meeting of the Moldovan government in Comrat, and called upon the members
of the People’s Assembly to declare a moratorium on their disagreements.32
International mediation between the various conflicting parties in Moldova
continued. In early April 2002, the OSCE HCNM paid a three-day visit to Mol-
dova, during which he commented on the turbulent political climate there, in-
cluding the situation in Gagauzia, and stressed the need for constructive dialogue
and appealed to all parties to respect the rule of law and to follow democratic
procedures.33
Soon after this, on 15 April, the Executive Committee of Gagauzia resigned vol-
untarily because, as it announced it could not fulfil its responsibilities due to the
30 See http://www.gagauz-press.narod.ru/o-31.htm; and http://www.gagauz-press.nar-

od.ru/o-31.htm.
31 See http://www.gagauz-press.narod.ru/o-29.htm.
32 See Nezavisimaya Moldova, 21 March 2002, at http://www.nm.md/daily/
news/2002/03/21.html#1; see a more detailed account of the meetings at http://www.
gagauz-press.narod.ru/013-20-03.htm.
33 See http://www.osce.org/news/generate.php3?news_id=2395.
split in the People’s Assembly.34 According to the law, the Bashkan had to appoint
a new Executive Committee and send it to the People’s Assembly for approval.
However, on 24 May, this approval became unlikely when it appeared that the op-
position in the People’s Assembly had acquired three additional members, which
secured the necessary majority to dismiss the Speaker and to elect the leader of
the opposition into that office.
D Future Prospects
The prospects of the autonomy arrangement in Gagauzia for the near future de-
pend considerably on how Chisinau and Comrat can resolve their disagreements;
in particular, how the legislative acts bearing on the special status of Gagauzia
can be harmonized. The thorniest issue seems to be the amendments to the Mol-
dovan Constitution that have reached the Constitutional Court before they have
been processed by the parliament. They have also been reviewed by the Venice
Commission. Draft amendments provided by Chisinau and Comrat were funda-
mentally different. While Chisinau tried to write the autonomy of Gagauzia into
the Constitution, though not very convincingly according to the Gagauzian side,
the latter proposed amendments that would constitute Moldova as a federation.
As a result, neither the OSCE nor the Moldovan side supported the Gagauzian
draft, which was rejected and did not reach the Constitutional Court. It was ex-
pected that the Court would give its opinion of the Moldovan draft in April 2002,
but this did not happen. In July 2002, the negotiations between Moldova and
Transdniestria took a new turn as international actors, such as the OSCE, intro-
duced a notion of a federal state into the negotiations on the status of Transdnies-
tria. In February 2003, the president of Moldova proposed to Transdniestria the
drafting of a new Constitution of the Republic of Moldova. (International Crisis
Group 2003: 8–11) This was, however, not implemented. Whether Gagauzia be-
comes a member of the Moldovan federation or remains an autonomous unit will
be decided by the negotiations on the federalization of Moldova. Gagauzia has
not been invited to participate in these negotiations. However, it can be assumed
that if a new Moldovan constitution is adopted, the status of Gagauzia will be
given due attention and the problems of power-sharing that have overshadowed
the relations between Gagauzia and the central government so far will be avoided
by more sophisticated constitutional stipulations.
As previously mentioned, during the harmonization of Moldovan laws with
the legal status of Gagauzia, the new law on administrative-territorial organiza-
tion needs to be adjusted in order not to turn the Gagauz autonomous region into
an empty concept.
Finally, it would be helpful if the court cases brought against the leaders of
Gagauzia and the president of Moldova could reach amicable solutions so as to
avoid fuelling new conflicts.
34 See Moldavskie vedomosti, 24 April 2002, http://www.vedomosti.md/index.asp?doc=1_

2&nom=450.
342 Priit Järve
Fortunately, the relations of different ethnic groups in Moldova have tradition-

ally been peaceful. However, sociological evaluation of interethnic tolerance in
Moldova in 2000 showed that, while the majority of respondents considered in-
terethnic relations as normal and good, one third of the respondents considered
these relations to be hostile and tense. Therefore, a latent psychological potential
for aggravation of interethnic relations does exist and these relations have to han-
dled very carefully (UNDP 2001a). It means that the parties to the power-sharing
agreement should not push their demands too far or too forcefully, and, most
importantly, they should avoid ethnic mobilization at all costs.
VII Concluding Remarks

A post-communist state faces formidable challenges. It is supposed to simultane-
ously transform itself into a multiparty democracy, to carry out ownership re-
form together with the introduction of a market economy, to build up new state
institutions, often from scratch, and last, but not least, to enforce the rule of law.
In theory, these transitions are deemed to support one another because democ-
racy, a market economy, and the rule of law are all sine qua non elements of
the Western states where they do function together. In post-communist practice,
however, the early phases of transition to these essential elements do not produce
much synergy. On the contrary, the simultaneous transitions are often at cross-
purposes. For example, the participants in privatization, the infamous ‘new’ Rus-
sians, Estonians, Moldovans, etc., are not interested in transparency, or in the
rule of law; while trying to maximize its political and economic power, no group
is prepared to share this power with others, let alone with ethnic minorities. In
short, rules of formation of a socio-political system are quite different from the
rules governing its functioning.
The frictions between these different transitions produce a weak state, which
may seek to overcome its weakness by centralizing power and financial resources.
Inefficient collection of taxes allowing for massive tax avoidance, big regional
differences and starving social programmes, which all ask for state-level action,
are only a few reasons driving a strong tendency towards centralization and a
reluctance to share power with regions or others.
Therefore, power-sharing in post-communist, transitional settings will not be
ensured by the precision or succinctness of the wording of a law or of an agree-
ment. Nor will it be entrenched by noble appeals of high-minded and enlightened
personalities to promote democracy. Rather, it will be accepted only in those ex-
ceptional cases, if at all, when it appears to be the only way to avoid the dangers
of social disruption, armed conflict, and chaos. Moreover, when the imminent
dangers seem to have vanished, the implementation of power-sharing may well
stall and new tensions develop, which is precisely what has happened to the pow-
er-sharing arrangement between Gagauzia and Moldova at times.
In 2002, after almost eight years since its enactment, the Gagauzian autono-
mous region was facing challenges not only in its relations with the central gov-
ernment but also internally because of the political confrontation between the
People’s Assembly and the Executive Committee of Gagauzia. The internal con-
flict developed step by step under the influence of the external one with Chisinau.
The authorities of Moldova, obviously dissatisfied with the recurrent demands
of the Gagauzian leadership to amend the Constitution, let alone national leg-
islation, and irritated by Gagauzia’s ongoing cooperation with Transdniestria,
brought charges of misuse of public funds against the leaders of Gagauzia and
appealed to the population and lawmakers of the autonomous region to decide
the fate of their leaders at a local referendum. This referendum, prohibited by the
Gagauzian authorities, failed to produce any convincing results. However, the co-
operation between the Executive Committee of Gagauzia and People’s Assembly
was paralyzed, as the majority of the latter had become opposed to the Execu-
tive. To avoid further escalation of this confrontation, the mediation of foreign
diplomats in Chisinau was needed to restore the communication between the
President of Moldova and the Bashkan.
At the national level, with the help of international bodies such as the Par-
liamentary Assembly of the Council of Europe and the European Commission
for Democracy Through Law (Venice Commission), efforts were made to ensure
stability in the region and to defuse chronic disagreements over compatibility of
national laws with the 1994 Law. In these efforts, an important role went to the
discussion and refinement of additions to the Constitution of Moldova in order
to create explicit constitutional guarantees for the Gagauzian autonomy and to
resolve the contradictions between the 1994 Law and some articles of the Con-
stitution.
The experience of power-sharing between Moldova and its autonomous region
of Gagauzia has displayed a pattern of recurrent need of international assistance
at different levels and of various magnitudes whenever the disagreements be-
tween the two sides tended to escalate into serious dispute due to contradictory
interpretations of the 1994 Law and/or because its implementation was perceived
as wrong by one side. These disagreements have stemmed to a large extent from
the ambiguously formulated articles of the 1994 law on the status of Gagauzia,
which gave rise to incompatible expectations and attitudes, particularly on eco-
nomic and financial matters. Sometimes, insufficient administrative capacity, ar-
bitrary decision-making and improvised action have considerably aggravated the
existing disagreements and transformed them into serious confrontations. These
two factors – shortcomings in the power-sharing agreement and an imperfect
administrative (and political) culture – can be held responsible for the uneasy
development of power-sharing between Moldova and its autonomous region of
Gagauzia. The availability of international assistance during the turbulent peri-
ods of this development has been crucial for its continuation.
Chapter 11
Case Study of the Conflict in South Ossetia
Ketevan Tsikhelashvili and Natasha Ubilava*
I Introduction
Following nearly three years of irregular armed confrontation, the breakaway re-
gion of South Ossetia has been enjoying relative peace brought about by a cease-
fire in 1992. Unfortunately, formal peacemaking has ceased there. The conflict
has remained politically ‘frozen’ as negotiations have failed to yield any genuine
settlement for more than a decade. In the absence of a formal settlement, how-
ever, a rather peculiar informal peace arrangement emerged out of local and in-
ternational initiatives and needs-driven transactions.
Initially, segmented functional interactions were maintained by previously op-
posed groups. Over time, internationally-sponsored bilateral contacts and joint
activities of all conflicting sides have gradually forged multifaceted interactions
that are based upon an informal/semi-formal power-sharing principle. This ‘bot-
* The authors would like to acknowledge the valuable comments from Emmanuel An-
quetil (Project Manager, European Commission Delegation in Georgia), Jonathon
Cohen (Caucasus Programme Manager, Conciliation Resources, London), Bruno
Coppieters, (Associate Professor and Head of the Department of Political Science,
Vrije Universiteit Brussel [Free University of Brussels]), S. Neil MacFarlane, (Lester
B. Pearson Professor of International Relations, University of Oxford), Klaus Ras-
mussen (Political Officer, OSCE Mission to Georgia), and Ermina Van Hoye (Special
Assistant to the Special Representative of the Secretary-General for Georgia UN-
OMIG).
South Ossetia is a breakaway region of the territory of Georgia. It is also referred
to as “Tskhinvali Region,” “Samachablo” or “Shida Kartli” in different sources, each
evoking tense debates between Ossetians and Georgians. Ossetians argue for “South
Ossetia,” which emphasizes the desired territorial and ethnocultural political affinity
with North Ossetia. Georgians prefer “Samachablo” or “Tskhinvali region,” which is
used in official documentation. As “South Ossetia” is most frequently used in com-
mon and diplomatic parlance, it will be used hereafter without political or any other
connotation. Tskhinvali itself is the regional capital.
346 Ketevan Tsikhelashvili and Natasha Ubilava
tom-up’ development of power-sharing, from incremental cooperation at the

‘grassroots’ level towards more formalized arrangements, may ultimately pave
the way for a congruous settlement, possibly embracing ‘vertical’ components of
conflict resolution. Currently, however, the existing network of field-specific ar-
rangements exclusively targets ‘horizontal’ aspects of self-regulation. Provision-
ally, such a pattern could be described as ‘grassroots’, ‘horizontal’, or ‘informal/
semi-formal’ power-sharing.
The South Ossetian experience suggests that a formal agreement may not be
a necessary prerequisite for meaningful movement towards dispute settlement.
It demonstrates that genuine and practical grassroots cooperation can function
without an overarching political settlement spelling out the terms of engagement.
In South Ossetia, these terms of engagement tend to be dictated by the needs of
daily life that transcend ethnic barriers and cannot wait for a long-overdue politi-
cal resolution. Mostly unwritten and largely assumed, these rules govern an intri-
cate and delicate pattern of needs-based and project-specific interactions at infor-
mal or semi-formal levels. What remains puzzling and compelling about this case
is the exceptional combination of internal and external factors that has prompted
significant low-level cooperation without a comprehensive settlement.
South Ossetia can be viewed as unique in comparison to the other case stud-
ies in the Resolving Self-determination Disputes using Complex Power-sharing
project, where officially recognized complex power-sharing arrangements have
been implemented or are being implemented. This case thus may challenge the
conventional understanding of resolving self-determination conflicts. It goes be-
yond the existing framework that complex power-sharing theory has offered for
dispute settlement. Perhaps, it may even offer a novel way of addressing self-de-
termination issues, at least in the short term.
II Approach
Because of its unique qualities, South Ossetia does not readily fit into the tradi-
tional analysis of complex power-sharing arrangements. To date, there has been
no systemic exploration of conflict settlement with a focus on practical achieve-
ments at the horizontal level such as in South Ossetia. Research thus far has mainly
concentrated on the ebb and flow of political developments with regard to a long-
awaited but still uncompleted formal solution. The lack of systematic data on infor-
mal power-sharing practices in South Ossetia further complicates the analysis.
A tailored approach was developed for the exploration of the South Ossetian
case. During field research in Georgia in the summer of 2001, information was
collected from various primary sources such as official documentation and in-
terviews with officials, independent experts and representatives of international
as well as local governmental and nongovernmental organizations. The objective
was twofold: firstly, to identify case-specific factors, whether internal or external,
that have enabled and continue to facilitate bilateral or multilateral cooperation;
and, secondly, to examine the areas, channels, and patterns of the existing power-
sharing practices.
11 Case Study of the Conflict in South Ossetia 347
To address the first research objective, this chapter begins with a brief account
of the origins, scope, and dynamics of the conflict in South Ossetia. It not only
gives a general context but also examines the reasons for the failed political pro-
cess. The following section, predictably, explores the second research objective
and provides a detailed study of grassroots power-sharing that focuses on the
participants as well as on the level and scope of their involvement. It attempts
to analyze which parties, internal or external, were best-positioned to capitalize
on ‘relative peace’ and what were the most urgent and politically least sensitive
issues that triggered cooperation. The findings of this section ultimately inform
a discussion about potential long-term conflict resolution in the region. Perhaps
more importantly, though, they may provide powerful implications for the con-
cept and application of complex power-sharing.
Comparison with Abkhazia, another unresolved ethnic dispute in Georgia,
highlights the unique character of the South Ossetia experience. Whereas joint
efforts produced practical arrangements in the latter case, peaceful initiatives
yielded neither political nor pragmatic rapprochement in the former. It is par-
ticularly interesting to explore the reasons for the different post-conflict develop-
ment in two cases that are similar in many other aspects.
III The Conflict in South Ossetia

Little in the origins of the conflict in South Ossetia suggested an extraordinari-
ly complex and unique outcome, as the hostilities initially unfolded in a simi-
lar fashion to other self-determination disputes erupting in the post-Soviet era.
The major source of confrontation can be found in the destructive effects of the
Kremlin’s ‘nationality policies’ that created political and economic inequality
among the USSR’s many ethnic groups and further accentuated any pre-existing
conflicts. Underlying tensions became extremely sensitive after the collapse of
the Soviet Union, as the central government could no longer keep a lid on the
boiling cauldron of ethnic frustrations. Domestically, ultranationalist policies of
the Georgian federal government clashed badly with uncompromising separatist
ambitions in South Ossetia. Externally, Russia’s sedulous endeavours to maintain
influence over the Caucasus region largely came at the cost of instigating destabi-
lization and further exacerbated the confrontations.
A favourable combination of certain internal and external factors helped South
Ossetia escape the trappings of other ethnic disputes that still remain unresolved
in the post-Soviet area. Some of these features were inherent to the region where
Georgians and Ossetians had maintained a largely peaceful coexistence prior
to Soviet rule. As a result, the armed confrontation was smaller both in terms
of scale and intensity, producing less physical and emotional damage than in a
number of other post-Soviet conflicts. The change in political regimes in both
capitals – Tbilisi and Tskhinvali – also helped to bury the hatchet as the new
In September 1993, the more moderate Ludwig Chibirov replaced incumbent ‘Presi-
dent’ Torez Kulumbegov, who was the leader of the Ossetian extremist movement.
governments could blame their predecessors for past mistakes and embark on
fresh ideas for collaboration.
As the conflict had evolved, a number of external factors came into play that
ultimately reinforced the positive achievements of local peace initiatives and cre-
ated new platforms for cooperation. In this respect, international organizations
proved vital as they managed to largely stay out of the political controversy and
focus more on diplomacy and cooperation on practical issues. Their flexibility
and less-politicized approach allowed for progress in other spheres once it be-
came clear that political negotiations were deadlocked. Furthermore, it appears
that there were few external powers with strong vested interests in destabilizing
the region. Unlike the Abkhazian nationalists, South Ossetian leaders found only
limited support for their secessionist cause abroad. Russia, which played a nega-
tive role in Abkhazia, ultimately acted more ambiguously towards South Ossetia:
stirring up hostilities, on the one hand, while effectively maintaining a ceasefire,
on the other. Even their direct kinsmen in North Ossetia somewhat distanced
themselves from the South Ossetian extremists and supported the peace pro-
cess.
The following section examines the various factors in order to understand how
they emerged and ultimately contributed to the development of horizontal pow-
er-sharing in South Ossetia.
A Internal Factors
1 Conflict Origins
The origins of the present day conflict in South Ossetia closely followed the blue-
print of ethnic self-determination disputes masterminded by the Soviet ‘nation-
ality policy’ strategists. Dating back to the early days of Soviet rule, this policy
was part of an overall ‘dual’ strategy that ensured tight control over artificially
created Matryoshka-type federal territorial arrangements. It granted separate
territorial political status to utterly different nations and ethnicities. This dual
strategy combined the policies of ‘homogenization’ at the macro level (aiming
at the integration of all ethnicities into the ‘Common Soviet People’) and the so-
called ‘nationalities policies’ at the micro level (implying institutionalization of
peripheral ethnopolitical and ethnocultural identities). The latter, in essence, was
a practical application of the age-old divide et impera (divide-and-rule) principle.
Cunningly, the ethnic differences were reified to the extent where units would
Chibirov was in office until December 2001. President Zviad Gamsakhurdia’s ultra-
nationalist government in Tbilisi was ousted in January 1992. Eduard Shevardnadze,
who came to power in March 1992, remained president until 23 November 2003.
The Matryoshka (nesting doll) principle implied a complicated ethnofederal subor-
dination of units and sub-units within the Soviet Union. The fifteen Union Republics
had the highest rank status, followed by the Autonomous Republics, and then the
Autonomous Regions/Oblasts (Nodia 1998: 22-23; Coppieters 1999: 16-23).
continuously compete with and counterbalance one another so that none of them
could gain too much strength.
The damaging results of this ‘nationality policy’ could hardly be overstated as
these Matryoshka units began to explode one-by-one following the collapse of
the Soviet Union. Glasnost and Perestroika introduced previously taboo freedoms
and rights that incited the claims for self-determination of the Union Republics
and their autonomous sub-units. The full impact of the divide-and-rule policy
pursued by the Kremlin was displayed upon the collapse of the Soviet empire
when the ruler disappeared but the divided community lines remained. The pow-
er vacuum at the top left power-hungry ‘titular nations’ scrambling for whatever
influence and resources were suddenly available at the bottom.
South Ossetia was fertile soil for this kind of power struggle. Throughout the
years of Soviet rule, Ossetians had felt disadvantaged in several respects. Politi-
cally, they held the status of an autonomous oblast within Georgia, thus ranked
a step below the status of the self-governing autonomous republics of Abkhazia
and Adjara. Accordingly, South Ossetia’s legal powers were inferior to those of
Abkhazia and Adjara. South Ossetia had suffered economically as well. While
hardly an area of significant economic potential, the region fell short of the pros-
perity and well-developed infrastructure enjoyed by its immediate rivals – Ab-
khazia and Adjara.
In the aftermath of Georgian independence, South Ossetia promptly attempt-
ed to upgrade its inferior political status to a level equal to that of Abkhazia,
which was thought to enjoy the greatest sovereignty rights in the country. Tbilisi
denied this request as it represented a potential threat to the region’s Georgian
population. Underlying ethnic tensions were further fuelled by an ultranation-
alistic language programme propagated by the newly-elected federal govern-
ment in Tbilisi. South Ossetian authorities retaliated with equally nationalistic,
if not discriminatory, policies, which brought Georgians onto the streets and,
eventually, into the first clashes with local militia on 23 November 1989. The
first year of conflict was marked by sporadic violence, but by December 1990
full-scale military confrontation became inevitable as Georgians and Ossetians
grew increasingly antagonistic. The Georgian Supreme Court barred the South
Ossetian People’s Shrine – a secessionist political party – from participating in
the forthcoming first parliamentary elections. In return, Ossetians raised their
demands from enhanced autonomy to the status of Union Republic, which meant
The right to self-determination of the Union Republics was nominally protected by

the Constitution of the Soviet Union but its practice was never allowed.
Each national group that was granted the right to a separate territorial unit was rec-
ognized as a ‘titular nation’ and, correspondingly, enjoyed political privileges and
access to power.
Six people were killed and twenty-seven injured during the following two days (Kom-
somolskaya Pravda, 26 December 1990; Human Rights Watch/Helsinki 1992: 7).
The South Ossetian People’s Shrine actively supported preceding Abkhaz demands
for secession (Fuller 1991: 21).
secession from Georgia. The ultranationalist Georgian government responded by

abolishing all autonomy rights for South Ossetia and introducing militia to the
Tskhinvali and Djava regions.
This episode led to renewed armed confrontation, and the following months
turned the Tskhinvali regional centre and surrounding territories into a theatre
of brutal street battles, barricades, and chaotic shooting. From that point, the
armed dispute between the Ossetians and Georgians lasted, with interludes, until
the July 1992 ceasefire. The conflict brought about more than 1,000 deaths and
around 60,000 internally displaced persons (IDPs) and refugees – both ethnic
Ossets, who fled mostly to North Ossetia, and Georgians, who sought shelter in
Georgia proper. The war damage in South Ossetia was further aggravated by a
coincident strong earthquake, epicentred in the Djava-Tskhinvali region, which
seriously damaged economic infrastructure and destroyed housing.
2 Past Relations
In general, Ossetians and Georgians have a record of largely peaceful coexistence,
especially when compared to the turbulent Abkhaz-Georgian relations. High
rates of intermarriage and large numbers of Ossetians living outside of South Os-
setia testify to moderate relations between Georgians and Ossetians.
Interestingly, and perhaps even ironically, the first incident of interethnic con-
frontation coincided with Georgia’s brief independence in 1918–1921. Georgia’s
Menshevik government accused Ossetians of collaborating with Russian and
Georgian Bolsheviks, who fought to usurp their power and reunite Georgia with
Russia (MacFarlane, Minear, and Shenfield 1996: 14). In 1918–1920, the series of
strong uprisings of Ossetians claiming independence were deemed as Bolshevik
provocations and an ‘aperture’ to Russian annexation of Georgia in 1921. In the
aftermath, Georgians blamed Ossetians for contributing largely to the loss of in-
dependence, while Ossetians, for their part, were embittered by the experience of
Georgian-led supressions. The period gave rise to ill feelings between the com-
munities.
A similar confrontation occurred in Georgian-Abkhaz relations in the same
period. However, unlike those, Georgian-Osset relations eventually returned
to normal. Here, the Kremlin intrigues failed to further magnify tensions at the
intercommunal level as they did in Abkhazia. Because of its greater economic
and strategic appeal, Moscow was more involved in, and, thus, achieved greater
impact in Abkhazia. Abkhazia’s riches, whether innate or inflated by Moscow’s
favouritism, fuelled among its ethnic groups a bitter power struggle for ultimate
control of resources. In contrast, South Ossetia was hardly a region of signifi-
Figures on casualties differ, although most agree on around 1,000 deaths (Cvetkovski
1996: para 4.1; MacFarlane, Minear, and Shenfield 1996: 8).
Figures on refugees and IDPs vary considerably across different sources. The above
figures reflect the data provided by UNOCHA (March 2001) and UNHCR (June
2001), which estimate that approximately 60,000 people, mostly Ossets, were dis-
placed.
cant economic or strategic value and, hence, political stakes were lower both for
Moscow and local ethnic groups. Furthermore, South Ossetia proved less sus-
ceptible to political manipulation from above. Ossetians, as a sizeable majority in
their autonomous unit, felt less vulnerable and sensitive to Soviet ethnic policy.
Abkhazians, however, were a small minority and thus appeared more exposed.
Consequently, ethnic relations in South Ossetia suffered from latent tension to a
lesser degree than in Abkhazia.
A relatively ‘clean sheet’ of local interethnic relations prior to and after the
1918–1921 incident helps explain the lower intensity of hostilities in the Georgian-
Ossetian confrontation. Firstly, it signals a lack of deep-seated rancour among
the two ethnic groups. It also allows for the externalization of the causes of the
conflict, at least partially. Given the relatively short history of the conflict, it may
be easier to identify the circumstances that produced the initial confrontation
and, thereby, to prevent the spread of resentment to other areas of interethnic
relations.
3 Demographics
The ethno-demographic balance in pre- as well as post-conflict South Ossetia
has provided a conducive environment for potential settlement. Ethnic Ossetians
comprised 3% of the total population of Georgia, and before the war 60% of Os-
setians were dispersed over the rest of the country rather than living in their
autonomous province. Here, the remaining 40% of Ossetians enjoyed a majority
of 66%.10 This figure presumably grew following the exodus of Georgian refugees
during the fighting.
The large proportion of Ossetians living outside South Ossetia further indicates
the lack of significant tensions prior to the conflict. Within South Ossetia itself,
their numerical dominance led to little sense of insecurity among the Ossetians.
They were less preoccupied with alleged threats of assimilation or the so-called
‘Georgianization’ than the Abkhaz (Hewitt 1996: 202–210). The demographic bal-
ance made Ossetians less vulnerable in this regard and created more space for
constructive conflict regulation. This would still apply even if full repatriation of
the Georgian refugees occurred. In fact, some displaced Georgians have already
returned to their homes in South Ossetia. At this stage, the repatriation process
is hampered more by the depressed socioeconomic conditions and legal ambi-
guities regarding the restoration of property rights than by political and security
problems.
Abkhazia paints a greatly contrasting picture. Mass expulsion of Georgians
from the region has converted Abkhazians from a minority (17%) into the absolute
majority. Hence, the So(u)hkumi11 de facto authorities undoubtedly understand
that the return of Georgian refugees would restore the original demographic bal-
10 Figure taken from the 1989 population census. More recent official data is not avail-
able, as the 2001 population census in Georgia did not cover the two zones of conflict
in Abkhazia and South Ossetia.
11 So(u)khumi is the capital of Abkhazia.
ance, which was regarded as unfavourable to the Abkhazians. Moreover, it would

challenge, if not deprive, the incumbent regime of public legitimacy, as it would
no longer be representative of its constituency. The So(u)khumi fears of losing
power and reverting to a minority are so strong that they make the repatriation
of IDPs a major obstacle in Georgian-Abkhaz negotiations. Together with other
hindrances, such as sporadic violence or ineffectiveness of external involvement,
the impeded repatriation process undermines peace efforts both in the political
and public domains. Thus, social reintegration and reconciliation of the two com-
munities in Abkhazia are not in sight.
More importantly, there is another significant difference between these two
cases. South Ossetia did not suffer a deliberate ‘ethnic cleansing’12 policy, which
resulted in the substantial expulsion of one ethnic population. Approximately
266,000 Georgians fled their homes in Abkhazia within a shockingly short time.13
They constituted almost the entire Georgian population of the region, with the
exception of the few villages in Gali rayon that border the Georgian interior. The
devastating results of such a massive exodus can hardly be overstated. Unlike in
South Ossetia, in Abkhazia several counties have been fully deserted by Geor-
gians, thereby undermining and effectively eliminating an opportunity for coop-
eration, especially at the grassroots level. With the prospects of their repatriation
bleak, there is little hope that Abkhazia will be able to enjoy the benefits of infor-
mal, spontaneous power-sharing arrangements.
4 Sporadic Warfare
The warfare in South Ossetia was of a smaller scale and intensity as well as more
sporadic and disorganized than that in Abkhazia. Additional Georgian militia
groups were officially deployed in the region for only a year after the first clashes
between Georgian civilians and Ossetian forces in November 1989. There were
few large military operations that involved a massive build-up of equipment and
combatants from either side, again unlike those in Abkhazia.
12 The mass expulsion and harassment of Georgians in Abkhazia has been qualified
as ‘ethnic cleansing’ by the Georgians. The OSCE, in its Declarations of the OSCE
Summit in Lisbon (1 December 1996: 20) and the OSCE Summit in Istanbul (19 No-
vember 1999: 17) also use this term. The UN has never officially recognized the fact
of ‘ethnic cleansing’ of Georgians, although the UN Security Council has many times
reiterated its “deep concern about the humanitarian suffering” and “numerous no-
tifications on ‘ethnic cleansing’ and other serious violations of the international hu-
manitarian law” (UN Security Council Resolution 876/1993).
13 UN Economic and Social Council, Specific Groups and Individuals, Commission on
Human Rights, 57th session, 25 January 2001. However, in different sources the num-
bers for IDPs and refugees vary. The Ministry of Refugees and Accommodation of
Georgia (MRA) originally registered 259,656 IDPs, which coincides with UNHCR es-
timates. However, in 1997 MRA increased this up to approximately 268,072. The US
State Department estimates range from some 230,000 to 250,000 people displaced
from Abkhazia by 1993 (US State Department 1993: 877, 881).
To separate combatants and prevent a resumption of hostilities, 500 Russian/

Soviet interior soldiers were deployed in 1990. These troops stayed in the region
for a year (Human Rights Watch/Helsinki 1992: 13). There is evidence to suggest
that mercenaries from abroad, namely North Caucasian or Russian, were also
involved, in addition to armed units of the direct conflict parties The scale of their
participation, though, was significantly lower in Ossetia than in Abkhazia, where
entire squads were made up of foreign fighters (Human Rights Watch 1995: 19, 33,
42-44). The local population also took part in sporadic street clashes in Tskhin-
vali. Although there were Ossetian and Georgian partisan movements, in South
Ossetia guerrillas never became entrenched in a single stronghold from where
they could continue fighting even after the ceasefire. Also, because of the lack of
central command and coordination between small paramilitary groupings, the
lines between criminal and military activities were blurred on both sides.14
The lower scale of hostilities had a similar effect on the South Ossetian conflict
as the fairly good record of ethnic relations did. It spared the region from the po-
tentially heavy burden of physical and emotional destruction that more intense
warfare would have caused. Although heartbreaking in itself, the level of destruc-
tion and human tragedy in South Ossetia pales in comparison to the devastation
and the loss of life in Abkhazia. The war in South Ossetia produced more than
1,000 casualties and around 60,000 IDPs of both ethnicities, while the Abkhazian
death toll stands at 10,000 and the number of IDPs is an astounding 266,000.
Perhaps it is fair to say that peacemaking in Abkhazia is as many times harder as
there are so many more victims.
5 Change of Wartime Elites and the ‘Personal’ Factor

In the aftermath of the conflict, radical nationalists in both South Ossetia and
Georgia were replaced by more moderate political forces. The South Ossetian
leader Ludwig Chibirov, who came to power in 1993, was elected ‘president’ in
1996 and held his post until December 2001. Chibirov adhered to a much more
balanced and moderate policy in the conflict resolution process than his prede-
cessors – Snaur Gassiyev (1991–1992) and the leader of the Ossetian extremist
movement Torez Kulumbegov (1992–1993). Georgian officials, as well as interna-
tional actors, often stress that progress in the peace process was largely due to the
temperate policy of the ruling elite in Tskhinvali.
Simultaneously, the wartime elite changed in Georgia. The ousting of Presi-
dent Zviad Gamsakhurdia (1992), who failed to tackle the South Ossetia issue
diplomatically, brought to power Shevardnadze’s government. Shevardnadze re-
versed the tactics and admitted that the escalation of the conflict in South Ossetia
was the “biggest mistake” of the Georgian government, implying that they really
meant Gamsakhurdia’s leadership (Zverev 1996: 47). Such a political gesture no-
tably defused wartime tensions and further facilitated reconciliation between the
two ruling elites. South Ossetian authorities say that they associated the name of
14 Hansen (1998: 11) claims it to be one of the common features of warfare conduct in
the Caucasus region.
Shevardnadze “with the continuation of the settlement of the Georgian-Ossetian

conflict.”15
The change of leadership and subsequent moderation of policies pushed the
whole negotiation process in a more constructive direction than could have
been expected had the wartime elites remained in power. Both Georgian and
South Ossetian authorities now emphasized the importance of good collegial
and friendly interpersonal relationships between the negotiators. This applied to
plenipotentiaries as well as the regional and local administration representatives
who demonstrated the goodwill, trust, and confidence necessary for constructive
cooperation. In contrast, there was no such change of wartime elites following
the Abkhazia conflict. Therefore, a strong antagonism, inherited from the times
of war, burdens the process of Georgian-Abkhazian dialogue.
However, there remain political groups in South Ossetia that would like to
exploit ethnic platforms. The moderate course of Chibirov’s team raised dis-
content among radical groups in South Ossetia. In their struggle for power, the
opposition leaders started to manipulate extremist sentiments.16 In December
2001, a more radical leader – Eduar Kokoid, a successful Russian businessman
– came to power17 on the wave of these sentiments. His ‘election’ brought a sig-
nificant dissonance into the Georgian-Ossetian peace negotiations. He hinted
that priority would be given to further integration with North Ossetia/Alanya,
while relationships with Georgia would be contractual. Markedly, Kokoid’s first
political gesture upon taking office was meant to demonstrate the changed ap-
proach of the Tskhinvali new administration: Kokoid demanded personal and
public apologies for the conflict in South Ossetia from Eduard Shevardnadze,
then president of Georgia. The recent changes in the political regime in Tbilisi
have further alarmed the South Ossetian rulers. Public protests forced Shevard-
nadze to step down on 23 November 2003 and Mikheil Saakasvhili, the leader of
Georgia’s United Nationalist Movement, won the extraordinary presidential elec-
tions on 4 January 2004.18 Certainly aware of the sensitivities in Tskhinvali, the
newly elected president has pledged to use peaceful means to restore the territo-
rial integrity of Georgia. Even if high-level political relations remain uncertain,
the progressive pace of Georgian-South Ossetian rapprochement will most likely
be maintained, at least at the community and informal/semi-formal level because
these relationships are less politicized and mutually beneficial. It is encouraging
15 Ludwig Chibirov at the press conference after a meeting with then State Minister of
Georgia Vazha Lortkipanidze, on 23 February 2000 (OSCE 2000a).
16 The claim for Chibirov’s immediate resignation ‘without bloodshed’ was a vivid il-
lustration of opposition pressure by the local radical movement “Hope of Ossetia,”
although it failed to attain that goal (OSCE 1999a).
17 Kokoid was the winner of the last ‘presidential elections’, which neither Georgia nor
international observers have accepted.
18 Only certain parts of the South Ossetia conflict zone participated in national presi-
dential elections on 4 January 2004.
that up to now most of the existing power-sharing arrangements at the informal

and grassroots level have been largely maintained.
6 Community Modus Vivendi

Interethnic relations at the community level are fairly smooth now. A number
of factors discussed above (history of peaceful coexistence, high rate of inter-
marriages, lower scale of conflict in comparison with Abkhazia, etc.), as well as
the mutually beneficial trade intercourse, have contributed to a workable modus
vivendi at this level. The results of a 1999 social inquiry (mainly among Osse-
tians living in Tkhsinvali) confirm this (Kontakt 1999: 12). Nearly half of the re-
spondents assessed the situation as “stable enough” and 42% as having “already
significantly improved,” with some 20% of them convinced that Georgian-South
Ossetian relations will keep progressing in future.
B External Factors
1 Multinational Organizations
Initially, international reaction to the South Ossetia conflict was at best margin-
al, if not utterly passive. This is particularly true for the Western community,
which had three reasons for its limited and delayed involvement. Firstly, it did
not perceive the Caucasus as a source of immediate security threat, unlike the
Balkans,19 for instance, where such a threat was more obvious. Secondly, the West
avoided entering Russia’s ‘backyard’ immediately following the collapse of the So-
viet Union. Moscow reinforced this restraint when it obtained an international
mandate for its self-asserted “special powers as a guarantor of peace and stability
in the region of the former Union.”20 Lastly, the larger scale of recent conflicts in
Abkhazia and Nagorno-Karabakh overshadowed events in South Ossetia and,
subsequently, diverted the international focus away from it.
For these reasons, external involvement was rather limited. By the end of 1992,
i.e., more than three years into the conflict, there were only three major actors:
the Russian Federation, North Ossetia/Alanya within the Russian Federation, and
the Commission on Security and Co-operation (CSCE – later the Organization
for Security and Co-operation in Europe (OSCE)). Other international organi-
zations, such as the European Union, the United Nations High Commissioner
on Refugees (UNHCR), the United Nations Development Programme (UNDP)
and the Norwegian Refugee Council (NRC), only became involved later. The
scope and pattern of their involvement was intended to follow the multiple-actor
19 Although, in the case of numerous ethnic wars in the former Yugoslavia, the interna-
tional community still acted with significant and, at times, regrettable delays.
20 On 28 February 1993, Russian President Boris Yeltsin, in a speech to the Civic Un-
ion Coalition, declared that “the moment has come when responsible international
organizations, including the United Nations, should grant Russia special powers as
a guarantor of peace and stability in the region of the former union” (Human Rights
Watch/Helsinki 1993; Khachikian 2000: 28).
model commonly exercised in modern conflict resolution worldwide. However,

Western unresponsiveness and Russia’s virtual patronage over the conflict un-
dermined the initial remit and power of international actors in South Ossetia. At
first, international organizations were restricted to ‘soft’ and ‘normative’ actions.
These mainly included fact-finding missions (‘soft actions’) or reaffirmation of
the principles of international law and support of other peace initiatives (‘norma-
tive actions’) (Paye and Remacle 1996: 105–110). Notably, in terms of ‘normative
actions’, international actors unanimously condemned the South Ossetian claim
for secession.
The aftermath of the conflict revealed rather different circumstances, where
the role of international organizations had to be extended beyond the traditional
monitoring of the ceasefire and political mediation. As talks over comprehen-
sive settlement froze, the political leadership in both Georgia and South Ossetia
redirected their efforts towards practical issues of immediate or short-term con-
cern. The series of high-ranking meetings between Georgian and South Ossetian
authorities were a clear example of this new focus on practical cooperation. Al-
though they did not produce a comprehensive settlement, these summits suc-
cessfully separated the long-term need for a political solution and the short-term
requirements for regulated interactions in practical areas of daily life. Despite
international, namely the OSCE, efforts to stimulate a shift from mere manifesta-
tions of goodwill to specific political proposals, leaders remained more focused
on the horizontal aspects of pragmatic interaction (OSCE 1998). International or-
ganizations, especially the OSCE, became an integral part of these negotiations,
and subsequently reinforced these processes.
At a communal level, international organizations were besieged with grassroots
and semi-formal activities that emerged spontaneously and largely ran unsystem-
atically. There was a need to set a formal, if not legal, framework around these
activities and international organizations were best placed to serve this purpose.
Furthermore, through a series of public diplomacy projects, international orga-
nizations themselves introduced a string of new activities that aimed exclusively
at horizontal aspects of self-governance and thus stretched international involve-
ment beyond its traditional scope.
Despite initial limitations, international organizations proved willing and flex-
ible to adapt to the local context in South Ossetia and to capitalize on the posi-
tive factors of the post-conflict détente. Consequently, they have created a more
effective mechanism that delivers consistent and productive input into the peace
process. Its notable efficacy rests on two underlying principles: firstly, the clear
distribution of powers and competencies among the external agents without a
large overlap of interests and strategies; and, secondly, a de-politicized approach
focussing on practical issues of immediate and short-term priority.
The first separation of responsibilities occurred in 1993 when the UN took the
lead in Abkhazia, leaving South Ossetia to the OSCE. The OSCE remains the
chief political mediator in the region. Together with Russia, it acts as the ma-
jor international peacekeeper and participant in the Joint Control Commission
(JCC). Interestingly, the authority and activities of these two do not clash sig-
nificantly; rather, they counterbalance and even complement each other. Even
though Russia’s immense influence on the overall peace process remains widely
recognized, the powers on the ground are cleared distributed. It may also be the
consensus-based decision making at the OSCE – with Russia represented equally
with other members – that enables the organization to act as a truly independent
party.
The OSCE in South Ossetia proved more flexible when compared to the UN
in Abkhazia. It managed to move away from purely political mediation towards
more active participation in horizontal initiatives on the ground. On the one
hand, the scope of priority actions of the OSCE itself was expanded. In 1994, the
Permanent Council decided to add the Human Dimensions component to the
mandate of its Mission to Georgia. Human rights, democratic institution-build-
ing, promotion of free and independent media, and facilitation of cooperation at
different communal levels became focal issues. Since 2002, the Office of Econom-
ic and Environmental Issues has been dealing with the matters of special security
relevance in these fields. In 1996, the OSCE Tskhinvali field office, with political
and military representation, was established to complement the regular field mis-
sions to the conflict zone and to permanently monitor and report on the situation
on the ground. These institutional changes, together with a gradual growth in the
size and budget of the Mission, created more space for modification and diversi-
fication of field activities. On the other hand, the pragmatic nature of the internal
conflict regulation capacities stimulated the OSCE shift towards more practical,
non-politicized forums. Therefore, credit should be given to the flexibility of both
the organization that capitalized on and further cultivated positive potential as
well as to the internal actors who allowed international facilitation to develop in
these areas of mutual interest.
As enshrined in its mandate, the OSCE continues to facilitate political negotia-
tions through active participation in and support to the Georgian and Ossetian
plenipotentiary delegations and direct political dialogue at high levels. It also car-
ries on its initial commitment to liaise with and monitor the Joint Peacekeeping
Force (JPKF). Moreover, the OSCE substantially assists the JPKF to carry out its
goals. Since 2000, the Mission has supported the voluntary handover programme
for small arms and light weapons in the conflict zone by nominating compensa-
tion levels and financing small community projects in exchange for weapons. The
OSCE performs most extensively in practical forums such as the JCC, the weekly
administrative and law-enforcement meetings in Tskhinvali, the joint law-en-
forcement bodies, and nearly all other internationally sponsored projects and the
public diplomacy channels in South Ossetia. These fora are instrumental in the
OSCE’s response to the immediate concerns of both Georgians and Ossetians in
the region They allow the Organization to bypass sensitive political uncertainties
and, in parallel, materially assist direct dialogue between the parties at different
layers of authority, which might ultimately lay the ground for political negotia-
tions.
More recently, the European Union (EU) entered the fray through economic
rehabilitation and development projects and financial-technical support to JCC
activities. The EU now participates in the economic working group of the JCC
and has also supported the creation of JCC permanent secretariats since 2001.
The OSCE is an implementing agent for this project. The UNDP, the NRC, and
the UNHCR launched interim humanitarian economic rehabilitation and devel-
opment projects, which have set up instrumental platforms for projects related to
bilateral cooperation on different horizontal aspects of regulation.
The flexible and constructive approach of the international actors in South Os-
setia is in stark contrast to the Abkhazian experience, where such international
actors are yet to produce a significant improvement. International involvement
there is less action-oriented and less adapted to local context. The endless pre-
eminence of political questions over all other aspects of regulation blocks the
conflict resolution process. The Abkhazian impasse indeed largely accounts for
the rigidity of internal actors but also for the inadequacy of external efforts to
push the process forward by different, perhaps more resolute, measures of in-
volvement.
2 Russia’s Dual Role

The Russian Federation has played a dual role in the Ossetian conflict. On the
one hand, its biased and interventionist policies further destabilized the region, at
least from the Georgian perspective. Russia even drew Georgia to the edge of war
when it openly bombed Georgian villages in retaliation for what it claimed was
genocide against Ossetians.21 On the other hand, Russia has emerged as the effec-
tive guardian of stability in the region. Through both independent and joint activi-
ties, it continues to assert itself as the dominant facilitator in the peace process.
Russia uses direct as well as indirect measures to maintain its exclusive posi-
tion in South Ossetia. Most overtly, it has established a military presence, initially
using a temporary dispatch that was later integrated into the Joint Peacekeeping
Force (JPKF). Since the early days of their involvement in 1990, Russian troops
have come under severe criticism for exceeding their designated responsibilities
to secure the ceasefire agreement. They often misused their authority to suppress
any fighting against Georgian combatants and civilians (Human Rights Watch/
Helsinki 1992: 13). There was great suspicion about Russian troops leaving their
armament and heavy weaponry to Ossetians, thus backing them even after their
withdrawal.
The 1992 agreement on the JPKF attempted to balance military power in the
region by giving equal representation to all three vested parties – Georgians, Os-
setians, and Russians. In practice, however, Russia took the lead, partially because
of the lack of volunteers and the poor equipment of the Georgian and South Os-
setian ‘peacekeepers’. This reinforced Russia’s dominant position as the primary,
and now the only, external military force in the conflict zone.
21 Chairman of Russia’s Supreme Soviet Ruslan Khazbulatov demonstrated this in a

statement in June 1992 (Sammut and Cvetkovski 1996: 13–14). Russia expressed its
interest in getting involved and warned that Georgian actions could ‘urge’ Moscow
to seriously consider Ossietians’ request to join the Russian Federation.
The JPKF served Russia in another respect – it made South Ossetia the testing
ground for Russia’s newly institutionalized CIS peacekeeping concept (Leanen
1998: 92). However, there was a significant discrepancy between the CIS peace-
keeping idea22 and its practice in South Ossetia. Firstly, here the CIS did not au-
thorize the deployment of its forces, and then a bilateral agreement was reached
between Russia and Georgia, which at that time was not even a member of the
CIS. Secondly, the JPKF structure was confined to direct protagonists of the war
as peacekeepers and no third-party forces from the CIS were allowed. Thirdly,
the posture of the JPKF was closer to peacemaking than to peacekeeping, as pre-
scribed in the 1992 agreement. Thus, the JPKF in South Ossetia was shaped in a
manner completely at odds with the requirements of the CIS peacekeeping con-
cept.
It should be noted that, overall, the JPKF did serve the purpose of maintaining
stability in the region. The surprising success of this otherwise controversial de-
vice could be explained by Russia’s relative passivity in South Ossetia as well as a
certain degree of cooperation between Georgians and Ossetians. Russia, whether
militarily or politically, was less incendiary in South Ossetia than it had been in
Abkhazia. In the absence of such direct negative influence from Moscow, Geor-
gians and South Ossetians were able to reach mutual control even though their
participation in the JPKF was limited. Also, because of its less inflammatory in-
volvement, the Georgians and Ossetians may have felt less hesitant to let Russia
act as the custodian of stability under the JPKF.
With regard to more covert measures, Russia continues to exploit the ambigu-
ous political terrain of conflict-ridden Georgia. Although an external party to the
conflict, it insists on participating in almost every debate on dispute settlement
and even created a Russian-Georgian intergovernmental forum specifically to
deal with South Ossetia. Moreover, in addition to certain aggressive declarations
of some Russian authorities (OSCE 2000c: [I/II]; OSCE 2000b: [II]), the Russian
Duma on several occasions has discussed and issued statements on Abkhazia
and South Ossetia. The Georgians have often found these actions provocative.
From the perspective of international law, it is unclear why Russia has at times
assumed the same level of authority as the Georgian federal government.
It is difficult to explain Russia’s somewhat inconsistent involvement in South Os-
setia, and it is even more challenging to reconcile it with Russia’s decisively nega-
tive role in Abkhazia. Some Georgian academics and officials claim that Russia has
had differing strategic interests in the two cases.23 According to this view, Russia
saw higher stakes in Abkhazia than in South Ossetia and for this reason its often
22 The CIS peacekeeping concept was first enshrined in the “Agreement on Group of
Military Observers and Collective Peacekeeping Forces in the CIS”, signed by ten
member states at the Fourth CIS Summit in Kiev, Ukraine, March 1992.
23 Interviews with Ghia Nodia, the director of the Caucasian Institute for Peace, De-
mocracy and Development; Alexander Rondeli, the president of the Georgian Foun-
dation for Strategic and International Studies; David Darchiashvili, the head of the
Research Center, the Parliament of Georgia, July–August 2001.
contradictory policy had a greater role in the former. The argument is only partially
acceptable as it risks making an error of attribution. On the one hand, it is hardly
disputable that Russia, as the key external actor, established herself in a unique
position to influence the conflict regulation processes both in Abkhazia and South
Ossetia. Thus, Russia’s uneven strategic interests may explain the different degrees
of international involvement in the two conflicts. On the other hand, a share of re-
sponsibility goes to the joint ventures of other external actors as well because they
appear unable to counterbalance the negative exogenous interference in the peace
process, if or when they occur, be it from Russia’s or any other side.
3 North Ossetia/Alanya’s Moderate Policies

Introducing North Ossetia as an independent external party to the Georgia–South
Ossetia conflict is another example of Russia’s indirect involvement.24 The move
hoped to exploit the ethno-cultural kinship between this Russian autonomous re-
gion, situated on the northern slopes of the Caucasus Mountains, and South Os-
setia, located on the opposite side. However, unlike their kinsmen in Tskhinvali,
leaders in Vladikavkaz (North Ossetia) displayed less nationalist and political am-
bitions for south-bound expansion. North Ossetian statesmen managed to main-
tain a delicate balance between ethnic affinity and political affiliation with South
Ossetia. While broad economic, social, and cultural cooperation was encouraged
as a sign of ethnic solidarity, there were no stated ambitions for outright political
union. Furthermore, North Ossetian leaders, Galazov (until 1998) and Dzassok-
hov (after 1998), officially called upon Ludwig Chibirov in South Ossetia to pursue
“normal relations with Georgia” making it clear that “unification of the two Os-
setias is unrealizable at the present stage” (Centre for Russian Studies 2001).
The rationale for such a moderate approach appears to be financial as well
– North Ossetia was neither willing nor able to absorb or sustain its economically
depressed neighbour. Whether because of the lack of economic motivation or
presence of political vision or both, Vladikavkaz distanced itself somewhat from
the radicals in Tskhinvali and thus contributed to a constructive environment for
negotiations. With the prospects for reunification with their northern kinfolk be-
ing unpromising, the secessionist cause in the South had less legitimacy and the
leaders who advocated it possessed less bargaining power.
Abkhazia is again at the other end of the spectrum. Lacking any political or eth-
nic kin, Abkhazian leaders identify with some insurgent fundamentalists in the
North Caucasus. Through alliances with other similar breakaway entities in the
post-Soviet area, they aim to secure political legitimacy for these unrecognized
regimes. Abkhazia is part of bi- and multilateral structures and agreements with
Nagorno-Karabakh, South Ossetia, and Transdniestria.25 While South Ossetia’s
24 Northern Ossetia is an autonomous oblast in Russia. It lies on the other (northern)

slopes of the Caucasus.
25 In November 2000, Abkhazia, Nagorno-Karabakh, South Ossetia, and Transdnies-
tria formed a Conference of Foreign Ministers (ITAR-TASS 2000). Later, on 23 Au-
participation in these liaisons is nominal, for Abkhazia they represent an impor-

tant tool in coordinating its policies and ultimately achieving independence.
4 Abkhazia as a Diversion
The concentration of wider international attention on the Abkhazia case has, in
a way, a double effect on South Ossetia’s post-conflict development. On the one
hand, it hampers advancement, since South Ossetia’s political status is virtually
tied to that of Abkhazia. Tskhinvali seeks the same settlements that So(u)khumi
can negotiate with Georgia. For Ossetians, a ‘lesser offer’ is unacceptable and
thus Georgian-Ossetian negotiations are locked in a vicious game of wait-and-
see. On the other hand, the same disposition in South Ossetia helps to divert the
focus of the conflicting parties from highly vulnerable political questions towards
‘less vulnerable’ functional cooperative arrangements. The day-to-day pragmatic
interaction at the grassroots and informal/semi-formal level might point to a lon-
ger but more effective route towards a mutually acceptable comprehensive settle-
ment in future.
IV Power-sharing Arrangements in South Ossetia

South Ossetia could be considered as a ‘deviant’ case in the practice of complex
power-sharing. There is no comprehensive settlement aimed at resolving the
conflict and/or formalizing the existing public and private discourse. However,
the lack of a comprehensive agreement has not prevented the development of
genuine and even substantial cooperation at the grassroots level. To make the
case even more peculiar, there is little legal framework to support the existing
arrangement itself. Numerous political efforts, whether bilateral or multilateral,
have largely been confined to administrative and organizational issues. The ma-
jority of the interactions, especially economic, are rooted in an ‘implicit’ social
contract. Only a few interactions, mainly with international exposure, are gov-
erned by written agreements. Although impressive in itself, this under-defined
and under-regulated practice is only an embryonic version of what could become
a transparent system based on a successful complex power-sharing principle.
South Ossetia failed to capitalize on its unique combination of favourable in-
ternal and external factors when the time for serious peacemaking arrived. Al-
though these factors originally restrained the conflict somewhat, they appeared
insufficient to produce a viable long-term settlement. Another set of powerful
reasons forestalled a full-scale settlement and reduced the scope of negotiations
to immediate concerns, essentially avoiding sensitive political issues. As has of-
ten happened in South Ossetia, this setback did not obstruct peacemaking in
other areas, which have proved flexible enough to survive without a general ac-
cord. This ability to move forward, even if against the odds, later became advan-
tageous to the ultimate emergence of grassroots power-sharing in the region.
gust 2001, Abkhazia and Transdniestria signed a bilateral agreement reiterating full
support for each other’s politics (Radio Free Europe/Radio Liberty 2001).
With significant international facilitation, the current mechanism is still effec-

tive. On the one hand, it leaves room for future peace talks and, on the other, it
defines the procedures for specific arrangements in the areas of administration,
economic rehabilitation, and refugee repatriation. These agreements constitute
South Ossetia’s only political settlement device and serve as a link to spontaneous
interactions that have come to dominate the region.
Functional interactions run through a complex structure of multiple parties
liaising at different levels on various issues. Hence, the three dimensions of the
following analysis: participants, level of involvement, and sectors of cooperation.
The participants in these spontaneous interactions range from high-ranking of-
ficials to private individuals from both sides. They form multilayered relations at
informal, semi-formal, and formal levels. The scope of their interaction, however,
is less diverse. Intercommunal contacts are more intensive across economic and
administrative sectors, especially at informal and semi-formal levels. Humani-
tarian and rehabilitation projects are mainly coordinated by international donor
organizations, while social and human rights issues have a rather weak represen-
tation at any level.
The existing grassroots arrangements have gradually emerged from fragment-
ed interactions, mostly prompted by urgent needs of post-war living conditions
and the coincident earthquake. An effective ceasefire allowed both Georgians and
Ossetians in the region to act according to their economic self-interests and in
meeting their daily requirements. The resulting intercommunal trade gave impe-
tus to further cooperation in other related areas such as transportation. Moreover,
the tightly interrelated economy prior to the conflict also encouraged extensive
cooperation. Because the South Ossetian market has low purchasing power and
can quickly reach saturation, private business activities thus must cross regional
barriers. Consequently, the ‘transborder’ commerce along the main trade route
in South Ossetia running northwards to the Russian Federation has become the
major, if not the only, income source for local Georgian and Ossetian communi-
ties. The flourishing trade rests upon significant collaboration between interest
groups from both sides, even if inner regulations are largely implicit. Despite se-
rious legal shortfalls, these virtually ‘duty-free’ transactions have generated con-
siderable positive value for intercommunal rapprochement. Namely, the informal
commerce has prompted a certain degree of interaction in the fields of security,
administration, and transportation.
International facilitation sought to formalize this spontaneous discourse
to some extent by introducing official negotiations and agreements. External
mediation has been crucial in the establishment, as well as the functioning, of
the region’s first effective administrative body – the Joint Control Commission
(JCC).26 Through its working groups and an ad hoc committee, the JCC develops
and coordinates joint endeavours on military, economic, and social issues. More
importantly, multilateral organizations appear to be in the best position to deal
26 The Joint Control Commission was established in accordance with the Yeltsin-Shev-
ardnadze agreement on 24 June 1992.
with the politically more sensitive and yet critical issues such as the repatriation
of refugees and the rebuilding of infrastructure. Internationally sponsored pro-
grammes have broadened the spectrum of bilateral relations to include public
diplomacy and confidence-building projects.
The aforementioned development style allows us to provisionally differenti-
ate between two types of horizontal power-sharing arrangements – structural
and instrumental. Structural arrangements rest on needs-based considerations
and issues of the parties’ mutual interest. Originally, they grew out of indigenous
factors of post-conflict development and gained further momentum through
international support. Instrumental arrangements, on the other hand, emerged
from exogenous efforts of external factors with the purpose of encouraging rap-
prochement and re-integration between the two sides. By establishing power-
sharing as a prerequisite for funding, international organizations hoped that the
arrangements they created would become instrumental in bringing about more
comprehensive changes.
South Ossetia is compelling not only because it gave rise to unique grassroots
cooperation but also because it managed to maintain and invigorate its experi-
ence over the years. The notable efficacy of this mechanism and the key to its
survival lie in the three-way separation of issues, actors, and hierarchy. Political
and specific practical issues are successfully channelled into independent par-
allel tracks involving different actors at different levels. High-ranking negotia-
tions remain focused on an overall solution while local efforts can concentrate
on immediate concerns. Hence, political settlement is no longer a precondition
for cooperation on practical short-term issues, even if with longer-term implica-
tions. Neither is there a harmful overlap of domains of interest and competencies
among the many actors involved. Participants, whether domestic or foreign, have
demarcated their spheres of involvement fairly clearly. South Ossetia benefits
from a structure that defines in a non-competitive manner who is responsible
for what and when. Thereby, in South Ossetia it is possible to negotiate different
aspects of regulation without a single comprehensive package. This experience
is in sharp contrast with Abkhazia, where political uncertainty blocks all other
issues from being negotiated, let alone resolved.
The following section examines the practical cases of power-sharing in South
Ossetia. It analyzes the reasons for failed and failing political settlement and
whether the existing practices have matured enough to be transformed into for-
malized agreements. Furthermore, it attempts to identify other potential areas of
cooperation that still remain open to complex power-sharing.
A Political Settlement Process

1 Conflict Settlement Mechanism
Russia and Georgia made the first formal attempts to reach a durable peace in
South Ossetia. The rounds of unofficial and official negotiations in 1991–1992
brought several fragile ceasefires before reaching the 14 July 1992 agreement. The
effective ceasefire and the deployment of mixed peacekeeping forces are the only
tangible results of this relationship. Procedural delays and political tensions be-
tween Russia and Georgia have obstructed the implementation of many bilateral
agreements27 throughout the years.
Cooperation between the two countries proved more effective under OSCE
mediation. The significant achievement of Russian and OSCE facilitation lies in
the endorsement of internationally sponsored bodies that function on the basis
of the complex power-sharing principle. In terms of instrumental devices, these
mainly include the Joint Control Commission (JCC) and its subordinate institu-
tions, which constitute the core of the conflict settlement machinery. The JCC
comprises three working groups each addressing issues related to: 1) the military
and security; 2) the economy; and 3) refugees and IDPs. The latter has a further
ad hoc committee, which is enhanced by the participation of the UNHCR. The
OSCE provides political guidance and technical support to the JCC and its sec-
retariats.
The current version of the JCC is a product of long negotiations as well as a
testament to the flexibility of the South Ossetian peace process, most notably its
chief facilitator, the OSCE. Initially, the OSCE strove to reconcile Georgia’s terri-
torial integrity and South Ossetia’s self-governance. In August 1994, it introduced
a draft proposal on the constitutional status of South Ossetia within Georgia.
Even though the document suggested various options based on the power-shar-
ing principle, the parties found it difficult to reconcile their political interests.
As talks stalled, the JCC was transformed into a platform for reconciliation and
confidence-building both at the communal and elite levels. The Memorandum on
Measures to Provide Security and Strengthening Mutual Trust Between the Sides
in the Georgia-Ossetian Conflict (1996) marked the first tangible achievement of
the JCC efforts to stimulate bilateral negotiations.28 The agreement was reached
at the JCC forum and signed by the JCC members: Georgia, South Ossetia, Rus-
sia. and the OSCE. The Memorandum itself said very little about political settle-
ment; rather, it denounced the use of force, reiterated the need to ensure the safe
return of refugees and to support further peace negotiations by encouraging civic
dialogue. However, the Memorandum set the tone for the following high-profile
meetings as well as the internal working sessions in terms of marking the start
of a détente in Georgian-Ossetian relations: it actually separated long-term po-
litical concerns from short-term needs. This shift towards more practical coop-
eration became especially apparent during the Chibirov–Shevardnadze talks in
27 Pursuant to the 14 September 1993 “Agreement between the Government of Rus-

sian Federation and the Government of Georgia on Economic Rehabilitation of the
Regions in the Zone of Georgia-South Ossetia Conflict,” Moscow pledged to fund
one third of emergency assistance, totalling an estimated rub 34.2 billion, whereas
Tbilisi promised to pick up the rest of the bill. However, these financial commitments
have not been fully adhered to, either within the initial terms of 1993-1997 or subse-
quently.
28 The full text of the Memorandum is available at http://www.ecmi.de/cps/docu-
ments_gum_memo.html.
1996–1998. Although these talks did not produce any tangible political outcome,
they encouraged a whole series of sector-specific arrangements and pledges of
support for economic cooperation and facilitation of international donor proj-
ects and rehabilitation.
The OSCE, through the JCC, reacted to the changes in the political climate in
South Ossetia by institutionalizing and, thereby, facilitating them. It strongly sup-
ported the creation of and the activities of Georgian and South Ossetian Plenipo-
tentiary Delegations in 1997, which consolidated all preceding, concurrent, and
future efforts for conflict regulation under their coordination. The plenipoten-
tiaries, with their expert groups, are vested with special authority to negotiate
full-scale resolution and to work in close cooperation with Russia, North Os-
setia, and the OSCE. Recognizing it as a pillar of peacemaking, other interna-
tional organizations have joined the OSCE in their support of the JCC-faciliated
platforms. Markedly, the plenipotentiary activities enjoy considerable financial
assistance from the European Union. The plenipotentiary forum generated a
benign platform for constructive thinking and political rapprochement that has
been adequately reflected over the course of the peace process since 1997. It was
at the Expert Groups meeting in Baden/Vienna in 2000 that the most significant
agreement (after the 1996 Memorandum) was reached. According to the OSCE
assessment, the “Declaration on the Basis of State-Legal Relations between the
Sides of Georgia-South Ossetia,” is a serious attempt “to bridge the gap – to the
maximum extent possible – between the positions of the sides” (OSCE 2000b). It
was the first time that the Georgian and South Ossetian sides directly discussed
future constitutional relationship issues. Three items were deemed appropriate
for further consideration and inclusion into a single package: 1) state-legal ar-
rangements within internationally recognized borders; 2) the attributes of South
Ossetia’s future status; and 3) the mechanisms of international guarantees. The
document did not stipulate any specific political and territorial arrangement in
itself, but it made concrete steps towards institutionalizing grassroots and semi-
formal power-sharing elements in South Ossetia. Among the horizontal aspects
agreed upon were economic rehabilitation and reconstruction projects, juridical
support for law enforcement, and business cooperation through joint ventures,
as well as joint fundraising and control over allocation and use of international
funds for South Ossetian rehabilitation. Interestingly, much of the progress in
this forum is due to good interpersonal relations, which have evolved through
regular meetings and, in turn, have created a favourable environment.29
Active engagement of all levels of public decision-making is another reason
for the success of the conflict settlement mechanism. Each Thursday, local law
enforcement and administrative authorities hold meetings at the JPKF Russian
Command Centre in Tskhinvali under the chairmanship of the Russian Com-
mander of the JPKF, and with the participation of the OSCE Tskhinvali and Tbili-
29 Interviews with Irakli Machavariani, head of the Georgian Plenipotentiary Delega-

tion and Archil Gambashidze, member of the Georgian Plenipotentiary Delegation,
July 2001.
si office representatives. Whereas high level negotiations prepare the ground for
a comprehensive settlement to the conflict, the local administrative weekly meet-
ings settle day-to-day practical questions of a lower profile. The regular ‘Thurs-
day meetings’, as they are commonly known, in reality perform joint trilateral
administrative functions (Georgia, South Ossetia, and the JPKF, with the active
participation of the OSCE), and serve as an important backbone to the liaison
between the parties.30
2 The Failed Overall Political Settlement

In South Ossetia, the conflict settlement process has been more stable and prom-
ising than that in Abkhazia, but it was not substantial and viable enough to pro-
duce a genuine political settlement. Ossetians found it difficult to give up their
newly found de facto independence from Georgia and engage in serious political
talks where Georgia’s territorial integrity was uncontested. Georgians, for their
part, never accepted South Ossetia’s self-proclaimed independence and thus did
not want to negotiate any agreement outside Georgia’s territorial integrity. With
this fundamental issue unresolved, it became impossible to discuss any compro-
mise solutions within the framework of a certain federal state. For a decade now,
both sides seem to have been trapped by external and internal forces that prevent
them from capitalizing on the positive achievements of both high-profile nego-
tiations and the grassroots cooperation.
Firstly, there is an ‘Abkhazian factor’ that greatly affects the South Ossetian
political settlement process. In particular, the South Ossetian authorities seem
unwilling to secure any full-scale solution before the Abkhazia issue is resolved.
They are playing a ‘waiting game’ and prefer to follow the Abkhazia path. South
Ossetians believe that the political stakes are higher in the Georgian-Abkhazian
conflict and, accordingly, Abkhazians can bargain for a more favourable settle-
ment. Tskhinvali authorities, thus, aim to reach the best possible result, asserting
that “whatever is the political solution to the Abkhazia conflict, South Ossetia
will accept nothing less.”31
Secondly, the ‘Russian factor’ is often regarded as a major determinant for
both the Abkhazian and the South Ossetian peace processes. Notably, the re-
sults of a social inquiry conducted in South Ossetia in 1999 demonstrated that
the overwhelming majority of Ossetian respondents (82.4%) believe any change
of the state of the conflict and its solution ultimately depends on Russia’s policy
(Kontakt 1999: 12). The immensity of Russia’s influence on the Transcaucasus in
general, and in South Ossetia in particular, is no longer contested in Tbilisi either.
30 JPKF and OSCE participants of the Thursday meetings both underscored the prag-
matic and instrumental importance of such a regular interaction of the local admin-
istrative bodies for maintaining stability within the region. Interviews with General
Churaev, Commander-in-Chief of JPKF, and interviews with the members of the
OSCE field office in Tskhinvali, July 2001.
31 Interview with Konstantin Dzugaev, Speaker of the ‘Parliament’ of South Ossetia,
July 2001.
However, the nature of Moscow’s immediate political and strategic interests is

a subject of great speculation. Its often contradictory official policy and perfor-
mance raise suspicions that the current status quo might serve Russia’s aim to
maintain an instrumental platform (through peacekeeping and facilitation) for its
influence and control over the whole region, and in Georgia particularly. Alterna-
tively, one could explain Moscow’s ambiguous politics with a wish to eventually
absorb South Ossetia within Russia, as a pretext for this already exists.32 In this
case, Moscow is often accused of deliberately impeding a political settlement, or,
in more moderate assessments, making insufficient efforts to use all the tools at
hand to put an end to the dispute.33 It is difficult to imagine a constructive peace
process in South Ossetia without either deposing Russia as the regional ‘benefac-
tor’ or Russia itself adopting a more productive approach.
Thirdly, there are interest groups on both sides who prefer to maintain the cur-
rent quasi-peace, as they benefit greatly from widespread ‘duty-free’ commerce
and smuggling. Unsurprisingly, these groups are not interested in a formalized
and regulated market that would deprive them of their illegal income sources.
Unfortunately, the Georgian government cannot offer a powerful alternative in
terms of economic incentives that would make a political trade-off more appeal-
ing. Georgia is not able to cover even those modest financial allotments for the
Tskhinvali region rehabilitation projects that are earmarked for this purpose in
the annual state budget. Georgian plenipotentiaries commonly point to the lack
of strong economic leverages as one of the major drawbacks for an effective move
towards a final settlement. Peace in South Ossetia will only be possible if it will be
economically more profitable to all parties involved than the current situation.
B Structural Power-sharing Arrangements

1 Security and Law Enforcement
The JPKF, established by the July 1992 ceasefire agreement, was the first – and a
fairly effective – instrument for cooperation in the area of security and law-en-
forcement. The simultaneous creation of an ad hoc Working Group on Military
and Security Issues (with Georgian, Ossetian, JPKF, and OSCE participation)
under the JCC introduced another multilateral device to prevent, investigate
and combat crime in the region. Due to increasing progress in law-enforcement
collaboration, certain functions of the JPKF were gradually delegated to local
administrative bodies.34 At the beginning of 2000, the long collaboration and
facilitation efforts of the OSCE yielded a permanent bilateral joint decision-mak-
32 South Ossetians are ethnically kindred with North Ossetia, which is within the Rus-
sian Federation, and Tskhinvali authorities often call for unification with it.
33 Apart from secondary sources, the argument is based on the interviews with Kon-
stantin Dzugaev, Gia Gvenetadze, Irakli Machavariani, and Alexander Rondeli.
34 In “Protocol on the Interaction of Law-Enforcement Bodies in the Georgian-South
Ossetian Conflict Zone,” JCC Working Group on Security and Military Issues meet-
ing, Djava, South Ossetia, 26 September 1997.
ing body – the Joint Law-enforcement Co-ordination Center – with JPKF par-
ticipation and under JCC supervision.35 Together with its subordinated groups,36
the Joint Law-enforcement Co-ordination Center and local ‘Thursday meetings’
form a powerful platform, not only for crime prevention, but also for overall de-
militarization of the conflict zone. In addition, the Co-ordination Center assists
with search operations for missing persons, as well as with investigations of war
crime allegations.
The operational capability of the joint ventures has proved to be satisfactory.
At least, according to OSCE and the JPKF assessments, there are no more ethni-
cally motivated criminal activities in the conflict area37 and stability in general is
maintained. The system is not without shortcomings though. While the existing
mechanism is effective in policing crime, it may not be adequate for administering
justice. Control and investigation are carried out jointly, but for political reasons
the competencies for arbitration are distributed in accordance with the ethnic
settlement pattern of the region. The Tkhinvali regime does not have full control
of the territory, which is defined as South Ossetia or the Georgian-Ossetian con-
flict zone. Some administrative units do regularly take part in the countrywide
elections. There are some purely Ossetian and some purely Georgian populated
settlements, but over 40% of the area is mixed. The ‘territorial principle’ of po-
licing crime is based on informal agreement between local Georgian and Osse-
tian law enforcement agents to demarcate their spheres of influence according to
ethnic lines. Georgian police control the predominantly Georgian villages, while
Ossetians supervise the mainly Ossetian villages. Uncertainty emerges in mixed
villages where such lines are blurred.38 The cases that fall within heterogeneous
settlements are delegated to one or the other side according to the place of origin
of the matter regardless of the ethnic affinity of an offender. The JPKF collects
and updates data on the ethnic proportions in these so-called ‘grey zones’. The
sides become responsible for the arbitration of the cases under their exclusive
competence.
2 Trade
In parallel with security concerns, economic devastation caused by the war and
a series of earthquakes necessitated the initial emergency-based transactions,
which later expanded to structural cooperation. The original informal channels
35 The Joint Law-Enforcement Co-ordination Center became operational on 29 Febru-

ary 2000, pursuant to the Resolution of the 10th Meeting of the JCC of 23 July 1999.
36 The Co-ordination Center comprises the Council of Plenipotentiary Representatives
of Georgian and Ossetian executives, as well as three subordinated groups: an inves-
tigation and operations group, a group for organizing interaction between law-en-
forcement officers, and a data processing centre.
37 Interviews with General Churaev, JPKF Commander-in-Chief; Jan Kukuk, head of
the OSCE Tskhinvali Office, July 2001.
38 For the present ethnic settlement pattern in the South Ossetian conflict zone, see the
map of Tskhinvali region at http://www.ewers-iccn.ge/map.html.
for intercommunal economic interactions were revitalized soon after the cessa-
tion of hostilities. The communal trade and trans-border commerce along the
important trade route running north through South Ossetia soon turned into the
only income-generating business for the larger part of South Ossetia’s population
and its neighbouring regions of Georgia proper. The absence of formal customs
and tolls along that perimeter created favourable conditions for virtually ‘duty-
free’ transactions. The unofficial tariff on the illegal ‘Roki pathway’39 amounts to
only 3%40 of the gross value of goods and is at least five times less than the tariff
applied at one of the Georgian-controlled Larsi customs post (another mountain
tunnel across the Georgian-Russian border). Therefore, many traders found the
South Ossetia trade route to be the cheapest and most profitable way to transport
goods to and from the Russian Federation, particularly unlicensed cigarettes and
petrol. Unsurprisingly, the passage circumventing official customs turned into a
thoroughfare for booming drug smuggling, contraband trade, and trafficking in
both directions.
The informal trade marked the initial alleviation of antagonism and shaped im-
plicit arrangements on the distribution of benefits and resources. The enormous
so-called Falloy Market located by the main Gori-Tskhinvali road adjacent to the
‘border check point’ on the Ossetian side is, in a way, the stage for a grassroots
power sharing practice. The Ergneti market, where prices are about half of that in
Georgia, is at first sight an abnormal self-regulated shadow economic machine.
In fact, the regulation of transactions and profit distribution is in the hands of
certain communal groups beyond the official control of either Georgian or South
Ossetian authorities. This, however, does not necessarily exclude the possibility
of unofficial participation of some of the official bodies41 who prefer to keep the
inner workings of intercommunal trade relations as obscure as it seems from the
outside.42
Notwithstanding its curious positive impact on reconciliation and peace de-
velopment, the establishment of freewheeling illegal transborder commerce rais-
es concerns for Tbilisi officials. Firstly, the absence of formal customs and tolls
regulations over South Ossetia trade route has created favourable conditions for
smuggling and criminal transactions. Secondly, the profits completely evade the
Georgian budget and generate little benefit to the South Ossetian budget either.
The loss of revenue is especially severe, as South Ossetia remains the only land
route connected to the lucrative Russian market. The post-Shevardnadze admin-
39 The South Ossetian authorities claim the pathway and tariffs are legal because South
Ossetia’s ‘customs code’ and respective laws on tariffs regulate it. The ‘pathway’ is a
tunnel under the Caucasus mountain range.
40 Interview with a member of the OSCE Mission to Georgia, Tbilisi, June 2001.
41 The assessment is based on confidential interviews.
42 Two Georgian newspaper correspondents who tried to shed light on these details
were captured while taking photos at Falloy Market in June 2000. They were released
after two days of detention by a Georgian and South Ossetian joint police operation
(interview with one of the detained from the Georgian newspaper Kviris Palitra).
istration in Georgia pledges and has already taken steps to curb the shadow econ-
omy that flourishes along this route. However, the details of regulation remain to
be considered given the political vulnerability of the region and the interests of
ordinary local populations involved, whose only earnings depend on this daily
commerce.
International organizations have been involved in efforts to stop this massive
tax evasion and smuggling scheme for some years. Several proposals, including
the European Commission’s initiative on joint customs regulations and its later
proposal of a ‘magistral tax disk system’ under the JCC, are still to be elaborated,
discussed, and agreed by the sides. The project intends to introduce special pre-
paid licence stamps for the vehicles that transit and import specific goods. This is
an alternative to the EC’s initial idea of having joint customs posts on the Geor-
gian-Russian border, which the parties found difficult to accept. Although more
modest than the original proposal, the system would be, in fact, a substitution
of the grassroots commercial arrangement with an official power-sharing settle-
ment – a joint tax regulation policy – more beneficial to both the Tbilisi and
Tskhinvali budgets. Because of its political connotations and implications, the
joint customs issue is the subject of cautious ‘cost-benefit calculations’ in Tskh-
invali. Although concrete projects are not being discussed officially, South Os-
setian senior authorities do not exclude the possibility of agreeing to a common
customs policy,43 but they intend to bargain with Tbilisi for an equal sharing of
customs revenues.44 However, it is questionable whether such a deal would serve
the latter’s interests.
3 Transportation
The spontaneous emergence of informal trade effectively opened the ‘borders’
between Georgia and the insurgent region. The three checkpoints (Georgian,
Russian, and South Ossetian) on the Gori-Tskhinvali highway regulate the move-
ment of public transportation and private vehicles both with Georgian or South
Ossetian licence plates. The number of checkpoints along the entire route up to
the Georgian-Russian border creates a bureaucratic nightmare and fosters ram-
pant corruption among Georgian, Russian, and Ossetian law enforcers who may
desire to divert transport-generated revenues.
According to Georgian sources, there is a new initiative to expand this trans-
portation corridor and devise more flexible rules for road regulation based on a
special arrangement between the Ministry of Transportation of Georgia and the
43 South Ossetian ‘President’ Ludwig Chibirov articulated such a possibility during a

visit by State Minister of Georgia Vazha Lortkipanidze to Tskhinvali in February
2000. More recently, in October 2002, the Plenipotentiary Experts Groups agreed
on the basic principle to introduce a joint system to increase revenues from the trade
road.
44 As stated by Merab Chigoev, the ‘head of government’ of South Ossetia (OSCE
2000b: [II] 6).
relevant South Ossetian body.45 Still, after passing the short distance from the
Georgian to the South Ossetian border post, the clocks have to be set back an
hour to Moscow time, which the South Ossetian authorities prefer to maintain as
a ‘reminder’ of their ‘independence’ from Georgia.
Apart from the spontaneously emerging functional grassroots arrangements,
there are a number of contacts that have developed between large private compa-
nies and individual entrepreneurs, namely in satellite telecommunications, small
business, and production sectors. Such links, which in some cases have brought
about the establishment of joint Georgian-South Ossetian enterprises in Tskhin-
vali, are positively viewed at high-level official meetings as complementary efforts
to widen the constructive cooperation of the parties. Ideally, these contacts might
establish preconditions for the successful development of more official attempts
at structural cooperation; they may form a template for more official attempts at
cooperation. However, at this stage, private business enterprise has not emerged
as a powerful enough incentive for such an augmentation.
C Instrumental Power-sharing Arrangements

1 The UNDP Tskhinvali Rehabilitation Project
The UNDP Rehabilitation of the Tskhinvali Region project (GEO/96/019) was
the first operative economic programme in South Ossetia. When it was started
in early 1997, negotiations were still effectively deadlocked and the UNDP had
to ‘break the ice’ of mutual distrust before launching the project itself.46 Hence,
the organization embarked on an integrated strategy of responding to both eco-
nomic needs and the necessity of launching a confidence-building process. These
two directions served the corresponding goals of reconstruction and rapproche-
ment.
In terms of organizational and operational arrangements, the joint device rest-
ed exclusively on the equal power-sharing principle. Georgia and South Ossetia
acted as direct decision-making and implementation agents while the UNDP re-
mained a neutral facilitator. Under the UNDP requirements, project priorities
and particulars were identified, agreed upon and implemented through consen-
sual decisions of the parties to the conflict (Hansen 2000: 4). In this manner,
proportionality and consensus became the ground rules for any joint exercise
within the project framework.
The UNDP facilitation policies had important political implications. Namely,
compliant with the international and, in particular, the UN position, the UNDP
supports Georgia’s territorial integrity and jurisdiction over the conflict zone.
Hence, the UNDP appointed Georgia’s federal administration (and not the Tskh-
invali de facto regime) as the direct recipient of the project funds. South Ossetian
45 Interview with Irine Skhirtladze, Georgian Plenipotentiary Delegation on Georgia–

South Ossetia Conflict Regulation, July 2001.
46 Interview with Marco Borsotti, UNDP Resident Representative in Georgia, July
2001.
authorities, despite initial resentment, faced the necessity to accept the Tbilisi-
intermediated delivery of badly needed resources. The joint apparatus consisted
of three different structures that coordinated and complemented each other at
all levels of functional interactions, from high-profile official executives to indi-
vidual local agents. The decision process ‘trickled down’ from the Steering Com-
mittee, where Georgian Foreign Ministry delegation and the ‘Special Advisor to
the President of South Ossetia’ defined the priority sectors for the optimal use of
the UNDP funds. The OSCE was entitled to hold the status of observer/facilita-
tor (Hansen 2000: 9). At the next level, senior experts and ministry authorities of
the two sides elaborated detailed proposals and implementation under the Sec-
tor Working Groups, instituted both in Georgia proper and in South Ossetia.
Finally, two Project Support Units, comprised of technical experts from both
sides, monitored organizational matters and ensured the Project’s implementa-
tion transparency.
On the whole, the joint decision-making demonstrated a substantial degree
of pragmatism regarding the rational use of the UNDP resources. The construc-
tive dynamism of the joint mechanism, for its part, was largely defined by the
UNDP-vested ‘compulsory’ formula of ‘either consent or nothing’. Initially, dur-
ing somewhat ‘obligatory’ joint efforts within the project framework, the par-
ties realized the value and, perhaps, the inevitability of sharing powers and re-
sources rather than contesting their monopolization – which would carry equal
risks and costs of failure for each side. Later, mutual benefits turned this almost
forced joint exercise into ‘voluntary’ bilateral cooperation. As a result, the UNDP
project implementation process (1996–1999) became the driving force behind in-
cremental softening of initially hardline positions. Furthermore, it succeeded in
removing ambivalent political aspirations from pragmatic interactions, as much
as the circumstances permitted, and project-related cooperative ventures served
as constructive inputs into the political negotiations.
2 The EC Rehabilitation Project

The EC rehabilitation project consists of two major financial donations made
under the EC 1997 and 1998 budgets. A third transfer was earmarked in the Eu-
ropean Union 1999 budget. The stated goals of the EC programme are similar
to those of the UNDP project. Also, similarly to the UNDP funding principles,
the EC delegation and the Georgian government signed an agreement endorsing
Tbilisi as an immediate beneficiary of all project-related transactions.47 Despite
essential similarities, the EC and UNDP projects opted for different approaches
regarding how and when power-sharing among the parties to the conflict was ap-
plied. Strategic issues, such as priority areas for funding, were designated by the
47 Emmanuel Anquetil (project manager at the EC Delegation to Georgia) mentioned

that the Georgian side ardently advocated the introduction of an ‘EC-Tbilisi-Tskhin-
vali’ intermediary formula as a prerequisite for starting the project-related negotia-
tions with the South Ossetian authorities (interview, July 2001).
EC itself. Project implementation was delegated to local entities, strictly under

the EC-developed procedures and guidelines.
This ‘design centrally, implement locally’ approach was already present at the
beginning of the project when its Joint Co-ordination Body was created. Its found-
ing moderators, the EC and Hilfswerk Austria48 were in charge of defining priority
directions as well as technical procedures, whereas Georgian and South Ossetian
delegates enjoyed only limited powers and, thus, assumed limited responsibilities
in the process. This mechanism left little space for consensus-building among the
conflicting parties, as they were on the receiving end of pre-determined rules of
engagement. At the same time, and again similar to the UNDP project, the EC’s
strict requirements drove both sides into more congruent ways of interaction in
order to secure much-needed financial or technical assistance. The parties were
almost forced to cooperate by following a strict set of instructions before they
could obtain any funding. However, unlike the UNDP project, the mechanism
proved less effective in overcoming negotiation deadlocks and procedural delays.
Because local parties were not privy to the project inception process, they found
it easier to excuse themselves from responsibility over its ultimate progress. In
the case of South Ossetia, this general imbalance was further compounded by the
inflexibility of the Hilfswerk organization.
While the EC initial projects prompted merely short-lived ‘unavoidable’ bilat-
eral interaction during the project implementation, the second installment aimed
to foster more direct and sustainable cooperation that would lead to economic
reintegration and interdependence. The “Platform for an Agreement” was a single
comprehensive package internally organized as a set of new projects addressing
vital economic issues. 49 Georgian ministerial officials, together with their de facto
counterparts in Tskhinvali and local enterprises, were to implement the project
under the supervision of the OSCE and the EC. All contractors were bound by
a common imperative to distribute powers, commitments, and responsibilities
proportionally. This allowed the EC to synchronize actions across the board and
to use certain financial projects as leverage to achieve success where progress
was not forthcoming. The more ambitious third instalment envisages two major
projects: common customs regulations across the South Ossetia trade route and
the reconstruction of transportation infrastructure. The EC has linked the badly
needed road reconstruction to the sensitive joint customs policy issue. Tskhinvali
authorities thus find themselves tempted to make political concessions over the
latter in order to earn the desired funding for the road works.
The EC endeavours have not yielded genuine power-sharing practices thus far,
but they may yet pave the way for sustainable power-sharing arrangements. Most
of the EC rehabilitation projects target the construction of interrelated and in-
terdependent infrastructure between South Ossetia and inland Georgia, which
may unite the local and central systems. In return, structural integration might
48 Hilfswerk Austria is a nongovernmental organization that won the EC tender for this
project.
49 The second rehabilitation program sponsored by the EC.
guarantee that the present arrangements are not only maintained but also ex-
panded beyond the current project scope, focused on transportation, energy, and
communication sectors. The projects on reconstructing the Tskhinvali under-
ground gas pipeline, restoring electricity supply and rebuilding the Gori-Tskhin-
vali railway have developed precisely in this manner. The EC strategy here aims
at the transformation of the Georgian-South Ossetian state-economic relations
into well-integrated and coherent networks. In addition, the EC devised project-
specific financial responsibilities that strongly promote the Georgian Lari – the
official currency, which is currently squeezed out by the Russian Ruble in the
Tskhinvali region, as it is in Abkhazia – to re-establish it in the breakaway region.
Such a policy plays an instrumental role in shaping preconditions for the possible
restoration of a common monetary space.
3 Georgian-Russian Intergovernmental Forum

The Georgian-Russian Economic Rehabilitation Programme (2000) was de-
signed to serve as an effective mechanism for addressing the vital economic and
infrastructure rehabilitation issues in order to create the requisite conditions for
refugee repatriation. Much like the preceding project of a Russian-Georgian in-
tergovernmental forum (1993), this programme has failed to honour its financial
obligations due to a lack of resources on both sides. However, it has successfully
expanded the previous scope of agreements beyond mere pledges for bilateral
financial assistance and towards more complex cooperation in designing joint
structures and implementing common economic policies. The JCC was autho-
rized to set up joint decision-making structures within this programme frame-
work – the Georgian-Russian intergovernmental body, the Georgian-Russian
Programme Executive Committee and subordinated working groups – to pro-
vide the necessary institutional mechanisms for the required legislative work and
project implementation (JCC 2001a: Appendix 2).50 The programme bodies were
commissioned to report to the JCC and, under their mandate, draft normative
acts and develop procedures for the programme-related arrangements, such as
infrastructure reconstruction and transport.
Consensual decision-making over the use, distribution, and allocation of the
programme funds and their coherent management was delegated to the Joint
Georgian-Ossetian body, which would coordinate with the Georgian-Russian in-
tergovernmental body and the JCC. In 2001, the Georgian-Ossetian forum was
the only functioning part of this complex structure. It was actively working with
the JCC working groups to define the programme priorities for 2002 and decide
on subsequent project financing (JCC 2001b).51 The implementation of the other
50 “Decision of the Joint Control Commission on the Georgia-South Ossetia Conflict

Regulation,” in Protocol of the 11th Plenary Meeting of the Joint Control Commission,
Vladikavkaz (North Ossetia-Alanya within Russian Federation), Appendix No.2, (21
April, 2001).
51 Protocol on the Elaboration of a Russian-Georgian Intergovernmental Programme
on Cooperation for Economic Rehabilitation of the Georgia-South Ossetia Conflict
formal arrangements of the forum have failed, due largely to turbulent Georgian-
Russian relations and the avoidance of financial commitments by both sides.
4 Confidence-building Measures
An informal forum for dialogue has been set up as a supplementary track to for-
mal political negotiations. One of the most successful initiatives in this direction
was promoted and implemented by the US Conflict Management Group (CMG)
in cooperation with the Norwegian Refugee Council (NRC) in Georgia. The
Georgian–South Ossetian Dialogue project (1995–1999) involved different lev-
els of authority together with civic actors. It envisaged the organization of joint
brainstorming sessions in order to delineate alternative proposals and recom-
mendations for a comprehensive solution. There were several bilateral meetings
convened within the CMG-NRC project framework, which provided a forum
for constructive thinking and dialogue. Although the parties found it difficult to
translate written agreements on several non-political issues into practice, these
activities were thought to have contributed considerably to reconciliation and
confidence-building.
Along with political and practice-oriented negotiations, multilateral endea-
vours fostered effective public diplomacy channels via mass media, NGO, and
social-academic networking. A number of significant attainments in these areas
have linked the two sides across ethnic community lines. The first ‘ice-breaking’
meetings between youth organizations were initiated by VERTIC, a British NGO.
Thereafter, a series of journalists’ meetings and several other civic dialogue proj-
ects were sponsored by the OSCE. These efforts spawned some tangible results
in terms of generating effective joint enterprises, such as the Public-Information
Center for the Georgia-South Ossetia Conflict Regulation, which is comprised
of nongovernmental news agencies of the two sides. The range of joint activities
also included ‘radio bridges’, collective production of documentary films,52 and
the creation of the NGO Network Forum.
Furthermore, the United Nations Volunteers (UNV), in cooperation with the
UNICEF and the UNHCR, promote interactive projects at the school level. They
produce a children’s interactive magazine and distribute it in the Georgian- and
Ossetian-populated regions of the conflict zone. The UNV also brings children
of different ethnic communities together at summer schools. In the absence of
joint education policies, these efforts are crucial in terms of building trust and
confidence in the younger generations.
Zone, Meeting of Georgian and South Ossetian Representatives of the JCC Working
Group on Economic Issues, Mtskheta, Georgia (19 July 2001).
52 Georgian television programme Studio Re, along with its South Ossetian partners,
filmed a documentary on the UNDP project implementation in South Ossetia in
1999. The organizers say the atmosphere was very friendly and collaborative (inter-
view with Mikheil Mirziashvili, Studio Re, June 2001).
V Conclusion
The horizontal power-sharing practices enacted in South Ossetia are undoubt-
edly a positive development in many ways. Primarily, they signal the cessation
of hostilities and the painful loss of human life and destruction of property. Per-
haps, they may also encourage the re-building of peaceful co-existence among
Georgians, Ossetians, and the other ethnicities in the area. The synergy of prac-
tical achievements creates favourable conditions for the advancement of a full-
scale settlement. To exploit this momentum, two questions must be recognized
and addressed: 1) Why is it imperative to formalise this grassroots cooperation?
2) How can the factors involved, whether internal or external, encourage and
regulate the necessary transformation? The answer to the first question should
help identify the crucial areas where informal power-sharing still fails to deliver
a transparent and viable long-term solution, while the answers to the second
question should delineate the path towards institutionalized settlement where
such a solution could be possible. Together these questions should also inform
a wider debate on the implications of horizontal power-sharing, not only for the
prospects of long-term peace in the region, but also for the concept of complex
power-sharing itself.
There are several compelling reasons for questioning the current power-shar-
ing practice in South Ossetia. It suffers from systemic failures, which undermine
its legitimacy and viability as well as limit its potential to resolve future prob-
lems. Namely, in the short run, South Ossetian power-sharing does not deliver a
transparent and democratic system based on lawful institutions and rules. At its
best, the current mechanism allows some functioning of ‘normal’ life and fosters
confidence-building. At its worst, it endorses a non-transparent and repressive
regime. In the long run, the existing grassroots cooperation offers no solution of
the conflict in the region because it avoids the underlying issue that has prompted
the conflict in the first place. No matter how encouraging the goodwill among the
ruling elites or how strong the intercommunal relationships, the political status
of South Ossetia is still largely ignored. In response to our first question, the fol-
lowing analysis argues that these structural distortions are fundamental and that
only an institutionalized power-sharing settlement can resolve them.
The formal framework surrounding and governing the present power-sharing
practices is either ambiguous or deficient. There are several formal agreements
to regulate military and economic issues but. because of their narrow remit, they
cannot bring necessary legitimacy to the overall arrangement. The current sys-
tem remains unconstitutional and without legal basis because it has not been
endorsed by any democratic and transparent plebiscite.53 The majority of private
or public persons join in this arrangement passively, merely acquiescing to the
objective reality. It is unclear if their participation in this ambiguous system could
53 Georgian authorities and the OSCE observers have declared the 2001 presidential
elections in South Ossetia illegal, as with all previous ‘elections’ conducted by the
Tskhinvali de facto regime.
be considered a public contract. As long as this remains the case, the existing
horizontal power-sharing in South Ossetia should at least be regarded with some
caution.
The present horizontal power-sharing in South Ossetia not only avoids the
underlying political conflict but it also fails to create a methodical approach for
resolving both political and non-political issues in the region. Several important
structural aspects remain ignored or inadequately addressed by the current ar-
rangement. These are either highly politicized (such as foreign and trans-border
relations, minority rights), ethno-culturally sensitive (such as language and edu-
cation policies), or inadequately financed (such as healthcare, social services).
These further undermine South Ossetia’s ability to present an integrated complex
power-sharing model even at the horizontal level. Rather, they unveil the exist-
ing practice as a set of fragmented interactions that may create an illusion of a
systemic approach largely due to the multiplicity of actors involved and to the
structures apparently created.
Unfortunately, the current variety of power-sharing in South Ossetia has built-
in limitations that may prevent it from evolving into a peaceful long-term solu-
tion. These limitations are either inherent in the mechanism itself or stem from
external influences. Internally, the hope for institutionalized power-sharing in
South Ossetia is severely compromised by the authoritative interest groups that
benefit from the status quo and thus are determined to maintain it. This is par-
ticularly true of the region’s flourishing shadow economy, which relies precisely
on the lack of transparency and legitimacy. The absence of democratic processes
also self-perpetuates the existing regime. It prevents the larger community from
exercising its rightful participation in public discourse. Without access to power,
this larger community not only remains without representation but also incapa-
ble of initiating much-needed transformation. Lastly, in the minds of many South
Ossetians, who continue to vote for secession, an institutionalized power-sharing
is barely considered to be a solution to their independence claim. These Ossetians
view horizontal cooperation as a means to reap practical (short-term) benefits,
with little bearing on their long-term political aspirations. As a result, those who
may endorse complex power-sharing have a modest profile, while those with a
higher profile still choose a quasi-legal state today with the hope of secession
tomorrow. Despite the various palpable achievements of grassroots coopera-
tion, the gap between horizontal power-sharing and political settlement further
undermines the current practice and its potential to lead to institutionalization
from within.
As for the external influences compromising power-sharing in South Osse-
tia, the overall conflict settlement seems to be locked in a vicious circle of other
post-Soviet ethnic conflicts, most specifically that in Abkhazia. Tskhinvali looks
to So(u)khumi for a precedent and does not want to accept anything less than
what the latter gets from Tbilisi. When analyzing the South Ossetia case, many
consider that, realistically, the key to both conflicts in Georgia has to be sought
in Abkhazia.54 Regrettably, the political dialogue in Abkhazia is progressing at an

alarmingly slow pace, thereby causing further delays in the conflict negotiations
in South Ossetia.
Given its inability to generate change from within, the existing power-sharing
structure can only rely on its flexibility in accepting outside pressure if it hopes
to evolve into a formal settlement. As mentioned above, the current system has
not established long-term peace but it can certainly be argued to have laid down
some preconditions to permit it. Confidence-building and flexibility are the two
achievements that could be exploited using a more rigourous strategy for politi-
cal reconciliation. Indeed, rapprochement at both elite and intercommunal levels
may be a platform for the combined use of consociational and integrational ap-
proaches as the basis for comprehensive power-sharing arrangements in future.
The challenge lies in identifying where and how this outside pressure can be ap-
plied effectively, and without deleterious effects.
This crucial external involvement could come in two forms – directly, as con-
sistent and explicit policy of international actors, and indirectly, as a consequence
of a conflict settlement precedent elsewhere in the troubled post-Soviet area. The
latter is currently less likely. The post-Soviet conflicts that South Ossetia is associ-
ated with unfortunately remain deadlocked and thus hardly suggest a settlement
prospect in the foreseeable future. Furthermore, because of the unique features,
past and present, that have shaped the South Ossetian power-sharing experience,
it is difficult to imagine a conflict-settlement mechanism that could be easily and
readily imported into such peculiar circumstances. Consequently, a customized
policy and an integrated response of multinational organizations could be deci-
sive. The need and importance of bold international action is recognized on the
ground as well. In the previously described social inquiry conducted by the Tskh-
invali Public Information Center in South Ossetia in 1999, international organi-
zations were ranked as the most important factor for the present progress in the
Georgian–South Ossetian conflict management and regulation.55 Capitalizing on
their political repute and financial/military might, these organizations could: 1)
either impose, implement, and guarantee a specific and comprehensive arrange-
ment (top-down solution); or 2) embark on invigorated facilitation and media-
tion of preparatory arrangements. This would mean starting out and backing up
more dynamic and diversified channels of bilateral or multilateral interactions
,which ultimately would combine into a self-propelling complex power-sharing
mechanism (bottom-up solution).
54 Interviews demonstrated that many on the Georgian side, as well as some Ossetians,
believe that as long as the Abkhazia question remains unresolved political break-
through in South Ossetia is unfeasible (interviews with Alexander Rondeli, Konstan-
tin Dzugaev, and Ghia Nodia, summer 2001.)
55 The role of the international community and international organizations were ranked
first (rating 167) followed by the role of law enforcement (139) and the JPKF (107); “in-
creasing cultural and economic contacts” (65) and “success of public diplomacy” (43)
were less prominent ( Kontakt 1999: 13).
The disposition of various factors in South Ossetia suggests better prospects

for bottom up settlement than for the top-down solution. Firstly, none of the
international actors involved in the South Ossetia peace process at this stage are
willing or have enough powers to forcibly impose and administer an ultimate
top-down solution that would address highly sensitive political-territorial issues.
Secondly, enforcement of such an arrangement would certainly compromise the
interests of one or another party to the conflict and thus undermine all posi-
tive accomplishments of the long efforts in confidence-building, rapprochement,
and cooperation. As a forcible top-down solution appears both unfeasible and
unsuitable for South Ossetia, the bottom-up strategy remains the only workable
option. A promising start in this regard already exists. This would require the in-
ternational organizations to mobilize and invigorate their efforts to push forward
more genuine and sustainable cooperation at different levels and in different di-
rections. In this case, international programmes could serve as cooperation and
capacity-building instruments rather than as temporary remedies to (re)vitalize
project-related power- and resources-sharing practices. Such an approach might
foster the development of diversified power-sharing networks at grassroots,
semi-formal and formal levels, which may ultimately turn into a self-propelling
mechanism, per se requiring some form of institutionalized comprehensive regu-
lation. In this manner the result would be internally generated regulation with
strong international facilitation and backup. Fortunately, the activities of inter-
national organizations in South Ossetia’s peace process, especially those of the
OSCE and the EU as well as specialized UN agencies and other humanitarian
organizations, are gradually expanding. Even if the current dynamism is not suf-
ficient to vigourously advance a bottom-up solution, it certainly gives hope for
leading in this direction.
As for the dependence on Abkhazia, only the aforementioned rigourous inter-
national mediation and far-sighted leadership of both Georgian and Abkhazian
elites could reverse the direction of causality. Unfortunately, the present disposi-
tion shows no prospects for such an outcome. Hypothetically, however, South
Ossetia has the potential to become a precedent-setting case due to the matu-
rity of certain practical arrangements favouring further advancement towards a
genuine solution.
Finally, the analysis of post-conflict development in South Ossetia suggests
powerful conceptual implications. Because of its structural failures and limita-
tions for long-term peacemaking, horizontal power-sharing cannot be consid-
ered a substitute for institutionalized complex power-sharing practice. However,
it can be viewed as a compliment to the overall concept, one that even challenges
the traditional sequence of conflict resolution. The case of South Ossetia demon-
strates that a formal agreement may not be necessary to initiate dispute settle-
ment and that grassroots power-sharing could move the settlement process for-
ward, even without a political resolution. The horizontal power-sharing in South
Ossetia is a functionally valuable achievement that can serve as an intermediary
platform from where both Georgian and Ossetian sides can develop a long-term
perspective on how to arrange their lives together.
Interviews (conducted in July–August 2001)

1 Georgian officials and scholars
Cheishvili, Vasil. Head of the Department of Program Coordination, Ministry of Health
and Social Services of Georgia.
Darchiashvili, David. Head of the Georgian Plenipotentiary Delegation on Georgia-South
Ossetia Conflict Regulation.
Geradze, Levan. Senior Advisor, Legal Department, Ministry of Special Affairs of Geor-
gia.
Gogolashvili, Kakha. Head of the Department of the European Union, Ministry of Foreign
Affairs of Georgia.
Gvenetadze, Ghia. Head of the Department for Russian Federation, Ministry of Foreign
Affairs of Georgia.
Ivanidze, Nugzar. Director, International Organization for Fair Elections and Democracy.
Kakabadze, Malkhaz. Minister of Special Affairs of Georgia.
Khutsishvili, George. Chairman of Board and Director, International Center on Conflict
and Negotiation.
Kopaleishvili, Mamuka. Secretary of the Central Election Commission of Georgia.
Machavariani, Irakli. Special Envoy of the President of Georgia on Georgia-South Ossetia
Conflict Regulation, Head of the Georgian Plenipotentiary Delegation.
Mirziashvili, Mikheil. Director, Studio “RE,” Tbilisi, Georgia.
Nodia, Ghia. Chairman of Board, Caucasian Institute for Peace, Democracy and Develop-
ment (CIPDD).
Pataraia, Tamar. Researcher, CIPDD.
Rondeli, Alexander. President of the Georgian Foundation for Strategic and International
Studies.
Siamashvili, Mikheil. Deputy Head of the Department of the United Nations, Ministry of
Foreign Affairs of Georgia.
Skhirtladze, Irine. Member of the Georgian Plenipotentiary Delegation on Georgia-South
Ossetia Conflict Regulation.
Tsintsadze, Irina. Director, “Alternativa”/Public-Information Center on Regulation of
Georgia-South Ossetia Conflict (Tbilisi, Tskhinvali, and Gori offices).
2 Russian officials
General Churaev. Commander in Chief, Joint Peacekeeping Force.
3 Ossetian officials and experts

Dzugaev, Konstantin. De facto Speaker, the ‘Parliament’ of South Ossetia.
Parastaev, Alan. Director, Caucasus Forum, Network of NGOs for Peace and Sustainable
Development.
Pliev, Alan. De facto Deputy Minister of ‘Foreign Relations of the Republic of South Os-
setia’.
Pliev, Igor. Member of the Center on Humanitarian Initiatives and Research of South Os-
setia.
4 International organization officials

Anquetil, Emmanuel. Project Manager, European Commission Delegation in Georgia.
Backlund, Ulla. Resident Representative in Georgia, Norwegian Refugee Council (NRC).
Boden, Dieter. Special Representative of the UN Secretary-General for Georgia, Head of
UN Observer Mission in Georgia (UNOMIG).
Borsotti, Marco. UN Resident Coordinator in Georgia, UNDP Resident Representative in
Georgia.
Johnson, Cate. Caucasus Regional Director – Office of Democracy and Governance, US
Agency for International Development (USAID).
Kharashvili, Julia. Project Coordinator, United Nations Volunteers (UNV) in Georgia.
Martirosyan, Arthur. US Caucasus Project Coordinator, Conflict Management Group
(CMG), Cambridge MA. (Phone-interview, April, 2001).
Olson, Lara. Project Coordinator (in 1996-1998) Georgian-South Ossetian Dialogue
Project, NRC. (Phone-interview, April 2001).
Parekh, Rajen S. Field Adviser in Georgia, UN Office for the Coordination of Humanitar-
ian Affairs (OCHA).
Tsoneva, Valentina. Protection Officer in Georgia, UN High Commissioner for Refugees
(UNHCR).
Van Hoye, Ermina. Special Assistant to the Special Representative of the Secretary-Gen-
eral for Georgia UNOMIG.
Confidential interviews with members of the Organization for Security and Co-operation
in Europe Mission to Georgia and members of the OSCE Tskhinvali Field Office.
Section Two
Cross-cutting Analysis
Part Three
Vertical Power-sharing
Chapter 12
Addressing the Self-determination Dispute
Marc Weller
Self-determination disputes have traditionally been resolved through the force of

arms. Either the government would decisively defeat an attempted secession (Ka-
tanga), or low-intensity conflict might be carried out over decades. The question
arises whether recent practice evidences a trend to address and actually resolve
the self-determination issue directly in internationalized power-sharing settle-
ments.
An initial review of the cases surveyed would suggest that there are four ap-
proaches to the self-determination issue in complex power-sharing agreements.
These are:
• Denying self-determination to the secessionist entity and compensating it
with complex power-sharing provisions;
• Ratifying secession and compensating non-dominant groups in the new
state through complex power-sharing;
• Confirming the possibility of self-determination including secession, but de-
ferring the act of self-determination to a future date and possibly making the
exercize of this act dependent on the fulfilment of certain conditions. Com-
plex power-sharing will provide for interim administration in such cases and
may also protect non-dominant groups in case of an ultimate decision in
favour of secession; and
• Suspending the issue of self-determination for the moment and instead
promising the creation of a mechanism through which this issue may be
addressed in future. Complex power-sharing will be used for interim admin-
istration in such cases, although there may be an expectation that this will
consolidate into a permanent internal settlement over time.
In considering each of these types of cases, it will be convenient to consider

whether a settlement resolves the issue of who is the self-determination unit (i.e.,
what is the ‘people’ entitled to self-determination), what, if any, mechanism is
established for the exercise of the right to self-determination and whether this
exercise is conditional. Moreover, the basic structure of the complex power-shar-
388 Marc Weller
ing design will be highlighted. This may help answer the question whether certain
types of self-determination settlements tend to attract certain models of complex
power-sharing designs.
While mainly drawing upon the cases selected for this study, a number of oth-
ers have also been included. These fit the criteria for this study and were ad-
dressed after the project was too advanced to include them formally through case
studies.
I Secession Denied
A number of settlements will seek to trade the claim to secession for a com-
plex power-sharing arrangement. Often this will take the form of a re-negotia-
tion of the state constitution. In the cases without significant international in-
volvement in achieving or implementing a settlement – excluded from this study
– decentralization will typically be provided (North Mali), or more far reaching
autonomy arrangements will be established. Another example is furnished by the
change of the status of Corsica that was proposed by the central authorities of
France. More recently, a framework agreement was obtained in relation to the
Aceh conflict in Indonesia, providing for self-government of the territory based
on the democratic will of its population, but clearly within the framework of
Indonesian legislation and uncontested sovereignty. Among presently ongoing
negotiations, it is likely that a settlement for Sri Lanka may be forthcoming that
will grant very extensive separate administration to the areas of Northeastern Sri
Lanka, at present controlled by Tamil Tiger forces, in exchange for a pledge of not
pursuing secession.
At times, a federal structure will be proposed in this type of settlement. How-
ever, the recent consolidation of the doctrine of constitutional self-determination,
and especially its rather open-ended formulation by the Badinter Commission,
have made such a solution less attractive in a number of circumstances. Central
authorities tend to fear that a federal solution might provide a stepping-stone
towards an entitlement to external self-determination. After all, on the basis of
its new federal status, the entity might claim constitutional self-determination.
Accordingly, the governments of Moldova and Georgia have recently clarified
their position, offering asymmetrical autonomy instead of federalization of the
overall state.
The proposed settlement for Cyprus that failed to be approved by a majority in
the mainly Greek populated part of the island in 2004 exemplifies an attempt to
overcome the fear of generating an entitlement to secession in a document that
is intended to consolidate or re-establish the territorial integrity of a state. The
Foundation Agreement, tying together a number of elements of the draft settle-
These are Aceh, Chechnya, and Sudan. The situation in Sri Lanka has also been mon-
itored, although no settlement was obtained at the time of writing.
Cessation of Hostilities Framework Agreement between the Government of the Re-
public of Indonesia and the Free Aceh Movement, 9 December 2002, Preamble.
12 Addressing the Self-determination Dispute 389
ment, described Cyprus as one independent state, enjoying one single interna-
tional legal personality and sovereignty. However, the state would be composed
of two equal entities, described as constituent states. While the common na-
tional government ‘sovereignly’ exercises the powers specified in a constitution,
the constituent states also ‘sovereignly’ exercise their powers. These powers are
not vested by the Constitution in the common, national government. They are
derived from the “inherent constitutive power, by our free and democratic, sepa-
rately expressed common will”, to be made manifest in two separate referenda,
both of which must be positive in order to bring the agreement into effect.
The location of all authority not specifically granted to the common govern-
ment in the two equal sovereign constituent states would indicate that they are
the sources of original sovereignty, despite the language that also describes the
central common powers as sovereign. This would, in principle, suggest a union or
confederation of sovereign entities, or perhaps a very loose confederation. Under
the Badinter principles, a constitutional entitlement to self-determination might
be implied. Hence, the Foundation Agreement states that “any unilateral change
to the state of affairs established by the Agreement, in particular union of Cyprus
in whole or in part with any other country or any form of partition or secession,
shall be prohibited.”A similar provision is contained in the draft constitution,
which was part of the package of the draft agreement, adding that the independ-
ence, territorial integrity, security, and constitutional order of Cyprus shall be
safeguarded and respected by all.
Russia’s recent action in relation to Chechnya reflects a similar attempt to grant
federal-type powers without implying a future entitlement to constitutional self-
determination. A referendum was held by the Russian authorities on 23 Febru-
ary 2003. Given that Russia is structured as a Federation, and that Chechnya is
designated as a republic within the Federation in the 1993 Russian Federal Con-
stitution, the Federal status of that entity could not be in doubt in the settlement
that is at present being imposed by the Moscow authorities. Hence, the refer-
endum invited the Chechen population to endorse a new Chechen constitution
that would provide for self-government, but expressly within the ‘unquestion-
able’ sovereignty of the Russian Federation.
Out of the instances covered by case studies, four have attempted to go the
route of trading claims to independence for enhanced self-government. These
are Georgia; Moldova; Macedonia; and, in its internal aspects, Bosnia and Herze-
govina.
In relation to Georgia, only rudimentary framework agreements of 1994 con-
cerning Abkhazia and South Ossetia, respectively, exist. In the Declaration on
Measures for a Political Settlement of the Georgian/Abkhaz conflict signed on
4 April 1994 by the parties in the presence of the representatives of the UN, the
Article 1(6).
Article 1.
The external aspect of that latter case, i.e., the question of the independence of Bos-
nia and Herzegovina, will be considered in the section that follows.
390 Marc Weller
Russian Federation and the then CSCE, it is agreed that Abkhazia shall have its
own constitution and legislation and appropriate State symbols, such as an an-
them, emblem, and flag. In that version of the draft, there was specific confirma-
tion of the territorial unity of Georgia to balance this description of Abkhazia as
a ‘state’. However, the continued territorial integrity of Georgia was protected by a
series of UN Security Council resolutions on the conflict to which the agreement
expressly refers. The agreement also provided for a provision on the ‘distribution
of powers’, providing for ‘joint action’ in the fields of foreign policy, foreign eco-
nomic ties, border guard arrangements, energy, transport and communications,
and the preservation of human and minority rights.
Hence, it appears that a rather nominal incorporation of the territory into the
legal order of Georgia was foreseen, with exclusive and original jurisdiction of
Abkhazia over its own affairs and joint powers in relation to certain issues of
common concern. Another explanation would be that it was not, in fact, possible
to clarify the issue of status sufficiently to provide a basis for the development of
a substantive power-sharing arrangement. This might explain the lack of progress
on this issue for close to a decade.
The Memorandum on Measures to Provide Security and Strengthen Mutual
Trust between Sides in the Georgian Ossetian Conflict of 17 April 1996, con-
cluded by Georgia, ‘the South Ossetian side’ with the involvement of Russia, the
Republic of North Ossetia, and the OSCE, is also quite general in its terms. How-
ever, it does at least refer to the concepts of territorial integrity and self-determi-
nation directly. In its preamble, it confirms the commitment of the parties to the
provisions of the UN Charters, the basic principles and decisions of the OSCE,
and generally recognized norms of international law. This would ordinarily be
understood to support a restrictive approach, supporting continued territorial
unity. However, in the following preambular paragraph, the parties profess to be
“guided by the principles of territorial integrity of states, and the right of nations
for self-determination”. Hence, once again, the underlying self-determination is-
sue is not actually resolved in the agreement itself, advancing with equal vigour
the two competing concepts that would result in opposing solutions on the issue
of status. Given that no further agreements obtained to implement the initial
framework accord, the structure of a power-sharing settlement cannot be consid-
ered as a means of shedding light on the status issue. Nevertheless, as is the case
in Abkhazia, the principle of territorial integrity has also been privileged (in this
instance by the UN Security Council) over the claim to self-determination and
independence. Hence, this case has been provisionally listed under the heading
of ‘self-determination denied’. However, it must be acknowledged that, in actual
fact, both entities are administering themselves separately. The lack of substan-
tive progress on a political settlement and formal power-sharing has necessitated
the alternative development of informal power-sharing, revealed through the
case study above. Since then, the Georgian government has advanced a number
Article 4.
See Chapter 11.
of further initiatives, effectively disowning the 1996 text. Instead, Georgia has
advanced several new designs more clearly relying on authonomy since then.
Moldova, on the other hand, did achieve a formal settlement. However, this
settlement is contained in a piece of domestic legislation, the Law on the Special
Legal Status of Gagauzia, adopted on 23 December 1994, generated with the in-
volvement of the OSCE. The law defines Gagauzia as:
[A]n autonomous territorial unit with a special status as a form of self-determination of

the Gagauzes, which constitutes an integral part of the Republic of Moldova.
The law clarifies that the legal powers enjoyed by Gagauzia are devolved, rather
than original, confirming that Gagauzia shall be governed on the basis of the na-
tional constitution and national laws. The public powers, including the legislative
powers of Gagauzia, are broad, but probably exclusively defined. They are subject
to review by the Moldova Constitutional Court, which acts as dispute settlement
mechanism with the involvement of the state parliament and president. Gagauz
legislation is valid only to the extent that it does not conflict with the legislation of
Moldova. The entity has its own symbols, to be used alongside the state symbols
of Moldova.
Overall, this would appear at first sight to be a typical settlement where the
acknowledgement of the right to self-determination has been traded for devolved
autonomy. Or, to put it in other words, self-determination has been acknowl-
edged on the strict condition that it cannot be actualized through secession. The
right has been spent in the acceptance of integration with Moldova on the basis
of an autonomy settlement. Nevertheless, the express invocation of the right to
self-determination does make this case somewhat special. Moreover, this right
is not merely a notional one, as it can, according to the law, be actualized in one
circumstance:
In the case of a change of the status of the Republic of Moldova as an independent state,
the people of Gagauzia shall have the right of external self-determination.
Hence, one might wonder whether this settlement is one of association rather
than integration. Gagauzia does obtain a self-determination status that is action-
able, although under a very narrowly defined condition. Otherwise, the settle-
ment is heavily dominated by territorial autonomy elements, although this has
not been fully implemented.
While not covered in a full case study, this project also peripherally considered
the case of Transdniestria. A memorandum on the Bases for Normalization of
Relations between the Republic of Moldova and Transdniestria was achieved on
8 March 1997, signed by the Republic of Moldova and Transdniestria, with the
Russian Federation acting as guarantor state, and in the presence of the OSCE
Article 1(1).
Article 1(4).
392 Marc Weller
chairman in office. As was the case in the Abkhazia document, the preamble
refers to the principles of the UN, the OSCE, and generally recognized interna-
tional norms of international law. However, there is no reference to self-determi-
nation to balance these provisions which tend to favour the doctrine of territorial
unity. In Article 2, the agreement foresees the “establishment between them of
state-legal relations”. A formal document on status including the division “and
delegation” of competences is to be negotiated immediately. This might suggest
a far-reaching separate identity for Transdniestria, but within the overall legal
framework of Moldova. Article 3 establishes the right of the entity to establish
unilateral international contact in the areas of science and culture. This too could
be consistent with integration on a quasi-federal basis. Moreover, that article also
indicates that Transdniestria shall “participate in the conduct of foreign policy of
the Republic of Moldova – a subject of international law – on questions touching
its interest.” This implies that Transdniestria is not a subject of international law.
However, its strong internal legal identity is reflected in a power of co-decision on
foreign affairs issues of special concern to it.
A unique feature is provided in a joint statement of the presidents of the Rus-
sian Federation and Ukraine in connection with the signing of the memorandum.
In this joint statement, Russia, the guarantor of the principal agreement, and the
Ukraine, a mediator, declare that it will not be interpreted or acted upon in con-
tradiction with existing agreements and OSCE decisions “which recognize the
sovereignty and territorial integrity of the Republic of Moldova.”
Since then, however, the Moldovan government has moved away from con-
templating a quasi-federal or even confederal solution for Transdniestria. In-
stead, since 2004, it has emphasized the concept of autonomy.
The case of the Philippines provides a more complex settlement. The Tripoli
agreement of 23 December 1976 already provided for the “establishment of au-
tonomy in the Southern Philippines”, i.e., the areas mainly inhabited by the Mus-
lim population. This was to take place expressly within the realm of the sover-
eignty and territorial integrity of the Republic of the Philippines.10 It took twenty
years to reach the Manila agreement on the implementation of the Tripoli ac-
cord, adopted on 2 September 1996. That highly detailed instrument provided for
a three-year period during which a southern ‘Zone of Peace and Development’
and other interim arrangements were foreseen. There would then follow a sec-
ond phase involving the repeal of the existing constitutional provisions (Organic
Act) and the decision by the “people of the concerned areas in a plebiscite to
determine the establishment of a new autonomous government and the specific
area of autonomy thereof.”11 Interim government was to be arranged through a
devolved Southern Philippines Council for Peace and Development and local
governance, along with the integration of the armed forces. In Phase II, there
was provision for self-government through an Executive Council, a Legislative
Assembly, and Administrative System, and provision for representation in the
10 Tripoli Agreement, first principle.

11 Section I.
national government. The construction of state authority locates all public au-
thority in the Southern Region, other than that specifically retained by the cen-
tre (foreign affairs, defence and security, fiscal and monetary policy, transport
and communications). There is also provision for a special regional security force
controlled from within the region. Detailed provisions for education, economic
self-management and fiscal relations (an Islamic ‘banking unit’ to be established,
but within the framework of the National Bank), and provisions for revenue shar-
ing were also provided.
Implementation of the agreement was delayed and, as the case study chapter
reveals, in many respects flawed. Hence, the case study chapter expects a new
attempt to establish or implement a settlement.12 However, its basic shape as a
settlement that denies external self-determination is unlikely to be changed.
The most advanced settlement based on a denial of the option of secession is
represented by the case of Macedonia. The armed campaign of the ethnic Alba-
nian NLA fighters was not directed towards independence. Instead, it sought a
more pronounced recognition of the identity of ethnic Albanians in Macedo-
nia, including territorial autonomy. However, the developments in Macedonia
in the first half of 2001 were widely seen to be potentially part of a pan-Albanian
agenda, also involving Kosovo. Certainly from the perspective of the central gov-
ernment of Macedonia, the NLA appeared to be mainly supported from Kosovo,
and the possibility of detaching ethnic Albanian populated areas and joining it to
Kosovo was strongly feared.
The Ohrid Agreement of 13 August 2001 was signed by the Macedonian politi-
cal parties alone, despite the very strong international involvement in its crea-
tion. The agreement confirms the territorial integrity of Macedonia, stating that
“Macedonia’s sovereignty and territorial integrity, and the unitary character of
the state are inviolable and must be preserved. There are no territorial solutions
to ethnic issues.” Consistent with this aim of preserving the unitary character
of the state, the Framework Agreement does not formally establish territorial
autonomy for land bordering Kosovo and mainly inhabited by ethnic Albanians.
Instead, however, it establishes a disguised form of territorial autonomy through
the means of “enhanced local self-government” of mainly ethnic Albanian inhab-
ited municipalities, including the possibility of the coordination of the adminis-
tration of these areas. A sophisticated mechanism of co-decision in the central
parliamentary organs is provided for, avoiding the automatic veto mechanism
that has dogged the Bosnian settlement. Through the Committee on Inter-com-
munity Relations, a potentially effective mechanism of dispute settlement provi-
sions has been introduced, backed up by the role of the Constitutional Court.
There is also provision for equitable representation of all communities in the
state organs, including the judiciary and police. Legislation has been introduced
to enhance educational and language rights of the communities. While there was
strong international involvement in the achievement of the settlement, the inter-
national role in implementation is more limited. Formally, there is international
12 See Chapter 6.
394 Marc Weller
involvement in the crucial issue of holding a census and in police training. Infor-
mally, however, the EU mission in Macedonia has played a key role in mediating
the adoption of the constitutional changes required by the Ohrid agreement and
in the adoption of the special legislative measures that were envisaged.
The Dayton Agreement on Bosnia and Herzegovina of 14 December 1995, in
its internal aspects, provides the most complex example in this category. The set-
tlement of course confirms the statehood and continued legal personality of the
state of Bosnia and Herzegovina. However, the settlement is more noteworthy for
its treatment of the Srpska Republic and its ethnic Croat counterpart. Both were
left in existence, the former as one of the two constituent entities out of which
Bosnia and Herzegovina is composed, the other as a constituent unit of the other
entity, the ‘Federation’. There is a fairly pronounced entitlement of the entities
“to establish special parallel relationships with neighbouring states” but this is
subject to the “sovereignty and territorial integrity of Bosnia and Herzegovina.”
Hence, the claim to self-determination in the sense of secession of both entities
is precluded, a fact also confirmed in the attendant Chapter VII Security Council
resolutions.13 Instead, a highly fractionated and complex power sharing arrange-
ment is provided, locating most powers at the level of the Srpska entity and the
Muslim/Croat Federation, respectively. Within the Federation, authority is again
split, locating it mainly with associations of the mainly ethnic Muslim and Croat
cantons.
In terms of state construction, the centre only enjoys the authority specifically
assigned to it; other authorities are exercised by the entities or, in the Federa-
tion, by cantons. Weak central authority is balanced with extensive human rights
provisions, which are, however, not always effectively enforceable at the central
level. Very extensive consociational mechanisms of co-decision, disproportion-
ate representation, and veto provisions are not balanced by an effective dispute
resolution mechanism. Instead, in practice, decision-making at the international
level has had to take on this role. There is also provision for excessive representa-
tion in the government and in executive agencies at the central level, and separate
agencies of governance at the entity, federal and cantonal level. There are even
separate armed forces of both entities.14
II Ratification of Secession
In stark contrast to the previous category of cases, a settlement can provide for
a direct right to self-determination of nominated units, expressly including the
right to self-determination. Essentially, this amounts to a ratification of a seces-
sion that has already occurred.
In a few cases, a constitutional right to self-determination will have already
existed, including the right to secession. For instance, the USSR constitution con-
tained an express right to self-determination and secession for Union Republics.
13 Article 3(2)(a).
14 For a detailed analysis, see Chapter 7.
When the Baltic republics sought to activate their independence in 1991,15 this
was nevertheless resisted by the central USSR authorities.16 While the authorities
did not deny the right to self-determination, they sought to subject it to central
approval through a protracted process of agreement by all other Union Republics
and the national Congress of Peoples’ Deputies.17 This process would have taken
some eight years and could have been brought to a halt by a negative decision
taken outside of the republic seeking secession at any stage. Effectively, it would
have nullified the constitutional right to secession.
Lithuania formally reclaimed its independence in March 1991 – a move re-
sisted by Moscow sought to maintain at least elements of effective control in the
republic.18 Only when the USSR itself collapsed after the abortive hardline coup
against President Gorbachev could independence be actualized, also in relation
to Estonia and Latvia, which adopted formal declarations concerning independ-
ence at that time. All three were recognized by the State Council of the Union
on 12 September 1991 (Crawford 1997). Soon afterwards, the USSR as a whole
dissolved, as was confirmed in the Alma Ata Declaration and related instruments
of December 1991.19
In another instance, several of the constituent republics of the Socialist Federal
Republic of Yugoslavia, led by Croatia and Slovenia, and followed by Macedonia
and Bosnia and Herzegovina, exercised what they claimed to be a constitution-
ally-based right to unilateral secession.20 While these claims were internationally
recognized and UN membership was granted in May 1992 (delayed for a year in
the case of Macedonia due to extraneous reasons), the then Federal Republic of
15 The Baltic Republics maintained that they had never been lawfully incorporated into
the USSR, having been annexed in 1940. A number of Western states also took the
view that the Baltic Republics had never de jure lost their independence. However,
they found it somewhat difficult to respond when confronted with Lithuania’s uni-
lateral claim to independence, which was only recognized by Iceland, until the USSR
collapsed altogether.
16 Lithuania led the way, formally reactivating its independence on 11 March 1991. Es-
tonia and Latvia, while declaring sovereignty, only formally and fully activated their
independence on 20 and 21 August 1991, respectively, when it became clear that the
central system of governance in the USSR was collapsing and a dissolution might be
likely.
17 See Chapter 2.
18 Black Beret forces were used, for instance, in an assault on the Lithuanian television
station to emphasize the continued claim to effective authority of the central govern-
ment.
19 31 ILM, 138 et seq.
20 The preamble of the 1974 SFRY constitution – “proceeding from the right of every
nation to self-determination, including the right to secession –” appeared to grant
such a right. Led by the Badinter opinions, the EC/EU considered this to be a case of
dissolution, rather than secession.
396 Marc Weller
Yugoslavia (Serbia and Montenegro) only ratified this result in the context of the
Dayton settlement of December 1995.21
Ratification of a constitutional claim to self-determination will not normally
attract a complex power-sharing arrangement. The unilaterally seceding entity
will give itself a constitution. This may well include human and minority rights,
minority protection provisions, and perhaps modest cultural or even territorial
autonomy elements. However, as the arrangements for the newly independent
state will not have been negotiated bilaterally or with international involvement,
these mechanisms tend not to be well developed. Hence, it was left to the EU to
press for subsequent provisions for Russian speakers in the Baltic republics and
for minority arrangements in Croatia and Macedonia. This was followed some
years later by Council of Europe supported drafting of minority legislation in the
Federal Republic of Yugoslavia and autonomy proposals concerning Vojvodina.
The exception is Bosnia and Herzegovina, which was subjected to conflict for two
and a half years after it had been internationally recognized and ultimately to an
internationalized settlement that was already noted above. The Dayton agree-
ment confirms that Bosnia and Herzegovina “shall continue its legal existence
under international law as a state, with its internal structure modified as provided
herein and with its present internationally recognized borders.”22
III Deferred Possible Secession

In a number of instances, the central government will come to accept that the
secessionist unit is in fact entitled to self-determination. However, rather than
providing for an immediate act of self-determination through a referendum, a
standstill period will be provided for in a settlement agreement. Broadly, one can
distinguish three different political constellations that may trigger the application
of this device. These are instances of inevitable independence, possible indepen-
dence, and unlikely independence or balancing of claims.
A Inevitable Independence
In the first type of case, both parties are fully aware that secession will occur.
However, the standstill period is meant to facilitate negotiations on succession
issues and the establishment of independent structures of government in the en-
tity in question. Hence, when a change in government occurred that brought into
power the previous opposition forces, it was agreed that Eritrea’s long-standing
campaign for secession should now be satisfied, subject to a referendum to be
held after a period of three years. As the eventual result of 99.8% in favour of
independence proved, there could never have been any doubt as to the outcome
of the referendum (Crawford 1997). The granting of self-determination status was
therefore identical with the granting of independence.
21 See Chapter 17.

22 Annex 4, Article 1(1).
A further example of deferred secession with a certain outcome was provid-

ed by Chechnya. In 1991, that entity had declared unilateral independence from
what was to become the Russian Federation. It had withstood a military assault
from the Federation, impelling the central government to agree to a settlement.
On 31 August 1996, General Lebed of the Russian Federation and Mr Maskhadov
representing Chechnya adopted a joint declaration in the presence of an OSCE
representative. The declaration proceeded from “the universally recognized right
of peoples to self-determination, the principles of equality, voluntary and free
expression of will” and determined that future relations of both entities should be
determined in accordance with universally recognized principles and norms of
international law by 31 December 2001. While currency, financial and budgetary
‘interrelations’ would be restored in the meantime, Chechnya would enjoy powers
of legislation limited only by the observance of human and civil rights, the right
of peoples to self-determination, the principles of equality among nationalities,
and interethnic accord. On 12 May 1997, a formal Treaty on Principles of Inter-
relation between the Russian Federation and the Chechen Republic Ichkeria was
concluded between Russian Federal President Boris Yeltsin and Mr Maskharov.
In the treaty, both sides agreed to develop their relations on the basis of generally
recognized principles and norms of international law, thus confirming the inter-
national legal personality of Chechnya. Essentially, therefore, the Russian Federa-
tion had settled for statehood, although deferred for a period of four to five years.
However, citing the involvement of Chechen groups in acts of terrorism, Moscow
unilaterally abrogated this commitment and set about to re-conquer the territory.
This action was not significantly resisted by international bodies or states. Instead
of defending the identity of Chechnya as a pre-state entity with international legal
personality, they limited themselves to the demand for a cessation of an armed
campaign that paid little regard to humanitarian law.
B Possible Independence
A second type of deferred secession concerns cases where independence after a
standstill period is perhaps likely, but not inevitable. Deferral is a conflict man-
agement tool, allowing both parties to claim to their constituents that they have
not caved in to the demands of the other side. Instead, they can maintain that the
transitional period will consolidate their position. After interim governance, the
central government may argue (or, in the case of Kosovo, the governments in-
volved in the administration of the crisis thought), the pressure for independence
will have subsided. The power-sharing element of the interim governance struc-
ture, it is hoped, will have demonstrated to the secessionist entity that its identity
can be fully reflected within the legal order of the state. Leaving that structure,
on the other hand, may bring economic and other risks with it. The secessionist
party, conversely, will take success in a referendum to be a foregone conclusion.
A very good example for this approach is represented by the Machakos Proto-
col of 20 July 2002. That framework document, which was supplemented in 2005
by a more detailed settlement, was concluded by the Government of Sudan and
398 Marc Weller
the Sudanese People’s Liberation Army/Movement. Its signature was witnessed

by an envoy of the long-standing IGAD international mediation process (led then
by Kenya’s president Daniel Arap Moi). The negotiations were also strongly sup-
ported by the US government, which, since October 2001, had applied certain
pressures upon the parties.
The preamble of the agreement at first appears to point to an attempt to inte-
grate the country after prolonged conflict between the mainly Muslim North and
the mainly Christian South. Division is to be overcome by correcting historical
injustices and inequalities in development between both regions and by estab-
lishing a framework for governance through which power and wealth are equita-
bly shared and human rights guaranteed. Article 1.1 of the agreement accordingly
proclaims that the unity of the Sudan “is and shall be the priority of the parties
and that it is possible to redress the grievances of the people of South Sudan and
to meet their aspirations within such a framework”. However, this strongly inte-
grative provision is immediately countered by Article 1.3, which states clearly and
unambiguously that the people of South Sudan have the right to self-determina-
tion, inter alia, through a referendum to determine their future status. Article 1.10
concludes that the design and implementation of the Peace Agreement is to be
performed so as “to make the unity of the Sudan an attractive option especially to
the people of South Sudan”.
The agreement then provides for a transition process to apply during an in-
terim period lasting for six years. At the end of that period:
There shall be an internationally monitored referendum, organized jointly by the GOS

and the SPLM/A, for the people of South Sudan to: confirm the unity of the Sudan by
voting to adopt the system of government established under the Peace Agreement; or
to vote for secession.23
There then follows an agreed text on the right of self-determination for the peo-
ple of South Sudan. This provides for a mid-term review of the implementation of
the peace agreement by the parties and international representatives with a view
to improving the institutions and arrangements created under the Agreement
and making the unity of Sudan attractive to the people of South Sudan.
In terms of state structure, the Protocol provides that the national Constitution
of the Sudan shall be the supreme law throughout the Sudan. The Constitution is
taken to be the source of all public authority, suggesting a devolved interim pow-
er-sharing arrangement. However, the Constitution is to be amended even before
the transition period to take account of the elements to be agreed in the definite
settlement. This includes a provision that limits the authority of the national gov-
ernment to the exercise of such functions “as must necessarily be exercised by a
sovereign state at national level”. There is also provision for the exemption of the
South from legislation inspired by the Shari’a. Instead, legislation of national ap-
plication is to take account of the diversity in Sudan.
23 Article 2.5.
A whole bundle of additional settlements based on the Machakos Protocols

were completed during 2004, followed by a rapidly drafted new constitution for
Sudan in 2005. The basic design of deferred secession was left untouched.
The Bougainville agreement of 30 August 2001 represents perhaps the most
interesting case of deferred possible secession. The agreement establishes an au-
tonomy regime for Bougainville that is described in detail in Chapter 5. The au-
tonomy is to be exercised under a “home-grown Bougainville constitution with
a right to assume increasing control over a wide range of powers, functions and
personnel and resources”. The agreement states one of its aims as being the pro-
motion of the unity of Papua New Guinea and the maintenance of a mutually
acceptable balance of interests between the interests of Bougainville and Papua
New Guinea as a whole.24
Section C of the agreement contains detailed principles on a referendum. No
earlier than ten years and no later than fifteen years after the election of the first
Autonomous Bougainville Government, a referendum shall be held, unless the
Bougainville government waives this entitlement. The referendum pledge is a
conditional one, depending on the achievement of good governance and weap-
ons disposal. A dispute settlement mechanism is to be established to address
divergences relating to the referendum. A UN mission is to help stabilize the
transitionary process.
In terms of state structure, the Bougainville Constitution is the supreme law
within the territory in relation to all matters that fall into its jurisdiction. The
Constitution may be changed by Bougainville, according to special procedures.
The national government exercises competence over defence, foreign relations,
transport and communication, and some other areas. The Bougainville authori-
ties enjoy authority over all other matters, provided they have been established in
a list of powers to be developed. Unlisted items remain initially with the national
government, although there is a procedure to address claims to the exercise of
authority in these areas by either entity. Listed powers will be transferred gradu-
ally to Bougainville. While a Bougainville court system is to be developed, the
national Supreme Court remains the final court of appeal for Bougainville, in-
cluding constitutional matters.
The rather cryptic Accord on Principles in Relations between Serbia and Mon-
tenegro of 14 March 2002 on the future of ‘a common state’, probably taking the
form of a confederation, provides for a ‘sovereign identity’ of both constituent
republics. The agreement was adopted under very strong pressure from the Euro-
pean Union. The underlying assumption appears to be that Montenegro does en-
joy a residual right of secession, probably based on its former status as a constitu-
ent republic of the SFRY. However, the agreement obliges the parties not to seek
independence for a period of three years. Should Montenegro leave the common
state at that time, Serbia would automatically become the successor state to it.
On 6 December 2002, a Constitutional Charter was generated, to be adopted
by both entities and the federal structures of the former FRY. The state was de-
24 Section B.1.4.
400 Marc Weller
scribed as the ‘Union of Serbia and Montenegro’, having the aims of fostering in-
tegration with the EU, human rights and the maintenance of an ‘internal market.’
An Assembly of Serbia and Montenegro enjoyed supreme authority in relation to
military issues, international affairs, and free movement of goods and services.
There was a blocking mechanism in the Assembly, requiring a majority of the
representatives of both entities. There was a president and a Council of Ministers,
with provision for representation from both Serbia and Montenegro. A common
defence force was controlled by a Defence Council, where no decisions could be
taken against the wishes of either unit. A dispute settlement mechanism existed
in the shape of a Court of Serbia and Montenegro with equal membership from
both entities. As was provided in the March agreement, after a three-year pe-
riod, either entity could initiate a procedure for the withdrawal from the union
through a referendum. The state not withdrawing was to inherit the legal per-
sonality of the union. This is, in fact, what occurred in 2005, after Montenegro
narrowly passed a referendum on independence.
A final and interesting case on deferred possible secession is furnished by the
Gagauz agreement already mentioned above. The structure of the agreement is
very much a devolution-type settlement. Autonomy is granted to Gagauzia, but
within limitations strictly set out in the agreement. As was already noted, the
dominance of the legal order of Moldova as a whole is confirmed and there is in
general no suggestion that Gagauzia might enjoy a legal identity as a constitu-
tional self-determination unit. However, the agreement does provide for the pos-
sibility of secession for Gagauzia should Moldova change its political status (i.e.,
join neighbouring Romania, perhaps with Transdniestria joining the Ukraine). In
this instance, a political development external to Gagauzia is constitutive of its
legal identity as a self-determined entity should that development arise.
C Balancing of Claims
A further case might be one where independence is theoretically possible, but
not likely in the foreseeable future. Hence, what is after all the granting of a self-
determination status satisfies the demands of secessionist groups. This, on the
other hand, is acceptable to the central government, as it appears that secession
is not really possible, given the realities of the situation, especially including the
population balance.
In the 1998 Belfast Agreement on Northern Ireland, the Republic of Ireland
and the UK “recognize the legitimacy of whatever choice is freely exercised by a
majority of the people of Northern Ireland with regard to its status, whether they
prefer to continue to support the Union with Great Britain or a sovereign united
Ireland.” The Agreement continues:
[I]t is for the people of the island of Ireland alone, by agreement between the two parts
respectively and without external impediment, to exercise their right to self-determina-
tion on the basis of consent, freely and concurrently given, North and South, to bring
about a united Ireland, if that is their wish, accepting that this right must be achieved
and exercised with and subject to the agreement and consent of a majority of the people
of Northern Ireland.
This language might lend itself to differing interpretations. However, it appears

that the Agreement moves towards a recognition that the actual self-determina-
tion unit is the island of Ireland as a whole. But the right of self-determination is
to be exercised separately by the North and South, establishing a blocking mech-
anism for the mainly Unionist population of the North.
The draft clauses/schedules provided in the Agreement for incorporation into
British legislation provide:
1 (1) It is hereby declared that Northern Ireland in its entirety remains part of the Unit-
ed Kingdom and shall not cease to be so without the consent of the majority of the
people of Northern Ireland voting in a poll held for the purposes of this section in
accordance with Schedule 1.
(2) But if the wish expressed by a majority in such a poll is that Northern Ireland
should cease to be part of the United Kingdom and form part of a united Ireland,
the Secretary of State shall lay before Parliament such proposals to give effect to
that wish as may be agreed between Her Majesty’s Government in the United
Kingdom and the Government of Ireland.
This clause could be seen formally to establish a constitutional right to self-de-

termination. While previously the United Kingdom had indicated that it would
comply with the wish of the population of Northern Ireland to join the Republic
of Ireland should this be made manifest in a referendum, this has now been made
express (subject to the doctrine of parliamentary sovereignty). However, one
must note that this is not a right of self-determination in the full sense. Instead,
it is a right to opt for a specific territorial change, moving sovereignty in relation
to Northern Ireland from the United Kingdom to the Republic of Ireland. Other
options, such as independence, are not available.
In terms of process, Schedule 1, to which reference is made in the above provi-
sion, requires the United Kingdom Secretary of State to order the holding of a
poll on joining the Republic of Ireland “if at any time it appears likely to him that a
majority of those voting would express a wish that Northern Ireland should cease
to be part of the United Kingdom and form part of a united Ireland”.25 Under this
clause, the Secretary is required to hold a poll (shall), but this mandatory require-
ment only comes into force subject to the exercise of his or her appreciation of
popular will – the very fact that can really only be assessed through the poll. No
further poll is to be held for seven years after a referendum that has rejected a
change in territorial status.
This provision makes a change virtually impossible until a significant demo-
graphic shift has occurred, or until popular sentiment in the North has changed
dramatically due to the experience of the new complex power-sharing arrange-
25 Schedule 1 (1 and 2).

402 Marc Weller
ments introduced by the accord. That arrangement provides for multi-layered

governance from the local level to the level of Northern Ireland, the level of UK
authority, and joint Anglo-Irish mechanisms.
IV Agreement on Process
A fourth approach to the self-determination issue is to leave the possibility of se-
cession unaddressed and instead provide for a mechanism to address the issue at
a future stage. This kind of case is perhaps exemplified by the Rambouillet agree-
ment that was left unimplemented, and possibly the situation that has emerged in
the context of Security Council Resolution 1244 (1999).26 Kosovo, an autonomous
province within Serbia that nevertheless had full federal status, declared itself
independent on 22 September 1991. Kosovo argued that it, too, should be entitled
to constitutional self-determination, in line with the claims of some of the Yugo-
slav republics. A decisive international attempt to achieve a settlement was made
in February/March 1999. The parties were presented with so-called non-nego-
tiable principles put forth by an international contact group. These principles
sought to balance wide-ranging self-government for Kosovo with the preserva-
tion of the unity of the Federal Republic of Yugoslavia. This compromise had been
consistently supported by the UN Security Council, for instance in Chapter VII
Resolution 1199 (1998).
The draft agreement that resulted from talks conducted at Rambouillet and
Paris confirmed “the commitment of the international community to the sov-
ereignty and territorial integrity of the Federal Republic of Yugoslavia.”27 While
the settlement was presented as an interim agreement, it was in fact virtually
permanent, as its duration was not formally limited and changes could only be
obtained with the consent of both parties – a most unlikely condition given their
diametrically opposed interests. On the other hand, in a crucial provision, it was
stated that:
Three years after the entry into force of this Agreement, an international meeting shall
be convened to determine a mechanism for a final settlement for Kosovo, on the basis
of the will of the people, opinions of relevant authorities, each Party’s efforts regarding
the implementation of this Agreement, and the Helsinki Final Act, and to undertake a
comprehensive assessment of the implementation of this Agreement and to consider
proposals by any Party for additional measures.28
26 Another view would be to classify this case as one of secession denied, given the
strong references to the continued territorial integrity of the Federal Republic of Yu-
goslavia in both the Rambouillet Agreement and Resolution 1244 (1999). However,
in view of the provisions invoked here, it was felt more appropriate to list this case
under this heading.
27 Preamble.
28 Chapter 8, Article 1(3).
This provision was of course not free of deliberate ambiguity. The reference to the
Helsinki Final Act, which is generally taken to emphasize territorial unity over
self-determination in the sense of secession, was taken to strengthen the position
of the Yugoslav government. The prospect of the final settlement, “on the basis of
the will of the people”, was understood to imply the possibility of a change in the
status of Kosovo on the basis of a referendum. This latter point was confirmed
in a confidential side-letter issued to Kosovo by the US delegation.29 The condi-
tionality element, based on performance by the parties of their obligations under
the agreement, also seemed to support the view that a change in status could be
contemplated, depending on the conduct of the parties.
While there may be dispute about the substance of the provision in terms of a
possible change of status for Kosovo, it is of course principally a requirement of
process. In mandatory terms, it provides that a meeting “shall” be convened three
years after entry into force of the Agreement. That meeting would be “interna-
tional”, presumably led once more by the international contact group. However,
the mandate of the meeting might be quite limited. Rather than achieving a final
settlement for Kosovo, the meeting was to determine the mechanism that would
be applied towards achieving such a settlement.
In terms of state construction, the Rambouillet draft was strongly geared to-
wards the position of the Yugoslav side. The legal identity of Kosovo was frac-
tionated, principal residual authority being based at the level of local communes.
These ethnically compact entities would therefore enjoy all powers not assigned
expressly to other levels of government. In addition, in a novel development, cer-
tain powers were allocated to national communities (i.e., the Serb minority in the
territory). The powers of the Kosovo institutions, on the other hand, were sub-
stantial, but delimited by express grants of authority in the agreement. Significant
powers also remained with the Serb government.
The Rambouillet agreement was, after some hesitation, formally accepted by
the Kosovo delegation. Although the Yugoslav delegation had initially signalled
that it might support the political aspects of the proposed settlement, ultimately
it refused to sign. After the termination of an armed campaign conducted by
NATO in relation to this conflict, the UN Security Council adopted Resolution
1244 (1999). That resolution also provides for a fairly complex approach to the
underlying self-determination issue. Again, the resolution restates the commit-
ments of UN member states to the sovereignty and territorial integrity of the
Federal Republic of Yugoslavia. It also recalls its previous call for substantial au-
tonomy and meaningful self-administration for Kosovo. This is to be achieved
through an international interim transitional administration overseeing the de-
velopment of, “pending a final settlement”, substantial autonomy and self-govern-
ment, and a transfer of authority to these agencies of self-government.30
The UN mandate, adopted under Chapter VII, also extends to “facilitating a
political process designed to determine Kosovo’s future status, taking into ac-
29 On the drafting history and all other relevant materials, see Weller 1999 Ch 8.
30 Para 11(a) and (c).
404 Marc Weller
count the Rambouillet accords.”31 Annex 1 of the Resolution consists of the G-8
statement concluded at Petersberg, Germany, on 6 May 1999, when attempts to
bring the conflict to a close were ongoing. That statement refers to:
A political process towards the establishment of an interim political framework agree-

ment providing for a substantial self-government for Kosovo, taking full account of the
Rambouillet accords and the principles of sovereignty and territorial integrity of the
Federal Republic of Yugoslavia and the other countries of the region, and the demilita-
rization of the KLA.
Annex 2, which contains a list of conditions for the termination of hostilities that
was accepted by the Federal Republic of Yugoslavia, restates this provision, add-
ing that “[n]egotiations between the parties for a settlement should not delay or
disrupt the establishment of democratic self-governing institutions.”
The meaning of these provisions is of course subject to dispute. The Federal
Republic of Yugoslavia has pointed to the strong reference to territorial integrity
and political independence. Instead of the more oblique invocation of the Hel-
sinki Final Act in the provision of the Rambouillet Agreement, this requirement
now appears expressly to limit the remit of the “political process” that is to oc-
cur. This limitation is no longer balanced by a reference to the “will of the peo-
ple”. Still, there is a cross-referral to the Rambouillet accords, which shall be fully
taken into account. Hence, it might be argued that the pledges contained therein
have been incorporated into the regime of Resolution 1244 (1999).
A more careful reading of the provisions in the Resolution and its annexes,
however, reveals a somewhat more surprising result. The conditions established in
the provisions cited above relate merely to the political process aiming to achieve
an “interim political framework agreement”. That is to say, they are not focused
on final status negotiations, but instead establish a limitation for an interim set-
tlement in advance of a determination of final status (much like Rambouillet).
Hence, the separate mandate for the facilitation of a political process designed
to determine Kosovo’s future status established in paragraph 11(e) is unrestricted
by any condition bar the requirement “to take into account” the Rambouillet ac-
cords. As was noted above, these are ambiguous, referring both to the Helsinki
accords and also to the exercise of the will of the people, presumably through a
referendum.
In actual practice, the UN Mission in Kosovo has in fact proceeded quite swift-
ly in establishing an interim arrangement for self-government. A constitutional
framework document for provisional self-government was promulgated on 15
May 2001. That document was mainly drafted by the UN mission, although both
parties were invited to have input into the draft. This option was exercised by the
parties in Kosovo to some extent, but mainly refused by the Federal Republic of
Yugoslavia/Serbia. There is no express reference to the continued territorial integ-
rity of the Federal Republic of Yugoslavia in that document. Similarly, when the
31 Para 11(e).
UN appointed mediator Martti Ahtisaari presented his comprehensive proposal

for the final status of Kosovo in February 2007, after a full year of negotiations
between the parties for the final status, there was no reference to any constraints
on independence. In fact, in his recommendations accompanying the package,
President Ahtisaari recommended supervised independence as the only realistic
solution – an option strongly resisted by Serbia and the Russian Federation.
V Conclusion
The self-determination rule in classical international law cannot resolve persis-
tent claims to self-determination that have arisen outside of the colonial context.
After all, the legal right of self-determination was expressly framed so as to ex-
clude such cases. By strongly privileging the doctrine of territorial unity over the
implementation of the will of a population that has expressed itself in favour of
secession, it only leaves the options of struggle or surrender to such populations.
Even in a colonial or post-colonial context, the doctrine of uti possidetis similarly
disenfranchises populations that do not wish to be part of the newly-independent
state. Again, no remedy is provided at the international legal level, and unprivi-
leged armed struggle appears to be the only possible avenue towards potential
independence.
This study has identified four approaches to the settlement of self-determina-
tion conflicts that previously appeared unresolvable. These are: (1) the formal
trading of a claim to secession for an enhanced identity of the secessionist entity
within the state; (2) ratification of secession; (3) the acceptance of the identity of
the secessionist unit as a self-determination unit, coupled with a deferral of the
act of self-determination; or (4) the attempt to defer the issue of status altogether.
This diversity of approaches is noteworthy. Moreover, in all four types of cases,
complex power-sharing plays an important role. Option (1) represents a gradual
evolution from the adoption of autonomy settlements to the use of complex pow-
er-sharing. Option (2) is a fairly novel development, representing an increasing
acceptance that it may be necessary, after all, to grant independence even outside
of the colonial context. This realization is only gradually being balanced with an
insistence of power-sharing tools in relation to ethnically complex situations in
the newly independent states. In cases of deferred secession, the third category
of cases, complex power-sharing is employed towards two apparently opposed
ends, depending on the perspective of the parties. The secessionist entity will
view complex power-sharing as a means of establishing substantive self-govern-
ance in advance of full independence. The central government, on the other hand,
may hope that complex power-sharing will demonstrate to the population that it
can best express its identity within the framework of the existing state. Finally,
the suspension of the self-determination issue by merely pledging a mechanism
to address the matter in the future reflects the hope of the international actors in-
volved in such instances that the matter can be best addressed at another time. In
the meantime, however, complex power-sharing is intended to generate stability
406 Marc Weller
and to promote post-conflict reconciliation, to facilitate a substantive settlement

at a subsequent stage.
Chapter 13
Power-sharing and the Vertical Layering of Authority:
A Review of Current Practices
Stefan Wolff
I The Context of Self-determination Conflicts and Power-sharing

Self-determination conflicts are, at one level, also conflicts between competing
views of how decision-making powers should be allocated to different layers of
authority within a state, and thus how the state as a whole should be constructed.
Traditionally, powers have been, and still are, distributed between different ver-
tical layers of authority, for example, between central and local governments in
unitary states, or between federal and federated governments in federations and
federacies. In addition to such vertical layering of authority, the resolution of self-
determination conflicts often requires additional mechanisms of power-sharing
that are horizontal, i.e., where power is (mandatorily) shared between different
parties at one-and-the-same level. While the level of such horizontal power-shar-
ing can be the region and/or the central government, the precise mechanisms and
rules of such horizontal power-sharing differ from case to case, and can be rigidly
consociational, as in Northern Ireland, or voluntary, as in Macedonia, where they
are facilitated by the ethnic demography, the structure of the country’s party sys-
tem, and the rules of the electoral system.
Depending on the complexity of a given power-sharing system, it comprises
one or more of the following mechanisms: co-decision making (e.g., executive
power-sharing), split decision-making (e.g., territorial arrangements, such as fed-
eracy or federalism; or corporate arrangements, such as cultural autonomy), and
a range of pre-determined decisions (e.g., proportionality rules for representa-
tion for different communities in legislatures and the civil service) or pre-deter-
mined procedures (e.g., qualified majority voting or parallel consent regulations
in legislatures).
Within systems of power-sharing, the vertical layering of authority acquires
extra importance, often necessitating specific new institutional structures, thus
adding to the overall complexity of the process and outcome of state construc-
tion. However, the vertical layering of authority also provides opportunities for
instituting formal and informal mechanisms of power-sharing at different levels
408 Stefan Wolff
of the political process – from the central government level down to that of lo-
cal communities. Yet, it can only accomplish this if two conditions are fulfilled:
the institutions and institutional structures created must be internally viable and
externally recognizable. That is, they must be capable in a technical sense of de-
livering the outcomes they are set up to achieve (e.g., effectiveness and represen-
tativeness of the political process), and the institutions and outcomes must be
recognized by the agents participating in them as, if not desirable, at least prefer-
able over continued violent conflict. Under these conditions, vertically layered
power-sharing institutions and the individual agents operating in and through
them will be capable of establishing a political process that is predictable and sta-
ble. This in turn will facilitate, and over time be facilitated by, an increasing belief
in the authority of the institutions and institutional structures thus created.
With these preliminary considerations in mind, this chapter analyzes state con-
struction in complex power-sharing systems from the perspective of how author-
ity is distributed at and between vertical layers of authority. The empirical basis
for this analysis is provided by eight recent cases of self-determination conflicts
where attempts have been made to resolve them by establishing complex power-
sharing institutions. Examining the vertical layering of authority in Bosnia and
Herzegovina, Bougainville, Gagauzia, Kosovo, Macedonia, Mindanao, Northern
Ireland, and South Ossetia, I initially evaluate the particular vertical structures of
state construction in each case and contextualize them briefly in the nature and
dynamics of each individual self-determination conflict. The sequence of case
studies is determined by the complexity of the institutional structures. I begin
with Bosnia and Herzegovina, where power-sharing exists at regional and central
levels, and is complemented by elements of devolution of powers to cantonal
and municipal levels. At the next level of complexity, regional consociations exist
in Bougainville, Mindanao, and Northern Ireland. Here two traditional conflict
resolution techniques combine – territorial autonomy and consociational power-
sharing. Although similar in this particular aspect, the three cases can be further
distinguished according to their institutional structures. The arrangements for
Bougainville include limited power-sharing (co-decision making) between the
regional and central authorities; in Northern Ireland, extensive arrangements for
cross-border cooperation between Northern Ireland and the Republic of Ireland,
as well as between these two entities and a range of others within the British Isles
form part of the 1998 Agreement; and, in Mindanao, the co-optation of regional
officials to corresponding central institutions again provides for limited co-de-
cision making. The remaining four cases are examples of territorial autonomy
(Gagauzia), enhanced local self-administration (Macedonia and, pending the
resolution of its final status, Kosovo), and a quasi-sovereign parallel entity (South
Ossetia).
Following this description of the empirical basis of this chapter, I then assess
the relevance of the vertical layering of authority within complex power-sharing
systems by comparing and contrasting all eight cases from the perspective of
the types of institutional structures; the combination of vertical and horizontal
power-sharing mechanisms; the distribution of powers at and between different
13 Power-sharing and the Vertical Layering of Authority: A Review of Current Practices 409
vertical layers of authority; the types of coordination between different vertical

layers of authority; the constitutional and legal entrenchment of the institutions
created; and territory and population as boundaries of authority. Following this
thematic comparison, I examine three common and potentially problematic is-
sues relating to the vertical layering of public authority in complex power-sharing
systems: the relationship between vertical and horizontal layers of power-shar-
ing, the coordination of government activities at and between these different lay-
ers, and the overall political institutional settlement within which vertically and
horizontally structured institutions have to operate. Synthesizing this discussion,
I conclude by outlining the role that the vertical layering of authority can play
as part of a power-sharing ‘toolkit’ by examining the main types of institutional
structures and mechanisms of policy coordination and by assessing their context-
dependence.
II Multiple Layers of Authority in Practice: A Brief Review of Eight Case

Studies of Complex Power-sharing
In order to assess similarities and differences and their significance in the vertical
construction of state institutions, empirical data are required. Given the focus of
this chapter, the following case studies do not attempt to provide a comprehen-
sive analysis of each conflict and its settlement, but rather concentrate on how
executive, legislative, and judicial institutions are constructed in the state overall
and at multiple levels of authority.
A Bosnia and Herzegovina

The war in Bosnia and Herzegovina was part of a wider regional conflict – the
disintegration of Yugoslavia. Over three-and-a-half years, between 1992 and 1995,
three main conflict parties – Serbs in Bosnia (and their supporters in Serbia),
Croats in Bosnia (and their supporters in Croatia), and Bosnian Muslims – fought
each other in shifting alliances with different aims. Serbs sought secession and
unification with Serbia, as they felt threatened in a state potentially dominated
by a Muslim or Muslim/Croat majority hostile to them. To some extent, Cro-
ats shared this goal of secession and unification (with Croatia), while Muslims
fought to prevent the disintegration of what they perceived as their ancestral
homeland. The intensity of the conflict prompted the UN to declare six safe areas
for Muslims and to despatch a peacekeeping force for their protection. Following
the breakdown of a four-month ceasefire between Muslims and Serbs, the lat-
ter launched an intensive campaign against Muslim safe areas between May and
August 1994 in which large numbers of civilians were deliberately targeted and
killed. In response, NATO intensified its air strikes against the (regular and ir-
For detailed background on the eight case studies of self-determination conflicts,

refer to Pat Two of this volume.
410 Stefan Wolff
regular) armed forces of the Bosnian Serbs and eventually forced all three conflict
parties to the negotiating table in Dayton, Ohio in September 1994.
The Dayton Peace Agreement of 1995 provides the legal foundation upon which
the post-war Bosnian state has been constructed. It establishes several layers of
authority: principally, the state level, the entity level, and the local level. Within
the Bosnian-Croat Federation, cantons provide a further layer of authority. All
four layers of authority have their competences clearly laid out in the Dayton
Peace Agreement, its various annexes and follow-on documents, as well as vari-
ous subsequent amendments. A significant change to this structure was made
in 1997 when the so-called Peace Implementation Council, uniting almost sixty
states and governmental and nongovernmental organizations involved in the
implementation of the Dayton Peace Agreement, decided to endow the Office
of the High Representative with the authority to dismiss elected and unelected
officials in Bosnia and Herzegovina if they were deemed to obstruct the imple-
mentation of the Dayton Peace Agreement, and to make legally binding decisions
(i.e., to pass laws) in any area in which the state or entity parliaments were unable
or unwilling to legislate. This establishes the High Representative not only as the
ultimate arbiter in any cases of difficulties in implementing the Dayton Peace
Agreement and in coordinating policy between the institutions it established,
but endows the office, similar to that of the Special Representative of the UN Sec-
retary-General in Kosovo, with significant legislative and executive powers. This
is comparable only to the powers of the Secretary of State for Northern Ireland,
who, through subsequent amendment to the original Agreement, is able to sus-
pend Northern Ireland’s power-sharing institutions and assume their executive
and most of their legislative powers.
What is striking about the construction of the Bosnian state is the almost exces-
sive degree of decentralization (Figure 1). Powers at the level of state institutions
are very few. They include foreign relations, foreign trade, customs, monetary
policy, immigration, international and inter-entity criminal law enforcement,
communications infrastructure, inter-entity transportation, air traffic control,
and inter-entity coordination. Any other power or part thereof, not explicitly
mentioned, is by default allocated to the entities, which thus become the sources
of original authority. Whereas there is further devolution to cantons and even-
tually to municipal authorities in the Bosnian-Croat Federation, the Republika
Srpska is an almost oddly centralized entity in the context of Bosnia, retaining
most powers at the level of the entity government and endowing municipalities
primarily with administrative functions in the areas of development, urban plan-
ning, budget, local infrastructure, and specific local needs of citizens in the areas
of culture, education, health, and social welfare, etc.
Office of the High Representative
State Institutions Parliametary Assembly (House of Repre- Presidency Constitutional Central Bank
sentatives and House of Peoples Council of Ministers Court
Entity Institutions Parliament of the Presidency of the Judicial System National Assembly Presidency of the Judicial System
Bosnian-Croat Fed- Bosnian-Croat of the Republica Republica Srpsla
eration (House of Federation Govern- Srpsla Government of the
Representatives and mentof the Bosnian- Republica Srpsla
House of Peoples Croat Federation
Cantonal Institutions Cantonal Cantonal Judicial System

Legislatures President
Cantonal
Government
Municipal Institutions Municipal Govern- Municipal Judicial System Municipalities

ing Council Executive
Figure 1 Vertical Layers of Public Authority in Bosnia and Herzegovina
13 Power-sharing and the Vertical Layering of Authority: A Review of Current Practices
411
412 Stefan Wolff
This particular layering of authority reflects the balance of power within Bosnia
and Herzegovina as a whole and within the two entities. The entities gained wide-
ranging autonomy in almost all functions of government, including defence and a
significant part of foreign policy, and were thus able to assert their independence
from a weak central government at the state level. This distribution of power at
the state-entity nexus is mirrored within the Bosnian-Croat Federation: cantonal
and municipal authorities are strengthened at the expense of the Federation gov-
ernment.
Bosnia and Herzegovina, thus, could be characterized as an asymmetric feder-
ation in the sense that there are significant differences in how authority is layered
within each entity. Under different circumstances, the existence of cantons in the
Federation but not in the Republika Srpska, might have resulted in policy coordi-
nation problems within the overall structure of the Bosnian state, similar to those
in Gagauzia and Mindanao. However, the fact that very few powers remain at the
centre, which would require coordination, counteracts this problem, as does the
existence of the Office of the High Representative, who retains key powers that
enable him to enforce compliance of, and coordination between, different layers
of authority within each entity and across entities, if only by dismissing elected
and unelected officials unable or unwilling to cooperate and by passing by decree
laws and regulations deemed necessary for the functioning of institutions at all
layers of public authority in Bosnia and Herzegovina.
In the context of the ethnic demography of Bosnia and Herzegovina, this layer-
ing of public authority has several implications. Firstly, most powers are located
at the interethnically least contentious level – the entity in the case of Repub-
lika Srpska, the cantons in the case of the federation. This institutional design
absolves elites from substantive cooperation, as significant powers mostly lie at
levels where there is significant ethnic homogeneity. For this very reason, there
is, secondly, little or no need for segmental autonomy: virtually monoethnic lev-
els of government have authority over (usually segmented) policy areas such as
culture and education. Thirdly, because of the degree to which power has been
retained at the entity level, especially in the federation, and because of the fact
that there remain certain powers in the competence of the state-level institu-
tions, there is a greater need for mandatory horizontal elements of power-sharing
(proportionality, qualified majority voting procedures in legislative assemblies,
etc.), which are provided for in great detail in the Dayton Peace Agreement and
other relevant constitutional documents and their subsequent amendments. The
devolution of power to ever-lower levels of government, approximating almost
perfect subsidiarity in the federation, is not a substitute for horizontal power-
sharing, in contrast to what can be observed in relation to Gagauzia, Kosovo, and
Macedonia (see below). On the one hand, the need for horizontal power-sharing
exists as long as there is a politically significant ethnic mix of the population at
the relevant level of government. Related to this, on the other hand, the issue of
who exercises authority in which policy area does not become less problematic if
power is devolved further and further down from the centre. For the exercise of
power to be legitimate, i.e., for power to become authority, the very institutions
of government need to be recognized as legitimate – and this is as true at the

level of municipal government as it is at that of state government in Bosnia and
Herzegovina and elsewhere.
B Bougainville
Similar to the conflict in Mindanao and Northern Ireland, and, with some quali-
fications, Kosovo, the self-determination dispute in Bougainville has two dimen-
sions – one internal to Bougainville, and one between Bougainville and Papua
New Guinea (of which Bougainville is still a part). The internal dimension of the
conflict is characterized by the, at times, violent confrontation between those
who favour the integration of Bougainville into Papua New Guinea, albeit with
substantial autonomy granted to the province, and those who argue for indepen-
dence. Because of the ethnic and linguistic mix of population groups in Bougain-
ville, it is difficult to classify this internal dispute as an ethnic conflict, particu-
larly as there is evidence that an overarching Bougainvillean ‘ethnic’ identity has
developed over past decades that bridges the existing cultural (but not politi-
cal) differences. The conflict between Bougainville and Papua New Guinea is a
manifestation of a secessionist conflict – a peripheral region seeking independent
statehood – with the national government rejecting this demand for economic
reasons (when it was operating, the Bougainville copper mine generated about
one sixth of Papua New Guinea’s tax revenue), as well as out of fear that accepting
the secessionists’ demands might unleash further secessionist claims elsewhere
in the country.
Intense conflict from the late 1980s to the mid-1990s, initially between seces-
sionist Bougainville forces and national government security forces, and sub-
sequently between pro- and anti-independence groups in Bougainville, led to
several thousand people being killed and many times more forcibly displaced.
Even though peace initiatives were launched almost from the day the conflict
escalated violently, it was particularly from 1994 onwards that measurable prog-
ress was made in negotiating a comprehensive agreement with the help of the
United Nations and some key regional states (primarily Australia, New Zealand,
and the Solomon Islands). Yet, because of the duality of the internal and external
conflicts, and the complexities arising from this, the Bougainville Peace Agree-
ment was only concluded in 2001. It provides for an autonomous Bougainville
government with clearly defined powers entrenched in and limited by the na-
tional constitution and operating according to principles established by a future
Bougainville constitution. While the institutional structures thereby provided are
the central concern of this chapter, it is also worth noting that the agreement
specifies a future, albeit non-binding, referendum on independence in Bougain-
ville and a weapons disposal plan to be co-monitored by the United Nations Ob-
server Mission on Bougainville. Bougainville thus shares with Kosovo and South
Ossetia the characteristic that a final status decision has been postponed or not
made. However, as a formal agreement exists between national and provincial
leaders laying down the principles of institutional design for the interim period,
414 Stefan Wolff
Bougainville’s situation more closely resembles that of Northern Ireland, where

the Agreement of 1998 provides for referenda to be held on the so-called ‘border
question’, i.e., whether a majority of people in Northern Ireland and the Republic
of Ireland express their desire for a change in sovereignty in separate, parallel
referenda.
Examining the institutional structures relevant to the question of how a resolu-
tion of the Bougainville self-determination conflict was attempted requires exam-
ination of three different dimensions: the overall construction of the Papua New
Guinean state, the distribution of powers between, and how they are exercised by,
the central and provincial governments, and the design of provincial institutions
aimed at resolving intra-Bougainville conflict between pro- and anti-secessionist
factions. Although details of the last will depend on the outcome of constitutional
negotiations among different political and other groups in Bougainville, the main
parameters of the post-conflict Bougainville government have been laid down in
the Bougainville Peace Agreement. Of particular relevance for this chapter is the
fact that the national constitution of Papua New Guinea (as revised in accordance
with the Bougainville Peace Agreement) will retain primacy over any Bougainville
constitution in the sense that Bougainville constitutional law has to be compat-
ible with Papua New Guinea constitutional law. Consequently, the head of state
will only endorse the Bougainville constitution once he or she is satisfied that the
proposed regional constitution is compatible with the national one and with any
international obligations into which Papua New Guinea has previously entered.
However, the Bougainville constitution will have supreme status in all areas of
Bougainville’s jurisdiction and will be enforceable in the Supreme Court of Papua
New Guinea and, to the extent that it provides for this, any Bougainville courts.
Although the precise nature of the institutional arrangements in Bougainville
itself are thus not yet clear at the time of writing (June 2003), the division of
functions between national and autonomous provincial government is laid down
in significant detail in the Bougainville Peace Agreement. Leaving aside the fact
that there is a (largely ceremonial) head of state (i.e., the reigning monarch of the
United Kingdom of Great Britain and Northern Ireland) who appoints a Gov-
ernor-General, there is a three-tiered system of public authority in Papua New
Guinea consisting of the national government, provincial governments, and local
governments (Figure 2). As a province in its own right, Bougainville therefore fits
neatly into the existing (symmetric) structure of the state of Papua New Guinea,
but acquires, as a result of the Bougainville Peace Agreement, significant addition-
al powers compared to those enjoyed by other provincial governments. Although
there is very little evidence to the contrary at this early stage of implementation
of the Agreement, it is unlikely that Bougainville will be plagued by the same un-
certainties and difficulties in coordinating the exercise of powers and functions
by different layers of authority that have been identified in the Gagauz and Min-
danao cases, where additional layers of authority were created that did not fit in
the pre-existing structure of the respective states. This is all the more likely as the
Bougainville Peace Agreement foresees quite extensive arbitration mechanisms
in case of disputes between the national and provincial government.
Figure 2 Vertical Layers of Public Authority in Papua New Guinea
Head of State of Papua New Guinea

Governor General of Papua New Guinea
State Institutions National Parlia- National Judicial

ment of Papua Executive System
New Guinea
Bougainville Legislation Executive Judicial

Autonomy Institutions System
Local Government
(Structures to be determined)
The distribution of powers and functions between the national and autonomous
Bougainville governments is regulated according to a system of two comprehen-
sive lists. The government list details the following powers and functions that
are to be exercised exclusively by the national government: defence, foreign rela-
tions, immigration, highly migratory and straddling fish stocks, central banking,
currency, international civil aviation, international shipping, international trade,
posts, telecommunications, and all other powers either assigned to the national
government under the Bougainville Peace Agreement or which it requires in the
process of implementation. The list for the autonomous Bougainville govern-
ment is less explicit and more ‘open-ended’, but simply includes all known or
identifiable powers not on the national government’s list. In the area of foreign
relations, special provisions are made to account for the distinct relationship be-
tween the central government and the Bougainville autonomous government, as
well as for particular interests that Bougainville has. Thus, representatives of the
Bougainville government may be included in delegations of Papua New Guinea
to regional meetings and organizations; and before the conclusion of any future
treaties affecting Bougainville directly or indirectly, such as border issues or fish-
ing rights, a consultation process between the national and Bougainvillean gov-
ernments has to be conducted.
The list of powers and functions available to either government can be amend-
ed or altered in the future. Amendments are likely to occur whenever new powers
and functions are identified, and in such cases the parties are obliged to inform
one another of their intention to claim a particular function or power and to sub-
mit to agreed arbitration mechanisms if no consensus can be reached. Further
alterations are possible where one of the two governments delegates powers and
416 Stefan Wolff
functions originally allocated to it to the other. Initially, all powers and functions
are held by the national government, and their devolution occurs only after a
process of notification and consultation initiated by the autonomous Bougain-
ville government. That is, the Bougainville government will assume the powers
and functions allocated to it in the peace agreement gradually, and there is only
a limited automatism in the transfer of powers in the sense that any power or
function not explicitly reserved for the national government has to be devolved
to Bougainville upon request from the autonomous government.
As the structure of institutions in Bougainville itself is subject to the outcome
of negotiations on a new provincial constitution, little can yet be said about the
layering of public authority in the autonomous province itself. Parallel local gov-
ernment structures had developed through the years of intra-Bougainville con-
flict (Council of Elders and Council of Chiefs) and will have to be integrated into
subsequent new structures of local authority. However, the Bougainville Peace
Agreement does not exactly prescribe this as necessary, but merely points out
that under a Bougainville constitution such local structures of government may
be created. What is, however, quite explicit in the agreement is that the Bougain-
ville government will consist of three branches – a legislature, an executive, and
a judicial system.
C Mindanao
The self-determination conflict in Mindanao is characterized by the parallelism
of internal and external conflicts. Within Mindanao, the conflict between a pre-
dominantly indigenous Muslim population and a predominantly migrant Chris-
tian population is about the control of what Muslims perceive as their ancestral
homelands. An additional complication in this territorial dispute arises from the
fact that Muslims have become a marginalized minority as a result of massive im-
migration of Christian settlers multiplying Mindanao’s population over the past
century by more than fifteen times and turning what was once a three-quarters
majority of Muslims into a less than one-quarter minority. Superimposed on this
ethno-religious regional conflict, there is a conflict between Muslim militants
and the national government of the Philippines over the secessionist aspirations
of a considerable section of Mindanao Muslims. Both conflicts combined from
the 1960s onwards into a civil war and general state of lawlessness in Mindanao
that saw tens thousands of people killed and hundreds of thousands displaced or
turned into refugees. The two main insurgent agents are the Moro National Lib-
eration Front (MNLF) and one of its break-away groups, the Moro Islamic Lib-
eration Front (MILF). This has added a third dimension to the conflict, namely an
intra-Muslim divide along ethnic and political lines.
Efforts to bring peace to Mindanao in the 1970s and 1980s failed, despite an
agreement being signed in Tripoli in 1976 and the 1987 approval of a new Consti-
tution by referendum providing Mindanao with its own autonomous structures.
The failure of these two earlier attempts to resolve the conflict peacefully was
primarily due to the obstructionist attitudes of the two main rebel groups, which,
among other things, resulted in only four out of thirteen eligible provinces in Min-
danao opting for membership in the Autonomous Region of Muslim Mindanao
(ARMM). Following new efforts by the Philippines government, an agreement
was concluded with the MNLF that was designed to bring the decades old violent
self-determination conflict to an end by fully implementing the 1976 agreement
and all subsequent legislation for autonomy arrangements in the ARMM. A sub-
sequent referendum on membership in the ARMM resulted in only one addi-
tional province and one further town joining the autonomous region.
As a layer of authority, the ARMM was ‘inserted’ into an existing four-tier
structure in the Philippines, consisting of the national government and pro-
vincial, municipal, and local authorities (Figure 3). At each level, the traditional
three branches of government – legislature, executive, and judicial system – ex-
ist, even though their competences extend to different areas. The particularity
of the autonomous region not only arises from the fact that it is an additional
structure in the Philippine political system or that it is made up of five, territori-
ally non-contiguous provinces, but also in that it prescribes a power-sharing re-
gional administration consisting of a cabinet and executive council, controlled by
a Regional Governor and a Deputy Governor. Equally, the Regional Assembly is
unique within the Philippines in that it comprises both popularly and corporately
elected members. The judicial system in the region is distinct from other national
judicial institutions in that it allows Muslim Shari’ah courts and tribal courts to
practice, albeit only in areas of family and religious matters, alongside national
judicial institutions. Authority between the national government and the ARMM
government is clearly divided according to policy areas. Foreign affairs, defence,
security, postal service, fiscal and monetary affairs, administration of justice, for-
eign trade, customs and tariffs, citizenship and immigration, communication and
auditing remain the exclusive domain of the national government, while health,
education, human resources, science and technology, people empowerment, in-
tra-regional communication, and economic development are in the domain of
the ARMM government.
418 Stefan Wolff
Figure 3 Vertical Layers of Public Authority in the Philippines

National Government Congress President Judicial
Government System
National Security
Council
Autonomous Region of Regional Assem- Regional Gov- Judicial

Muslim Mindanao bly (popularly ernor and Vice System
and corporately Governor
elected) Power-sharing
Cabinet
Power-sharing Ex-
ecutive Council
Provincial Government Provincial Provincial Judicial

Assembly Governor System
Municipal Government Municipal Mayor District Courts

Assembly
Local Government Local Barrangay (local Barrangay

Assembly community) Head Courts
The key problem that remains to be resolved is not the separation of competences
between the national government and the government of the autonomous re-
gion, but rather how authority is layered in the relationship between the ARMM
government and the three pre-existing lower levels of government upon which
the ARMM structure was superimposed, a situation that is similar to that of
Gagauzia in Moldova. The nature of this problem is one of coordination within
the specific system of institutions established with the creation of the autono-
mous region as an additional layer of authority. From this perspective, the al-
location of specific areas of competence to the ARMM government proves both
advantageous and disadvantageous at the same time. Its advantage is the clearly
defined degree of autonomy that the autonomous region has thus achieved from
the national government. However, this clear definition of powers at the same
time may prove limiting in its dealings with provincial, municipal, and local levels
of government, which have specific competences of their own that do not derive
from the autonomous region as their original source of authority, but from the
national government, and which, at the same time, report directly to the national
rather than to the regional government. Rather than establishing a clear vertical
division of power between national government and autonomous region, a dual
structure is beginning to emerge. The ARMM has autonomy from the national
government in some policy areas, but provincial, municipal, and local govern-
ments have competences in other areas where authority overall remains with the
national government, thus bypassing ARMM and being accountable directly to
the national government. As in Gagauzia, this situation has the potential to un-
dermine the whole idea of autonomy for a specific region and render it meaning-
less, thereby providing fuel for renewed conflict.
D Northern Ireland
The conflict in Northern Ireland is essentially about competing conceptions of
national belonging. The Nationalist vision of a united Ireland is diametrically op-
posed to the desire of a Unionist community to retain strong constitutional links
with Great Britain in the United Kingdom. Since its creation in 1921, the province
of Northern Ireland has seen more than one resounding failure of attempts to
resolve this self-determination conflict. For the first fifty years after the partition
of Ireland, the province enjoyed wide-ranging autonomy within the United King-
dom, having its own parliament, government, civil service, and judicial system.
This, however, exacerbated ethnic tensions more than it contributed to calming
them. Northern Ireland was run as a majoritarian democracy, with a Unionist
majority heavily discriminating in all spheres of public life – from employment
and housing to education and culture – against the Nationalist minority. When
the Nationalist community, from the mid-1960s onwards, began to push for
equal rights, the political system set up to protect the status and privileges of the
Unionist community proved utterly unsuitable to handling the ensuing conflict.
Escalating violence prompted the deployment of the British Army in the province
in 1969 and the later suspension of the system of self-government in Northern
Ireland in 1972. However, within a year of the suspension, the moderate political
parties in Northern Ireland and the British and Irish governments had agreed on
a new institutional framework that provided mechanisms for power-sharing be-
tween the two communities and a formal involvement of the Republic of Ireland
in the governing of the province as the Nationalist community’s kin-state, the
so-called Sunningdale Agreement, which failed within months of its inaugura-
tion. Throughout the next two-and-a-half decades, several initiatives to resolve
the conflict failed. It was only when the Belfast/Good Friday Agreement, which
replicates many of the key features of the 1973 Sunningdale Agreement, was con-
cluded in 1998 after an inclusive negotiation process that the prospect of sustain-
able conflict settlement arose anew.
The power-sharing institutions in Northern Ireland slot in between the central
government in Westminster and the twenty-six local councils within Northern
Ireland (Figure 4), and are, as a layer of public authority, by-and-large comparable
to the institutional structures established in Scotland and Wales since 1997. The
national government remains the residual source of all public authority. This ap-
pears to include, contrary to the original agreement of 1998, the power to suspend
the power-sharing institutions in Northern Ireland unilaterally. In this respect,
Northern Ireland is unique among the cases considered here in that its autonomy
420 Stefan Wolff
can be revoked at any time by the central government. When the power-sharing
institutions in Northern Ireland are operational, they have powers in all presently
devolved matters, i.e., economic development, education, health and social ser-
vices, agriculture, environment, and finance. Depending on a future assessment
by the British government, further powers may be devolved to the institutions
in Northern Ireland. These are, at the moment, so-called reserved matters and
include criminal law, criminal justice, and policing. A third category of powers is
to remain with the British government indefinitely. These excepted matters are
foreign and defence policy, the Crown, and monetary policy.
Figure 4 Vertical Layers of Public Authority in the United Kingdom in relation to

Northern Ireland
Head of State
National Government Houses of Government of Judicial

Parliament the UK System
(sovereign)
Power-sharing Insti- Assembly EXECUTIVE Judicial

tutions in Northern First and Deputy System
Ireland First Minister
Ministers
Executive Com-
mittees
Local Authorities Council Town Clerk and

Chief Executive
Similar to the provisions relating to Gagauzia and Bougainville, the 1998 agree-
ment foresees the possibility of constitutional change through a referendum.
Should a majority of the people of Northern Ireland express the wish to unite
with the Republic of Ireland at some stage in the future, both governments have
committed themselves to respect such an expression of the popular will and the
British government is to provide for referenda at regular intervals to gauge public
opinion on this issue. A crucial difference in the case of Northern Ireland, how-
ever, is that for Irish unification to happen, a majority of the population in the Re-
public of Ireland also needs to approve such a change in international boundaries
by referendum. Thus, in fact, secession itself is not an option, only an irredenta,
i.e., unification with the kin-state.
The structure of institutions in Northern Ireland mirrors the classical division

of powers between legislature, executive, and judiciary. The directly-elected As-
sembly has full legislative competence over all devolved matters. The power-shar-
ing executive, which enjoys full executive competence over all devolved matters,
is comprised of a First and Deputy First Minister (elected on a joint ticket) with
coordinating executive functions, ministers (selected according to the d’Hondt
rule) who formulate and execute policy and enact assembly legislation within the
remits of their portfolios, and Executive Committees who scrutinize ministerial
departments. Legislature and executive are complemented by an extensive judi-
cial system consisting of a High Court, County Courts, and Magistrates Courts,
an Attorney General, an Advocate General, a Public Prosecution Service, a Chief
Inspector of Criminal Justice, and a Law Commission.
The third layer of public authority relevant in the case of Northern Ireland is
that of local authorities. Here, twenty-six local councils, also referred to as bor-
oughs, have competences in a range of areas, including development, tourism,
community relations, and the environment. Local authorities have a directly-
elected Council and a Town Clerk and Chief Executive, who are responsible for
running day-to-day affairs.
E Gagauzia
The territory of today’s Republic of Gagauzia, an autonomous republic in the
Republic of Moldova, has belonged to a variety of rulers and states in its history.
It was part of the Ottoman empire until 1812 when it was annexed by Russia. In
the middle of the nineteenth century, it came under Romanian administration
as part of the Romanian province of Moldova. After the Second World War, it
became part of the Moldavian Soviet Republic and remained as such after the
dissolution of the Soviet Union. A Gagauz self-determination movement began
to develop from around 1980 onwards, but only gained significant momentum
when the Moldovan Supreme Soviet passed a discriminatory language law in
1989. In November 1989, the Gagauz, who make up more than 80% of the popula-
tion in their homelands, formed an ethnically based organization, Gagauz Halkî
(Gagauz People), to represent their particular interests. This process of increas-
ingly political self-assertion culminated in 1990 in the unilateral declaration of
an independent Republic of Gagauzia, which, although it was not recognized ei-
ther internally or externally, signified a new stage of escalation in the long-stand-
ing conflict. Although comparatively little violence was involved, this escalation
prompted intense negotiations between Gagauz and Moldovan officials over the
future status of their relationship. These were successfully concluded in 1994,
resulting in the status of the Republic of Gagauzia as an autonomous republic
within Moldova being formally recognized.
The layering of public authority in Moldova follows a three-tiered system (Fig-
ure 5). The central institutions of the state are the original source of all authority
except for those powers that are explicitly and fully devolved to Gagauz autono-
mous institutions or to the third layer of authority, the so-called rayons.
422 Stefan Wolff
Figure 5 Vertical Layers of Public Authority in Moldova

State Institutions Parliament of President of Judicial
the Republic of the Republic of System
Moldavia Moldavia
Government of
the Republic of
Moldavia
Gagauz Autonomy People’s Governor of Judicial

Institutions Assembly Gagauzia System
Executive
Committee of
Gaugazia
Districts Council Head of Administration

(Rayons) Elected, but not reporting Appointed by the Governor of
directly to the central Gagauzia (1998–2003)
government (2003–)
The autonomous institutions of Gagauzia have devolved powers in the areas of

science, culture, and education; housing and urban planning; health; sports; bud-
getary and taxation policy; economic policy; environment; industrial relations;
and social security. As in most other democratic systems, the institutional system
in Gagauzia is characterized by the separation of powers between the legisla-
ture (People’s Assembly), the executive (Governor and Executive Committee of
Gagauzia), and a judicial system (comprising the Tribunal of Gagauzia and lower-
order courts). The areas in which the elected People’s Assembly is competent to
legislate include most public policy areas, except defence and foreign affairs, as
well as matters relating to the territorial organization of Gagauzia and the orga-
nization of local administration, elections, and referenda. The directly-elected
Governor of Gagauzia issues decrees and regulations and proposes to the Assem-
bly the Executive Committee, which is charged with implementing and enforcing
laws passed by the Assembly, and with formulating and implementing policies
within the remit of Gagauzia’s competences.
The rayons in Moldova enjoy self-rule on local matters. Before 2003, the rayons
were run by a Head of Administration who was, in all rayons falling into the juris-
diction of Gagauzia, directly appointed by the Governor of Gagauzia. Since 2003,
the rayons have been run by an elected council that reports directly to the Mol-
dovan central government. At worst, this may potentially render the autonomous
institutions of Gagauzia superfluous. Even a slightly more optimistic perspective
promises at best a rather muddled arrangement and increases the demand for co-
ordination between the different layers of authority that are no longer ordered in
a clear hierarchy. Similar to the Autonomous Region of Muslim Mindanao, there
thus exists a problem with the autonomy arrangements in Moldova in that the
Republic of Gagauzia has now become a unique layer of authority in an otherwise
(still) unitary state, but it does not have the authority to structure itself and the
relations between its own layers of authority (i.e., Gagauz republican government
and local governments in the territory of Gagauzia), as Bougainville, for example,
is able to do by virtue of its own constitutional powers and as determined in the
Bougainville Peace Agreement.
Another feature of the Gagauz autonomy with direct relevance for the layering
of public authority is the fact that Gagauzia has, in law, no clear territorial demar-
cation. The fact that local communities can decide individually, by referendum,
whether they want to become part of the autonomous region has meant in prac-
tice that Gagauzia itself is not a territorially contiguous entity. This has not (yet)
caused any noticeable administrative difficulties. At a theoretical level, however,
it offers an interesting new perspective on the design of autonomous entities that
lies somewhere between, and yet combines elements of, territorial and personal
autonomy.
Finally, one should note that the process of state construction in Moldova itself
is far from complete. Another unresolved self-determination conflict (Transd-
niestria) will require substantial reform of the current state structure and in all
likelihood will also affect the status of Gagauzia. At present (mid-2003), the most
likely outcome of such state-wide institutional reform is a trilateral federation,
with Gagauzia as a constituent element in it, but probably less powerful and less
autonomous than the other two.
F Kosovo
Kosovo has a complex history with an array of different rulers and states claiming
sovereignty over the area over time. With the exception of a brief interlude dur-
ing the Second World War, Kosovo has been a province of Serbia for most of the
twentieth century, despite the fact that it is now inhabited by an overwhelming
majority of Albanians and borders the Republic of Albania. Throughout most of
this period, Kosovo had some form of autonomy within Serbia and/or the various
configurations of the Yugoslav state. Created on 3 September 1945 as a constitu-
ent part of Serbia, the status of Kosovo as an autonomous province with limited
self-government was confirmed in the Yugoslav constitutions of 1946, 1953, and
1963, and in the Serbian constitution of 1963, before the Yugoslav constitutional
reform of 1974 significantly enhanced the status of Kosovo as an autonomous
entity in Serbia and Yugoslavia and gave it de facto equality with all Yugoslav Re-
publics at the federal level, except for the right to secede. The interethnic tensions
between Serbs and Albanians in the province increased under the 1974 autonomy
arrangement, as it led to an ever-stronger perception among Serbs of discrimina-
tion at the hands of the Albanian majority in the province. After the death of Tito
in 1980, interethnic tensions across Yugoslavia began to increase, particularly in
Kosovo, where protesters began to demand republican status for the province
(with the implication of subsequent secession). These protests were quickly sup-
424 Stefan Wolff
pressed by Yugoslav security forces, yet tension continued and occasionally esca-
lated into violence. While more and more Serbs and Montenegrins left Kosovo,
Serbian repression of Albanians in Kosovo increased, and public sentiment in
Serbia turned increasingly anti-Albanian. By 1990, both sides had become even
more radicalized, so that the 1990 removal of all elements of sovereignty enjoyed
by Kosovo under the 1974 constitution and the (unrecognized) 1991 Kosovo Alba-
nian referendum on independence were only logical steps along a path of further
alienation between the two communities and of escalation of the Kosovo conflict.
This process culminated in the events of the second half of the 1990s, which saw
a violent ethnic conflict in Kosovo and eventually NATO’s air campaign against
Serbia establishing Kosovo as a quasi-protectorate of the UN in 1999.
Within the institutional structure that has emerged since 1999 (Figure 6), the
Special Representative of the Secretary-General of the United Nations (SRSG)
retains the full authority given by UN Security Council Resolution 1244. On this
basis, the SRSG also retains full decision-making authority regarding any aspect
of provisional self-government in Kosovo and can unilaterally effect any change
to the existing Constitutional Framework. While the powers of the international
community in Kosovo are thus broadly similar to, albeit somewhat more exten-
sive than, those in Bosnia and Herzegovina, the design of Kosovo’s self-govern-
ment institutions is less complex than there (Figure 1). Apart from the SRSG, a
two-layered system provides the backbone for the exercising of public author-
ity in Kosovo. The Kosovo central authority has a wide range of competences
in almost all sectors of public policy. These include: economic, financial, and
budgetary policy; customs; trade, industry, and investment; education, science,
and technology; youth, sport, and culture; health and family policy; agriculture;
environment; tourism; labour; social welfare; transport; media; communication;
statistics; spatial planning; good governance; non-resident affairs; local adminis-
tration; and judicial affairs. In coordination with the SRSG, the Kosovo Central
Authority also has certain competences in the area of external affairs.
Figure 6 Vertical Layers of Public Authority in Kosovo
Special Representative of the United Nations Secretary-General

Municipal Administrator
Kosovo Central Assembly President of Judicial

Authority Kosovo System
Government of
Kosovo
Independent Ombudsperson
Bodies and
Offices
Municipal Institutions Municipal As- President and Judicial

sembly Deputy Presidents System
of Municipality
Chief Executive
Officer and Board
of Directors of
Municipality
The directly-elected Assembly of Kosovo is, among other things, responsible for
the adoption of laws within the specified remit of Kosovo’s authority, for electing
the president of Kosovo, and for approving the prime minister, as well as for en-
dorsing any international agreements that fall within the remit of its competence.
The executive branch of government in Kosovo is comprised of the president,
who conducts Kosovo’s foreign policy in coordination with the SRSG, and the
government of Kosovo, which exercises executive authority and is charged with
implementing any laws passed by the Assembly. There is also a fully-developed
court system, which is made up of the Supreme Court of Kosovo, District Courts,
Municipal Courts, and Minor Offence Courts. In addition to these ‘traditional’
branches of government, Kosovo’s institutional structure also includes a set of
independent bodies and offices, such as the Central Election Commission, the
Judicial and Prosecutorial Council, the Auditor-General, the Banking and Pay-
ments Authority, the Media Commission, the Board of Public Broadcasters, and
the Housing and Property Directorate/Claims Commission. A specially-appoint-
ed Ombudsperson is charged with receiving and investigating complaints about
human rights violations and abuse of authority by any public body and thus pro-
vides an additional form of checks and balances at the central level of Kosovo.
The third layer of public authority in Kosovo are the municipalities which have
powers in all areas of local administration that are not expressly reserved for the
Kosovo Central Authority. Horizontally, power is divided between a municipal
assembly, an executive branch that comprises a president and deputy president
426 Stefan Wolff
of the municipality (who perform general oversight functions) and a Chief Ex-
ecutive Officer and Board of Directors (who implement all municipal decisions),
and a judicial system that consists of Municipal Courts and the so-called Minor
Offence Courts. The Municipal Assembly has competences in budgetary and fi-
nancial matters; is charged with the election of president and deputy president,
the appointment of officers and the establishment of committees; and is to re-
solve any disputes within the executive branch between chief executive officer
and president.
As a quasi-protectorate of the United Nations, Kosovo shares with a number
of the other cases studied the fact that its final status remains to be determined.
However, one crucial difference between Kosovo (and, for that matter, South Os-
setia), on the one hand, and Bougainville, Gagauzia, and Northern Ireland, on
the other, is that the latter three are institutionally integrated into larger state
structures for the period until a change in their political status may occur, while
Kosovo (and South Ossetia) are not. From an institutional perspective, this mere-
ly means that at some point in the future there may have to be changes to exist-
ing structures, but these are primarily technical matters. What is a potentially
more difficult issue therefore is not that Kosovo and South Ossetian institutions
may have to be reintegrated into a wider institutional framework of the Union
of Serbia and Montenegro, and Georgia, respectively, but that this process of
reintegration and the institutions emerging from it need to be recognized by the
relevant political agents in these now quasi-independent entities and need to be
technically viable at and between all levels of government.
G Macedonia
Although Macedonia’s independence from Yugoslavia came about peacefully, the
country has experienced serious ethnic tensions, in particular between ethnic
Macedonians and ethnic Albanians, but also between these two groups and the
country’s sizable Roma minority. While the latter tensions were relatively minor,
at least compared to other countries in Central and Eastern Europe with large
Roma populations, they have increased since the conclusion of the Ohrid Agree-
ment in 2001, which is seen by many Albanian leaders as threatening to establish
a bi-national state.
Albanians in Macedonia live territorially concentrated in the west of the state.
Upon Macedonian independence they organized an unofficial referendum, which,
at a turnout of 90% of the ethnic Albanian electorate in Macedonia, showed that
roughly three quarters supported the idea of their own political and territorial
autonomous structures. On this basis, ethnic Albanian parties argued for chang-
es in Macedonia’s constitution to elevate the ethnic Albanian population to the
status of a ‘constituent people’ of Macedonia, for improvements in the Albanian
language situation, the establishment of an Albanian university, and the inclusion
of ethnic Albanians in the administration. These tensions simmered for most of
I am grateful to Eben Friedman for pointing this out to me.

the 1990s at a level below the threshold of violence, but escalated in the after-
math of NATO’s intervention in the Kosovo conflict in neighbouring Yugoslavia.
The emergence of the ethnic-Albanian National Liberation Army significantly in-
creased the stakes in the latent conflict, which duly erupted into a short, but rela-
tively intense violent confrontation in 2001. Intervention by the European Union
and the facilitation of negotiations between ethnic Albanian representatives and
the Macedonian government resulted in an agreement on constitutional and ad-
ministrative changes to the structure of the Macedonian state, aimed at providing
greater autonomy to all local communities.
In addition to this internal dimension, the complex nature of the relationships
between Macedonia, on the one hand, and Albania, Bulgaria, and Greece, with
their various territorial, ethnic, and political claims, on the other, has compli-
cated the dispute between ethnic Albanians and ethnic Macedonians within the
country. At several stages, there appeared a very clear and imminent danger that
Macedonia would turn into a source of grave instability in an already volatile
region. The acceptability of the settlement reached with the 2001 Ohrid Agree-
ment to both the conflict parties in Macedonia and the relevant external agents
(EU, US, neighbouring states) was therefore as important as its technical viability,
i.e., the capacity of the new institutional structures thus created to live up to the
expectations that the two conflict parties in particular had.
According to the Ohrid Agreement, Macedonia retains its two-layered system
of authority. The powers between the two levels – the national government and
the municipalities – are now more clearly divided and the municipalities enjoy a
substantive degree of autonomy within this system (Figure 7). The national gov-
ernment – comprised of a unicameral assembly, an executive with a president
and cabinet government, and a judicial branch with a constitutional court and
lower-order courts, as well as a public attorney with decentralized offices at the
local level – is the residual source of all public authority in the country. At the
local level, 124 municipalities and the capital city of Skopje have enhanced lo-
cal self-administration powers in the areas of public services, culture, education,
social welfare, health care, environment, urban and rural planning, economic de-
velopment, and local finance. Municipal institutions comprise a council, whose
competences include the budget and other financial matters, the establishment
and control of public services, institutions and enterprises, and the establishment
and supervision of governing and administrative organs at the municipal level.
The municipal executive is made up of an elected mayor, who is responsible for
the appointment and dismissal of all officers of the governing and administrative
organs and services in the municipality and for their overall management. The
governing and administrative organs and services draft and implement individual
acts and supervise activities in their areas of competence.
428 Stefan Wolff
Figure 7 Vertical Layers of Public Authority in Macedonia
National Government Assembly President of Judicial

Macedonia System
Government of
Macedonia
Municipal Institutions Council Mayor Judicial

Governing and System
Administrative
Organs
Neighbourhood Self-
government
In addition to these two layers of public authority that exist throughout Mace-
donia, citizens have the opportunity to establish so-called ‘neighbourhood’ self-
governments within the municipalities in which they live. The precise nature of
their jurisdiction and organization depends on the by-laws of the respective mu-
nicipality, and thus leaves significant room to address specific local concerns in
ways that are felt most appropriate by those immediately concerned.
H South Ossetia
Of all the case studies reviewed here, South Ossetia is the one where formal
structures of vertically layered authority are the least developed. This is largely
due to the fact that no final settlement of the conflict has been agreed upon by
the main parties to the dispute – Georgia, South Ossetia, and Russia – and the
OSCE, which acts as the principal mediator.
The conflict in South Ossetia is best analyzed against the background of the
dissolution of the Soviet Union and the subsequent problems of state-building in
Georgia. A politically and economically disadvantaged part of the Georgian Soviet
Socialist Republic during the Soviet era, political leaders in South Ossetia sought
to use Georgian independence as a catalyst to improve the status of their region
within Georgia. Already confronted with one major threat to the emerging coun-
try’s territorial integrity in Abkhazia, the Georgian state was intent on strength-
ening the central government and forging a Georgian national identity along the
lines of an ethnically Georgian culture. Unsurprisingly, rhetoric escalated on both
sides into tit-for-tat discrimination and eventually violent confrontation. Initially
sporadic clashes developed into serious military confrontation, accompanied by
an increase in the stakes on both sides – South Ossetians began to demand repub-
lic status (implying secession), Georgia abolished the existing autonomy of South
Ossetia. A further two years of violence eventually subsided into a stalemate, but
not before around 1,000 people had been killed and some 60,000 been displaced
within South Ossetia and Georgia and across the border to North Ossetia. Kin-
ship links between South Ossetians and North Ossetians, who live in the Russian
Federation, and Russia’s own geopolitical interest in maintaining influence in the
Caucasus region after the end of the Soviet empire, added another dimension to
the conflict that initially proved to be of an escalating rather than calming nature.
Russia’s predominance in the region and concerns with regional instability due to
the conflict in Chechnya and the tensions in Dagestan were recognized by other
agents in the international community, which meant that, while the conflict was
noticed, no formal intervention took place during its violent phase.
Subsequently, however, international mediation efforts and development pro-
grammes have helped establish something akin to institutional structures that al-
low for a minimum of coordination between the conflict parties and for concrete
approaches to the pressing social, political, economic, and military problems of
the region. Probably most important of these bodies is the so-called Joint Control
Commission (JCC) and its various organs, including a Working Group on Mili-
tary and Security Issues and a Joint Law Enforcement Coordination Centre. The
JCC and its members (Georgia, South Ossetia, Russia, and the OSCE) provide the
only permanent administrative institution to coordinate issues of common inter-
est (trade, transportation, organized crime, etc.) among the conflict parties. The
Joint Peace-Keeping Force (Russia, Georgia, South Ossetia) established under the
1992 ceasefire agreement is a partner along with Georgia, South Ossetia, and the
OSCE in the Working Group on Military and Security Issues. In addition, there
are bilateral institutions, such as the Georgian-Russian Intergovernmental Body,
also created under the JCC. Muddled and informal as these structures may be,
they have been fairly successful in providing a minimum of regular coordination
and joint decision-making among the conflict parties, and most crucially they
have helped maintain the 1992 ceasefire for more than ten years. However, be-
cause of the nature of the arrangements on the ground, they cannot be classified
in any sense as a manifestation of layered authority. Rather, in the absence of a
comprehensive settlement, Georgia and South Ossetia are both quasi-sovereign
in the territories under their control, bearing in mind the fragility, weakness, and
often informality of institutional structures in both entities.
III The Vertical Layering of Authority and Its Relevance to Complex

Power-sharing Arrangements
The brief overview of institutions and institutional structures given above illus-
trates the broad range of institutional designs available to state builders in address-
ing the specific conditions of particular conflict situations, the varied interests of
conflict parties and other external agents connected with the conflict, individual
parties involved, or other agents with political, economic, or other interests in the
region. This is a true, albeit trivial, observation. However, what is more important
for conflict resolution is the examination of the nature of commonalities between
these cases, analyzing the impact of particularities on specific institutional designs
encountered, and to study recurring problems, and how, and with what degree of
430 Stefan Wolff
success, they have been addressed in different instances. The remaining two sec-
tions of this chapter will therefore first look at a number of general structural as-
pects of institutional design and then (to the extent that this is possible given the
often recent conclusion of some of the agreements) draw some conclusions as to
the role that the vertical layering of authority has in the power-sharing ‘toolkit’.
A Types of Institutional Structures

The first element to consider in this comparative analysis of how vertical author-
ity is layered in complex power-sharing systems is the number of layers of author-
ity that actually exist across the eight case studies (Table 1).
Table 1 Variation in the Vertical Layering of Authority

Two-layered Structures Three-layered Structures Multi-layered Structures
Macedonia Bougainvillei Bosnia and Herzegovina
Gagauzia Mindanao
Kosovo
Northern Irelandii
i This ignores the fact that, from a purely technical point of view, the monarch of the
United Kingdom of Great Britain and Northern Ireland is the official, albeit largely
ceremonial, head of state and is represented in Papua New Guinea by a Governor
General.
ii This ignores the fact that the monarch of the United Kingdom of Great Britain and
Northern Ireland is the official, albeit largely ceremonial, head of state.
Table 1 illustrates the predominance of three-layered structures of government in
more than half of the cases (bearing in mind that South Ossetia cannot be prop-
erly categorized as a result of the absence of proper formal institutional struc-
tures). In the cases of Bougainville, Gagauzia, and Northern Ireland, these three
layers are central, regional, and local government. In the case of Kosovo, central
government functions are presently exercised by the Special Representative of
the Secretary General, who derives his or her authority from UN Security Coun-
cil Resolution 1244 (1999). Unless Kosovo is granted independence in a future
final status settlement, the Special Representative will be replaced at some point,
and with significantly reduced powers, by the national government in Belgrade,
thus preserving the three-layered structure of authority.
In Macedonia, the middle level of government – the region – is missing, re-
ducing the levels of government to two, namely a central government and local
governments, which are both prescribed in the constitution and whose functions
and powers are detailed there and in relevant legislation. There also exists a legal-
ly guaranteed opportunity for citizens to develop a further layer of government
at the level of neighbourhoods, but this is regulated by by-laws of the individual
local governments and thus a matter of local decision-making rather than of state
construction.
In the cases of Mindanao and Bosnia and Herzegovina, multiple levels of
government exist. In Bosnia and Herzegovina, this is a result of the interplay of
domestic, regional, and international factors in the process of state creation at
Dayton, leading to a complex federal-confederal structure of the state. In the case
of Mindanao, an existing four-layered structure of government was altered with
the creation of a specific and unique fifth layer – the legal-political entity of the
Autonomous Region of Muslim Mindanao – to which powers were devolved in
an effort to resolve the underlying self-determination conflict.
Another way of looking at structural types of vertically layered authority is
to examine the degree to which the eight cases represent institutions that are
structurally and/or functionally symmetric or asymmetric (Tables 2–4), as this
perspective provides a more comprehensive picture of the structure of the entire
polity concerned and the place and status of complex power-sharing institutions
within it.
Table 2 Structural Symmetry and Asymmetry of Institutions

Structural Symmetry Structural Asymmetry
Single Asymmetry Multiple Asymmetry
Macedonia Gagauzia Bosnia and Herzegovina
Bougainville Mindanao Northern Ireland
Kosovo
Table 3 Functional Symmetry and Asymmetry of Institutions

Functional Symmetry Functional Asymmetry
Kosovo Bosnia and Herzegovina
Macedonia Bougainville
Gagauzia
Mindanao
Northern Ireland
432 Stefan Wolff
Table 4 Structural and Functional Symmetry and Asymmetry

of Institutions Compared
Structures Functions
Symmetric Single Multiple Symmetric Asymmetric
asymmetric asymmetric
Bosnia and Herze-
X X
govina
Bougainville X X
Gagauzia X X
Kosovo X X
Macedonia X X
Mindanao X X
Northern Ireland X X
Tables 2 and 3 indicate that there is no clear-cut predominance of symmetric or

asymmetric forms of institutional structures across the case studies (again, leav-
ing aside South Ossetia), but that from a functional perspective, i.e., the way in
which powers and functions are distributed horizontally at the relevant levels
of government in a polity, asymmetry is more frequent than symmetry. In other
words, the vertical layering of authority, regardless of whether it is structurally
‘coherent’ across a given state or not, facilitates asymmetric distribution of powers
and functions, thus enabling central governments and specific regions to create a
special relationship in the sense that more powers and functions or parts thereof
are devolved to a particular region, which thereby acquires greater autonomy in
a wider range of policy areas compared to other territorial entities in the same
country. This is also demonstrated in Table 4, which illustrates that, while sym-
metric structures and symmetric functions correlate more frequently (Kosovo,
Macedonia), symmetric structures do not preclude asymmetric functional capaci-
ties (Bougainville).
B The Combination of Vertical and Horizontal Power-sharing

One element of the complexity of power-sharing as a mechanism to resolve self-
determination conflicts stems from the fact that constitutional engineers have
developed innovative ways to combine traditional structures of horizontal and
vertical power-sharing. While all the cases examined in this chapter are examples
of state structures characterized by multiple vertical layers of authority, formal
horizontal structures of power-sharing exist only in some of them, and where
they do exist they always involve power-sharing at the regional level (Table 5).
Table 5 Combinations of Horizontal and Vertical Power-sharing

No horizontal Horizontal power- Horizontal power- Horizontal power-
power-sharing sharing at national sharing at regional sharing at national
level only level only and regional level
Kosovo Macedoniaiii Northern Ireland Bosnia and Herze-
South Ossetia Moldovaiv govinav
Bougainvillevi
Mindanaovii
iii Even though there is no mandatory power-sharing at any level in Macedonia, the
power balance of national politics makes coalitions at the centre between ethnic
Macedonian and ethnic Albanian parties highly likely, and Macedonia has been gov-
erned by such coalitions for several years.
iv To the extent that certain members of the Gagauz government are co-opted into
structures of the national government, there is a certain degree of power-sharing at
the national level.
v Mandatory power-sharing at regional level only applies to the federation.
vi A regional constitution is yet to determine the details of horizontal power-sharing in
Bougainville.
vii To the extent that certain members of the government of the Autonomous Region
of Muslim Mindanao are co-opted into structures of the national government, there
is a certain degree of power-sharing at the national level as well as the mandatory
power-sharing at regional level.
As the cases of Macedonia, Mindanao and Moldova demonstrate, the absence of

formal structures of power-sharing at the national level does not preclude power
nevertheless being shared to some extent. In Macedonia, this is more obvious,
as the country’s demographic balances, structure of the party system, and elec-
toral formula combine in a way that make the formation of government coalitions
between ethnic Macedonian and ethnic Albanian parties likely (and they have
been a reality since 1992). In Mindanao and Moldova, on the other hand, there
is a somewhat greater degree of formality in power-sharing arrangements at the
centre, as members of the regional governments are co-opted into respective
branches of the national government. Co-optation, however, limits the extent of
the influence that can be exercised by the region at the centre, as in both cases re-
gional co-optees are outnumbered by other members of the national government
and have little, if any, leverage compared to situations in which a regional party is
a member of a governing coalition.
Formal horizontal power-sharing at the regional level exists in all those cases
where there is significant ethnic or other diversity within the region, i.e., where
mere devolution of powers to a lower level of authority would simply replicate the
conflict at the national level. This is clearly the case in Bosnia (Federation level),
Mindanao, and Northern Ireland. The absence of regional horizontal power-shar-
434 Stefan Wolff
ing in Kosovo, Macedonia, and Moldova has different causes. In Kosovo, apart
from the fact that no final status has been agreed yet, the only numerically signifi-
cant minority group – ethnic Serbs – is relatively concentrated in the northern
parts of Kosovo, and thus benefits directly from the significant degree of author-
ity located at the municipal level. Additionally, reserved seats in the Kosovo As-
sembly for Serbs and members of other minorities ensures their presence and
influence in the legislature. In Macedonia, the territorial concentration of ethnic
Albanians, the range of powers devolved to the municipal level and the oppor-
tunity for citizens to establish a further layer of authority at the neighbourhood
level addresses a wide range of self-government concerns among ethnic Alba-
nians. In addition, the numerical strength of ethnic Albanians in the Macedonian
polity and the structure of its party and electoral systems guarantee significant
representation of ethnic Albanian parties in the national parliament and makes
their participation in a coalition government at least highly likely. This strength
of Albanians that allows them to benefit fully from the implementation of local
autonomy as foreseen in the Ohrid Agreement (once relevant national legislation
has been passed), is another explanation for the absence of horizontal power-
sharing: the geographical concentration and size of the minority make a federal
solution less attractive for ethnic Macedonians, as it could be construed as a first
step to the partition of the country. In Moldova, the relative ethnic homogene-
ity of Gagauzia, the ability of residents in districts to determine by referendum
whether they want to be part of the autonomous territory, and the fact that lo-
cal affairs in these districts are run locally are all meant to combine to provide
sufficient autonomy for individuals and communities to make formal regional
power-sharing unnecessary. This means that under certain conditions – rela-
tive territorial concentration of ethnic communities, sufficient levels of devolu-
tion, and a minimum degree of representation at the centre – vertical layering of
authority can function as a useful substitute for formal structures of horizontal
power-sharing, both at national and regional level, and suffice in addressing insti-
tutional dimensions of power (re)distribution in self-determination conflicts. The
fact that vertically layered authority can only substitute for horizontal levels of
power-sharing under very specific conditions is also highlighted by the example
of Bosnia and Herzegovina, where, despite wide-ranging devolution, horizontal
power-sharing remains mandatory at the level of state institutions and at the level
of the Bosnian-Croat Federation.
C Distribution of Powers
One of the key questions to ask of any vertically layered system of authority is
where powers rest; i.e., how different competences are allocated to different
layers of authority and whether they are their exclusive domain or have to be
shared between different layers of authority. As with other dimensions in this
analysis, there is a certain degree of context-dependent variation across the eight
cases of complex power-sharing under examination. Variation exists primarily
with regard to the way in which powers are allocated and the degree of flexibil-
ity concerning new fields of policy-making not relevant or not included at the
time a specific agreement was concluded. The principle mechanism to handle the
distribution of powers is the drawing up of lists that enumerate precisely which
powers are allocated to which levels of authority and/or which are to be shared
between different such levels. These lists can be very specific for each layer of
authority (Bougainville, Mindanao, Northern Ireland) or they can be specific for
one or more layers and ‘open-ended’ for others (Bosnia and Herzegovina, Gagau-
zia/Moldova, Kosovo, Macedonia). The key difference in the latter case is which
layer of authority has an ‘open-ended’ list, i.e., which layer holds original author-
ity competence for any partly devolved power or any other policy area not explic-
itly allocated elsewhere (Table 6). South Ossetia, given the particularities of the
situation there, cannot yet be categorized in any of these mechanisms of power
distribution.
Table 6 Power Allocation in Vertically Layered Systems of Public Authority

Specific Lists Combination of Specific and ‘Open-ended’ Lists
Open-ended list at centre Specific list at centre
Bougainville Moldova Bosnia and Herzegovina
Mindanao Macedonia Kosovo
Northern Irelandviii
viii In case the Assembly in Northern Ireland asks for it, the regional power-sharing
institutions could enjoy an open-ended list of powers allocated to them, with only
specifically excepted matters retained by the Westminster government.
In Gagauzia/Moldova and Macedonia, the national level holds original authority

over all matters not expressly devolved to the lower layers of authority, while in
Bosnia and Herzegovina the two entities retain all the competences not explicitly
delegated to the level of state institutions (with the qualification that in the Fed-
eration cantonal institutions assume most of these powers from the Federation
entity). Similarly, in Kosovo, the current system is more ‘open-ended’ with regard
to the municipal institutions that retain powers in all areas of local administra-
tion not expressly reserved to the Central Authority.
In the Philippines/Mindanao, the multi-layered system of public authority that
is in place there has very specific lists of powers allocated to the individual levels
within it, even though the central government remains the original source of all
authority. This is also the case in Northern Ireland, but here the system of allo-
cating powers operates on the basis of three different lists enumerating devolved,
reserved (with the future possibility of devolution), and excepted (without the
future possibility of devolution) matters. In Bougainville/Papua New Guinea,
which also operates a system of specific power allocation to the different layers
of public authority, an additional feature is that there are specific arrangements
as to how to deal with emerging policy areas (a joint commission that will resolve
436 Stefan Wolff
disputes over the allocation of new powers). Another distinctive feature of the
Bougainvillean system is that initially all powers allocated to the autonomous
province are retained at the central level and are, albeit almost automatically,
devolved to Bougainville upon application to the central authorities by the pro-
vincial authorities.
An unambiguous distribution of powers should indicate that the most impor-
tant issues of the underlying self-determination conflict have either been resolved
or postponed (e.g., future referendum on independence) and that the potential
for conflict re-erupting should be minimal and limited to disputes over emerging
new policy areas not covered by the provisions of the original agreement be-
tween the conflict parties (provided that the institutions established discharge
their functions properly). Where such an unambiguous allocation of powers is
missing in the sense that one layer of authority automatically retains all powers
not explicitly allocated elsewhere (and thus implicitly also the competence over
all emerging new policy areas in the future), renewed conflict over the distribu-
tion of power between different layers of authority is more likely, even though
there is no automatism in this. In cases where the central authority retains all the
powers not expressly devolved, autonomous areas may, over time, seek renego-
tiation of past agreements or the allocation of additional powers. In the reverse
case, central authorities may be continuously weakened, potentially leading to
the break-up of the central state. This is obviously not on the agenda in the near
future for any of the three cases where no such unambiguous allocation of pow-
ers has taken place (Bougainville, Mindanao, Northern Ireland), but it is an issue
worthy of consideration in the construction of states within complex power-shar-
ing institutional frameworks.
D Types of Coordination
Coordination of law and policy-making and their implementation is an impor-
tant issue in the operation of any multi-layered system of government. In the con-
text of self-determination conflicts and power-sharing institutions, it assumes
additional significance, as coordination failures not only have an impact on the
effectiveness of government but also have repercussions for the perception of a
particular institutional structure designed to resolve a self-determination con-
flict. The eight cases studied in this analysis suggest that, although there is a wide
spectrum of individual coordination mechanisms, these can be grouped into four
distinct categories: co-optation, joint committees and implementation bodies,
judicial review and arbitration processes, and direct intervention by the interna-
tional community (Table 7).
Table 7 Coordination Mechanisms between Different Layers of Public

Authority in Complex Power-sharing Systems
Co-optation Joint Committees and Judicial Review and Direct Intervention
Implementation Bod- Arbitration by the International
ies Community
(including ad hoc
bodies sponsored by
international organi-
zations)
Mindanao Bougainville Bosnia and Herze- Bosnia and Herze-
Gagauzia Macedonia govina govina
Mindanao Bougainville Kosovo
Gagauzia Kosovo
Northern Ireland Macedonia
South Ossetia Mindanao
Gagauzia
Northern Ireland
As Table 7 indicates, with the exception of South Ossetia, all the case studies ex-
hibit at least two different coordination mechanisms, with one of them always be-
ing judicial review and arbitration processes. This suggests that there is a strong
reliance upon the legal regulation of the relationships between different layers of
public authority. This is similar to any other country that has adopted the rule of
law as a basic principle of running its own affairs. It is therefore more interesting
to consider the other three types of coordination mechanisms in greater detail
with a view to examining the degree to which they are the specific results of
adopting complex power-sharing institutions as settlements for self-determina-
tion conflicts. Co-optation, adopted in the Philippines and Moldova, is a mech-
anism to ensure the representation of regional officials (from the ARMM and
Gagauzia, respectively) at the centre. In both cases, the regional governors and
officials of their executive are ex officio members of relevant national government
departments. This arrangement is, on the one hand, symbolic, and emphasizes
the special relationship between central government and autonomous region,
but, on the other hand, also necessary as in both cases the autonomous entities
are artificial constructions from an administrative point of view and do not fit
into the pre-existing structures of authority in either country. Co-optation thus
becomes a mechanism to overcome this kind of administrative ‘abnormality’ and
ensure that the special circumstances of the autonomous regions are borne in
mind in the process of national law and policy-making.
In the context of coordination between different vertical layers of authority in
complex power-sharing structures, the need for joint committees and implemen-
tation bodies often arises from two sources – to find common interpretations
for specific aspects of agreements and regulations and to coordinate the imple-
mentation of specific policies at national and regional levels. An example of the
former is Bougainville, while the latter can be found in Macedonia (interethnic
438 Stefan Wolff
relations), Philippines/Mindanao (development), Moldova/Gagauzia (fiscal and

budgetary policy, property legislation), Northern Ireland (cooperation between
Northern Ireland and the Republic of Ireland, and among all entities party to
the British-Irish Council), and South Ossetia (transport, crime prevention, Geor-
gian-Russian cooperation). Such bodies can hold one-off (Moldova) or regular
meetings (Papua New Guinea/Bougainville, Macedonia, Philippines/Mindanao,
Northern Ireland); and they can be in their nature domestic, centre-periphery
bodies (Papua New Guinea/Bougainville, Macedonia, Philippines/Mindanao,
South Ossetia), or reflect the international dimension of a particular self-deter-
mination conflict (Northern Ireland, South Ossetia). They may be prescribed
in agreements between the conflict parties (Papua New Guinea/Bougainville,
Mindanao, Northern Ireland) or arise from actual needs (Macedonia, Moldova/
Gagauzia, South Ossetia).
Unique to two case studies – Kosovo and Bosnia and Herzegovina – is the di-
rect intervention of the international community as a mechanism to coordinate
law and policy-making. In both cases, powerful international officials retain sig-
nificant powers, enabling them to intervene directly into the political processes
of the two entities. This results primarily from the unprecedented involvement of
the international community in the process of resolving the two underlying self-
determination conflicts and the responsibility that international agents thereby
assumed for post-conflict state construction, as well as from the particularly bit-
ter nature of the disputes concerned.
E Entrenchment of Institutional Structures

Guarantees of institutional structures of horizontal and vertical power-sharing
are essential to prevent the arbitrary abrogation of devolved powers and thus to
ensure conflict parties of the relative permanence of the institutions they agreed
upon. Guarantees are particularly important for the relatively weaker party in
a self-determination dispute, i.e., a specific minority, to protect it from a state
reneging on earlier concessions. However, such guarantees are also valuable for
states in that they commit all parties to an agreed structure and, in most cases,
imply that there can be no unilateral change of recognized international bound-
aries.
In principle, guarantees can be either international or domestic, and in the lat-
ter case they can be part of a country’s constitution or other legislation (Table 8).
Given the complexity of many of today’s self-determination conflicts, guarantees
often exist at more than one level. In addition, international guarantees can take
the form of hard guarantees (international treaties) or of ‘soft’ guarantees (non-
binding standards and norms, declarations of intent, etc.).
Table 8 Guarantees of Power-sharing Institutions

International Guarantees Domestic Guarantees
‘Hard’ ‘Soft’ Constitutional Guarantees in Spe-
Guarantees cific Laws
Bosnia and Herze- Bougainville Bosnia and Herze- Bougainville
govina Macedonia govina Gagauzia
Kosovo Gagauzia Bougainville Macedonia
Northern Ireland Mindanao Gagauzia Mindanao
South Ossetia Kosovo Northern Ireland
Macedonia
Table 8 illustrates that there is great variance across the eight cases considered
here, and with only one exception (South Ossetia), guarantees exist at multiple
levels. In terms of the strength of the protection that they afford established hori-
zontal and vertical power-sharing structures, hard international guarantees are
preferable over other forms of guarantees, provided there is significant commit-
ment of the international community to uphold its guarantees. In Kosovo and
Bosnia and Herzegovina, this commitment is unquestionable, with the presence
of peacekeeping forces in both territories and with the investment that has been
made over the past years by the international community in order to foster eco-
nomic development, institution-building, and institutional reform. Whereas in
Bosnia and Herzegovina and Kosovo there exist international bodies with a clear
mandate (the multinational Peace Implementation Council and the UN Security
Council, respectively), the situation in Northern Ireland is such that the hard
international guarantee of the 1998 agreement exists in the form of a British-Irish
treaty. The crucial difference here is that for any violation of the treaty (as has ar-
guably occurred on several occasions with the unilateral suspension of the power-
sharing institutions by the UK government) to be addressed, one of the signatory
parties needs to bring a case before a relevant international legal institution (e.g.,
the European Court of Justice). If this does not happen, the protection theoreti-
cally afforded by the link between the agreement and an international bilateral
treaty remains an empty shell.
Soft international guarantees primarily manifest themselves in the form of the
involvement of international organizations in the negotiation, implementation,
and (potentially) operation of a particular peace agreement. While not of the
same legally binding and thus potentially enforceable status as hard international
guarantees, a significant presence of international agents is often instrumental
in shaping preference and opportunity structures for the conflict parties. In the
cases studies, this has taken different forms. In Bosnia and Herzegovina, similar
to Kosovo and Macedonia, an international troop presence, as well as the in-
volvement of various international governmental and nongovernmental organi-
zations on an unprecedented scale, have, for better or worse, been instrumental
in the implementation and operation of the respective agreements thus far. In
Bougainville, a UN Observer Mission has been crucial in facilitating demilitariza-
440 Stefan Wolff
tion; while in Gagauzia, the OSCE has played an important role in facilitating the
coordination of policies and laws between regional and national government. In
South Ossetia, the commitment of resources and personnel by UNHCR, UNDP,
the EU, and the OSCE have contributed to the maintenance of the 1992 ceasefire
agreement and the ensuing cooperation between the various parties to the con-
flict on a number of substantive issues.
At the level of domestic guarantees, constitutional guarantees are more en-
trenched than those that have their source in normal legislation. Incorporation of
specific provisions of peace agreements into national constitutions is a common
way of realizing constitutional guarantees and has occurred in Bosnia and Herze-
govina, Bougainville, Moldova, Macedonia, and, bearing in mind its provisional
status, in Kosovo. In the case of Bougainville, an additional safeguard exists in
that no changes to the agreed and constitutionally entrenched structure of the in-
stitutions created by the peace agreement is permissible except with the explicit
consent of at least two thirds of the members of the Bougainville parliament.
Guarantees through specific laws exist in the cases of Bougainville, Gagau-
zia, Macedonia, Mindanao, and Northern Ireland. In practice, they have proven
weakest in Northern Ireland, where, in the absence of a written constitution, an-
other law on the statute books has given the UK government the power to sus-
pend the power-sharing institutions at any given time.
F Territory and Population as Boundaries of Authority

Authority as the legitimate exercise of political power has two boundaries – it is
normally limited to a specific territory and/or a defined group of people. A national
government has the authority to exercise its power within the territorial confines
of the state it is governing and over the residents of this territory (with the excep-
tion of foreign diplomats, for example). Some elements of a national government’s
authority may also extend beyond the territorial boundaries of its state, but then
they will normally be limited to that particular state’s citizens, for example in the
field of tax collection. In terms of the territorial layering of authority, the extent
of these two limitations placed on the exercise of authority is similar. Regional,
territorial autonomies are spatially confined. The powers devolved to a regional
government only apply within the territorial boundaries of the region and, by ex-
tension, only to (permanent) residents of the region. An analogue to authority
extending beyond territorial boundaries are instances of personal autonomy in
which the autonomous body has authority over all individuals belonging to it no
matter where they live in the territory of the state or region concerned.
In reverse, this means that all members of the ethnic group concerned can enjoy
the rights accorded to them in the autonomy arrangement anywhere in the terri-
tory of the relevant state. This form of autonomy is particularly useful in instances
where groups are more dispersed. It is also used to complement territorial forms of
autonomy in specific policy areas (culture, religious affairs, education, etc.) when
autonomous territories are ethnically heterogeneous.
These observations are relevant in two of the cases examined here. The ter-
ritories of Gagauzia and of the Autonomous Region of Muslim Mindanao were
determined by referendum at the local level, giving the population an opportu-
nity to express in a free vote whether they want to live under the authority of a
newly created regional government or want to continue being governed within
the existing structure of vertically layered institutions. In Mindanao, this vote
took place at the level of provinces and towns, and in Moldova at that of local
communities, thus allowing for a much more ‘precise’ gauging of popular will.
In both cases, the result was that the autonomous territory thus created is not in
fact a contiguous area, but is made up of a number of patches of territory. Early
indications suggest that this is not necessarily detrimental to the exercise of au-
thority at the level of the autonomous territory. A degree of personal autonomy
exists in the Autonomous Region of Muslim Mindanao with regard to judicial af-
fairs, as Shari’ah and tribal courts have authority alongside lower-order courts of
the national judicial system in religious and family affairs to cater for the specific
needs of the different religious, ethnic, and tribal communities in these areas.
This suggests that there is an additional degree of differentiation available that
goes beyond the traditional territorial delimitation of authority in that it incorpo-
rates a public consultation process for the definition of the territorial boundaries
of the autonomous area. If combined with levels of personal autonomy in specific
policy areas, the range of authority that an autonomous entity enjoys can be tai-
lored to the specific demographic and geographic situation, taking account of
settlement patterns and ethnic, religious, cultural, and other types of heterogene-
ity. While such ‘fine-tuning’ increases the complexity of vertical and horizontal
power-sharing mechanisms, it may also make them more suitable to particular
contexts and thus more acceptable. In other words, careful territorial and per-
sonal delimitation of autonomy potentially increases the belief in the authority of
the institutions established among those governed by them and is thus likely to
contribute to greater stability of these same institutions and the political process
of which they are part. However, as I have previously indicated, adding a further
layer of authority to those already existing within the structure of an established
state increases the complexity of institutional design, places greater demands on
policy coordination and has the potential to undermine the authority of the ter-
ritorial entity created specifically to increase the autonomy of a population group
seeking a higher degree of self-governance.
However, what is striking about the arrangements in both Gagauzia and Min-
danao is the fact that while the relevant local government units can decide in a
referendum on whether they want to belong to the newly created autonomous
entity, there seems to be no provision for the reverse process, i.e., units leaving
the autonomous entity. In case of radical changes in the population balance in
one or more such units, a new minority would be created within the autono-
mous entity (whose demands would have to be accommodated). Demographic
developments always have implications for security perceptions and the stability
442 Stefan Wolff
of settlements of complex self-determination conflicts, but it is reasonable to as-

sume that their implications would be even more severe in cases where territorial
(re)arrangements are recent, precisely because they will imply a degree of fluidity
that is threatening to majorities and minorities at the same time. On the other
hand, given reasonably and rationally acting political elites, there is nothing to
say that significant demographic shifts could not be addressed constructively.
G The Functioning and Stability of Institutions in Vertically Layered

Complex Power-sharing Systems: Design vs. Context
Now that the key structural aspects of the case studies have been compared and
contrasted, I will address three sets of issues that have a bearing on the func-
tioning and stability of institutions within a system of vertically layered author-
ity. Two of these issues – the relationship between vertical and horizontal layers
of power-sharing institutions and the coordination of government activities at
and between different vertical layers of authority – are structural issues and thus
more easily subject to modification. However, the third one – the overall politi-
cal institutional settlement within which vertically and horizontally structured
institutions have to operate – is more of a contextual aspect, which depends on
the interplay of a larger number of factors. Although contexts can also be shaped
by interventions, the complexity of their make-up often means that interventions
have intended and unintended consequences. In some instances, contexts will
have to be changed in order to make any kind of conflict settlement at all possible,
but more often it will be institutional structures that have to be adapted. While
these three sets of issues are clearly inter-related, they are sufficiently distinct
from one another to warrant separate treatment.
In all the case studies, the vertical structures of authority are different in one
of two, and sometimes both, ways: there are differences in the numbers of layers
of authority in the polity concerned, and in the way in which powers are distrib-
uted between them. The specificity of each individual self-determination conflict
partly accounts for the variance encountered, but this does not mean that there
are no general lessons to be drawn.
Inasmuch as any practice has developed yet, the number of layers of authority
seems to be less critical than the question of whether additional layers are super-
imposed on pre-existing ones, as is the case with Mindanao and Gagauzia. In or-
der for these layers of authority to be meaningful, vertical hierarchies have to be
restructured, making lower levels of authority accountable to the newly created
autonomous governments and subjecting them to their political agendas, rather
than those of the centre. This is quite problematic in ethnically heterogeneous
regions, especially where there are no formal or informal structures of power-
sharing at the regional level. This is equally true in cases in which the traditional
vertical structure of the state has been preserved, but the competences of one
entity in it have been significantly increased, leading to a ‘special relationship’
I am grateful to Tom Trier for pointing this out to me.

between this entity and the centre, as is the case in Bougainville, and to some
extent in Northern Ireland. While in both cases (additional layer of authority
and increased competences for a pre-existing layer) the distribution of powers
between centre and region is more or less a technical matter, it raises the issue
of what status this particular region assumes within the overall structure of the
state and how it relates to other layers of authority, both vertically and horizon-
tally. Within such a context of domestic political dynamics, asymmetrical state
construction, either in terms of vertical layers and/or in terms of different com-
petences devolved from the central government to regional authorities, may not
be sustainable, especially if less autonomous regions demand equity in the distri-
bution of powers; or if additional layers of (regional) government created within
pre-existing structures of a state are undermined in their authority because lower
levels of government are unwilling to give up powers previously held, to submit
to a new regional authority, or maintain ‘parallel’ reporting structures within the
traditional vertical hierarchy.
The second set of issues is quite closely related to this last point, namely the
need for adequate coordination of government activity at and between different
layers of authority. Again, the key issue here is to retain institutional legitimacy
and ensure a smooth process of government throughout a given polity. The two
are obviously not mutually exclusive, but what to some may be a matter of effi-
ciency-driven decision-making may appear to others as (and may in fact be) un-
dermining institutions created to resolve a self-determination conflict. The first
aspect to consider here is the powers that central governments (or the equivalent
international institutions in Bosnia and Herzegovina and Kosovo) have to enforce
coordination and compliance of lower-level layers of authority. Suspending the
power-sharing institutions in Northern Ireland on several occasions may have
prevented their collapse but did little to create a sense of responsibility among the
political elites or to assure people in either community of the sustainability of the
peace process. The almost excessive powers that the High Representative in Bos-
nia and Herzegovina and the Special Representative of the Secretary-General in
Kosovo have, and use, creates similar problems in that the very idea of democracy,
i.e., government by elected officials who are accountable to the people who voted
them into office, is distorted. This is not to say that such interventions are not
justifiable or, in fact, necessary to facilitate the growth of genuinely home-grown
democratic institutions and their proper functioning. However, it is important
that such interventions by the international community can be subjected to a
legal review process. Where accountability and transparency are missing, institu-
tional legitimacy within a vertical hierarchy of power-sharing institutions may be
fatally undermined. This potentially includes all those cases where coordination
of government policy extends to far-reaching powers on the part of the central
government (or equivalent) to interfere with policies of lower-level authorities,
as in Northern Ireland, Kosovo, and Bosnia and Herzegovina, because regional
elites are unlikely to develop a full sense of responsibility for sustaining their own
institutions and/or their electorates may become increasingly frustrated and dis-
444 Stefan Wolff
enchanted with the lack of ‘real’ autonomy of the institutions that they accepted
as part of a conflict settlement.
Likewise, the restructuring of the Moldovan polity, which resulted in Gagauzia
becoming an additional layer of authority without equivalent elsewhere in Mol-
dova, may have been a reasonable decision from the perspective of institutional
efficiency, but it is questionable as to whether it has not begun to undermine
Gagauzia as an autonomous region in its own right. The resulting coordination
problems are not unique to Moldova, and Mindanao may face problems similar
to those Gagauzia is encountering. In Macedonia, on the other hand, the fact that
there is a very simple two-layered institutional structure, and hence straightfor-
ward and uniform mechanisms of coordination, bodes well for retaining insti-
tutional legitimacy from this particular point of view. It is, however, important
to bear in mind that this is very much facilitated by the territorial concentration
of the Albanian community in the west of the country and the fact that elec-
toral demographics encourage interethnic power-sharing at the centre. Thus,
the Macedonian experience cannot readily be translated into a recommendation
to transfer the relevant structures and coordination mechanisms to other cases.
Rather, the important point to emphasize is that what is needed is both legitimacy
and technical viability of the institutions established to ensure their sustainability
and their ability to contribute to the settlement of a specific self-determination
conflict.
Finally, there can be no question that the overall political institutional design
of the polity in question and the presence or absence of a final settlement of the
conflict has a bearing on the structure of vertical layers of authority and the coor-
dination between them, as well as their stability. Where institutional hierarchies
are muddled and competences not clearly assigned (as is the case in Moldova
and the Philippines), there is a danger that malign (future) governments will use
the opportunities thus created in order to undermine regional autonomies. This
emphasizes, once again, the need for a clear hierarchy of vertically layered insti-
tutions and the legal entrenchment of their powers.
Equally serious is the lack of a final overall settlement for a particular conflict
(Table 9). Especially where interim structures are set up (as in Kosovo) or develop
out of a conflict stalemate (as in South Ossetia) that give regional authorities sig-
nificant powers, there are serious long-term political implications for achieving
a final settlement, i.e., to construct a viable state out of constituent components
that had a prior existence as ‘quasi-states’. Post-conflict state-building following
state failure or disintegration requires the creation of a power base at the centre
to establish the authority of the state internally as well as externally. This is not to
advocate strong centralized, unitary states, but to point out that, in the context of
self-determination conflicts (i.e., where the structure and boundaries of the state
are contested), the reverse procedure, namely to create strong regional power
bases before a national or central one, often implies renewed conflict or bears the
potential of ultimate state disintegration (just imagine Bosnia and Herzegovina
without the presence and involvement of the international community). Why
would regional elites give up a degree of independence achieved and submit to a
central government, regardless of whether they would share power at the central
level or not? Interim status, and especially the longer it persists and the more
permanent it becomes, is therefore potentially detrimental to conflict resolution
and, by extension, state construction and survival. Similarly, the future option
of a constitutional change by way of referendum may have equally destabilizing
consequences, as such options are interpreted as both a promise and a threat and
cannot but have a polarizing effect on the communities affected by them.
Table 9 Provisions for Final Status Settlement

Final Status Agreed Final Status to be Determined by No Provisions
One-off ref- Referenda at Negotiations
erendum regular inter-
vals
Bosnia and Herzegovina Bougainville Northern Kosovo South Ossetia
Macedonia Gagauzia Ireland
Mindanao
However, one should also note that postponing final status settlement is a le-
gitimate strategy for conflict management. It enables political elites to focus on
crucial issues affecting people’s everyday lives without being side-tracked by as-
pects of the conflict in which finding compromises is exceedingly hard or close
to impossible. In such cases, and Kosovo and South Ossetia would fall into this
category, delaying a final status agreement is vital to establish political processes
that are, by-and-large, stable and free from violence. While this clearly does not
resolve a given conflict, it makes it more manageable and less costly for those
involved.
H Federation, Federacy, or Unitary State

All of the previous comparative issues in this section lead to one final question:
namely, what the relation between the different layers of authority is in the eight
case studies. In addition to unitary states, Elazar (n.d.: 9-10) distinguishes nine
different forms of states with federalist components: confederation, federation,
federacy, associated state, consociation, union, league, joint functional author-
ity, and condominium. Of these, only two are relevant for the discussion here:
federation and federacy. However, not all cases can be clearly categorized (Table
10). For only one is it straightforward: Macedonia is a unitary state. Bosnia and
Herzegovina is the case that most closely resembles a federation, i.e., a “polity
compounded of strong constituent entities and a strong general government each
possessing powers delegated to it by the people and empowered to deal directly
with the citizenry in the exercise of those powers” (Elazar n.d.: 10). Bosnia and
Herzegovina possesses all Elazar’s criteria for a federation, with the sole excep-
tion of having a strong central government. While the weakness of the central
government and the limited powers it has retained would normally point to a
446 Stefan Wolff
confederal arrangement, it fails the confederal test as it, strictly speaking, is not
made up of “pre-existing polities” (Elazar n.d.: 10).
Table 10 Forms of State

Unitary Federation Federacy Undetermined
Bosnia and Herzegovina X
Bougainville X
Gagauzia X
Kosovo X
Macedonia X
Mindanao X
Northern Ireland X
South Ossetia X
Elazar’s (n.d.: 10) definition of a federacy is that “a larger power and a smaller
power are linked asymmetrically in a federal relationship in which the latter has
substantial autonomy and in return has a minimal role in the governance of the
larger power” and that “the relationship between them can be resolved only by
mutual agreement.” This is most clearly the case for Gagauzia and Bougainville,
but arguably, and to a more limited extent, also for Northern Ireland. In the latter
case, the asymmetric link and substantial autonomy are clearly present. As for
the dissolution of the relationship only by mutual agreement, matters are more
complicated. The secession of Northern Ireland from the United Kingdom can
only happen as the result of a referendum in the province (and a corresponding
referendum in the Republic of Ireland), which then will require acceptance by
the government in Westminster. However, it could also be argued that suspend-
ing the autonomous power-sharing institutions in Northern Ireland constitutes a
case of dissolving this special kind of federalist relationship. In this case, no mat-
ter what the reading of the British-Irish Agreement, Northern Ireland’s consent
is not required. As discussed above, the limited and insufficient constitutional
entrenchment of Northern Ireland’s status as an autonomous entity in the United
Kingdom gives it a weaker position than both Bougainville and Gagauzia. Nev-
ertheless, Northern Ireland has a different position from that of a region in a de-
centralized unitary state, in that it has a full system of governing institutions and
original authority in a range of policy fields, neither of which is enjoyed by regions
in a decentralized unitary state. Depending on how the criterion of “a minimal
role in the governance of the larger power” is interpreted, the fact that Northern
Ireland sends eighteen representatives to the House of Commons in Westminster
could be seen as such minimal involvement. This, however, is clearly a lesser form
of participation than in Gagauzia (co-optation) and Bougainville (joint [execu-
tive] committees). Mindanao’s status as a federacy is similarly ambiguous. Un-
like Northern Ireland, Mindanao’s participation in central government is more
formally regulated through co-optation arrangements between the regional and
national government. However, like Northern Ireland and unlike Gagauzia and
Bougainville, Mindanao lacks full constitutional guarantees of its status, but a
comprehensive peace agreement negotiated and signed with international in-
volvement provides some compensation for this. Its status as the Philippines’s
‘fifteenth province’ is regulated in an organic law and endows the area with signif-
icant autonomous powers that are different and go beyond those enjoyed by any
other province in the country. From this perspective, the Autonomous Region of
Muslim Mindanao fulfills all but one of Elazar’s criteria for a federacy.
The yet-undetermined status of Kosovo and South Ossetia stems from the fact
that no final overall legal-political arrangements have been agreed in either of
these cases. The most likely scenario for Kosovo is that of a future status as a
constituent component of a federation of Serbia, Montenegro, and Kosovo. For
South Ossetia, a federacy arrangement with the Georgian central government is
one possibility, as is its participation in a future Georgian federation, with Abkha-
zia as another possible constituent entity in such an arrangement.
IV The Vertical Layering of Authority as Part of the Power-sharing Toolkit

Clearly, there is no single blueprint for the design of a system of vertically layered
authority that could be applied to all self-determination conflicts alike. At the
same time, these case studies highlight that constitutional designers have a wide
range of different options at their disposal for the construction of technically vi-
able institutional structures in response to self-determination conflicts that may
be recognized as legitimate by the conflict parties.
In such designs, the vertical layering of authority is a key component of complex
power-sharing settlements, i.e., horizontal power-sharing alone is not sufficient
for the establishment of stable political and institutional processes conducive to
resolving self-determination conflicts. In fact, vertical layering of authority is a
necessary condition in a number of cases for horizontal power-sharing: unless a
region (or regions) becomes a locus of power, no power can be shared at the sub-
national level. Power-sharing in the Bosnian-Croat Federation, in Bougainville,
in the Autonomous Region of Muslim Mindanao, and in Northern Ireland would
not be possible if these regions had not been established as legal-political entities
and powers had not subsequently been devolved to them. At the same time, it is
important to bear in mind that horizontal power-sharing at the regional level is
not a necessary consequence of vertically layered authority. In Gagauzia, Kosovo,
Macedonia, and the Republika Srpska, sub-national layers of authority have not
led to the establishment of formal power-sharing institutions at these levels.
With some qualifications pertaining to the lack of formalized structures, this is
also the case in South Ossetia. The difference between regions with horizontal
A future federation may also be on the cards for Moldova, in which case Gagauzia
and Transdniestria may become equal partners alongside Moldova.
This, however, does not preclude the emergence over time of informal or spontane-
ous forms of power-sharing at the regional level.
448 Stefan Wolff
structures of power-sharing and those without is first of all one of the degree of
ethnic (or other) heterogeneity. The bipolar ethnic and/or political demography
of the Bosnian-Croat Federation, Bougainville, and of Northern Ireland, as well
as the religious and tribal mix in the provinces that opted for membership in the
Autonomous Region of Muslim Mindanao, required constitutional designers to
devise mechanisms of conflict regulation below the national level and beyond
traditional notions of subsidiarity and devolution. Context-dependence contin-
ues at this level as well, as becomes evident from, among others, the differences
in strength that regional power-sharing authorities have in all these cases and
the degree of power that lower levels of authority within them enjoy, such as the
cantons in the Bosnian-Croat Federation or the individual provinces that make
up the Autonomous Region of Muslim Mindanao.
Where regional or national (formal) horizontal structures of power-sharing
are missing in the case studies, demography and the vertical layering of authority
has combined favourably in ways that make them superfluous. In Kosovo, for ex-
ample, ethnic Serbs remain a sizable minority, but their territorial concentration
and the relative homogeneity of the areas in which they reside means that the de-
volution of substantial powers to municipal authorities affords ethnic Serbs a rea-
sonable degree of autonomy from the central authority in Kosovo, while a system
of reserved seats guarantees them representation in parliament at the centre.
In Moldova, where horizontal power-sharing is absent at the regional and, with
some qualifications, national levels, the creation of the Gagauz autonomous ter-
ritory and the fact that referenda were held at the district level over whether par-
ticular local communities wanted to belong to the Gagauz territory helped create
a reasonably homogeneous territory within which horizontal power-sharing was
not deemed necessary. Formalized structures of horizontal power-sharing at the
national level exist only in a limited way in the form of co-optation of regional
officials to corresponding national bodies.
In Macedonia, the territorial concentration of ethnic Albanians in the west
of the country, combined with a substantial degree of autonomy and power for
local communities is, by-and-large, sufficient to address the major concerns of
the minority community. Moreover, the fact that the demographic balance in the
country and the structure of its party system facilitate interethnic coalitions at
the centre contribute to the overall satisfaction that both major ethnic groups
derive from this settlement.
This degree of variation across the case studies suggests four important conclu-
sions for the role that the vertical layering of authority has in the power-sharing
toolkit. First, dividing power along a vertical structure of institutions can serve
as a useful substitute for formal horizontal power-sharing at either national or
regional levels, provided that national or regional ethnic demographies create
suitably homogeneous territories and that substantial powers are devolved from
the centre. In other words, such cases lend themselves to the application of forms
of territorial autonomy or of the subsidiarity principle, instead of the use of co-
decision making as foreseen by power-sharing institutions. Moreover, a certain
degree of representation of minority groups at the relevant central level (regional
in the case of Kosovo, national in the cases of Macedonia and Moldova), in ad-
dition to these other two conditions, also seems to facilitate this kind of institu-
tional structure.
Second, while the vertical layering of authority may under certain conditions
be able to substitute for the creation of horizontal power-sharing institutions, the
reverse is empirically not the case. Even where power is shared at the national
level (as in Bosnia and Herzegovina and Bougainville) or at the regional level
(Bosnian-Croat Federation, Bougainville, Mindanao, Northern Ireland), author-
ity remains vertically layered and lower levels of authority enjoy different degrees
of autonomy in the local decision-making process.
Third, coordination between different vertical layers of authority and the es-
tablishment of clear hierarchies are important to ensure that vertical layering of
authority remains meaningful and can contribute to the long-term sustainability
of a particular conflict settlement. Where there is a danger of eroding the auton-
omy of regions created as a particular layer of authority with the specific purpose
of conflict resolution (such as the Autonomous Region of Muslim Mindanao and
Gagauzia), conflict settlements may not be sustainable in the long term.
This means, finally, that without safeguards against arbitrary government
interference, it is unlikely that the conflict parties will develop a sense of satis-
factory permanence and predictability in relation to a particular conflict settle-
ment. The power to suspend devolved government in Northern Ireland may on
several occasions have prevented the collapse of the power-sharing institutions
in the province, but it has also reduced the willingness of the conflict parties
to commit to working together, and develop a shared sense of responsibility for
sustaining the institutional structures created by the 1998 agreement. Legal and
constitutional entrenchment, possibly alongside international guarantees, is thus
one important mechanism for the stabilization of institutional structures. How-
ever, from the perspective of the minority community another mechanism can be
equally important, namely the option to secede in case of major constitutional,
demographic, or political changes. Thus, if Moldova at some point decided to
unite with Romania, Gagauzia has the opportunity to hold a referendum on its
independence; Bougainville has a future option for a referendum on its indepen-
dence from Papua New Guinea; and in Northern Ireland the popular will regard-
ing unification with the Republic of Ireland is to be gauged at regular intervals.
These two observations on entrenchment and popular consultation also un-
derscore that the preservation of democratic procedures is a key factor for stabi-
lizing institutional structures created for the purpose of resolving self-determina-
tion conflicts, because it is through this longevity that institutions acquire their
legitimacy. While democratic institutions in themselves are not necessarily and
automatically technically viable, compliance with rules and regulations agreed
between all conflict parties and their democratic accountability to voters in-
creases the survival chances of smooth and efficient institutional processes. Any
form of complex power-sharing will always modify and constrain majoritarian
forms of democracy, but this does not mean that power-sharing institutions can
or should be run without popular support. Power-sharing does depend upon on
450 Stefan Wolff
the willingness and ability of elites to cooperate and make compromises, but it
also depends on the willingness of the people to support their respective elites in
this process and to uphold a settlement negotiated to bring about a non-violent,
stable and predictable political process.
Chapter 14
Electoral Arrangements in Systems of
Complex Power-sharing
Andrew Reynolds
I Why Electoral Arrangements Matter

As is noted throughout this book, the set of democratic institutions a nation
adopts are integral to the long-term prospects of an embryonic democracy, as
they structure the rules of political competition and determine whether minori-
ties and majorities are embedded within the democratic process. In fragile so-
cieties emerging from socially and politically devastating conflicts, the effects
of institutions are the most pronounced, as they are the building blocks of the
power-sharing settlements that ultimately give the divided nation its best chance
for healing and survival. Within the plethora of democratic institutions there is
no more important choice than which electoral system is to be used (see Lijphart
1977; 1985; 1994; Sartori 1996; Reynolds 1999; Reilly 2002; Taagepera and Shugart
1989; Horowitz 1991). Electoral systems have long been recognized as one of the
most important institutional mechanisms for shaping the nature of political com-
petition because they are “the most specific manipulable instrument of politics”
(Sartori 1968: 273) – that is, they can be purposively designed to achieve particu-
lar outcomes.
Different ways of voting structure the arena of political competition and party
system that develops, and offer incentives for elites to behave in certain ways
by rewarding those who respond to these incentives with electoral success. The
great potential of electoral system design for influencing political behaviour is
thus that it can reward particular types of behaviour and place constraints on
others. This is why electoral system design has been seized upon by many schol-
ars (see Horowitz 1991; Lijphart 1977; 1984; Reilly 2002; Reynolds 1999; Sartori
1968; 1994) as one of the chief levers of constitutional engineering to be used in
mitigating conflict within divided societies. As Lijphart notes: “If one wants to
change the nature of a particular democracy, the electoral system is likely to be
the most suitable and effective instrument for doing so” (1995: 412). While it is
true that electoral system design has not proved to be a panacea to communal
conflict in many places, the fact that electoral systems often preoccupy the minds
452 Andrew Reynolds
of negotiators and that ‘getting it wrong’ can be so devastating persuades us that

the electoral arrangements in our eight cases of complex power-sharing are more
than worthy of our attention.
In this chapter, I shall focus on two main aspects of electoral arrangements.
The bulk of attention will be focused on the national electoral systems of Papua
New Guinea (as it relates to Bougainville), Philippines (Mindanao), Bosnia and
Herzegovina, Kosovo, Macedonia, Moldova (Gagauzia), Georgia (South Ossetia),
and the province wide system in Northern Ireland: how they operate and their
consequences for representation of minorities and power-sharing arrangements.
Subsequently, I address the impact of electoral administration in these cases. The
nature of and independence of the body responsible for running elections and
provisions made for returning refugees and displaced persons.
II Electoral Systems and Power-sharing

In any state, consolidated or emerging, an electoral system exists to do three main
jobs. First, it will translate the votes cast into seats won in the legislature. The sys-
tem may give more weight to proportionality between votes cast and seats won,
or it may funnel the votes (however fragmented among parties) into a parliament
that contains two or three large parties representing more polarized views. Sec-
ond, electoral systems act as the conduit through which the people can hold their
elected representatives accountable. They have the opportunity to punish parties
or throw individuals out of power if they displease. Third, electoral systems serve
to structure the boundaries of ‘acceptable’ political discourse in different ways,
and give incentives for those competing for power to couch their appeals to the
electorate in distinct ways.
But the effects of any electoral system – especially in systems of complex pow-
er-sharing – are conditioned by the matrix of accommodations and incentives
beholden within the other institutions that combine to make the democratic
patchwork of the state. Just as in medicine, in constitutional design it is counter-
productive to treat a specific ailment (such as non-functioning elections) while
failing to engage in a broader diagnosis that seeks to treat the patient as a whole
(i.e., addressing where power lies, the viability of the judiciary, etc.). Distinct in-
stitutional prescriptions can work against each other in a harmful way (and exac-
erbate the illness) if the treatment regimen is not complimentary for the whole.
In medicine, one drug may react with another and retard the patient’s progress,
or the treatment of one ailment may create new problems in other parts of the
body.
In democratic design, especially in instances of complex power-sharing, a
seemingly inventive electoral system may combine with a power-laden presiden-
tial executive to shut large minorities out of all access to power. When the web of
political institutions do not work in concert together, pull in the same direction,
one can end up in a situation where political measures, which when taken indi-
vidually seem appropriate and positive, combine to produce an outcome far less
than the sum of their parts and in actuality make the situation worse than it may
14 Electoral Arrangements in Systems of Complex Power-sharing 453
have been without intervention. Holistic interpretations of political institutions

are somewhat akin to the notion of complementarities between institutions that
shape the political economy of a state. Hall and Gingerich (2001) define an insti-
tution as complimentary to another when “its presence raises the returns avail-
able from the other.” When it comes to institutions that combine to form complex
power-sharing structures, one might broaden that conception of holistic to where
the matrix raises the sum benefit and precludes a negative loss in outcome when
institutions interact.
III Functions of Electoral Systems in Divided Societies

Within post-conflict societies, where institutions are as much about peace settle-
ments between conflictual groups as they are about democratic processes, then
the electoral system takes on three highlighted roles. First, they provide the chief
building blocks of power-sharing. The electoral system needs to facilitate a di-
verse legislature through which the representatives of majorities and minorities
can operate power-sharing, vetoes, and resource allocation. This is especially true
of consociational structures, which are often focused on the allocation and shar-
ing of legislative power: that power is shared between the parties that make it
into parliament. If the electoral system does not allow for the presence of the
salient groups within the representative legislature then power-sharing can still
take place but it would be non-democratic and of a very different type to what we
aspire to when we imagine successful complex power-sharing arrangements.
Second, a related point, is that the electoral system in a divided society needs
to produce a parliament that is ‘descriptively representative’, sufficient to reas-
sure minorities that their voices are being respected. The inclusion of the diver-
sity of majorities and minorities within legislatures can reduce group alienation
and violence in those divided societies where politics is often viewed as a win or
lose game. Many peace settlements since 1990 have revolved around reserved
seats for communal groups as part of broader power-sharing constructs. Others,
such as South Africa, have shown that the descriptive representation of alienated
minorities within a representative political system helps to mitigate anti-system
violence and engenders an air of cooperation
Last, there is the question of whether the system can be crafted to encourage
political elites to behave in an accommodatory way come election time. In terms
of deeply ethnically-divided societies, for example, where ethnicity represents
a fundamental political cleavage, particular electoral systems can reward can-
didates and parties who act in a cooperative, accommodatory manner to rival
groups; or they can punish these candidates and instead reward those who appeal
only to their own ethnic group. However, the ‘spin’ that an electoral system gives
to the system is ultimately contextual and will depend on the specific cleavages
and divisions within any given society.
454 Andrew Reynolds
IV Methods for Achieving Those Goals

At the legislative level, different electoral systems help fulfill the functions out-
lined above in a variety of ways. Unfortunately, it is rare that any one system fulfills
all the needs of democratization, conflict management, and representativeness.
However, providing the building blocks of executive and legislative power sharing
– i.e., producing a diverse and inclusive multi-party legislature that reflects the
array of majority and minority groups within society – is best facilitated by those
systems classified as proportional. On one base level, closed party list systems
with low thresholds for legislative entry allow for even the smallest groups to gain
some representation and for party bosses to control and select their parliamentary
caucus members. Consociationalists argue that what is lost to legislative detach-
ment at the individual MP level is gained by the creation of cohesive and stable
partisan party units, which are the component units of a functioning power-shar-
ing dispensation. Open lists, which allow some scope for voters to re-order the
party list presented to them on the ballot, may enhance voter choice (and indeed
democracy) but they retard some of the ability for ‘manufactured’ diversity in the
legislature. The difference between Kosovo’s municipal and state-wide elections
is demonstrative. Kosovo’s municipal elections (held before the Assembly elec-
tions) were conducted with ‘open’ lists. The UN administrators judged that these
elections were marked by voter confusion, the undue ‘personalization of politics’,
and the suppression of gender diversity through the pushing down of female can-
didates on the party lists. Thus, for the following assembly elections, the system
was changed to one of closed party lists. Bosnia moved the other way – changing
from closed to open lists for their elections of October 2002.
Mixed Member Proportional (MMP) systems, as used in Germany and New
Zealand, also facilitate the proportionality required for power-sharing systems
but they use single member districts to select half to two thirds of the MPs. Paral-
lel systems (as used in Macedonia, Georgia, and the Philippines) reduce the ‘mir-
roring’ of ethnic and partisan diversity by only allowing for some of the legislators
to be elected by proportional means. Under any type of proportional system, the
nationally imposed threshold of representation is going to effect deeply the legis-
lative system and propensity for fully blown power-sharing. If the threshold is set
at say 5% then minorities smaller than that figure will be shut out of the legislature
and thus the more formal political aspects of power-sharing arrangements.
An alternative (or supplement) to PR electoral systems to provide the basis for
power-sharing are reserved seats for identified communities and/or the deliber-
ate over-representation of geographical areas where minority communities dom-
inate. Historically, the practice of seat reservation, or gerrymandering for distinct
communal groups, has long existed. The recognition and desire for some degree
of descriptive representation is not a new thing – nor is it a construct unencum-
bered by past misuse. Manipulating the electoral system to produce descriptive
representation can take at least four forms: communal rolls, reserved seats, race
conscious districting, and the over-representation of given geographical areas.
Our eight cases demonstrate most prominently elements of the first two of these
strategies.
Bosnia and Herzegovina in effect uses communal roles in the way that it sepa-
rates and segments ethnic groups into voting blocks. The Dayton accords, signed
in November 1995, finally brought a type of closure to the civil war that had raged
throughout the territory of the former Yugoslavia since 1992, costing over 100,000
lives and the violent expulsion of over half of the Bosnian population from their
homes. But the hopes for peace born in Dayton, Ohio came at the cost of political
structures, which did all they could to separate, segment, delimit, and ultimately
make all politics in Bosnia about ethnic identity and political-physical separation.
Virtually no space was left for the evolution of moderate multiethnic parties that
might sow the seeds of a return to shared nationhood in the Balkans.
The Dayton accords divided Bosnia into two entities. The Federation (Bosniacs
and Croats), comprising 51% of the land, and the Republika Srpska, comprising
the rest. The national parliament has an upper house with five members from
each of the three ethnic groups, while the lower house has twenty-eight members
elected from the Federation and fourteen from the Serb Republic. Each com-
munity has an effective legislative veto over any matter they choose to delegate
“destructive of a vital interest.” There is balancing at almost every level of gov-
ernment but it is rooted in ethnic identity and segmentation – elections are, by
design, ethnic censuses in Bosnia.
This vulgar, identity-freezing form of consociation may well have been neces-
sary to get signatures on paper in Dayton and to stem the blood-letting in Bosnia.
But the preoccupation with short term fixes has meant that longer term integra-
tion and democratic consolidation is a much bleaker prospect. The institutions
gave little incentive to parties to moderate their ethnically-based appeals or work
together once elected. Elections in 1996 and 1999 were unsurprisingly character-
ized by the sight of all major parties crafting appeals on the basis of hostile and
polarizing notions of ethnic community and difference. The elections of 2002
actually saw a swing to even more extreme nationalist parties in both the Bosniak
and Serb entities.
The strategy of reserving seats for minorities to ensure their presence in na-
tional parliaments can have more of a peripheral impact on democratization
but nevertheless remains an important symbol of inclusion (if not acceptance).
Most dramatically, ten seats in the parliament of Kosovo are reserved for Serb-
Kosovans, which, along with the twelve seats the Serb party (KP) was able to win
in general competition, meant that Serbs won twenty-two seats in the 120 mem-
ber assembly – 18.3%. The Serb minority is estimated as being 5.9% of the total
population. As Table 1 shows, seats were also reserved for the Roma, Ashkali,
Egyptians, Bosniaks, Turks, and a Gorani.
Unfortunately, communal and reserved seats usually rest upon someone’s pre-
determined view of who, and how large, the various social groups are and this
can create a rigid and illiberal foundation that retards the growth of non-ascrip-
tive cross-cutting political movements. Even the Northern Irish Assembly, which
does not pre-determine voters, does require parliamentarians to identify them-
456 Andrew Reynolds
selves for voting purposes as Unionist/Protestant, Nationalist/Catholic, or non-

aligned. The key conundrum inherent within pre-determination is how to choose
which groups are awarded seats or are recognized as eligible for special treat-
ment. What are the building blocks for the power-sharing settlement and who
has (or needs) special rights? Do ethnic groups have to be small and oppressed
or can they be small and powerful, and what constitutes a clearly defined ethnic
group anyway? In Kosovo, Egyptians made claims that they should be treated on
a par with minority Serbs.
Pre-determination also excludes the recognition of groups that are less vis-
ibly identified as ‘groups’; communities of interest, which may still be identity
based but are not classically envisioned ethnic groups. Communities linked by
sexual politics, gender issues, youth or aged issues are almost never included as
valid building blocks of a polity. The dangers of precluding non-ethnic identities
for ‘special recognition’ was demonstrated by farcical scenes in Northern Ireland
in November 2001 when three members of the non-sectarian Alliance Party of
Northern Ireland had to re-designate themselves as ‘Unionists’ in order to ensure
the election of David Trimble as First Minister and the continuation of the As-
sembly. Lastly, pre-defined communal groups and fixed reserved seats are not
open to the inevitable flux in numbers of majority and minorities in nation states.
Rarely are there clauses in electoral acts for the periodic review of the minority
group size and its related number of reserved seats.
When it comes to the question of encouraging political elites to behave in
moderate and accommodatory ways, it has been argued that preference voting
crafts incentives for parties to appeal to the lower order preference votes of voters
from other communities (see Horowitz 1985; Reilly 2002). Preference voting in
single member districts is known as the alternative vote (AV) and requires a high
degree of intermixing of polarized ethnic groups for its electoral levers to kick in.
Preference voting in multi-member districts (as used in Northern Ireland) is the
Single Transferable Vote (STV). STV produces a more proportionate outcome
(thus facilitating consociationalism) but it lessens the incentives for multiethnic
vote pooling or the development of cross-cutting electoral coalitions when com-
pared to AV in an optimum demographic context.
At the executive level – if there is to be a directly elected chief executive/presi-
dent – there are a few institutional strategies for encouraging accommodating
and moderate leaders less beholden to ethnic chauvinisms or hostilities. First is
to require an absolutely majority system – either the French style run-off or the
alternative vote as used in Sri Lanka – to ensure that no single candidate/ethnic
group can win with a minority of the popular vote. Alternatively, one can impose
a super-majority requirement, where the winning candidate must win, not mere-
ly a national majority, but surmount a threshold in all regions of the country (as
in Nigeria and Kenya). Unfortunately, the prospects for these innovative methods
are beholden to the demographic realities of the country in question. In order
to elect a president beholden to inclusive nation-building, rather than ethnically
divisive exclusion, no one group can be in the absolute majority if the voting rule
is preferential (in Sri Lanka, lower preferences have never been needed). Super-
majority distributional formulas, in theory, better ensure that the president has
multiethnic support, but, in reality, they have proved less than successful when
used in Kenya and Nigeria.
V Descriptive Overview of Electoral Systems in the Cases
A Legislatures
At the legislative (state-wide) level, five of our eight cases use PR systems (four list,
one single transferable vote), while two use parallel systems (Macedonia switched
from a parallel to a list PR system in 2002). Only Papua New Guinea, of our cases
of complex power-sharing, uses a majoritarian first past the post system. There
is a degree of variation within both the list PR and parallel systems used. All bar
one used closed lists (Bosnia changed to open lists for their election of October
2002) but Kosovo and Moldova elect their MPs from very large national districts
(100 and 101 members in size, respectively) while Macedonia and Georgia split
their districts into much smaller entities (of 20 and 5-24 members, respectively).
The smaller districts make it more difficult for tiny nationally diffused parties
to win representation but do favour more geographically concentrated parties
or candidates. The threshold for representation varies from 7% in Georgia to no
legally imposed threshold in Bosnia and Kosovo. The parallel systems of Georgia
and the Philippines (and the system used in 2001 in Macedonia) vary even more
dramatically. The proportional element in the Philippines and Georgia domi-
nates (80% and 64%) while it was a much smaller aspect of the 2001 Macedonian
system (only 29% were elected from party lists). In the single member districts,
the Philippines use first past the post while Georgia and Macedonia use the two
round run off system.
458 Andrew Reynolds
Table 1 Electoral Systems (Legislative)

Assembly Electoral List District Threshold Reserved/
size System Magni- Commu-
tude nal
Seats
Northern 108 STV (PR) Open 6 x 18 - None
Ireland
Bougain- (??) FPTP - 1 - None
ville 109
(PNG)
Mindanao (??) Parallel Closed 1 x 208 2% None
(Philip- 260 52 x 1
pines)
Bosnia 42 List PR Closed 1 x 14 - 14 Serb

and Her- 1 x 28 14 Bosniak
zegovina 14 Croat
Kosovo 120 List PR Closed 1 x 100 - 10 Serbs
1 x 20 4 Roma/
Ashkali/
Egyptian
3 Bosniak
2 Turks
1 Gorani
Moldova 101 List PR Closed 101 4% None
Macedo- 120 Parallel Closed 1 x 85 5% None
nia 35 x 1
2001
Macedo- 120 List PR Closed 20 x 6 5% None
nia
2002
Georgia 235 Parallel Closed 10 x 5-24 7% 12 Abkha-
85 x 1 zia
B Sub-national Bodies
Four of the cases contain representative bodies elected from sub-national com-
ponent parts of the state and these legislative entities are, or have the capacity
to become, important parts of the power-sharing structures. It should be noted,
however, that the ‘People’s Assembly’ of Abkhazia is not recognized by Georgia
or any international bodies. Interestingly, the three sub-national bodies not pre-
viously discussed all revert to majoritarian electoral systems. This makes intuitive
sense on a couple of levels: first, the geographical areas are much smaller and
place more of a primacy on village/town/ward representation; and, second, the
pressures for inclusivity of majorities and minorities becomes less of an issue
as these bodies specifically serve the minority enclaves. Perhaps the most inter-
esting proposed regional body is the Autonomous Region of Muslim Mindanao
(ARMM) in the Philippines, which in fact covers approximately 15% of Mindanao.
The proposed ARMM has twenty-four members drawn from eight, three mem-
ber districts, along with ‘sectoral’ representatives from agriculture, labour, urban
poor, the disabled, youth, women, and indigenous cultural communities, who can
constitute up to 15% of the regularly elected members.
Table 2 Electoral Systems (Sub-national Territorial Bodies)

Assembly Size Electoral District Mag-
System nitude
Northern Ire- 108 STV (PR) 6 x 18
land
ARMM 24? List? 3x8

(Philippines)
Gagauzia ?? TRS 1
(Moldova)
Abkhazia 35 TRS 1
(Georgia)
C Executives
In Northern Ireland, Kosovo, and Papua New Guinea, executives are chosen by
democratically elected legislatures although the powers of each prime minister/
first minister/president vary considerably. In Bosnia, there is a three person na-
tional presidency, with the election compartmentalized by ethnic group: Serbs
elect the Serb member, Croats elect the Croat, and Bosniaks the Bosniak. This
method clearly gives the least space for the building of multiethnic coalitions
of interest (as encouraged by preference voting or distribution requirement sys-
tems) but does provide a solid segmental block of elites for consociationalism.
Four of the cases have directly elected presidents – Georgia, Macedonia, Mol-
dova, and the Philippines. The first three use a two round majority run off system
while the Philippines simply use a plurality method. Ironically, the East European
cases have not needed the run off requirement (presidential candidates winning
by well over 50% on the first round in each case) while the Philippines might
have been well served by an absolute majority system, with Presidents Ramos and
Estrada having weak legitimacy after winning with 23.6% and 39.9% in 1992 and
1998, respectively.
460 Andrew Reynolds
Table 3 Electoral Systems (Executive)

Executive type Electoral Sys- Further Concur- Term of
tem proce- rent? office
dure
Northern Ire- Parliamentary/ Leaders of larg- - - -
land First & Second est party within
Minister each community
Bougainville Parliamentary Chosen by legis- - - -
(PNG) lature
Mindanao Presidential Plurality - no 6
(Philippines)
Bosnia and Her- Presidential Plurality (within - yes 4
zegovina 3 member community)
Kosovo Parliamentary/ Chosen by legis- - - -
Presidential lature
Moldova Presidential Majority Runoff no 4
Macedonia Presidential Majority Runoff no 5
Georgia Presidential Majority Runoff no 5
VI Consequences of Electoral Systems

Conventional wisdom asserts that the two most important areas where the elec-
toral system is believed to have major determinant power are the degree of dis-
proportionality (ID) that the system produces between the votes won by parties
and the seats they win, and the number and respective strengths of parties in
legislatures (Effective Number of Parliamentary Parties (ENPP)). In fact, party
fragmentation is itself in part determined by the disproportionality of the system
and thus these variables are often collapsed together with party fragmentation
becoming the focal point. Maurice Duverger was the first in a long line of in-
stitutionalists to seize on “the electoral system’s permissiveness to small party
representation” as the chief explanation of fragmentation (Taagepera 1999: 531).
To summarize Duverger’s oft-repeated propositions:
(1) PR tends to a system of multiple, rigid and independent parties; (2) the majority sec-
ond ballot system [tends to] a system of multiple, flexible and independent parties; (3)
the majority [plurality] single ballot system [tends to] party dualism (Duverger 1954).
In the 1980s, Powell (1984) introduced another explanatory variable when it came
to the number of parties produced – that of social heterogeneity or fragmenta-
tion. That is, it was not merely electoral systems alone that determined party
system structure, but rather the ethnic, linguistic, religious, or regional-cultural
divisors of any given society. Powell used an additive measure of electoral system
and social heterogeneity factors to explain the effective number of political par-
ties; later work by Ordeshook and Shetsova (1994) and Amorim Neto and Cox
(1997) found that an interaction term of the two variables was more powerful.
The interplay between societal and institutional variables was found to be the
driver of heterogeneity. Taagepera articulates the state of current wisdom:
Low heterogeneity puts a lid on the number of parties even in the presence of a very
permissive electoral system, because there will be no demand for many parties. Similar-
ly, a low magnitude (single member districts [FPTP]) tends to put a lid on the number
of parties even in the presence of strong heterogeneity, because few parties can gain
representation. This mutual limitation can be obtained by multiplying the two effects,
but not by adding them. (1999: 535).
However, when we look at the indicators of electoral system consequences in

Table 4, it is marked how little our eight cases follow the expected patterns of
electoral system consequences. There is no obvious electoral system effect in
translating votes into seats or determining the type of party system that emerges.
The first to fall is Duverger’s law, as the effective number of parliamentary par-
ties in our eight cases flies counter to what we would expect. Papua New Guinea
has the most fragmented party system despite its first past the post system, while
Moldova has the least fragmented system despite its large district closed list PR
system. On average, the three parallel cases have an ENPP of 3.6, the four list PR
cases 3.9, Northern Ireland’s STV 5.5, and PNG’s FPTP 8.8. The eight cases also
fail to follow the rules of disproportionality. While the list PR cases are marginally
more proportional than the FPTP case, STV produces the most proportional out-
come and the Parallel cases demonstrate the highest degrees of disproportional-
ity, with Macedonia in 1998 being extremely disproportional. One would also ex-
pect the largest/dominant parties to be found in majoritarian (FPTP) or semi-PR
(Parallel) systems and the smallest ‘largest’ parties to occur in PR systems (List
and STV). But in our cases we find that the largest parliamentary party in PNG
is only 17.4% and list and parallel systems average largest parties that constitute
47% of the legislature.
When it comes to the representation of women, the electoral systems do ap-
pear to have exerted influence in more expected ways. There is a multitude of
evidence to show that list PR systems best facilitate the election of women, and
prospects are enhanced further if some type of gender quota is applied as well
(see Reynolds 1999). The average number of women elected in our four list PR
cases is 19.1% with a high of 28.3% in Kosovo. Kosovo is one of three cases where
a gender quota was applied – one out of every three candidates on a party’s list
had to be a woman. In Bosnia, the mandate was weakened but still had some ef-
fect. One candidate of the under-represented sex had to be included in the first
two candidates on the list, two among the first five, three among the first eight
candidates. However, the fact that the October 2002 elections were conducted
with open lists allowed some male candidates to be pushed above higher placed
women. In Macedonia, a minimum of 30% of each sex should be represented on
party candidate lists. This led to 18.3% being elected in 2002.
462 Andrew Reynolds
Table 4 Electoral System Outcomes

Electoral ENPP ID Largest Women Spoilt
System
Northern Ireland STV (PR) 5.5 3.4 25.9 13.9 na
Bougainville FPTP 8.8 11.1 17.4 0.9 1.8i
(PNG)
Mindanao Parallel 3.3 4.3 40.0 17.8 20.9
(Philippines)
Bosnia and Her- List PR 6.1 5.9 23.8 16.7 na
zegovina
Kosovo List PR 4.1 7.5 39.7 28.3 2.5
Moldova List PR 2.6 9.3 76.2 12.9 2.5
Macedonia Parallel 4.0 27.8 47.5 6.7 na
1998
Macedonia List PR 2.9 10.4 49.2 18.3 1.8
2002
Georgia Parallel 3.4 14.0 55.3 7.2 na
i From 1982.
NB: Electoral System = see Reynolds and Reilly 1997. Women = % of assembly, last elec-
tion. Largest = seat share of largest party. ID = Gallagher’s Least Squares Index of Dis-
proportionality. ENPP = Effective Number of Parliamentary Parties. Largest = Largest
party (seat share) last election. Spoilt = Spoilt ballot rate last election. Na = not avail-
able.
Data: Northern Ireland 1998, Philippines 1995 and 1998, PNG 1977-1997, Moldova 1994
and 1998, Georgia 1995 and 1999, Kosovo 2001, Bosnia 1996, 1998, and 2002.
These indicators demonstrate that our cases of complex power-sharing are em-
bryonic multi-party polities where the pressures of institutional arrangements
have not yet outweighed the social rigidity of deeply divided societies. The socio-
political historical pressures within our eight cases are overwhelming the mar-
ginal incentives that electoral systems can place on voter behaviour. Political or
ethnic tribalism overwhelms the incentives for strategic voting, and the parties
that might have fragmented under list PR systems retain enough glue to keep
themselves whole – at least for the time being. There is a temporal aspect to
electoral system consequences and it may be decades before their real effects on
party development are apparent.
When it comes to the more psychological, and less mechanical, effects of elec-
toral systems – i.e., the ‘spin’ they impart on campaigning and voter behaviour
– the cases of Northern Ireland (preferential voting-STV), Bosnia (segmented
closed lists), Papua New Guinea (a history of preferential voting-AV and FPTP),
and Kosovo (PR with reserved seats) are demonstrative in their various effects.
Preference voting STV was introduced in the 1970s in Northern Ireland precisely
because it both combined proportionality in outcomes with incentives for par-

ties to appeal outside of their own exclusive community for the lower preferences
of other voters. Thus, the system was intended to encourage moderation, inter-
ethnic bridge-building and create the space for cross-cutting parties to emerge.
However, in 1973 and 1982, the system failed to produce cross-community par-
ties, moderated ethnic parties, or vote-pooling between Protestant and Catho-
lics. Dixon estimates that only one quarter of 1% of all vote transfers crossed the
divide in 1973 (see Dixon 1997). This pattern of centrifugal political forces push-
ing parties to the extremes was repeated in 1982.
But for the first time in the Northern Irish Assembly elections of 1998 STV
appeared to engineer, or at least facilitate, some of the positive elements of mod-
eration that were hoped for. This time, transfers were far more apparent among
pro-agreement parties, and the major parties (including the UIP and Sinn Fein)
were encouraged to moderate their stances to appeal for lower order votes from
both their own communities and the other. Indeed, the UUP received a signifi-
cant ‘seat bonus’ through transfers by winning 26% of the seats with only 21 %
of the votes. O’Leary argues that the 1998 Assembly initially succeeded because
STV, and its capacity for the expression and effect of lower order preferences,
reduced the number of seats won by ‘anti-agreement’ parties such as the DUP
(see O’Leary 2002a: 313).
Papua New Guinea also provides a fascinating case study of the psychological
effects of electoral systems because of its switch from preference voting in sin-
gle member districts (AV) used for three pre-independence elections, to classic
FPTP used since independence in 1975. While the system had little observable ef-
fect on the growth or ebbing of secessionist fervour in Bougainville, it did have an
impact on interethnic accommodation and hostility in the state as a whole. Reilly
(2001) provides evidence that the use of AV between 1964 and 1975 encouraged a
degree of party building, moderation, and cross-ethnic alliances, which fostered
the fledgling party system and lessened the excesses of polarizing elections in
such a highly socially fragmented polity. The “standard practice was to cast a first
preference for a local clan candidate, and then a second preference for a knowl-
edgeable outsider” and “candidates who managed to establish broad, cross-tribal
support bases increasingly utilized the mechanics of preference distribution to
defeat opponents who would have easily won a FPTP contest” (Reilly 2001: 70,
76). When AV was replaced by FPTP upon independence, it removed incentives
for interethnic alliances, led to rising levels of violence at election time and ‘dum-
my candidates’ solely on the ballot as spoilers to split the vote. Reilly goes as far
as to argue that traditional divisions between clans have been reformulated and
reinforced by FPTP (Reilly 2001: 81-88).
Both Kosovo and Bosnia make use of closed list PR electoral systems, but their
respective adaptations of the system provide somewhat different psychologi-
cal impacts upon the development of each state’s party systems. In Kosovo, the
twenty seats reserved for Serbs and other minorities (along with those communi-
ties ability to win a share of the open seats) engineer a dramatic over-represen-
tation of the Serb catch all party and make it much more difficult for Albanian
464 Andrew Reynolds
parties to seize a working majority of the legislature. There was never likely to be
much enthusiasm for cross-ethnic voting, but the list PR system encouraged the
fragmentation of the majority Albanian ethnic vote and the consolidation of the
threatened Serbian minority vote. Campaigning behaviour was also impacted by
the shift from an open list system (in the municipal elections) to a closed list sys-
tem for the Assembly elections, which strengthened central party organizations
and suppressed some of the personalization inherent in the campaigning at the
municipal level. It might be argued that one of the failures of the Kosovo system
was that it created a climate that froze ethnic loyalties, while at the same time
providing the space for the fragmentation of the Albanian vote and thus the party
system as a whole. However, it is difficult to see how Serb minority interests could
have been assuaged through any system that did not over-represent them both in
the executive and the legislature. As noted earlier, Bosnia’s adaptation of the list
PR system was less about the over-representation of minorities and more about
the segmentation of communities into separate voting blocks. Thus, there is no
space for multiethnic movements to grow, and each ethnic group is corralled into
an electoral fight between moderates and extremists.
Electoral systems are an important part of complex power-sharing arrange-
ments but their role is contingent on the type of socio-political division present.
National electoral system design matters more in cases of geographical intersper-
sion or balance and a little less in cases where the minority is heavily concentrat-
ed in a given area and their needs can be better addressed through autonomies or
secession. Thus, elections and national representation matter more in Northern
Ireland, Bosnia, Kosovo, and Macedonia than say in Bougainville/PNG, Mind-
anao/Philippines, Gagauzia/Moldova, and South Ossetia/Georgia. For obvious
reasons the hard edge of power sharing in cases like Bougainville, Mindanao,
Gagauzia, and South Ossetia exists in arrangements for autonomy and self-gov-
ernment in those specific regions and by specific ethnicities. The reassurance of
minority access is not focused on national decision-making structures but rather
local separations from the dictates of the national majority.
VII The Importance of Electoral Administration

It has taken a while, but academics are finally beginning to wake up to a realization
that struck many policy makers and NGOs a decade before, i.e., that the quality of
electoral administration has a direct impact on the way in which elections in the
developing world and their outcomes are regarded, not merely by international
observers, but also – and more importantly – by domestic actors such as voters,
parties, media, and local observers. The administration of the vote, its compe-
tence and independence, dramatically impacts the public view of election results
and the legitimacy of governments that are produced. This is particularly true in
divided multiethnic societies clawing their way towards stability and democracy.
Our eight cases represent a wide variety of approaches to running elections
– and demonstrate varying degrees of success. Northern Irish elections are ad-
ministered (as in the rest of Britain) by the civil service wing of government.
However, the long serving Chief Electoral Officer of Northern Ireland Pat Brad-
ley, a Catholic from Derry, is one of the most broadly respected electoral offi-
cers in the world and his organization is praised on all sides for its efficiency
and independence. Similarly, the government of Papua New Guinea appoints
an electoral commission, and despite potential charges of bias this body is also
widely respected for its fairness (see Reilly 2002). The East European cases of
Moldova, Georgia, and Macedonia have appointed electoral commissions – and,
in the case of Georgia members appointed by majority and minority parties, but
these commissions have fared less well in issues of competence and the percep-
tion of fairness. The Philippines Electoral Commission is popularly elected, but
again their success in crafting a level playing field and managing to run elections
free of ballot rigging and intimidation has been decidedly mixed. Perhaps the
most interesting cases of the impact of electoral administration of the building of
democratic legitimacy come in the two post-conflict former Yugoslav states. In
Bosnia and Kosovo, elections were initially run by the OSCE and United Nations
– a domestic elections commission was set up for the Bosnian elections of 2002.
However, despite foreign ownership of the process, elections proved to be far
from legitimate in Bosnia in the 1990s. As Elklit and Svensson note, before and
on polling day, many freedoms were grossly violated, among them freedom of
movement, voter registration, freedom of speech, media access and the freedom
to vote. While by and large polling was smooth at the counting stage “rumors of
fraud began to circulate.” (Elklit and Svensson 1997: 44).
466 Andrew Reynolds
Table 5 Organizational Context of Elections

Responsibility Selection Reputation Observers
Northern Ire- Northern Irish Government Independent/ Domestic
land Office, UK Appointees Competent
Bougainville Electoral Com- Government Independent/ Commonwealth
(PNG) mission Appointees Competent
Mindanao Commission on Elected Independent/ Domestic
(Philippines) Elections Semi-compe-
tent
Bosnia and OSCE OSCE Appoin- Semi-indepen- Domestic and
Herzegovina tees dent International
Semi-compe-
tent
Bosnia and Electoral Com- Parliament/ Independent Domestic and
Herzegovina mission Government Semi-compe- International
2002 Appointees tent
Kosovo UN UN Appointees Semi-indepen- Domestic and
dent International
Semi-compe-
tent
Moldova Electoral Com- Parliament/ Domestic and
mission Government International
Appointees
Macedonia Electoral Com- Government Semi-indepen- Domestic and
mission Appointees dent International
Semi-compe-
tent
Georgia Electoral Com- Parliament Domestic and
mission – Majority & OSCE
Minority
Chapter 15
Third-party Involvement in Self-determination
Conflicts
Ulrich Schneckener
I The International Dimension of Self-determination Conflicts

Self-determination or separatist conflicts have implications for international
politics in at least two ways. They challenge international norms and they aim to
change international borders. Separatism points to the inherent tension between
the territorial integrity and sovereignty of states, both granted by the Charter
of the United Nations, and the right to self-determination, as enshrined, for ex-
ample, in the International Covenant on Civil and Political Rights. Furthermore,
the international community of states is deeply concerned about any attempt to
change borders, as it often endangers the stability of a region and may be regard-
ed as a threat to international peace. Self-determination conflicts raise a set of
legal, political, economic, and security issues that go far beyond the local struggle
between a central government and a separatist movement – they are by no means
a purely internal affair.
In terms of actors, self-determination conflicts can also no longer be adequate-
ly described as internal or civil wars. In fact, they are highly internationalized.
Since the early 1990s, we can observe an increasing number of external actors
and the spread of all kinds of cross-border activities. In order to reflect this de-
velopment, new labels have been coined: “wars of the third kind” (Holsti 1996),
“new wars” (Kaldor 1999); or “network wars” (Duffield 2001). They all refer, inter
alia, to the fact that these conflicts are characterized by the involvement of a large
number of external actors that often affect the level of violence and the strategies
of warfare. At least three different types of external actors can be distinguished:
allies, spoilers, and third parties.
First, secessionist groups and central governments as the key conflicting par-
ties themselves maintain cross-border contacts with external allies in order to
strengthen their political position as well as their military potential. Often, kin
states or powerful diaspora communities provide secessionists with military
assistance, economic aid, and/or political backing. Examples are Serbia in the
case of the Bosnian Serbs, the Albanian diaspora in Western Europe and North
468 Ulrich Schneckener
America in the case of Kosovo, or the American-Irish community in the case of

Northern Ireland. Sometimes also non-kin actors – such as neighbouring states,
great powers, or non-state actors – are willing and able to support one of the war-
ring sides (e.g., Arab sponsorship for the Palestinians, US support for Israel).
Second, due to the armed conflict and the collapse of public order, various
external actors are able to cross borders and to exploit the situation for their
own purposes and profits without being necessarily invited by any local actor.
They have an interest in keeping the conflict alive, insofar as they work as spoil-
ers from outside who have no interest in any kind of settlement. They often link
themselves to one of the parties or to fringe groups, most often local warlords,
terrorists, or criminals. In some cases, they may even control these local groups.
Examples here are trans-national criminal networks that use continuing conflicts
for all kinds of illegal activities, such as the smuggling of weapons, drug dealing,
or trafficking of women (Bosnia, Kosovo, Macedonia, Moldova, Georgia, Colom-
bia), international terrorists who seek safe havens or financial resources (Georgia,
Afghanistan, Philippines, Indonesia, Bosnia), or states or trans-national compa-
nies who have major economic interests, in particular in cases of easily accessible
natural resources (Angola, Congo, Liberia, Sierra Leone, or, to a lesser extent,
Bougainville).
Third, in many cases external actors engage in order to achieve conflict resolu-
tion. Over the past fifteen years, most peace agreements have been brokered and/
or have been implemented with the assistance of third parties. Their intervention
often proved to be crucial in managing and settling self-determination conflicts.
A whole range of actors may function as third parties, notably international orga-
nizations or multilateral fora, but also other states, international NGOs, or well-
respected individuals (e.g., former statesmen).
This chapter focuses on this third category of international involvement. How-
ever, it should be noted that the behaviour of the other two types of external
actors is of great importance when it comes to conflict resolution. Thus, third
parties will not only have to deal with the local conflicting parties, but also with
any external allies or spoilers. The focus is also on secessionist conflicts that have
been settled by complex power-sharing regimes and not by the change of inter-
national borders. There are basically two reasons why the international commu-
nity engages in such conflicts. First, obviously, the key parties involved – usually
a central government and a secessionist movement – are unable to solve their
disputes in a non-violent way and need help in various forms. Second, the inter-
national community itself is still very reluctant to accept and recognize self-de-
termination claims and therefore has an interest in preventing secession or the
dissolution of states by favouring some kind of internal solution, such as power-
sharing arrangements.
As a result, one can observe an increasing number of cases where international
actors play a vital role in achieving and maintaining power-sharing structures.
By contrast, prominent cases of successful or failed power-sharing of the past,
such as Switzerland, Belgium, the Netherlands, South Tyrol, Northern Ireland
(1973), Lebanon, Malaysia, or post-Apartheid South Africa are not characterized
15 Third-party Involvement in Self-determination Conflicts 469
by an international dimension – the only notable exception being the 1960 Cy-
prus solution, which was essentially brokered by the United Kingdom, Turkey,
and Greece and which later led to UN intervention. Consequently, the classic
literature on power-sharing or consociationalism hardly reflects the influence of
external actors.
This chapter tries to fill this gap. The key questions therefore are: what are
the different types of third party involvement? What is the role of international
actors in settling self-determination conflicts and implementing power-sharing
arrangements? Under which conditions can third parties have a positive impact
in making power-sharing settlements work? Accordingly, the paper is subdivided
into three sections: first, it will clarify more broadly the concept of third party
involvement and distinguish various types of that involvement. Second, it aims to
identify the specific activities and mechanisms of third party involvement found
in the eight case studies. Third, based on this empirical experience, the chapter
discusses some factors that may determine positive or negative outcomes.
II Concept and Typology of Third Party Involvement

Third party involvement in self-determination conflicts concerns the role played
by one or more external actors who are willing and able to assist local parties in
conflict resolution. A third party is, by definition, not a party to the conflict, but
an outsider who aims to solve the conflict via fair negotiations and agreements
and not by the victory of one side over the other. A third party has to be accept-
able to the relevant local actors (not necessarily to each and every fringe group)
and should be perceived as impartial. A third party, however, is not necessarily
neutral with regard to principles, procedures, and sometimes even outcomes (e.g.
non-acceptance of secession). Thus, third party involvement does not exclude
the application of pressure on one or both sides, including sanctions or the use of
force, in order to bring the parties to the negotiation table. Finally, a third party
can act in a formal or informal capacity; it can be invited by local actors or can en-
gage without invitation, but, for example, on behalf of the UN Security Council or
a regional organization. Clearly, the nature of the third party as well as the range
of activities at its disposal differ from case to case. Nevertheless, it is possible to
distinguish three categories of actors who may function as third parties:
See particularly Nordlinger 1972; McRae 1974; Lijphart 1977. Lijphart (1977: 67-68)
referred only to the possible effect of external threats and international conflicts in
unifying or antagonizing plural societies.
One classic definition of a third party comes from Young (1967: 34, quoted in Berco-
vitch 2002: 6). He defined a third party as “an actor that is not a direct party to the
crisis, that is designed to reduce or remove one or more of the problems of the bar-
gaining relationship, and therefore to facilitate the termination of the crisis itself.”
Others, however, stress the importance of neutrality for third parties. See Bercovitch
2002: 6-7.
• The first category refers to international organizations (IOs) or multilateral

actors such as the UN and its sub-organizations (in particular the UN High
Commissioner for Refugees (UNHCR) and UN Development Programme
(UNDP)), the World Bank and the International Monetary Fund (IMF), the
Organization for Security and Co-operation in Europe (OSCE), the EU and
NATO or other regional organizations (e.g., the Organization of African
States (OAS), the Organization of the Islamic Conference (OIC), the Arab
League, the Gulf Cooperation Council, the Economic Community Of West
African States (ECOWAS), or the Association of South East Asian Nations
(ASEAN)), but also loose multilateral fora such as the Balkan Stability Pact
or the G8 can be included.
• The second category covers the role of states, be it a single state or an ad hoc
coalition of states. A prime example of a state acting as a third party is the
USA in various conflicts in the world (e.g. the Middle East, Kashmir, Cyprus,
Bosnia, Kosovo, Northern Ireland), but also smaller states such as New Zea-
land, Switzerland, Sweden, or Norway have often played a significant role in
mediating conflicts (e.g., the Oslo Peace Process). Examples of coalitions of
states are the Bosnia and Kosovo Contact Group, which both aimed at co-
ordinating the policies of major powers towards these conflicts. Sometimes
even kin states may develop a third party-like approach when they move
from a partisan stance to a more impartial position. For instance, in the case
of Northern Ireland, both Ireland and Great Britain were able to establish a
common platform; their joint activities and proposals can therefore be por-
trayed as third party involvement.
• The third category includes various kinds of non-state actors, among them
international NGOs (e.g., Red Cross, the Quakers, International Alert), pri-
vate persons (e.g., ‘elder statesmen’), academic institutions and foundations,
churches, trade unions, or the business community. Prominent examples are
the efforts of the Carter Center in the USA, which has been part of many
conflict resolution activities around the globe, the role of Norwegian ac-
ademics prior to the negotiations of the Oslo Agreement (1993) between
Israel and the PLO, the mediation efforts of the Italian brotherhood Saint
Egidio in Mozambique or Kosovo, and the role of the Swiss-based Henry
Dunant Centre for Humanitarian Dialogue in the Aceh conflict. Another
typical non-state actor approach has been the problem-solving workshops
conducted by academics and practitioners with local actors that have been
used frequently in the case of Cyprus and Israel/Palestine (Kelman 2002).
In reality, one will always see a multiplicity of third parties involved in a single
conflict. Often they address different levels (high level, mid level, or grassroots
level), issues (political, social, cultural, or economic issues), as well as different lo-
cal actors (governments, parties, armed groups, civil society, etc.). In most cases,
Members of the Bosnia Contact Group were the USA, Russia, France, Great Britain,
and Germany. The Kosovo Contact Group also included Italy.
third parties do not only rely on persuasion, but on incentives and disincentives.
Incentives are, for example, security guarantees, monitoring and verification of
disarmament, support for confidence-building measures, assistance to institu-
tional reforms, humanitarian and development aid, as well as the long-term pros-
pect of advanced economic and political cooperation. Disincentives usually refer
to sanctions or military means, but may also include ‘softer’ measures such as
human rights monitoring, observer missions, or, more specifically, the establish-
ment of war crimes tribunals. In practice, third parties use a mix of instruments,
depending largely on their own capabilities, their political will, and the local situ-
ation.
However, two types of activities should be distinguished. Firstly, all efforts that
aim at achieving an agreement and stopping violent escalation may be termed
settlement activities. Secondly, all efforts that aim at securing, observing, imple-
menting, enforcing, and maintaining an agreement may be termed implemen-
tation activities – these can include temporary and permanent measures. Both
phases can be conducted by the same or by different third parties.
The distinctions between actors and activities can now be translated into a ty-
pology of third party involvement recognizing six forms of involvement (Table 1).
Table 1 Types of Third Party Involvement

Settlement activities Implementation activities
International organization 1 2
State(s) 3 4
Non-state actor(s) 5 6
This typology will be used to help classify the third party involvement in each of
the cases, and to summarize the third party activities seen in these conflicts.
A Northern Ireland
Since the Anglo-Irish Agreement (1985) and the establishment of the Intergov-
ernmental Conference (1985), the British and Irish governments were able to
overcome their differences step-by-step and issued various joint declarations
and proposals to the local conflict parties (e.g., Downing Street Declaration 1993,
Framework Documents 1995) that paved the way for the 1998 Good Friday Agree-
ment. These joint initiatives allowed both governments to go beyond their his-
torical partisan roles. Thus, in cases where they acted together or, at least, agreed
upon certain measures, both governments behaved as if they were a third party.
(Type 3 in Table 1). In the 1990s, the US government became more and more
involved. Although President Carter had already offered his assistance in 1977, it
was President Clinton who declared that he would “take a more active role” in the
Northern Irish conflict (Type 3). The US involvement, however, remained a mix-
ture of direct and indirect measures. From time to time, the US president himself
engaged actively (e.g., through phone calls, invitations to Washington, Clinton’s
visits to Northern Ireland), but more often the US administration used unofficial
envoys such as the Morrison Delegation in 1993–1994 (named after a former US
congressman). In a similar way, the former US Senator George Mitchell served
as an unofficial channel, but developed a far more independent role that made
him acceptable to all parties as the key mediator for the peace talks. He acted,
therefore, much more like a non-state actor than an official governmental envoy
(Type 5). For the implementation phase, state and non-state actors were again
the most relevant (Types 4 and 6): the British and Irish governments remain the
key actors, while the USA still plays an important role, in particular in cases of
crises. The use of non-state actors such as former foreign politicians and states-
men continued. Mitchell kept his role as mediator, de Chastelain became chair-
man of the Independent International Commission on Decommissioning (IICD),
and later the former Finnish President Martti Ahtisaari and the former South
African leader of the African National Congress (ANC) Cyril Ramaphosa were
appointed as international observers in order to verify decommissioning of the
Irish Republican Army (IRA). Finally, the Independent Commission on Policing,
headed by the British Conservative Chris Patten, comprised, inter alia, of inter-
national experts.
B Bougainville
During the settlement phase, the key third party role was played by New Zealand
and Australia (Type 3), as both took the lead in the so-called ‘Group of Friends’
which also included the Solomon Islands, Vanuatu, and Fiji. They supported the
peace negotiations and provided troops for peacekeeping efforts, firstly by the
New Zealand-led Truce Monitoring Group (October 1997–April 1998) after a
truce between the Papua New Guinea government and the secessionists, and sec-
ondly, by the Australia-led Peace Monitoring Force (from April 1998). Further-
more, the final settlement (Bougainville Peace Agreement, September 2001) was
facilitated by a small UN observer mission (Type 1), established in August 1998,
which co-chaired the peace talks with an Australian commander. Basically, the
same actors remain in place for the implementation period (Type 2 and 4). Aus-
Irish Times, 29 October 1992, quoted by Ruane and Todd 1996: 277.
The Morrison Delegation mainly put pressure on the Nationalist leadership to de-
nounce the use of violence, but it also paved the way for extending a US visa to Sinn
Fein leader Gerry Adams. See McGarry and O’Leary 1996: 330.
Mitchell was first appointed as member of the International Body on Arms (so-called
Mitchell Commission), together with the Canadian General John de Chastelain and
the former Finnish Prime Minister Harri Hokeri (since 1995).
For international involvement in Northern Ireland, see McGarry and O’Leary 1996;
Ruane and Todd 1996: 266-289.
tralia remains largely responsible for peacekeeping, while the UN has a formal
role in monitoring decommissioning. In addition, various development agencies
(New Zealand Official Development Assistance (NZODA), Australian Agency
for International Development (AusAID), UNDP) and other multilateral donors
(World Bank, IMF, Asian Development Bank) became engaged in Bougainville,
often in cooperation with numerous international NGOs, mainly from Australia
(Type 6).
C Mindanao
When compared to the other case studies, Mindanao is characterized by only
having seen low-level international involvement. For both settlement and imple-
mentation, the Organization of the Islamic Conference (OIC) can be identified
as the main third party (Type 1 and 2). As early as 1972, the OIC expressed its
concerns with the conflict, and in 1976, the OIC brokered the Tripoli Agreement
between the government and the Moro National Liberation Front (MNLF), the
largest secessionist movement. Again, in 1992-1993 and in 1996, the OIC assisted
peace talks and follow-up agreements in order to establish a functioning autono-
mous Muslim region in the Southern Philippines. In the 1996 agreement, the OIC
got a special role in sending representatives to the Joint Monitoring Committee,
which should review the implementation process. Moreover, via the OIC, vari-
ous states got involved (Type 3). In the 1970s, Libya played a key role, followed by
Indonesia in the 1990s, and by Malaysia since 2001. However, until now, all efforts
by third parties to settle the conflict and to implement the existing agreements
have not succeeded.10
D Bosnia
From the beginning of the war in Bosnia in 1992, the degree of international in-
volvement and the number of third parties involved was extremely high. The in-
ternational community, in particular the EU and UN, launched various peace ini-
tiatives to stop the killing and fighting (Type 1). Moreover, the UN got involved on
the ground quite early, with peacekeeping and humanitarian activities (UNPRO-
FOR11 I and II, UN safe havens). The first mediation effort was undertaken by the
EU (then EC) under the Portuguese presidency (Cultileiro plan in March 1992),
followed by joint EU-UN actions (Vance-Owen plan in March 1993 and Owen-
Stoltenberg plan in August 1993). After the failure of these plans, the Bosnian
Contact Group and the US government emerged more and more as the central
players in achieving a settlement (Type 3). The Contact Group plan (July 1994),
the US-sponsored Washington agreement between Croats and Muslims (1994)
For international involvement in Bougainville, see Smith 2002; Eagles 2002; Regan
2002.
10 For international involvement in Mindanao, see Turner 2003.
11 United Nations Protection Force.
and, finally, the increased military pressure by NATO can be seen as steps lead-
ing towards the Dayton Peace Agreement (DPA) (November 1995), which was
largely brokered by the USA.12 For implementing the DPA, international organi-
zations became the main agencies (Type 2). The EU, UN, United Nations High
Commissioner for Refugees (UNHCR), NATO, OSCE, and the Council of Europe
(particularly via the European Court of Human Rights) each got their own role
and responsibility in post-conflict peace-building and establishing a power-shar-
ing structure. Bosnia also became a market place for international NGOs that
received funding for civil society activities, for refugee return, for humanitarian
aid, for human rights, or for interethnic reconciliation (Type 6).13 However, major
Western states, in particular the USA, remain important throughout the imple-
mentation process (Type 4) – indirectly by supporting IOs and NGOs, as well as
directly by backing moderate and non-nationalist parties.14
E Kosovo
In the 1990s, the Kosovo issue received some international attention, mainly from
the OSCE/CSCE, the USA, and the EU, but this was not sufficient to prevent the
conflict from escalating in 1998-1999. During this escalation phase, the activities
of the international community to settle the conflict increased substantially. As in
the Bosnian case, international organizations and states combined their efforts.
Again, the Contact Group served as a key forum for multilateral consultations
(Type 1 and 3): international mediators and envoys were appointed (EU, USA,
Russia), observers were deployed (OSCE), the UN Security Council issued vari-
ous resolutions in 1998 and called upon the parties to stop the violence, interna-
tionally organized and sponsored negotiations took place and, finally, military
force was used (NATO) – to name just the most prominent actions. In ending the
war and implementing a political solution, the UN got a much more prominent
role than before (Type 2). Following UN Security Council Resolution 1244, the
UN established an interim administration (UNMIK), which includes OSCE and
EU participation, while NATO was mandated to ensure security (Kosovo Force,
KFOR). In addition, many international NGOs established projects on confi-
dence-building, civil society, or human rights (Type 6). Once again, the influence
of the USA and other states remains essential for any political progress (Type 4).
F Macedonia
The same multitude of actors can be found with regard to the settlement in Mace-
donia. Early on, the UN and the OSCE were involved in prevention efforts (Unit-
12 For international conflict management in Bosnia, see Gow 1997; Berg and Shoup
1999. For personal accounts of involved mediators, see Owen 1995; Holbrooke 1998.
13 For the role of NGOs in Bosnia, see Belloni 2001.
14 For the implementation of the DPA, see Chandler 1999; Cousens and Carter 2001;
Bose 2002; Schneckener 2002a: 290-299.
ed Nations Preventive Deployment Force/UNPREDEP, OSCE Spillover Mission),

and the EU granted financial and economic assistance to the country. When in
spring 2001 local fighting between Albanian extremists and the Macedonian se-
curity forces started, the EU, NATO, and the USA (Type 1 and 3) became active in
facilitating and in mediating the conflict, leading eventually to the Ohrid agree-
ment (August 2001). To a lesser extent than in Bosnia and Kosovo, external ac-
tors are involved in the implementation: the EU, NATO, and the OSCE assist and
monitor the peace process with a particular emphasis on decommissioning. At
the same time, international NGOs continue their efforts in interethnic dialogue
or media training, which could serve as a societal basis for power-sharing (Type
2 and 6).15
G Gagauzia
In comparison to the Balkan cases, international involvement has been rather
limited in Gagauzia. For settling the conflict, the OSCE via its Mission to Mol-
dova, the High Commissioner on National Minorities (HCNM), and, to a lesser
extent, the Council of Europe (Type 1) proved helpful in making proposals and
fostering the dialogue between the Moldovan government and local politicians
which finally led to an agreement in 1994. During the implementation process,
the same international organizations remained involved (Type 2), the OSCE play-
ing the most important role. Since 2001, a UNDP programme for Gagauzia sup-
plements the external efforts. There are, however, few significant international
NGO activities related to the political solution.
H South Ossetia
Third party involvement has been more widespread in South Ossetia, but still
much less prominent than in the Balkans. The Eltsin-Shevardnadze agreement
in 1992 was mainly achieved by OSCE moderation and Russian pressure (Type 1
and 3). Since then, the Russia-led Joint Peace-Keeping Force (JPKF) provides se-
curity in the region. Russia and the OSCE became members of the Joint Control
Commission (together with South Ossetia and the Georgian government), which
acts de facto as administration for the region until a final solution is found (Type
2 and 4). Moreover, the EU and UNDP (since 1997) started to provide economic
and financial assistance, while the UNHCR supports repatriation efforts. In order
to foster confidence-building at various levels, international NGOs are active in
South Ossetia (Type 6). One prominent example has been the Georgian-South
Ossetian Dialogue project (1995-1999) run by the US-based NGO Conflict Man-
agement Group and the Norwegian Refugee Council (NRC).
15 For international involvement in Macedonia, see Lund 2000; Daftary 2001; 2003;
Schneckener 2003.
1 Discussion
Table 2 summarizes the profiles of the case studies and allows for some inter-
esting observations. First, the highest level of international involvement can be
found in the Balkan cases, especially Bosnia and Kosovo. Third parties are less
prominent in Bougainville, South Ossetia, and Northern Ireland, while in the
cases of Gagauzia and Mindanao, their role has been rather limited or even mar-
ginal. Second, the sample of cases is characterized by a dominant role of inter-
national organizations (IOs). In all cases, with the notable exception of Northern
Ireland, IOs are involved in both settlement and implementation activities, most
often as key players. However, from this analysis it seems that IOs are particularly
relevant when it comes to implementation. Here they clearly act as lead agencies,
responsible for various aspects of establishing a power-sharing system (Bougain-
ville, Bosnia, Kosovo, Macedonia, Gagauzia, South Ossetia). Third, states as third
parties have been important in almost all cases for achieving a settlement. In
particular, great powers like the USA as well as regional powers (e.g., Australia,
Indonesia, Russia, West European states) have proven to be vital when diplomatic
or even military pressure was needed to bring the local parties to the negotiation
table. States are less prominent when it comes to implementing the power-shar-
ing agreement (only in five out of eight cases), and even then they often play a
more secondary, supplementing role and leave the day-to-day business to the
mandated IOs (e.g., Kosovo, Bosnia). Fourth, the contribution of non-state actors
is primarily in the field of implementation, and they rarely are part of settlement
activities. Here, again, the case of Northern Ireland is exceptional. International
NGOs are often used during the implementation phase by the leading IOs for
various purposes connected with the political solution: provision of expertise,
human and minority rights monitoring and education, media training, civil so-
ciety projects, support and establishment of local NGOs, interethnic or inter-
community dialogue, confidence-building and reconciliation efforts, etc. In other
words, they allow the IOs to ‘outsource’ some of their activities. In this respect,
some international NGOs cannot be regarded as independent actors. In general,
NGOs and other non-state actors tend to focus on grassroots or mid-levels in
order to widen public support for peace agreements and power-sharing (bottom-
up approach), while IOs and states address mainly the leadership, the key deci-
sion-makers and the group elites (top-down approach). For that reason, NGOs
have been excluded from Table 3, which concentrates on the most relevant third
parties for settlement and implementation identified in each case. Nevertheless,
as the case of Northern Ireland shows, non-state actors can become a key third
party at the top decision-making level, if they are asked by all sides to play that
role. In contrast to IOs and states, non-state actors are not able to ‘invite them-
selves’ – either by offering substantial aid and resources, by resolutions or state-
ments, by diplomatic means, or by the use of pressure.
Table 2 Profile of the Complex Power-sharing Cases

Types of third-party involvement
Conflict 1 2 3 4 5 6
Northern Ireland X X X X
Bougainville X X X X X
Mindanao X X X
Bosnia X X X X X
Kosovo X X X X X
Macedonia X X X X
Gagauzia X X
South Ossetia X X X X X
Table 3 Identified Key Third Parties

Conflict Settlement Implementation
Northern Ireland British-Irish initiatives, USA,
British-Irish initiatives, USA,
Mitchell Commission Mitchell, IICD, Patten Com-
mission
Bougainville Australia, New Zealand (Group New Zealand (NZODA), Aus-
of Friends), UN tralia (AusAID), UN, UNDP
Mindanao OIC, Indonesia, Libya, Ma- OIC, UNDP
laysia
Bosnia EU, UN, USA, Bosnia Contact UN, UNHCR, EU, OSCE, USA,
Group, NATO NATO, ECHR
Kosovo NATO, EU, USA, Kosovo Con- UN, UNHCR, EU, OSCE, USA,
tact Group, CSCE/OSCE, UN NATO
Macedonia EU, NATO, USA, OSCE, UN EU, NATO, OSCE
Gagauzia OSCE, HCNM, Council of OSCE, HCNM, Council of Eu-
Europe rope, UNDP
South Ossetia Russia, OSCE Russia, OSCE, EU, UNDP,
UNHCR
III Approaches of Third-party Involvement

What does third-party involvement look like in practice? What is the contribution
of third parties to complex power-sharing regimes? In all cases, the involvement
of third parties has been institutionalized in different ways. The third parties have
established special bodies, missions, offices, or commissions concerned with the
conflict (Table 4), they have deployed civilian as well as military personnel on the
ground in order to end violence, to control a ceasefire, to assist implementation,
etc. They are, thus, committed to stay in the region for a long period, in most
cases for years (e.g., Kosovo, Bosnia, Bougainville, Macedonia, South Ossetia,
Gagauzia).
Table 4 Institutionalized Third-party Involvement

Conflict Institutionalized Involvement
Northern Ireland • Intergovernmental Conference (since 1985)
• IICD (since 1995), international weapons inspectors
• Independent Commission on Policing for Northern Ireland
Bougainville • Truce Monitoring Group (October 1997 – April 1998)
• Peace Monitoring Force (since April 1998)
• UN Political Office (UNPOB, since August 1998)
• Peace Process Consultative Committee (since 1999)
Mindanao • Joint Monitoring Committee (includes OIC representatives)
Bosnia • UNPROFOR I and II (1992-1995)
• Peace Implementation Council (since 1995)
• Office of the High Representative (since 1995)
• NATO-led IFOR (1995) and SFOR (since 1996)
• OSCE Mission to Bosnia and Herzegovina
• UN Police Task Force (1996-2002)
• EU Police Mission (since 2003)
Kosovo • CSCE Mission on Long Duration (1992-1993)
• Kosovo Verification Mission (October 1998–March 1999)
• UNMIK (since June 1999)
• NATO-led KFOR (since June 1999)
• OSCE Mission to Kosovo
• EU Agency for Reconstruction
Macedonia • UNPREDEP (1992-1998)
• OSCE Spillover Monitor Mission (since 1992)
• NATO missions (Essential Harvest, Amber Fox, Allied Harmony
2001-2003)
• EU Military Mission (since March 2003)
• EU Monitoring Mission (since March 2001)
Gagauzia • OSCE Mission to Moldova (since 1993)
South Ossetia • Georgian-Russian Joint Peace-Keeping Force (since 1992)
• OSCE Mission to Georgia (since 1992)
• Joint Control Commission (since 1992)
These bodies, however, follow rather different approaches as they have to fulfill
distinct tasks, be it during the settlement or the implementation period. Gen-
erally, seven approaches or areas of activity can be distinguished, which are all
relevant for both settlement and implementation. This overlap is due to the fact
that the implementation process has to be understood as an ongoing negotiation
process, and that often follow-up agreements, clarifications, and further com-
promises are needed. In many cases, the peace process suffered serious set-backs
after a major agreement had been signed (Northern Ireland, Mindanao, Bosnia,
Kosovo, Macedonia, Gagauzia). Therefore, in many cases, these approaches are
still useful in the aftermath of an agreement when it comes to implementation.
A Aid and Assistance

The most basic and simple form of involvement is to offer aid and assistance to
the conflicting parties. This includes humanitarian and development aid, recon-
struction efforts, macro-financial and long-term economic support, preferential
trade relations, technical assistance, various forms of political cooperation, as
well as regional integration processes. Often international aid has been mobilized
by multilateral donor conferences, hosted by the EU or the World Bank (e.g.,
Bosnia, Kosovo, Macedonia). For example, in Macedonia, the issue of the do-
nor conference was explicitly made conditional by the EU on certain steps in the
implementation process being achieved or taken.16 In all cases, we can observe
these activities, which can work as incentives, in particular for national govern-
ments, to find peaceful solutions. The message given is: ‘if you are prepared to
keep peace, we as the international community are prepared to finance a ‘peace
dividend’ which will ideally improve social and economic conditions.’
In Northern Ireland, the EC/EU financed cross-community and cross-border
(North-South) projects for many years. After the Anglo-Irish Agreement in 1985,
the USA government set up an ‘International Fund for Ireland.’ In Bougainville,
the development agencies of Australia and New Zealand became active. Aus-
Aid as the largest donor spend aud 134 million (1997-2002) and NZODA gave
nzd 10.2 million (2000-2001) to Bougainville. Moreover, the EU has commit-
ted eur 9.1 million (2000-2001) for post-conflict rehabilitation and re-building
of infrastructure, channelled via the UNDP (Eagles 2002). In Mindanao, UNDP
serves as lead agency and offers a multilateral framework for bilateral donors;
in 1997 approximately usd 500 million had been pledged by donors. In South
Ossetia, the EU and UNDP set up their own rehabilitation projects and tried to
drive the local parties and Georgian government officials towards cooperation
in order to receive the funding. The EU projects are particularly directed to the
re-building of infrastructure between South Ossetia and the rest of Georgia (e.g.
transportation, energy, or communication). In Gagauzia, international donors
are less prominent – prior to the political solution in 1994, Turkey, for example,
promised investments and development assistance for the region.17 However, due
to various implementation problems in Moldova, it took until 2001 before local
authorities were able to agree with UNDP on a regional development programme
named ‘Gagauz-Eri.’
The largest amount of aid thus far has been spent in the Balkans, mainly by the
EU and, since 1999, largely via the Stability Pact for Southeastern Europe. During
16 After the Ohrid agreement, a donor conference was planned for late 2001. The EU,
however, cancelled the meeting because of delays in implementing the constitution-
al changes. The conference eventually took place in March 2002. See Schneckener
2003: 151.
17 In 1998, Turkish President Demirel finally announced that his government would
grant Moldova a usd 35 million credit (including usd 15 million for improving the
water supply system in Gagauzia) (Järve 2007).
the first years after the settlement (1996-1999), Bosnia alone received USD5.25
billion from international donors, roughly 60% was used for reconstruction, 14%
for economic reforms and recovery, and 16% for the establishment of institutions
(Cousens and Carter 2001: 88-89). In the case of Macedonia, the EU had already
transferred eur 452.3 million between 1992 and 2001 for supporting businesses,
the financial sector, infrastructure projects, and institution-building. During the
crisis in spring–summer 2001, the EU combined short-term humanitarian aid
with the long-term perspective of EU integration via the Community Assistance
for Reconstruction, Development and Stabilization programme (CARDS) and
the Stabilization and Association Agreement (SAA).18 In June 2001, the EU of-
fered a substantial financial aid package on the basis of the SAA (eur 42 million
aid plus a macro-financial assistance of eur 50 million) if the local parties are able
to solve the crisis peacefully (Schneckener 2003: 147). The example of the Mace-
donian crisis underscores the unique role of the EU with regard to the Balkans.
Here, the prospect of EU membership serves more and more as an incentive for
local elites to change their behaviour.
However, these cases also show that aid and assistance can have negative side-
effects or may even be counter-productive. First, aid may destroy local capacities
and local markets and thereby make the country or the region even more depen-
dent on international donors. Second, local elites may get accustomed to external
aid, which fosters a ‘rent-seeking’ mentality. Third, local politicians do not have
to act responsibly as long as the international community is prepared to pay for
the problems. Fourth, aid may be abused by local actors for black market activi-
ties, corruption, or crime; in particular, humanitarian aid is often used for wrong
purposes. Fifth, aid may increase the tensions between communities when one
side feels neglected or marginalized.19
B Facilitation
Third parties usually restrict themselves to providing ‘good offices’ as well as
technical or logistical assistance for dialogue, pre-negotiations, and, finally, of-
ficial negotiations between the conflicting sides. They may offer advice and ex-
pertise, but they do not try to dominate the agenda and to draft agreements.
Facilitation can be conducted in an informal or formal manner. The facilitator
can attend the negotiations or stay behind the scenes, and can also be used by
the conflict parties as a messenger, often via secret channels. Frequently, facilita-
18 While CARDS serves as the main channel for EU’s financial and technical coopera-
tion with Southeastern Europe, the SAA offers regular economic aid, improved trade
relations and an advanced political dialogue. Macedonia signed the SAA in April
2001. See Schneckener 2003: 140-141.
19 One example is the internationally financed programme in Mindanao, which was
directed exclusively to the re-integration of MNLF members and their families which
caused resistance and complaints by other groups and local authorities. See Turner
2003.
tion takes place outside the conflict region and the meetings are sponsored by
the host government or by an international body, and sometimes independent
research institutes or academies are used as locations for informal gatherings.
Facilitation thus provides room for ‘talks about talks’ or the exchange of unofficial
or unrecognized documents (so-called non-papers) that do not have a binding
character for both sides.
The Bougainville case may serve as a paradigmatic example: In early 1997, New
Zealand facilitated regular talks between the government and the rebels, which
led to the Burnham Truce (October 1997). The Australian government supported
these efforts with the Peace Transport Fund (PTF), which was set up to finance
the transportation of key actors and their advisors to meetings outside Bougain-
ville. Despite some problems in administration and increasing costs because of
the lengthy negotiations, the PTF proved essential, as Eagles (2002) notes: “the
PTF was flexible and responsive and it facilitated the movements of negotiat-
ing teams at short notice, which was crucial in maintaining the momentum for
peace and dialogue.” Similarly, in the case of Mindanao, the OIC always under-
stood itself to be a facilitator trying to initiate a dialogue between the government
in Manila and the MNLF. For instance, the OIC-sponsored exploratory talks in
Libya (October 1992) and in Indonesia (April 1993) can be viewed as typical fa-
cilitation efforts. The same is true for the more recent attempts by Malaysia to
establish a channel between the government and the more radical Moro Islamic
Liberation Front (MILF) (March 2001). In Northern Ireland, the so-called Mitch-
ell Commission played a facilitating role by producing reports and guidelines not
only for decommissioning, but also for a possible package solution. That material
was used as a basis for peace talks, which eventually led to the 1998 Agreement
(McGarry and O’Leary 1996: 358-364). In a similar way, external expertise has
been used in Gagauzia: the OSCE and the Council of Europe have issued vari-
ous reports and recommendations for Gagauz autonomy. This practice contin-
ued through the implementation process, which proved to be rather slow and
strewn with obstacles.20 In Ossetia, the OSCE facilitated the establishment of
various fora for a Georgia-South Ossetia dialogue (e.g., Plenipotentiary Forum
since 1997) or practical cooperation (e.g., Joint Law-Enforcement Coordination
Centre since 2000).
In general, facilitation pre-supposes that the parties concerned may be unable
to engage directly and without external help, but they are in principle willing to
negotiate and even to compromise. Facilitation also implies that the parties may
need some input from outsiders, but are still able and prepared to find their own
solutions. However, as most cases show, facilitation efforts are often not suffi-
cient for either achieving or for implementing a settlement (Mindanao, Northern
20 For example, in early 2002, the European Commission for Democracy Through Law
(Venice Commission), an international body of lawyers sponsored by the Council of
Europe, made an important contribution when it issued an opinion to the Moldovan
draft of the constitutional amendments for ensuring Gagauz autonomy. See Järve
2007.
Ireland, Gagauzia, or South Ossetia). Here, facilitation worked either as a starting

point for third party intervention or as one activity among others. This can also
be illustrated by the case of Macedonia. When the crisis started in early 2001,
the EU High Representative, Javier Solana, tried to restrict his role to facilita-
tion, stating: “The European Union is here to help, not to be a mediator in this
dialogue.”21 Solana, together with NATO Secretary General Lord Robertson, paid
several visits to Skopje in order to foster dialogue, but had to learn that the EU
had to take a more active and decisive approach, which finally led to the appoint-
ment of a EU Special Representative in June 2001 as a key mediator.
C Mediation
In contrast to facilitation, mediation is characterized by more active third party
engagement. Third parties actively attempt to broker agreements and do not
leave the solution solely to the conflicting parties. Therefore, as Bercovitch (2002:
5) points out, mediation “turns a dyadic relationship into a triadic interaction”,
whereas ideal-type facilitation just supports inter-group dialogue. In mediation
as understood here, third parties often prepare and chair the negotiations, they
table their own proposals, in some cases they draft settlements or even consti-
tutions and special laws. They try to manage and direct the peace process as a
whole, be it in a settlement or implementation phase. Mediation can have vari-
ous forms: it can be conducted by shuttle diplomacy or by a permanently present
mediator and negotiator, by multi-track diplomacy, or by a single forum. Media-
tion remains a non-coercive form of intervention, i.e., in the end, the local par-
ties cannot be forced to agree. Third parties, however, usually use incentives and
disincentives in order to foster cooperation.
The paradigmatic case here is Northern Ireland, where over many years vari-
ous mediation activities took place. Firstly, during the 1970s and 1980s, largely
conducted by the British government via the Northern Ireland Office, then with
the increasing inclusion of international components. In the end, US Senator
Mitchell served as the central mediator during the whole negotiation process
(1996-1998), supported by the Irish, British, and US governments. This setting
also shows a coordinated division of labour: The Irish and US governments acted
to moderate the Nationalist side, while the British government was primarily
concerned with the Unionists. These joint efforts have had to continue after the
Good Friday Agreement, since the implementation has been characterized by
various setbacks and crises, which have repeatedly led to the temporary suspen-
sion of the local power-sharing executive (Schneckener 2002a: 306-307). In Bou-
gainville, after the facilitation phase overseen by New Zealand and Australia, the
head of the UN Political Office (UNPOB), Ambassador Noel Sinclaire of Guyana,
and the Peace Monitoring Force Commander co-chaired the peace talks (March
2000–June 2001), which led to the Bougainville Peace Agreement of August
2001. UNPOB, invited to do so by the local parties, organized separate meetings
21 Solana at a press conference in Skopje (2 April 2001). See Schneckener 2003: 145.
with each side (in particular with hardliners), and round-table talks as well as
formal and informal rounds of negotiations. Moreover, the UNPOB established
and chaired the Peace Process Consultative Council (PPCC), which worked as
a confidence-building mechanism. During this process, the UNPOB from time
to time issued proposals and paved the way for a common understanding of the
problems. The role of UNPOB is captured in Eagles’ observation that “the parties
came to rely on the UN, but not to depend upon it” (Eagles 2002).
In a similar way, mediation has been conducted by the OSCE in Gagauzia (to-
gether with the HCNM) and in South Ossetia.22 As far as power-sharing agree-
ments are concerned, these efforts were more successful in the former than in the
latter case. In contrast, the mediation process in regard to Bosnia and Macedonia
became a rather complex undertaking, involving a whole range of external actors
and different initiatives. In Bosnia, the various joint EU and UN peace plans were
the most prominent, but unsuccessful, attempts to stop the war. A major reason
for this failure was that the joint mediation teams (Vance-Owen and Owen-Stol-
tenberg) did not always receive the necessary support from major powers (in
particular the USA, Russia, and France) and that the international community
could not agree to back these efforts with credible military threats (Gow 1997).
Each group (Croats, Serbs, and Muslims) could rely on certain ‘advocates’ on the
international stage and were thus able to play the third parties off against each
other. Only when the USA (together with the Contact Group) took the lead and
applied military force, was it possible to break the deadlock. In the case of Mace-
donia, the international community seemed to have learnt the lesson. There, after
some coordination problems at the beginning of the crisis, the EU and the USA
appointed the special envoys (former French Defence Minister Francois Léotard
and former US ambassador James Pardew) in June 2001, who worked together as
a team based on a common platform. They were supported by the OSCE’s special
representative Max van der Stoel, by international lawyers and, in particular, by
NATO’s so-called ‘technical contacts’ with the Albanian Kosovo Liberation Army
(Ushtria Çlirimtare E Kosoves, UÇK), who were critical in establishing an indirect
channel between the extremists and the Macedonian government (Schneckener
2003: 147-148). After a phase of pre-negotiations, the joint EU-USA team finally
chaired the successful Ohrid negotiations (August 2001). The comparison be-
tween Bosnia and cases such as Macedonia or Northern Ireland emphasizes that
it is crucial for mediation that the relevant third parties are able to combine their
efforts, to develop a common language towards the local actors and, thereby, ex-
22 According to its mandate, the main objective of the CSCE/OSCE mission in Georgia
was to promote negotiations between the conflicting parties in order to achieve a
political settlement. In March 1994, this mandate was amended to specifically inten-
sify discussions with all parties to the conflict, including the organization of round
tables, and to make recommendations regarding the convening of an international
conference under OSCE auspices, aimed at the resolution of the conflict, including
the definition of the political status of Southern Ossetia.
ert a maximum of diplomatic pressure during the settlement and implementation

phases without necessarily using the threat of force.
D Monitoring
Third parties observe the peace process as a whole or by concentrating on specific
elements. Typical examples are the monitoring of a ceasefire, of demobilization
and disarmament, of borders, of human rights, or of elections. They verify certain
steps in implementation, they report to the local as well as international actors
about violations of an agreement and delays in implementing it. Monitoring is
therefore mainly about fact-finding and information gathering. These efforts can
be understood as a response to the fact that conflicting parties often deeply mis-
trust each other and need a credible independent authority to observe the pro-
cess. Monitoring is thus a contribution to confidence-building and an assurance
for each side that no side can readily get away with disregarding agreed principles
and policies. Monitors often work as ombudspersons who collect complaints
about violations and seek dialogue with the authorities or ethnic communities
concerned to verify and/or rectify the situation.
The work of the OSCE field missions illustrates this approach. Most promi-
nently, the OSCE was responsible for post-agreement election monitoring in
five of the eight case studies (Gagauzia, Ossetia, Kosovo, Bosnia, and Macedo-
nia). Furthermore, in Gagauzia and South Ossetia, the OSCE is still responsible
for other monitoring tasks, tasks that have been gradually extended over time.
According to the renewed mandate (1999), the OSCE mission to Moldova with
regard to Gagauzia monitors interethnic relations, observes compliance with in-
ternational obligations and commitments regarding human and minority rights,
and cooperates with local authorities in cases of complaints in the ‘human dimen-
sion’. Due to serious implementation problems, the OSCE emphasized in 2000
its willingness to monitor “the political situation in the autonomous region of
Gagauz Yeri and the Moldovan-Gagauzian centre-region relations in the areas of
tax revenues, budget allocations, public property ownership, and adjustment of
legislation.”23 The OSCE mission in Georgia (since December 1992) has the tasks
of promoting respect for human rights and the development of the rule of law,
including the establishment of an independent judiciary in Georgia as a whole.
For the Ossetian conflict, the OSCE gathers information about the military situ-
ation, investigates violations of the ceasefire and maintains close contacts with
local authorities. In Macedonia, the OSCE has been present since 1992, when the
international community feared that the war in Bosnia could spill over to Mace-
donia. Later, in the wake of the crisis in 2001, the mission and its mandate were
significantly expanded in order to monitor the situation (up to 200 members,
starting in October 2001). During the Kosovo conflict, the OSCE was used twice
in order to observe human rights violations and stabilize the situation on the
ground. Firstly, in the early days of the conflict, the CSCE Mission of Long Du-
23 See OSCE Mission Report on Human Dimension Activities, October 2000.

ration in Kosovo, Sandžak, and Vojvodina (September 1992–June 1993) had re-
ported, inter alia, about the humanitarian and military situation in Kosovo before
the mission was expelled by the Yugoslav government. Secondly, at the peak of
the crisis in 1998, the Belgrade regime had to allow for an international presence
in Kosovo as a consequence of the so-called Holbrooke Agreement (13 October).
The OSCE Kosovo Verification Mission (KVM), which could have comprised up
to 2,000 unarmed observers, would have then been the largest OSCE mission
ever.24 The KVM had a fairly broad mandate for monitoring the ceasefire and
verifying compliance with the agreement, in particular with the military arrange-
ments made. Moreover, it was foreseen that the KVM would supervise elections
in Kosovo and assist in finding a political solution.
But a number of other actors are also involved in monitoring. In Mindanao,
the OIC was given an overall monitoring role for the 1996 agreement (para 12)
and a special representation within the Joint Monitoring Committee (para 13).
In Bougainville, the Truce Monitoring Group and, later, the Peace Monitoring
Force have been responsible for observing the military situation. According to
the Bougainville Peace Agreement, the small UNPOB (eight persons, including
four international staff members) shall monitor and verify the process of disar-
mament, as well as assist the re-integration and rehabilitation of ex-combatants.
In Northern Ireland, international monitors have similar tasks. The IICD has to
oversee the decommissioning of paramilitary weapons. Additionally, two former
statesmen were asked to inspect arms dumps of the IRA and report to the IICD
on progress in decommissioning. In a much more comprehensive and forceful
way, NATO played a monitoring role in Macedonia. Initially, NATO sent about
4,000 soldiers to collect and destroy the weapons of the Albanian extremists;
weapons which had to be handed over voluntarily (Operation Essential Harvest,
27 August–26 September 2001). NATO, via its smaller follow-up missions (Am-
ber Fox, 27 September 2001–16 December 2002, and Allied Harmony, 17 Decem-
ber 2002–31 March 2003), monitored the ceasefire, observed the general military
situation and secured the protection of the unarmed OSCE monitors.25
The case studies show that monitoring can be done in rather different and
flexible ways, is conducted almost entirely by international organizations, rang-
es from small observer teams to large, armed missions that are able to defend
themselves. Monitoring does imply observation and evaluation of compliance
with agreements (including ceasefires), but not the enforcement of compliance.
However, non-compliance, despite international monitoring, can still have con-
sequences for local parties. For example, monitors can publicly criticize certain
24 The KVM was supported by NATO air surveillance and by a NATO extraction force
which was deployed in Macedonia to ensure the protection and the evacuation of the
OSCE personnel if necessary. The KVM, however, did not reach the scale of 2,000
observers before it was removed in March 1999 due to the military intervention by
NATO.
25 Since April 2003, the EU has taken over NATO responsibilities by deploying the EU
Military Mission (Operation Concordia), comprising 350 soldiers.
violations or practices. Alternatively, monitors can try to persuade the party con-
cerned behind closed doors to change its behaviour. Finally, reports by monitors
have often been used by other third parties as a point of reference for taking cer-
tain actions, e.g., the application of sanctions or the use of force in cases of gross
human rights violations.
E Coercive Diplomacy
One or more local actors may be forced by third parties to stop violence, to come
to the negotiating table, to sign an agreement, or to comply with the settlement.
Third parties may apply severe economic, military and/or diplomatic sanctions,
e.g., no fly zones or arms embargoes. Furthermore, they may endorse the credible
threat of force, issue serious warnings to the local parties or set ultimatums, and
may finally, if necessary, intervene militarily (peace-enforcement or peace-mak-
ing). They may supervise the peace process by the provision of security guar-
antees and the deployment of troops on the ground (post-conflict peacekeep-
ing). These troops may then be authorized (and prepared) to use coercive means
against anybody who wants to end the peace process by returning to violence.
The exemplary cases here are Bosnia and Kosovo. In Bosnia, the USA finally
decided to use coercive means in order to pave the way for a political solution. In
summer 1995, they were able to change the military situation on the ground by
supporting the Croatian army, which then joined the combined Croat-Muslim
forces in Bosnia. In addition, NATO, with the permission of the UN Security
Council, ordered air strikes against Serbian forces around the besieged city of Sa-
rajevo (Operation Deliberate Force, 29 August–14 September 1995). In the State-
ment of Agreed Basic Principles (8 September 1995), the Serbian side accepted
certain premises for a settlement for the first time. During the Dayton peace ne-
gotiations (1–21 November 1995), the USA again put severe pressure on all sides
to reach an agreement, including the threat of military force. For implementing
the peace accords, the parties had to accept a robust international force led by
NATO. The Implementation Force (IFOR) deployed 60,000 soldiers, the follow-
up mission Stabilization Force (SFOR) eventually reduced the military personnel
to 12,000 (2002). In the case of Kosovo, the use of military force was even more
dramatic. The Holbrooke agreement with President Miloševic (13 October 1998)
has to be understood as an example of coercive diplomacy, as it was done under
a NATO ultimatum. NATO decided to apply limited air strikes and a phased air
campaign after a period of about ninety-six hours once the necessary activation
order (ACTORD) had been made. The ACTORD remained in force throughout
the following months, and therefore all diplomatic steps took place under the im-
minent threat of military means. In particular, the Rambouillet and Paris peace
talks (6–23 February and 15–19 March 1999), co-chaired by the USA, the EU, and
Russia, were conducted under the clear intent of NATO and the USA to intervene
militarily if the regime in Belgrade was not prepared to sign an agreement. The
failure of the peace talks subsequently led to full-scale NATO warfare against
Yugoslavia (24 March–10 June 1999), which ended with the withdrawal of the
Yugoslav army and police from Kosovo and the establishment of an international
regime, including a 60,000 strong force led by NATO (KFOR). In the other cases,
coercive elements were rather limited. The only notable exception may be South
Ossetia, where since 1992 a Russian-led peacekeeping mission would, in prin-
ciple, be able to act against paramilitary groups and may therefore deter others
from the use of violence.26 Coercive diplomacy is an extreme form of third party
intervention and happens rarely. Very few actors, such as powerful states or a
coalition of states, are willing and able to put their own soldiers at risk in order to
intervene military in a self-determination conflict. Coercive diplomacy involves
high costs for the third party and implies a long-term commitment. In contrast
to all other approaches, coercive diplomacy does not need any consent from the
local parties. The key question with regard to power-sharing systems, therefore,
is whether former enemies can indeed be forced from outside to share power.
The results in Bosnia and Kosovo thus far are not entirely convincing. More-
over, the use of force usually needs a special authorization by the international
community, notably by the UN Security Council. This was the case in Bosnia,
while NATO’s intervention in Kosovo did not get the blessing of the UN Security
Council (mainly due to Russian and Chinese opposition), so only KFOR operates
under a UN mandate. The Kosovo case highlights the general problem of this ap-
proach. Who decides when coercive diplomacy is needed and in what form?
F Arbitration
Third parties can make decisions when the local parties are not able to agree
upon certain regulations or policies. Arbitration can take place within a formal,
juridical procedure (e.g., courts) or political bodies (e.g., ad hoc or permanent
commissions). A crucial feature of arbitration is the binding character of the de-
cisions. The parties then have to comply with these decisions or face sanctions.
With regard to the case studies, the approach has only been relevant for Bosnia
and Kosovo. The Office of the High Representative (OHR) in Bosnia and UNMIK
in Kosovo have the right to take decisions on behalf of the local actors. In Decem-
ber 1997, the Peace Implementation Council granted the High Representative the
power to decree legally binding decisions in all areas where the parliamentary
assembly or the entity parliaments are not able to pass laws. The mandate of the
OHR was thus expanded because of various deadlocks in the implementation
process as the local parties had used the power-sharing structure not for cooper-
ating, but for blocking one another. The OHR made extensive use of this compe-
tence and took more than 100 decisions (Bieber 2007). Furthermore, the Bosnian
Constitutional Court acts as arbitrator in cases of disagreement from time to
time. The court has nine members, including three international judges appoint-
ed by the European Court of Human Rights. In July 2000, an important decision
26 In Macedonia, the presence of NATO might have had a similar deterring effect. The
missions, however, were not used by the international community as an instrument
of coercive diplomacy.
was made when the court declared with a majority vote that those articles of the
Constitution of the Republika Srpska that refer to the Serbs as the ‘constituent
nation’ of the entity are incompatible with the principles of the Dayton Constitu-
tion. The court therefore demanded changes, and after failed negotiations among
the parties this led to a final decision taken by the OHR.
As these cases show, the effects of arbitration may be ambivalent. On the one
hand, arbitration offers an opportunity to overcome deadlocks in the peace pro-
cess. It may even be a relief for the conflict parties to know that somebody else
will decide for them. On the other hand, the danger is that the conflict parties
leave almost every decision to the arbitrator in order not to make any compro-
mises. Arbitration by a third party, thus, may actively encourage hardline posi-
tions and, thereby, weaken power-sharing structures that aim at fostering coop-
eration among the former adversaries.
G Interim Administration
Here third parties implement the entire agreement or elements of it by them-
selves. They have executive and legislative competencies, they administer the
state or the region until powers are devolved to local parliaments or govern-
ments. The administration can be done by a purely international body (with some
local staff ) or by a mixed international-domestic institution where international
and local actors act jointly. Again, Bosnia and Kosovo serve as prime examples. In
Bosnia, both international authorities and mixed institutions are used to imple-
ment the Dayton Peace Accord (DPA). The OHR, supervised by the multilateral
Peace Implementation Council, acts as the key institution for civil implementa-
tion while NATO (via IFOR/SFOR) is responsible for the military aspects, the
UN for policing (International Police Task Force),27 the OSCE for elections, the
Council of Europe for human rights, the UNHCR for the return of refugees,
and the EU together with the World Bank and UNDP for economic reconstruc-
tion (Table 5). In addition, a number of institutions and bodies are dominated
by international actors: the Constitutional Court with its international judges,
the Central Bank (the first governor was chosen by the International Monetary
Fund), the Commission on Election (four of seven members from abroad), the
UNHCR-led Commission on Refugees, the Commission on Return of Properties
(three of nine members are appointed by the European Court of Human Rights),
and the Human Rights Commission, comprising an Ombudsperson (appointed
by the OSCE for five years) and a Chamber with eight of the fourteen members
nominated by the Council of Europe. The High Representative got the final au-
thority for interpreting the DPA, he also has the power to make, to change and to
annul laws, and he can dismiss officials, both civil servants and elected officials
if they are not willing to comply with the DPA. By 2002, over sixty officials have
been dismissed by the OHR (Bieber 2007).
27 Since 2003, the EU has taken over the responsibility for policing (EU Police Mis-
sion).
Table 5 Interim Administration in Bosnia

Task International actor
Military aspects NATO-led SFOR (former IFOR)
Civil implementation International High Representative
Regional stabilization OSCE, Stability Pact
International policing UN, European Union Police Mission
(EUPM) (since 2003)
Preparation and implementation of elec- OSCE
tions
Return of refugees UNHCR, EU
Protection and return of property European Court of Human Rights (ECHR)
Protection of human rights OSCE, Council of Europe, UNHCR,
ECHR, Human Rights Co-ordination Cen-
tre (HRCC)
Constitutional Court ECHR
Central Bank IMF
Protection of national monuments UNESCO
Economic reconstruction World Bank, EU, UNDP
Education policy OSCE, Stability Pact
In Kosovo, the interim administration (UNMIK) was established by UN Security

Council Resolution 1244 (1999). UNMIK was allocated a list of tasks: to perform
basic civilian administrative functions for as long as required, promote the estab-
lishment of substantial autonomy and self-government in Kosovo, organize and
oversee provisional local institutions, coordinate humanitarian relief, support
reconstruction of key infrastructure, promote human rights, assure the safe and
unimpeded return of refugees, maintain civil law and order (including interna-
tional policing), and, finally, supervise the successive transfer of its administra-
tive responsibilities to permanent, elected local institutions. In order to fulfill its
mandate the interim administration has been subdivided into ‘four pillars’. The
first pillar deals with police and justice (under UN leadership), after humanitar-
ian assistance (led by the UNHCR) was phased out in June 2000. The second pil-
lar concerns the UN civilian administration, the third pillar democratization and
institution-building (led by the OSCE), and the fourth pillar reconstruction and
economic development (led by the EU). UNMIK rules by regulation essentially.
In its first regulation (1999/1), UNMIK reserved for itself all “legislative and exec-
utive authority with respect to Kosovo, including the administration of the judi-
ciary.”28 In order to involve local actors, UNMIK established the Interim Admin-
istrative Council (IAC) in January 2000 – an eight member advisory body (four
UNMIK officials and four Kosovars, including one from the Serb minority). The
IAC could make recommendations and give guidelines to the twnty Joint Interim
28 See UNMIK Regulation 1999/1 of 25 July 1999, On the Authority of the Interim Ad-
ministration in Kosovo, Section 1.
Administrative Departments where Kosovars served as co-heads of department

along with international representatives. In May 2001, UNMIK went a step fur-
ther and issued the Constitutional Framework for Provisional Self-Government
(Regulation 2001/9) which foresees a power-sharing structure for Kosovo.29 The
framework was a pre-condition for the Kosovo-wide elections (November 2001)
and the establishment of local institutions (executive and parliament). Some tasks
have now been transferred, but according to the Constitutional Framework UN-
MIK still exercises a number of “reserved powers and responsibilities,” including
budget control, monetary policy, control of customs service, appointment and
removal of judges and prosecutors, authority over law enforcement, as well as
full authority to ensure that the rights of all communities are protected (Weller
2003). Table 6 summarizes the various tasks implemented or undertaken by dif-
ferent international actors.
Table 6 Interim Administration in Kosovo

Military aspects NATO-led KFOR
Civil implementation UNMIK, Special Representative of the
Secretary-General, Interim Administrative
Council
International policing, police education UNMIK, OSCE
tions
Institution-building, rule of law OSCE
Humanitarian aid OSCE, EU Humanitarian Aid Office
(ECHO)
Return of refugees UNHCR
Protection of human and minority rights UNHCR, Ombudsperson Institution
(OSCE)
Economic reconstruction and develop- EU (European Agency for Reconstruc-
ment tion), World Bank, USAID
Media, freedom of expression Temporary Media Commissioner
(OSCE, UN)
The mixed international-domestic regime in South Ossetia can also be regarded

as an interim administration, however at a somewhat lower level (Table 7). The
key body is the Joint Control Commission (JCC), which was set up following
the ceasefire in 1992. The JCC comprises representatives from the OSCE, Russia,
Georgia, and South Ossetia. As subordinate institutions, three working groups
deal with military aspects, economy, and repatriation, the latter included an ad
hoc committee with UNHCR participation. Moreover, the EU and UNDP are
29 See UNMIK Regulation 2001/9 of 15 May 2001, Constitutional Framework for Provi-
sional Self-Government.
involved in JCC economic policies. The JCC serves as a platform for both super-
vising day-to-day administrative work in the region and for improving the Geor-
gian-Ossetian relations in order to achieve a final settlement.
Table 7 Interim Administration in South Ossetia

Military aspects Georgian-Russian Joint Peace-Keeping
Force
Civil implementation JCC
tions
Institution-building, rule of law OSCE
Humanitarian aid EU, UNDP
Return of refugees UNHCR, Norwegian Refugee Council
Protection of human and minority rights UNHCR, Ombudsperson Institution
(OSCE)
Economic reconstruction and develop- EU, UNDP
ment
The three regimes differ to some extent. In Bosnia, the interim administration
is in charge of implementing a comprehensive power-sharing system and a con-
stitution, mainly written by outsiders, while in Kosovo and South Ossetia the
international bodies are based on interim or provisional arrangements and are
mandated to establish a permanent solution. In these latter cases, local actors
are more involved in the process of finding a solution than they are in Bosnia. In
Kosovo, after consulting local parties, UNMIK had the authority to set up a tem-
porary power-sharing structure pending a solution to the final status of Kosovo,
which is still disputed between Albanian Kosovars, the Serb minority, and Bel-
grade. By contrast, in South Ossetia the search for a solution has to be a joint
undertaking between international and local actors. In general, interim adminis-
trations are clearly the most far-reaching form of intervention in terms of power-
sharing. Here, the power is not only to be shared among local parties but also, at
least for an interim period, between local and international actors. In Bosnia and
Kosovo, local power-sharing can be overruled by international decisions. The ad-
vantage is that interim administrations provide a relatively reliable public order
and security framework that can be seen as a pre-condition for reconciliation
and confidence-building among the former local enemies. Moreover, the situa-
tion on the ground can be improved without waiting until the parties concerned
agree to certain policies. There are, however, serious problems with international
administrations. First, similar to coercive diplomacy, they involve comparatively
high costs and a large number of international personnel to be deployed on the
ground. Second, as with arbitration, as long as third parties take the decisions and
provide the framework, local politicians do not have to act responsibly and are
not accountable, they can afford to stick to their (hardline) positions and do not
need to compromise, as illustrated in particular by the Bosnian example. Basi-

cally, they are not forced to practice power-sharing.30 Third, local actors may feel
alienated from the political process and then organize protests and opposition
against the international administration, which may in extreme cases lead to ter-
rorist acts. Fourth, some decisions of the international actors may appear counter
to democratic standards and procedures (e.g., the dismissal of elected politicians,
the suspension of laws) and may be regarded as arbitrary by the population. Fifth,
interim administrations may have great difficulties in organizing the transfer of
power and choosing the right time for devolving competencies to local bodies.
Often they make themselves almost indispensable, and, unintentionally, do not
allow for local capacity-building. Finally, these administrations often do not have
a clear exit strategy, there exists no plan as to when and how to withdraw.
H Discussion
If we now look at all the approaches and all the case studies (Table 8), we can
make the following observations. Third parties act mainly as donors, assistants,
advisors, or monitors. Facilitation has also often been attempted, but has proved
to be insufficient for achieving or implementing an agreement in all cases – ei-
ther these efforts failed (e.g., Mindanao) or other measures had to be taken (e.g.,
Macedonia). Thus, in most cases, mediation was needed to set up power-shar-
ing arrangements. Only in a few cases were third parties prepared to use coer-
cive means. After the failure of non-military approaches in Bosnia and Kosovo,
military force was finally applied in order to end local warfare. The use of force,
however, increased the commitment and responsibility of international actors to
guarantee stability and a political settlement, which led to their role as arbitrator
and interim administrator. The cases of Bosnia and Kosovo are exceptional and
are extreme examples of third party involvement.
30 The former High Representative for Bosnia, Wolfgang Petritsch, made this point
very strongly when he said that: “In the long run the usage of the far-reaching powers
lead to a type of dependency syndrome. Local parties began to opportunistically rely
too much on the political intervention of the High Representative, especially when
it came to unpopular measures. They can behave despite being in government, as if
they were in opposition and defend their ethno-nationalist goals without [the] need
to compromise”.
Table 8 Approaches Adopted by Third Parties

Conflict Aid and Facilita- Media- Moni- Coercive Arbitra- Admini-
Assis- tion tion toring Diplo- tion stration
tance macy
N. Ireland X X X X
Bougainville X X X X
Mindanao X X X
Bosnia X X X X X
Kosovo X X X X X
Macedonia X X X X
Gagauzia X X X X
S. Ossetia X X X X X X
IV Favouring Third-party Involvement

Finally, there remains the question under what conditions third parties can expect
to significantly foster peace processes and achieve power-sharing arrangements.
Surely, it is difficult or almost impossible to measure accurately the specific con-
tribution of third parties where agreements have been successful and led to a
lasting peace. We do not always know to what degree the success of these agree-
ments is simply due to local factors and to what extent third parties are entitled
to claim credit. However, what can be stated is that there are certain factors that
favour third party involvement irrespective of whether their efforts have been
decisive for success or not. The underlying assumption is that if these favourable
conditions are present, then a positive contribution of third parties is more likely.
The minimum criteria for positive or successful third party involvement would
be that large-scale violence does not take place as long as key international actors
are on the ground.31 With regard to our case studies, this criteria is by-and-large
fulfilled. The only exception seems to be Mindanao, where the third party (OIC)
is not permanently involved and where violence still takes place. We have reason
to assume that third parties have had an impact without being able to pinpoint to
what degree they were responsible for stopping violence.
What are plausible factors that could determine the impact of third parties?
The following summarizes the main findings from the literature on success-
ful or failed peace agreements, but here we will focus on the role of third par-
ties (see in particular Zartman 1985; Hampson 1996; Walter 1997; Stedman and
Rothchild 1996; Stedman 1997; Doyle and Sambanis 2000; Stedman 2001; Sted-
man, Rothchild and Cousens 2002). These factors should be understood as no
more than a checklist and not as necessary or even sufficient conditions for posi-
tive outcomes. Three sets of factors can be distinguished: the first are related to
31 For similar efforts in defining successful third party involvement, see Downs and
Stedman (2002: 50-52).
the conflict itself, its key actors and its dynamics; the second to the terms of
settlement; and the third to the character of the third parties.
A Factors Related to the Conflict and its Key Proponents

• Number of warring parties: the more warring parties exist, the more dif-
ficult it becomes for outsiders to broker peace or to assist implementation.
The assumption is that a great number of parties will complicate negotia-
tions and make it almost impossible to have inclusive processes. Moreover,
if each side is highly fragmented internally then it is difficult to identify the
groups that may be representative and legitimized to negotiate. The more
parties involved, the greater the danger of violent fringe groups that are able
to derail any peace process. This factor played a prolonged role in the cases
of Northern Ireland and Bosnia. It still seems to be a problem in Mindanao,
where the Muslim minority is split in various groups and fractions.
• Existence of a mutually hurting stalemate: if a conflict is characterized by a
mutually hurting stalemate, i.e., no side is able to win militarily, then third
party intervention can be successful. The assumption is that third parties
often have to wait for the ripe moment before their efforts can have an im-
pact. A hurting stalemate can be seen as such a moment. Indeed, in the
Bosnian case it was important that the military situation changed on the
ground before diplomatic efforts could succeed. Here, third parties helped
to bring a stalemate about. A similar process took place in Kosovo (due to
Western intervention) and Ossetia (due to Russian intervention). In North-
ern Ireland, the major local parties themselves came to the conclusion that
in the end no side would be able to win militarily. In other cases, however,
there was no clear stalemate before third parties could help to broker peace
agreements (e.g., Gagauzia, Bougainville, Macedonia). These examples show
that the relevance of this factor is highly contested and has to be established
on a case-by-case basis.
• Level of violence: the lower the level of local violence, the better for third
party intervention. The earlier third parties get involved, the greater the
prospects for success. The paradox, however, is that if the level of violence
is low and not reported by international media, third party involvement is
usually unlikely. An exception to this rule seems to be the case of Gagauzia
where we could observe early involvement in a case of low-scale violence. To
a lesser extent, this was also true for Macedonia, where international actors
became involved from the early 1990s and again in the early days of the 2001
crisis. In all other cases, a certain level of escalation, unfortunately, seemed
to be necessary in order to attract international attention.
• Role and behaviour of local conflict parties: the more local elites are willing to
cooperate with international actors, the better for third party involvement.
In order to be successful, local parties have to be responsive to proposals,
advice, or assistance, but also to pressure from outside. In cases of self-de-
termination conflicts, however, governments often claim that the conflict is
an internal affair and rule out any international involvement. They are not
prepared to work with a third party (e.g., Kosovo) or, at least, they try to limit
their influence (e.g., Mindanao). This pattern can be observed in almost all
cases. Over time, the attitude of governments changed in some cases and the
level of cooperation with external actors gradually increased (e.g., Northern
Ireland, Bougainville, Gagauzia, South Ossetia, Macedonia). Militant sepa-
ratist groups do also not always welcome third parties as they generally rule
out secession as a solution. In extreme cases, they may perceive the interna-
tional actors as ‘foreign occupation’ (e.g., Kosovo, Bosnia).
• Role and behaviour of other external actors: the positive contribution of third
parties also depends on the behaviour of other external actors. As Hampson
(1996: 210) pointed out: “If key regional actors or outside great powers are
hostile to the peace process, third parties will find themselves in an uphill
battle in their efforts to make a settlement succeed.” The question is whether
they can be urged to play a constructive role and whether their interests are
considered in the peace process. This factor had been particularly relevant in
Bosnia, where major states supported local groups militarily and politically
(Croatia, Serbia, Islamic states). To a lesser degree, this was also the case in
Northern Ireland (role of the American-Irish diaspora), for Macedonia (role
of Kosovo-Albanians), and for South Ossetia (role of North Ossetia).
• Presence of spoilers: the outcome of third party activities can depend on the
presence and behaviour of certain spoilers, be they locals or outsiders. For
various reasons, spoilers have no interest in peace processes and try to un-
dermine any negotiation and settlement, often by the use of violence (e.g.,
terrorist acts) and/or by organized crime. In all cases, we can pinpoint spoil-
ers: terrorist groups are still active in Northern Ireland, Kosovo, and Mind-
anao; former warlords and criminals (including the influence of corruption)
characterize the political and economic situation in the Balkans, in Georgia
or in Moldova. The key question is whether third parties are able (in coop-
eration with the local elites) to either integrate these groups into the peace
process or to marginalize them in a way that removes their ability to prevent
progress. Both strategies have been applied in Northern Ireland with some
success, important paramilitary and terrorist groups were included, while
others were at least not able to undo the Good Friday Agreement, at least
thus far.
B Factors Related to the Settlement

• Indigenous solutions versus imported solutions: the more local parties can
identify themselves with the solution found, the better for third party in-
volvement. The settlement, assisted, brokered or facilitated by a third party,
is accepted by the local parties as their ‘own’ solution and regarded as their
own achievement. The agreement can thus not be denounced as an ‘alien
idea’ or as ‘forced upon us’. In Northern Ireland, Gagauzia, and Bougainville,
despite international involvement, the local parties finally had to take the
decision themselves. In contrast, in Bosnia or Kosovo the settlements are

largely seen as a result of international force and a significant number of
local actors have not fully accepted these solutions. To a lesser extent, we
can observe a similar situation in Macedonia and Ossetia. In both cases, the
central government particularly feels forced by the international community
to agree to solutions they otherwise would perhaps not have accepted.
• Inclusive versus exclusive solutions: the more a solution is inclusive, the
greater the opportunities for successful third party intervention. The ques-
tion is whether third parties are able to ensure a high level of participation
at the negotiation table or at least through indirect and informal means (e.g.
shuttle diplomacy, multi-track negotiations, secret channels). The aim has
to be that all relevant actors are part of the agreement (including external
actors such as neighbouring or kin states). The agreements or interim solu-
tions in Northern Ireland, Bosnia, Bougainville, Macedonia, and Ossetia can
indeed be seen as more-or-less inclusive, while in the cases of Mindanao
(e.g., MILF and other extreme Muslim groups) and Kosovo (e.g., Serb gov-
ernment, neighbouring states) important actors are thus far not part of the
settlement.
C Factors Related to the Third Parties

• Mandate and authority: the impact of third parties often depends largely
on their mandate and their authority. How and under what terms can they
get involved? What are they allowed to do? Under whose authority shall in-
ternational involvement take place? It is important that third parties have
a clearly defined role, a realistic mandate, and a certain degree of authority
and legitimacy. The latter can be assured by a mandate of an international
organization (UN, OSCE, EU) and/or by the support of great powers (e.g.
USA). If, however, third parties are invited by the local actors themselves,
their authority and legitimacy may generally be higher. Examples include
the cases of Northern Ireland (Mitchell), Bougainville (UN, New Zealand,
and Australia), Gagauzia (OSCE), and Ossetia (OSCE). In the cases of Bos-
nia and Kosovo, on the other hand, one could observe the failure of various
mandates that proved to be insufficient, too vague, or not supported by ma-
jor players (e.g., UN activities in Bosnia 1992-1995).
• Commitment of third parties: in order to be successful, third parties have to
show a clear and credible commitment when they engage in a conflict. This
can imply political, financial, economic, or military commitments, includ-
ing the risks for deployed international staff and soldiers. Often such com-
mitments are high when third parties themselves have or develop a deep
interest in resolving the conflict. For example, in the Balkans, the USA and
the European states became increasingly committed to stopping the bloody
wars and guaranteeing stability. This was particularly true of Macedonia,
where the international community made its commitment to a peaceful so-
lution very clear from the beginning. In a similar way, the USA and Australia
along with New Zealand were deeply committed to the processes in North-
ern Ireland and Bougainville, respectively. By contrast, the commitment of
the OIC in the Mindanao case seems to be somewhat lower profile and only
on an ad hoc basis.
• Resources and capabilities available to third parties: third parties need to
have the necessary resources and capabilities in order to broker peace or to
implement an agreement. This factor is of particular importance when in-
ternational actors have to engage in post-conflict peace-building, including
the provision of a security guarantee. If none of the third parties involved is
able to deliver basic needs, then it is most likely that the peace process will
fail. In Bougainville, for example, the small UN mission could succeed be-
cause Australia provided the security guarantee wanted by the local parties.
Often only states can offer such resources, most non-state actors and some
international organizations (e.g., UN, OSCE) are usually under-financed or
under-staffed, the EU is the notable exception in terms of its economic and
political clout, but is still largely unable to provide security guarantees. As
a rule, the more third parties have to engage, the more resources are neces-
sary. This can be illustrated by the Balkan cases, on the one hand, and by
Northern Ireland or Gagauzia, on the other hand.
• Coherence and coordination among third parties: since in most cases more
than one third party is involved, it is absolutely necessary to act as coher-
ently and in a as coordinated fashion as possible. This cooperation has to
take place at various levels, between headquarters as well as on the ground.
At the level of headquarters, it may include the regular exchange of infor-
mation and early warning signs of crisis situations and, most importantly,
the development of joint political approaches and platforms with regard to
the conflict. On the ground, it is often necessary to share resources and in-
formation, to develop an appropriate division of labour and to appoint a
leading third party. The case of Macedonia showed how a coordinated effort
among third parties (EU, USA, NATO, and OSCE) could stop escalation and
led to political solution. A similar coordinated approach by the main actors
could be observed in Northern Ireland, in Gagauzia, and in Bougainville,
whereas in Bosnia and, to a lesser extent, in Kosovo, mediation efforts by
different third parties were neither coherent nor coordinated until the Con-
tact Groups under firm USA leadership became the key players. In cases of
implementation, as again best illustrated by the Balkan cases, coherence and
coordination is a major requirement as the number of international actors
(including NGOs) increases rapidly.
V Conclusion
The settlement of self-determination conflicts usually requires international in-
volvement. Our case studies emphasize that third parties do indeed matter. They
may not be entirely responsible for the resolution, but they can heavily influence
the outcome – both in terms of the procedures as well as the contents. Since
third parties are usually not in favour of secession, power-sharing regimes have
become the preferred and most likely option for a political settlement. However,
due to the complexity of these arrangements, third parties have to be prepared
to engage for longer periods, sometimes even for decades, covering both settle-
ment and implementation phases. Moreover, third parties have to fulfill a num-
ber of different roles: assistant, facilitator, mediator, monitor, coercive diplomat,
arbitrator, or administrator. In a few cases, these roles can be played by the same
actor, but normally more than one party is engaged. As has been shown, the most
prominent and influential third parties are international organizations and states,
and often both types of actors are involved at the same time and have to coordi-
nate their approaches. The more pressure and force is needed, the more powerful
states come into play. The more assistance, monitoring, arbitration, and interim
administration is needed, the more likely IOs are to be in the driving seat. In the
area of facilitation and mediation, both IOs and states can play an equally signifi-
cant role, and in some rare cases non-state actors may also serve as a key third
party (Northern Ireland).
In many parts of the world, there is no alternative to third party intervention,
as otherwise the conflict would spread and escalate. As a rule, however, third
parties must be aware of the fact that every external engagement changes the
opportunity structures for local actors and effects their behaviour. This does not
mean that third parties are always beneficial and necessarily lead to positive out-
comes. As has been seen, this depends on a number of conditions, some that can
be managed by the third parties themselves, others that are beyond their reach.
Typically, third party involvement includes some difficulties that can not be easily
resolved but which the external actors should keep in mind before and as they
engage.
Firstly, international involvement is often caused by the use of violence by local
actors. Indeed, third parties have been very reluctant to engage where a conflict
is still characterized by the absence of violence or by small-scale violence. The
warring sides may get the impression that violence pays off economically (due
to international aid) as well as politically (due to international mediation), which
may trigger further violence.
Secondly, third parties may unintentionally increase the legitimacy of one side.
International involvement may initially appear to imply the recognition of sepa-
ratist movements and of militant groups as potential negotiation parties, a status
generally denied to them by most central governments.
Thirdly, third parties may be able to stop violence or to prevent further es-
calation, but very soon may run into serious problems and constraints when it
comes to implementation. Either they are not prepared or adequately equipped
to remain involved over the longer term. Furthermore, after a settlement, another
crisis elsewhere may attract more attention, leading to the premature withdrawal
of personnel and/or resources from still unstable regions.
Fourthly, long-term involvement of international actors usually has ambiva-
lent consequences. It can foster a subsidy mentality among the local actors when
everything is provided from outside. Particularly in cases of comprehensive in-
volvement, local elites are inclined not to take any responsibility, but can stick to
their radical demands and positions, which thus prevents progress in implemen-
tation and makes any power-sharing arrangement unworkable.
Fifthly, international actors simply cannot force local actors to share power.
Power-sharing arrangements may be introduced under international pressure as
in the Bosnian case, but they are unlikely to work under external pressure. De-
spite all these efforts, it is still up to the local parties to finally solve the conflict
themselves and to accept the settlement.
Lastly, there remains the great problem of exit. How and when should interna-
tional actors leave or, at least, significantly reduce their engagement? Surely, third
parties can not stay forever, but they should also not leave too early. Obviously,
there is no easy answer to this particular problem. As a general rule, the more
deeply involved international actors are, the longer they have to stay. More con-
cretely with regard to power-sharing, third parties should stay at least until the
power-sharing arrangement has passed its first serious tests – such as the second
round of elections, the change of laws, the first use of veto powers, or the change
of cabinet.
Part Four
Functional Power-sharing
Chapter 16
Education
Mark Turner
I Introduction
Education produces knowledge, skills, values and attitudes. It is essential for civic order
and for citizenship and for sustained economic growth and the reduction of poverty.
Education is also about culture; it is the main instrument for disseminating the accom-
plishments of human civilization. These multiple purposes make education a key area
of public policy in all countries.
(World Bank 1995: xi).
Every person – child, youth and adult – shall be able to benefit from educational op-
portunities designed to meet their basic learning needs.
(Article 1, World Declaration on Education for All, quoted in UNESCO 2000: 1)
These are compelling reasons for according education high priority in national
development. Economic, political, cultural, and human rights arguments all sup-
port extensive and sustained investment in education. It is a key to human welfare
in the broadest sense. In post-conflict situations, the urgency of investment in
education is considerable. The Machel Report on the impact of armed conflict on
children stresses that every effort should be made to maintain education systems
during conflict and that re-introduction of education programmes is a “prior-
ity strategy” for post-conflict situations (United Nations 1996: 47). Fragile peace
agreements and new political institutions need to be complemented by concrete
development initiatives that restore or improve the services expected by the citi-
zens of modern states. The speedy and efficient provision of these services can
give legitimacy to fledgling leaders and public sector managers and offer prom-
ises of a better future. Education bears a major responsibility in reconstructing
society, as it is that collection of institutions and organizations that develop “the
human resources that determine the way a society functions” (UNICEF 1999: 4).
There is no ‘one best way’ for the development of education in post-conflict
societies. Each conflict and territory is different and post-conflict environments
504 Mark Turner
vary enormously. In some cases, the task may simply be to put back what was
already there, but more likely it involves the reconfiguration of the system. In rare
cases, the conflict may have had little effect on education, but in many instances
there has been considerable destruction and disruption. Patterns of inequity in
the educational system may even have contributed to the eruption of the con-
flict. Where fighting has been widespread, authorities are usually anxious to get
children back into school as soon as possible. They want to reopen universities
and get on with vocational education. But repairing the infrastructure and get-
ting students into classrooms is only the first stage. Of equal, if not greater im-
portance, is the determination of what will be the system of education. In most
cases, the challenge is to simultaneously guarantee human rights while creating a
modern educational system that delivers the quality products desired by parents,
students, and employers. However, conditions for achieving such objectives are
frequently unfavourable. Negative economic growth, budgetary pressure, popu-
lation displacement are typical obstacles. In addition, peace agreements may not
be final, while existing education systems may have already been ill-equipped to
produce well-educated citizens to participate in a modern economy.
This chapter examines how complex power-sharing agreements impact on ed-
ucation – what arrangements are made and how are they implemented. The lead-
ing issue is whether educational arrangements encourage reconciliation among
conflicting parties or whether they reinforce ethnic divisions. There is invariably
tension, even contradiction, between the desire for peace, which encourages in-
tegration, and initiatives to guarantee ethnic identity. The latter can emphasize
ethnic divisions through policies such as educational segregation. The chapter
starts by presenting the broad socio-economic context of each case. Following
this are the individual case studies of educational arrangements in complex peace
agreements. Finally, the cases are reviewed to determine what we can learn from
them. There is a particular concern to identify lessons that can inform policy.
II Socio-economic Context
While all the case study countries and regions share the unfortunate legacy of
conflict and war, there is some diversity in their socio-economic situations. This
environment provides the context in which educational and other post-conflict
development takes place. Education policy and management must deal with the
very specific institutional, cultural, ethnic, and historical circumstances of each
case. However, other broader factors, such as economic situation and demogra-
phy, provide the more general context in which educational rebuilding and devel-
opment occur. They place limits on, and may give direction to, policy initiatives
in education.
16 Education 505
Table 1 Demographic Features and Human Development Index

Country/ Population Urbanization Life expec- HDI ranking
autonomous region 2001 2000 (%) tancy at 2002
birth 2000 (n = 173)
(years)
Bosnia and Herzegovina 3,900,000 43 73 99
Kosovo 1,860,000 N/A 74.1 82
Macedonia TFYR 2,000,000 62 73 65
Gagauzia (Moldova) 161,100 45 (M) 66 (M) 105 (M)
South Ossetia, and Ab- 65,200 (SO) 57 (G) 73 (G) 81 (G)
khazia (Georgia) 525,000 (A)
Northern Ireland (UK) 1,685,300 77.3 (UK) 77 (UK) 13 (UK)
Autonomous Region of 2,745,000 59 (P) 69 (P) 77 (P)
Muslim Mindanao (Phil-
ippines)
Bougainville 200,000 18 (PNG) 59 (PNG) 133 (PNG)
(Papua New Guinea)
Sources: World Bank 2003; UNDP 2002a, 2002b, 2002c.

When examining some basic socio-economic indicators for the case studies there
is no clear pattern that emerges. There are undoubted similarities between some
countries and regions but few, if any, generalizations that can be made to cover
all of them. Table 1 reveals the considerable demographic differences between the
countries – with populations ranging from 3.9 million in Bosnia and Herzegovina
to 65,000 in Abkhazia. The question this raises for educational autonomy is one
of scale and affordability. Can smaller populations actually support a distinctive
educational system with institutions at all levels and how far can that system vary
from the country in which it is located without jeopardizing the educational op-
portunities of students?
Urbanization figures range from an extremely low 18% in Papua New Guinea
(PNG) (and lower in Bougainville itself ) to 77.3% for the UK. The Northern Ire-
land figure will be close to that for the UK as a whole. In all cases except Northern
Ireland, there are substantial rural populations. Most figures in Table 1 are for the
countries rather than the region but it may well be that the level of urbanization
in affected regions is lower. For example, this is the case in Mindanao. Extend-
ing education and guaranteeing quality and conformity among dispersed rural
populations served by infrastructure of questionable quality is far more challeng-
ing than organizing and delivering education in urban areas. Inequities between
rural and urban education coverage and quality are common, even where conflict
is absent. Such differentiation can be highly politicized when it reflects the distri-
bution of ethnic groups.
The fundamental health indicator, life expectancy at birth, finds half the coun-
tries clustering at the 73-74 years mark. But three countries are significantly be-
low this level, with Papua New Guinea having a very poor 59 years. Poor health
506 Mark Turner
indicators are often associated with poverty and poor educational indicators.
Certainly, they suggest challenges in getting children to school and keeping them
there in a healthy state. Northern Ireland figures for life expectancy are far above
those of all the other countries which emphasizes the gulf that separates rich and
poor countries. When ranked on the UNDP’s Human Development Index (HDI),
most countries are located in the mid-range human development category. For
some countries, such as PNG, Moldova, and Georgia, the ranking has been de-
clining both because of conflict and difficulty in managing economic and social
transitions. This is not the most conducive environment in which educational
development policies can be framed and implemented. PNG is especially worry-
ing, as it has fallen into the low human development category. Once again, there
is considerable contrast between all the countries and Northern Ireland as rep-
resented by the UK HDI. It lies in the upper reaches of the high human develop-
ment category, placing it apart from the other case study territories.
Table 2 Economic Situation

Country/autonomous GNI per capita Growth of GDP Country econom-
region 2001 (usd) per capita 2000- ic status 2001
2001 (%)
Bosnia and Herzegovina 1,270 0.2 Lower middle in-
come
Kosovo 904 13 Lower middle in-
come
Macedonia 1,690 -4.7 Lower middle in-
come
Moldova (Gagauzia) 380 (M) 6.3 (M) Low income
South Ossetia and Ab- 580 (G) 4.4 (G) Low income
khazia (Georgia)
Northern Ireland (UK) 24,230 (UK) 1.9 (UK) High income
Autonomous Region of 1,050 (P) 0.6-2.1 (M) Lower middle in-
Muslim Mindanao (Phil- come
ippines)
Bougainville (Papua New 580 (PNG) -5.8 (PNG) Low income
Guinea)
Sources: World Bank 2003; UNDP 2002a; 2002b; 2002c.

Table 2 presents some basic economic indicators for the case study countries
and regions. The Gross National Incomes (GNIs) of the case study countries and
regions range widely, from the UK’s usd24,230 per capita to PNG’s usd580 per
capita. In general, the economic statistics make grim reading. Three countries
are in the World Bank’s bottom economic category of ‘low income’ while the
remainder, with the exception of ‘high income’ Northern Ireland, qualify for the
‘lower middle income’ group. They are all well short of the usd2,975 boundary
that separates them from ‘upper middle income’ countries. Some of the regions
16 Education 507
actually have GNIs below the average for the country, for example, Mindanao and
Bougainville. What is also worrying is that the Gross Domestic Product (GDP)
growth rates are often poor and sometimes negative. Poverty is associated with
such economic conditions and can reach extremely high incidences – such as
68% of the population in the Autonomous Region of Muslim Mindanao that the
Philippines government estimate live below the poverty line. Developing an edu-
cation system in conditions of negative or sluggish growth from an extremely
low level of economic development and with widespread poverty is always going
to be problematic. Even where significant growth is occurring, it may be aid-
driven and possibly unsustainable. Thus, the economic conditions found in most
of the case studies are not conducive to educational development. This does not
mean that educational development cannot take place. It does mean that ambi-
tious goals for educational development may be impossible to achieve. Where
post-conflict reconstruction is accompanied by rising expectations of welfare
improvements among the population, policymakers and planners may face tur-
bulent political conditions.
III Case Studies
A Bougainville
At the time that the dispute between the Papua New Guinea (PNG) government
and the Bougainville Revolutionary Army erupted into armed hostilities, the ed-
ucational indicators for Bougainville were already below the national averages,
and these in turn were below the average for lower middle income economies.
The PNG average for children in the 7-12 years age group enrolled in Grades 1-6
was 77%, while for Bougainville (North Solomons Province) the figure was 71.9%
(PNGDoE 1991). The average for lower middle income countries, which PNG was
at that time, was 101% (World Bank 1992). Transition rates from primary grade
6 into secondary grade 7 on Bougainville were 29%, well below the 36% national
average and distant from the average secondary enrolment figure for lower mid-
dle income economies. Quality of education varied throughout the country and
within Bougainville, and may have declined despite increased investment in the
1980s (Connell 1997; Turner 1990).
Already poor educational indicators plummeted to ever lower depths as war
took hold and continued in Bougainville. There was widespread destruction of
infrastructure and population movement. Supplies for schools dried up, teach-
ers were in short supply, such that only a makeshift education system continued
in the villages around the island – and then not for all villages. Secondary and
formal vocational education largely ceased to exist. There were no universities on
the island, only a university centre, which closed down by 1990. The situation was
one of severe dislocation to education over a protracted period, leading Bougain-
ville to the unenviable current status of having probably the worst educational
indicators of any province in PNG.
508 Mark Turner
The 2001 Bougainville Peace Agreement (BPA) does not make any specific
reference to education. It lists the functions that will be retained by the PNG
government and then states that the powers and functions of the autonomous
Bougainville government “include all known or identifiable powers not on the
National Government list” (BPA Art 5(52) ). As education is not on the national
government’s retained list, responsibility for it must therefore be invested in the
autonomous Bougainville government. But it is not clear how much responsibili-
ty will be shouldered by the provincial government. In the rest of PNG, provinces
deliver primary and lower secondary education but the national Department of
Education still retains certain functions concerning regulation, standards, and
curriculum. Universities are a national government responsibility. The Bougain-
ville Peace Agreement certainly allows for the Bougainville autonomous govern-
ment to assume more educational functions than would be the case for other
provinces in PNG. For example, the Bougainville government have been given
the authority to set up a “teaching service” and “an independent Salaries and
Remuneration Commission” of its own. However, it has so far chosen not to es-
tablish these bodies, preferring to remain under the equivalent agencies of the
national government of PNG.
The rehabilitation of educational facilities, and the recruitment and training
of teachers, is a major task for Bougainville, although the relatively small size of
the territory and its population should make rapid progress possible. Foreign aid,
either from bilateral donors or NGOs, has been crucial in providing resources for
the rehabilitation process. A school building programme has already resulted in
enrolment figures for primary education equalling or even exceeding pre-conflict
levels. There are now more secondary schools than before the conflict. Grades
11 and 12, which are mainly undertaken in specialized high schools in PNG, are
being tagged on to the normal grade 7-10 high schools. There are even some low-
level technical and vocational education centres operating on the island. How-
ever, given the current sad state of the PNG economy and the country’s budget-
ary constraints, continued assistance from the international community will be
essential for investment in education on Bougainville.
The question of language in education does not appear to be problematic,
despite Bougainville being home to approximately twenty different languages.
Before the conflict there was provision for some primary education in these in-
digenous languages. A few of these tok ples (literally ‘talk of the place’) schools
are still operating, but in general the language of instruction for primary and
secondary education is English. This situation is likely to prevail as translation
of materials and concepts and recruitment of sufficiently qualified teachers will
be difficult, if not impossible, when such small numbers of native speakers are
involved. Some of the teachers in the revitalized system have been recruited from
the PNG mainland and are unfamiliar with local Bougainvillean languages. There
could also be disputes about which language to use when secondary schools take
students from more than one language group. The inevitable parental desire to
give their children access to higher education and employment opportunities in
the formal economy will also work to bring Bougainville’s education system into
16 Education 509
line with that of mainland PNG in areas such as curriculum, as well as language.
The general PNG curriculum is currently being utilized in Bougainville and grad-
uating secondary students have started to trickle back into mainland institutions
of higher education. The missing component in Bougainville’s educational reha-
bilitation is the lack of special programmes for young persons who missed out on
their education due to the disruptions of war. While some can be found as over-
age students in the new schools, there are many others who have received little
or no formal education and have not attempted to catch up. At worst, they could
become a lost generation of underutilized potential if no special educational pro-
vision is made for them.
B Mindanao
Disruption to education in the Autonomous Region of Muslim Mindanao
(ARMM) has been sporadic in recent years. When conflict flares, such as in Ma-
guindanao and Lanao del Sur provinces in 2000 or in Basilan in 2001, then people
are displaced and schools temporarily close. Despite these violent episodes, there
has been little destruction and damage of educational institutions since the full-
scale hostilities of the 1970s. But the educational status of the ARMM popula-
tion remains low compared to national averages. In 1997, the literacy rate for the
Philippines was 95%, but for the ARMM it was 78% (NSO 2003). Another survey
found only 61% of persons aged 10-64 in ARMM were functionally literate as
against a national average of 84% (NSO 2001). The ARMM is the most disad-
vantaged region, in terms of education, in a country that places a high value on
education. The peace agreement (Republic Act 9054) reflects this nationwide be-
lief in the importance of education by promising to provide “a complete and inte-
grated system of quality education” as “a top priority” of regional government. To
accomplish this, the regional government is supposed to “adopt an educational
framework that is meaningful, relevant, and responsive to the needs, ideals, and
aspirations of the people in the region (RA 9054 Art 14(1)).”
The peace agreement contrasts with others in that it devotes several pages to
delineating aspects of education. In other peace agreements, there can be as little
as one line dealing specifically with education. Despite the decentralization of
many functions in the Philippines to sub-national governments, education re-
mains a responsibility of central government. However, under the Mindanao
peace agreement the responsibility for education in ARMM is moved to the re-
gional government. The agreement talks of “an educational subsystem”, which
will enjoy fiscal autonomy and academic freedom. A regional department of edu-
cation manages the educational affairs of the region. Its powers, functions, and
composition are defined by legislation in the elected regional assembly. The na-
tional Department of Education, Culture and Sports (DECS) maintains a regional
office in the ARMM.
The peace agreement sets down a list of principles and policies that should be
adopted by the regional government in determining educational policy. There is
a proviso that the regional government’s policies on education should be consis-
510 Mark Turner
tent with national policies. This is a fuzzy area. The degree to which the central
government can intervene is not precisely delineated. Similarly unclear is the de-
gree to which the central government must continue providing services if there
is a lack of regional capacity. The agreement provides a long list of ‘principles and
policies’ that should guide the region’s decision-makers. Some of the items are
not compulsory while others are necessarily vague about implementation. They
can be seen as reference points for policy making. The principles and policies in
the agreement include the promotion of Filipino and Islamic values, peace values,
culture, religious instruction, language of instruction, scholarships, vocational
education, community participation, ethnic identity, basic education structure,
materials, accreditation, teacher administration, curricula, and education man-
agement. Policy for education should be made in the Regional Assembly and by
the Governor with and through the regional department of education.
The regional structure of education follows the national structure – i.e., ba-
sic (primary) education seven years, secondary four years, non-degree tertiary
one-three years, and degree courses four-eight years. There is a provision for the
regional government to create its own structures. However, as the regional gov-
ernment must adopt the same “curricular years as those prescribed nationally”,
this limits the room for manoeuvre (RA 9054 Art 14(7) ). The structure of private
education follows that of the public education system. The religious schools of
the Muslim madrasah system are also a component of the regional structure of
education. The regional Department of Education is awarded the authority to
develop curricula. They should be ‘relevant’ to the economic, social, political,
cultural, moral, and spiritual needs of the region’s population. The capacity to
develop curricula in the region is limited while parental pressure will ensure that
curricula do not diverge greatly from national curricula as this might limit their
children’s qualification for entrance to higher education institutions. While the
marrasah schools are subject to the supervision of the DECS, they also appear
to develop their own curricula. Efforts are currently being made by representa-
tives of the major madaris to construct a unified curriculum and to integrate the
madaris into the national education system (Bagayaua 2002).
The production of textbooks is a joint responsibility of regional and central
governments. In connection with the preparation of materials, the legislation re-
fers to “agreed norms,” “academic freedom”, and “legal limits” but does not spec-
ify them. (RA 9054 Art 14(3)) There is emphasis on promoting Filipino, Muslim,
and tribal minorities’ values. The importance of science, technology, and the lat-
est advances in education are noted. Whether the region has much capacity for
making significant contributions to materials production remains to be seen. In
tight budgetary situations, it is possible that the national department will shoul-
der much responsibility. Donor-funded programmes may be utilized for region-
alizing the materials. There is specific mention of the subject contents in social
studies at secondary and tertiary levels with reference to the need to promote
Filipino and Islamic values.
The legislation clearly states that the provision of “learning aids and instruc-
tional materials” is a responsibility of the regional government. This appears to
16 Education 511
include items that are prescribed by the national government. Although not ex-
plicitly stated, it would appear that the regional government has the financial
responsibility for all items, whether regionally or nationally prescribed.
As elsewhere in the Philippines, the languages of instruction are Filipino and
English. What Filipino actually is remains debateable but is likely in practice to
be Tagalog, the language of Manila and central Luzon. Arabic is to be used as an
auxiliary language of instruction and will also be taught as a subject in all levels
as required for Muslims. It is optional for non-Muslims. For the madrasah edu-
cational system Arabic, is the recognized language of instruction. There is further
provision for the use of other regional languages for instruction – there would be
around ten of these languages. The development of a regional language is pro-
vided for but is highly unlikely to occur. In practice, there will be a continuing
demand for English as the medium of instruction at the secondary and tertiary
educational levels, as it enhances job opportunities both at home and overseas.
Funds for education are provided by the national government to the regional
government. The legislation decrees that the regional government will devote at
least 15% of the recurrent education budget to scholarships for “poor but deserv-
ing students.” It is not specified what these funds will actually pay for. It is also
not specified how the scholarships will be distributed among different types and
levels of education. The legislation urges the regional government to seek schol-
arships from other foreign and domestic donors and to provide tax breaks for
such donors. The regional government is also asked to provide assistance to “dis-
advantaged but deserving students.” There is no elaboration of how this provision
should be implemented and no indication of where the necessary finances might
come from. It remains to be seen whether the regional government can satisfy the
requirements of the legislation regarding scholarships and still pay the other bills
for the education system. As the ARMM has the highest incidence of poverty in
the Philippines, most students fall into the category of “poor but deserving.”
The legislation also makes reference to “the active participation of the home,
community, religious organization and other sectors in the total educative pro-
cess.” (RA 9054 Art 14(2)) How that is to be achieved is left to regional and local
stakeholders. Regional government is supposed to fund “local education pro-
grams” but what these are is not specified.
State colleges and universities are parts of the regional education subsystem.
The legislation guarantees their academic freedom and fiscal autonomy and au-
thorizes these institutions to continue under their existing charters. In July 2002,
nine state institutions of higher education were counted in the ARMM (CHED
2002). Overall, the ARMM has the lowest number of higher education institu-
tions of any region in the Philippines. The regional government has the right
of representation on the boards of state universities and colleges in the region.
There is mention of the possibility of a “tribal university system” but there is no
indication of where funds might come from or what such a ‘system’ might look
like.
The regional Department of Education is responsible for the supervision and
regulation of private schools. These schools can “organize themselves” and par-
512 Mark Turner
ticipate in deliberations of the department on relevant matters. There is provi-

sion for three private sector representatives, one each from the Christian schools,
non-sectarian schools, and tribal minority schools, to represent their respective
constituencies at these deliberations. The details of this formal consultative pro-
cess are not specified in the legislation. Furthermore, there is no mention of the
madrasah system, presumably as this is classified as part of the government sys-
tem.
The madrasah education system is for the Muslim majority in the ARMM.
The system already exists. The new legislation basically reiterates national and
regional government responsibilities. Accredited madaris are supervised by the
national Department of Education. There should be a monthly report to Con-
gress or Regional Assembly on the madaris and ‘periodic review’ by the national
department. The regional government is given responsibility for strengthening
the madrasah system and for conducting competitive qualifying examinations for
madari teachers. Such exams facilitate the teachers’ appointment in the regional
education system. The salaries of the madari teachers are paid by the regional
government. At present, the regional and national governments are attempting
to improve the skills of graduates from the madrasah system. Currently, their
graduates have qualifications that are not recognized by the government and of
little use to employers. Thus, the push is to “mainstream the madaris” by stan-
dardizing the curriculum, introducing new subjects and upgrading the teachers
(Bagayaua 2002). Government may also be worried about the potential or actual
role of madaris in ensuring they do not become “breeding grounds for terrorists”
(Bagayaua 2002).
Apart from the mention of a tribal university system, no other special arrange-
ments are specified with regard to the indigenous tribal communities. There are
many references in the peace agreement to respecting different cultures and to
relevant curricula that could be used to create some distinctiveness among the
schools for indigenous cultural communities. Religious education is optional in
public schools. This does not apply to the madrasah system and schools run by
Christian religious orders.
The regional government is responsible for the selection, recruitment, ap-
pointment, and promotion of teaching and non-teaching personnel. Personnel
are provided certain assurances regarding their positions. If they have appropri-
ate civil service eligibility, they may not be replaced, removed, or dismissed with-
out ‘just cause’. Also, while the region may define employment standards, these
cannot be lower than the nationally prescribed standards. National minimum
standards also apply to the selection, recruitment, and appointment processes
of teaching and non-teaching staff at all levels of education. Whether such stan-
dards can be applied is debatable in such provinces as Basilan, where recruiting
qualified teachers has been a major problem for years.
While it appears that much of the responsibility for education has been de-
centralized to the regional government, the centre still has the opportunity to
exert considerable control. Firstly, the budget for education comes from central
government as part of the General Appropriations Act. This gives all interested
16 Education 513
parties at national level, both elected and appointed, the opportunity to partici-
pate in determining the allocation of funds. Secondly, central government edu-
cation bodies have responsibility for monitoring compliance by the region with
national education policies, standards, and regulations. Thirdly, the civil service
guarantees regarding employment conditions apply in the ARMM. Fourthly, the
madaris are specifically identified as being under central scrutiny. However, all
of this monitoring, supervision, and compliance takes place in a context of state
weakness. The capacity of the central state to exert its authority in many parts of
the ARMM is very limited. This also applies to the regional government. Thus,
some of the provisions in the peace agreement that delineate the structure and
operation of the educational subsystem in ARMM may be illusory in practice.
C Gagauzia
According to the last census of population in Moldova, adult literacy was 96.4%
(WEF 2000a). Primary education was compulsory and enjoyed almost universal
coverage, while approximately 80% of the population aged 25-40 years had sec-
ondary and higher education qualifications. In short, Moldova had an education
system at least equivalent to those of other former Soviet Union republics, which
provided free education from pre-school to university level, which was judged
‘satisfactory’ and at least equivalent to its neighbours in coverage and quality. As
education in Gagauzia was part of the national system, it can be assumed that
school and tertiary enrolments and quality were roughly equivalent to national
averages.
Language represented a focal point for Gagauz nationalism and for the con-
flict between Moldova and Gagauzia. Although Gagauzians make up only 3.5% of
Moldova’s population, they comprise 81.4% of the 161,100 inhabitants of Gagauzia.
Most of them (92%) consider Gagauz their native language but 73% also use Rus-
sian as a second language (Järve 2001). Only 4% speak Moldovan, which is almost
identical to Romanian. The introduction of Romanian as the “state language” in
1989 and the classification of Russian as “the language of communication” raised
tensions in Gagauzia and Transdniestra, even though there were official guar-
antees concerning the protection and development of the Gagauz language and
provision for public servants in Gagauz areas to be competent in Gagauz. The
problem was that all Romanian speakers knew Russian but “not all Russophones
were able to speak Romanian” (Järve 2001: 4). The Gagauz felt under threat.
The ensuing struggles between the Gagauz and central state concluded with an
organic law that gave Gagauzia the status of an autonomous region. According to
the law, an elected People’s Assembly can issue laws, “within the limits of its com-
petence,” within the fields of science, culture, and education. Also in the law are
provisions for the Executive Committee of Gagauzia to draw up and implement
programmes in education, culture, and sports, and to promote the use of the na-
tional language and culture of Gagauzia. The law does not explicitly decentralize
all educational functions to Gagauzia. Indeed, the Gagauzian authorities must
follow national programmes and standards as prescribed in the Law on Educa-
514 Mark Turner
tion 1995. They are, however, responsible for appointing school heads and teach-
ers. A research unit has been established to elaborate the concept of ‘national
schools’ for Gagauz and other minorities, while university and teacher education
institutions in Comrat, the Gagauzian capital, are “preparing specialists for local
institutions of education and culture” (Järve 2001: 26).
While the spirit of the Law on Education is for a decentralized system of edu-
cation open to world values, adapted to the economic needs of the nation, and
compatible with the education systems in countries of the European Union, the
practical difficulties are immense. The pace of decentralization has been slow,
while investment to transform the education system has been inadequate. In Sep-
tember 2001, teachers in Gagauzia had not received their salaries for 3 months,
while electricity in some schools was disconnected (Järve 2001). Over the whole
of Moldova, the physical condition of schools has deteriorated as the government
is unable to find the necessary finance from its budget. One estimate claimed that
official allocations only covered 40% of budgeted current expenditures on edu-
cation in 1999-2000 (WEF 2000a). Furthermore, 50% of all schools in Moldova
needed repair work and about 33% of school libraries had been closed. There was
a shortage of textbooks and items such as computers were old and/or in very
short supply. Rural schools appear to be worst affected by these problems.
To address the financial woes of the education sector, there have been steps to
pass some of the monetary burden on to consumers. However, given the low in-
come status of the country and debilitating record of negative economic growth
(-9.5% per year in 1990–2000), it may prove difficult to shift financial responsi-
bility without discriminating against a large proportion of the population. For-
eign aid has filled some of the financial gap in education and has provided much
needed technical assistance. The World Bank, UNICEF, the European Training
Foundation, and the Soros Foundation are among the leading donors.
As expected, the language issue is prominent in debates and decisions on edu-
cation. The overarching Moldovan Law on the Use of Languages guarantees the
provision of secondary, non-specialized intermediate studies, specialized techni-
cal and intermediate studies, and higher education in Moldovan and Russian.
The provision on other languages, such as Gagauz, is less specific. The national
government is tasked with creating the necessary conditions for these other na-
tionalities to receive education in their own language.
There are practical difficulties with implementing the legal provisions on lan-
guage in education. Gagauz has a short history as a written language, dating back
only to 1957. Many adults do not have literacy in the language – only 37.8% in
1998. Furthermore, there is a strong preference for parents to want their chil-
dren’s education to be conducted in Russian (80.6%), presumably as it opens up
better opportunities for further studies and employment (Järve 2001). In order
to satisfy all the stakeholders in the language debate, schools in Gagauzia teach
three compulsory “local official languages” (Järve 2001). Children must also learn
either English or German as a foreign language. This results in a situation where
students spend large amounts of time studying languages but mostly fail to reach
a high degree of proficiency in any of them.
16 Education 515
D Abkhazia and South Ossetia

According to a UNESCO report, the educational system inherited by Georgia
from the former Soviet Union was “rather good” (WEF 2000b). Primary and sec-
ondary education were mandatory and free, teacher training was “well devel-
oped” and educational standards were “sufficiently high.” However, there were
concerns that a passive approach to learning and ideological conformity resulted
in a lack of creativity and critical thinking, and difficulty in the practical applica-
tion of knowledge. The situation in Abkhazia and South Ossetia would have been
little different than that prevailing over Georgia as a whole as long as the country
remained a Soviet Republic.
The ethno-nationalist conflicts in Abkhazia and South Ossetia resulted in mas-
sive disruption and destruction of the education system in these two territories.
Approximately 250,000 persons, mainly of Georgian ethnicity, were displaced
from Abkhazia, and about 60,000 Georgians and Ossetians from South Ossetia.
For many children, this meant a complete cessation of education. In the two ter-
ritories, the infrastructure of education was severely damaged. According to one
report, about 70% of Abkhazian educational institutions were burnt or looted,
and are in need of major repair work (Kamkia 1996). Text books were in short
supply, but even when available were outdated and inappropriate for a modern
education system. Prior to the hostilities there were 300 pre-schools but by 1996
only 20 were operating. Similarly, in South Ossetia, the education system lacks
the essential infrastructure and equipment to function effectively.
In Abkhazia, the Georgian education system was completely closed down.
Teaching of Georgian language, literature, and history were banned, and Geor-
gian schools ceased to operate. In 1992, there were 122 Georgian schools, 58 in
the Gali region, while today there are none according to Georgian authorities
(Global IDP c.2002). There are teachers working “in clandestine circumstances”,
reportedly at great personal risk to themselves and their students. A little prog-
ress has been made with the matter being discussed by the working group of the
UN-sponsored Coordinating Council and the reported introduction of ‘elective’
courses in Georgian into one Gali school (Global IDP c.2002).
The various ‘peace’ documents that have been signed over the past decade say
little about education. They have minimal details on how functions or powers
should be distributed between the different territories. They only go as far as
identifying or suggesting ‘target areas’ that can be discussed. The mechanisms
established for negotiation have failed to make any progress in power-sharing on
education or any other functional area concerning service delivery. As Tsikhelas-
hvili (2001) argues, there still seems to be a focus on how to monopolize power
rather than to share it. This stalemate takes place in the context of inadequate
funding for education and dismal economic performance. Many of the costs for
education, such as teachers’ salaries, are paid by parents or through the activi-
ties of international aid agencies. The budget is simply not available to attain the
objectives of educational reform and performance envisaged in Georgian policy
516 Mark Turner
documents. Agreement on who does what is an essential first step that has yet to
be taken.
E Macedonia
Macedonia’s education system is considered to be in reasonably good condition.
Starting from an extremely low base after the Second World War, there was con-
siderable expansion under the socialist regime. Universal primary education was
attained, access to secondary schooling was considerable, while there were good
opportunities for entry into tertiary education. Education was free and appar-
ently valued by the population. The gains have been maintained to the present,
with good educational coverage, high enrolment figures, and sustained efforts to
upgrade all levels of education to bring them into line with European standards.
The gross figures for enrolments do conceal some inequalities between the ma-
jority ethnic Macedonians and minority groups. The latter are under-represented
at secondary and university levels, and comprised only 16% of the university pop-
ulation of Skopje and Bitola in 1998-1999 despite the raising of university quotas
(Daftary 2001: 6). Inequality is even more pronounced among students studying
abroad, 89.2% of whom are ethnic Macedonian (MO c.2001).
The major divide is between ethnic Macedonians and Albanians. According to
statistics based on the 1994 census, Macedonians comprise 67% of the popula-
tion and Albanians 23%, although Albanians claim their numbers are understated
(MO c.2001). The other minority groups collectively comprise only 11% of the
population, the largest group being ethnic Turks at 4%. Language is at the core
of the dispute between the Macedonians and Albanians, with the latter wanting
their language to be recognized as an official language alongside Macedonian.
This Albanian demand was granted in the 2001 peace agreement (Framework
Agreement, FA) under the provision that any language spoken by “at least twenty
per cent of the population” (FA 2001 Art 6.5) qualified as an official language.
Albanian is the only qualifier from among the minorities. The Agreement also
affirms that “the official language throughout Macedonia and in the international
relations of Macedonia is the Macedonian language” (FA 2001 Art 6.4).
The Agreement states that primary and secondary education will be provided
by the state in the students’ native language(s). It would appear that the Macedo-
nian language will continue to be taught to other nationalities, although this is
not mentioned in the agreement, but is stated in Article 48 of the Constitution.
The Agreement appears to represent a reaffirmation of the constitutional provi-
sions that every person has a right to education and that education is available to
everyone under equal conditions (MMoES 2001). Article 48 of the Constitution,
which deals specifically with education and ethnic minorities, states that “the
representatives of the ethnic minorities have the right to education in their lan-
guages in the primary and secondary education in a way determined by law.” (FA
2001 Art 6.1). This will ensure the continued segregation of education according
to ethnicity. For example, in 1999-2000 there were ninety-two Macedonian high
16 Education 517
schools with 76,132 students, twenty-two Albanian schools with 14,353 students
and four Turkish schools with 598 students (MO c.2001).
Central control of education is asserted in the Framework Agreement through
the requirement that uniform standards for academic programmes will be applied
throughout Macedonia. The Ministry of Education and Science (MoES) proposes
regulations, decides on curricula, inspects institutions, enforces rules, and ap-
proves such things as textbooks. The bureaucracy of education is centralized.
There is community participation in the management of pre-schools and on the
school boards of primary schools. There is also an article (3.1) in the Agreement
that prescribes a revised Law on Local Self-Government. Such a law is intended
to ‘reinforce’ the powers of elected local officials and increase their ‘competen-
cies’. Among the listed competencies or functions is education. The intention ap-
pears to be the decentralization of authority and functions to sub-national units,
with the assurance that adequate funding will be available to cover the costs of
the new responsibilities. But decentralization is not popular among ethnic Mace-
donians, 66% of whom equate it with the eventual dissolution of the state (Daft-
ary 2001: 21). The revised Law on Local Self-Government is yet to be adopted and
“conditions for implementing any form of power-sharing in Macedonia are highly
unfavorable at present” (Daftary 2001: 19).
It appears that education will continue to be run along ethnic lines with con-
cessions to minorities, but with Macedonian re-affirmed as the dominant official
language. The centralized bureaucracy of the MoES will manage education and
maintain its array of policies to upgrade the system to bring it into line with Euro-
pean standards. In its UNESCO-sponsored account of education in Macedonia,
the MoES provides technical detail of the structure and operation of the educa-
tion system (MMoES 2001). It also utilizes the latest education terminology in
describing its policies and programmes for “education and learning strategies for
the 21st century” (UNESCO 2001: 27). But the report rarely, if ever, ventures into
the territory of power-sharing. Ethnicity is mentioned in passing, but as a techni-
cal matter and not as an issue concerning the determination of decisions about
education.
The report’s avoidance of ethnicity is in great contrast to the battles that con-
tinue to be fought over higher education in Macedonia. The two national univer-
sities have enrolled far fewer Albanians than quotas allow. This is due to most
courses being taught in Macedonian, except in teacher training. Macedonian is
only taught as a second language in Albanian high schools and often not par-
ticularly well. Indeed, the Albanian school system fails to attract the brightest
and best teachers because of the poor living conditions in rural areas where the
majority of Albanians reside. Albanian under-representation in universities is
compounded by high drop-out rates.
The Albanian solution was to set up their own university in Tetovo in the Alba-
nian-dominated northwest. The university was and remains privately funded by
the Albanian community and student fees. However, the government has refused
to recognize the legality of the university or its qualifications. Poor professional
qualifications of lecturing staff have been alleged, as have inadequate curricula.
518 Mark Turner
However, there appear to be deeper reasons for the impasse. The government has
been influenced by Macedonian fears of secession, the creation of a greater Alba-
nia, or simply by perceived threats to Macedonian society. One response, strongly
endorsed by foreign donors, has been to establish the South Eastern Europe Uni-
versity at Tetovo with instruction in Macedonian, Albanian, and English. Despite
substantial investment, the university has not attracted the anticipated numbers
of students. Meanwhile, the conflicts over the University of Tetovo continue, with
the government in 2003 indicating a possible willingness to move towards formal
recognition. Ironically, intra-Albanian political battles have contributed to slow
movement towards agreement.
F Bosnia and Herzegovina

The legacy of socialist rule in Bosnia and Herzegovina was an education system
that provided broad access to pre-school, primary, and secondary education, and
substantial participation in higher education (CoE 1999). In common with other
socialist systems, “curricula and teaching methods reflected socialist values that
stressed conformity and the acquisition of fact and political loyalty over critical
thinking” (CoE 1999: 5). However, the ‘self-management’ concept of the former
Yugoslavia differentiated it from its socialist neighbours and provided a partici-
patory model, which may have even encouraged nationalism. Although Bosnia
and Herzegovina had a tradition of centralized curricula, the conflict resulted
in educational fragmentation, with the combatants “promoting their ideological
and political goals through the provision of nationalistic textbooks and curricula”
(CoE 1999: 7). Often, one group’s material was offensive to another. Thus, the
1995 Dayton Agreement for peace reaffirmed an ethnically fragmented pattern
of education.
The Dayton Agreement identifies the functions belonging to national govern-
ment. It is a modest list that contains only those functions needed for Bosnia
and Herzegovina to operate as an internationally recognized state; for example,
foreign policy, customs, immigration, and air traffic control. Along with other
service functions, education is a responsibility of the two component ‘entities’,
the Federation of Bosnia and Herzegovina and the Republika Srpska. There are
no details in the Agreement, simply the statement that:
All governmental functions and powers not expressly assigned in this Constitution to
the institutions of Bosnia and Herzegovina shall be those of the Entities.
Thus, the way in which functions such as education are managed is for the Enti-
ties to decide. In the Federation of Bosnia and Herzegovina, responsibility for
education is devolved to the ten cantons. In some instances, there is provision
for further devolution to the municipal level, particularly where disputes arise
over access to education in a particular language. Each canton can determine all
educational matters, such as curricula, textbooks, and regulations. This has led
to the politicization of education along nationalist lines. It is alleged that many
16 Education 519
‘educational’ decisions are determined by nationalist considerations (CoE 1999:

9). Curricula can thus reflect the nationalist ideology of the dominant group in
a canton. One recent report states bluntly that “far too often schools in BiH are
still being used to spread ethnic hatred, intolerance and division” (OHR 2001d).
Another commentator noted that “despite some progress having been achieved
over the past year, discrimination, segregation and bias continue to dominate
education in BiH” (Global IDP 2001). But the structure of education and broad
curricula content remain remarkably similar and can be accessed by nationalist
minority students as long as they conform to the cultural and religious views of a
school’s dominant nationality.
The decentralization of education to the canton level was part of an overall pol-
icy of decentralization that seems to have been poorly designed. As has happened
elsewhere, decentralization is sometimes undertaken without due consideration
of the capacity of sub-national units to perform the decentralized functions effi-
ciently and effectively. Patterns of inequity may also be entrenched (Turner 1999).
In Bosnia and Herzegovina the cantons possess “a complex mix of institutions,
often with competing aims, roles, and functions and with limited technical and
financial expertise” (CoE 1999: 10). A further unsatisfactory aspect of decentral-
ization in Bosnia and Herzegovina, is “the lack of requirements for cooperation
and coordination and the lack of institutions which could facilitate these goals”
(CoE 1999: 10). In contemporary development policy and management, there is
much effort to promote partnerships between stakeholders and to foster inter-
organizational cooperation in order to enhance performance. These consider-
ations and practices appear to be absent in Bosnia and Herzegovina.
One of the outcomes of the Dayton Agreement is the extremely weak posi-
tion of the Federation Ministry of Education (FMoE). This contrasts with other
decentralizing countries where ministries of education often retain considerable
control or influence over such matters as curricula, standards, general policy and
inspection. The Dayton Agreement facilitated not only the transfer of the author-
ity for education to cantons, but also paved the way for the finance of education
through canton-level tax revenue. Thus, the FMoE can only aspire to a coordina-
tion role. In practice, even this is only possible in Bosniak-majority areas, and
then only to discuss such things as textbooks or common programmes. Croat in-
terests are pursued outside the FMoE. But, even with the Bosniak cantons, FMoE
simply duplicates their expertise while failing in normal national tasks such as
policy research, planning, and monitoring. The FMoE suffers from a familiar de-
centralization failing of being unable to adjust from being a central agency based
on directing activities and securing compliance to an agency that loosens its grip
in such matters while becoming a facilitator.
The organization of education in the Republika Srpska stands in great contrast
to Bosnia and Herzegovina. In Republika Srpska, responsibility for education is
centralized in a newly created Ministry of Education (MoE). What the cantons
do in Bosnia and Herzogovina the MoE does in Srpska. While the MoE is the
unchallenged formal authority in education, severe budgetary problems and ‘po-
litical discord’ in the eastern parts mean that its influence is much less in prac-
520 Mark Turner
tice than in theory. The curricula and textbooks used in Srpska ‘closely resemble’
those in the Republic of Serbia, although there is apparently little substantive
support from there (CoE 1999).
There has been considerable international assistance for education in Bosnia
and Herzegovina. Such aid was essential for the rehabilitation of physical infra-
structure. Over 60% of educational facilities were damaged or destroyed during
the war. International agencies such as the World Bank, UNESCO, the European
Union, and the Council of Europe, as well as NGOs have participated in the aid ef-
fort. High levels of investment in physical infrastructure have now been replaced
by projects concerned with improvements in teaching and systemic reform. For
example, the Office of the High Representative has listed essential ongoing re-
form efforts as textbook review, curriculum reform, freedom of movement in the
education system, efficiency in institutions of higher learning, and removal of
discrimination against minority groups (OHR 2001d).
The latest initiative in education has been the release of an educational reform
package subtitled “to give BiH a better future” (BiH 2002). It has been ‘applauded’
by the Organization for Security and Co-operation in Europe (OSCE) which co-
incidentally coordinated the work for it. The document consists of fine-sounding
pledges, which will assist in creating an educational system which is de-politi-
cized, equitable, modern, and of high quality. There are many laudable objec-
tives, but recent history suggests that programme designers and implementers
will have to overcome formidable obstacles to achieve them.
G Kosovo
In 1945, Kosovo’s Albanian population began to receive education in their own
language. Previously, there had only been limited education in Serbian. Under
the Socialist Federal Republic of Yugoslavia, there was considerable improvement
in educational indicators. 74% of Kosovo’s Albanian population were illiterate in
1948, but by the late 1980s the figure had been reduced to only 10% (OECD 2001).
But educational conditions for the Albanian population deteriorated quickly with
the removal of Kosovo’s autonomy and the assumption of all power by Kosovo
Serbs as “Albanian managers, professionals, teachers, doctors and academics
were removed from their posts or left the province” (OECD 2001: 7). The Serb
authorities imposed a Serbian curriculum, shut down Albanian textbook produc-
ers, ceased financing Albanian schools and sacked 23,000 Albanian teachers. To
fill the educational vacuum, the Albanian population organized an unofficial self-
financed ‘parallel’ system of education (Kostovicova 2002). In 1998, there were
266,400 primary school pupils, 58,700 secondary, and 16,000 university students
in the parallel system (OECD 2001: 7). Quality of education declined under these
makeshift arrangements and enrolments dropped at all levels. Even the Serbian
system received inadequate funding and many facilities were in poor condition.
The situation was exacerbated by the destruction of schools in the 1998-1999 hos-
tilities. 290 schools were destroyed and 800 suffered damage (OECD 2001).
16 Education 521
Attempts to rebuild the shattered education system began with the establish-
ment of the Department of Education and Science (DES) under the United Na-
tions Mission in Kosovo (UNMIK). A school rebuilding and repair programme
and the use of temporary facilities allowed most Kosovan students to return to
school in the 1999-2000 academic year. In order to coordinate the rehabilitation of
the education system, UNMIK adopted a strategy of ‘lead agencies’. This involved
international aid agencies taking responsibility for specific tasks under UNMIK’s
direction. In early 2002, in line with UN Security Council Resolution 1244, a new
Ministry of Education, Science and Technology (MEST) was created and man-
agement of most educational functions handed over to it (UNDP 2002b). The
MEST is divided into two departments – Education and Central Administrative
Services. The Education Department deals with matters of policy and operations
at all levels of the system while the Central Administrative Services Department
deals with finance, human resource management, outreach and information ac-
tivities, and infrastructure.
The peace agreement had little to say on education. The 2001 Constitutional
Framework for Provisional Self-Government (CFPSG) stresses freedom, human
rights, and the rights of ‘communities’. Among these rights is that the commu-
nities can establish educational institutions “in particular for schooling in their
own language and alphabet and in Community culture and history” (CFPSG
2001 Ch 4.4 j). There is a proviso that such initiatives should show tolerance and
respect for other communities. The provisional institutions of self-government
are awarded responsibilities in the fields of education, science, and technology,
and the assembly can pass laws in these field. Working out how these fields would
be organized and what policies would be put in place were matters left to the
peacetime authorities.
UNMIK got the ball rolling with its rehabilitation programme and with the
drafting and issue of various regulations to shape and manage the emerging sys-
tem. The establishment of the autonomous DEST in 2002 marks the assump-
tion of full responsibility for education by Kosovan authorities. However, because
of the difficulties of building a modern education system, it has been observed
that “the solution is to prepare the system for a long period of emergency status”
(OECD 2001: 12). The determination of a realistic and coherent education policy
is still in process. For example, the New Kosovo Curriculum Framework is strong
on good intentions but short on how they will be achieved. The new curriculum
seeks to reflect the multicultural composition of Kosovo, accommodate differ-
ence among students, link to new development in science, and use up-to-date
teaching methodologies. The New Kosovo Curriculum Framework does contain
a core curriculum, which will be subject to external evaluation and will provide
the preconditions for further studies. It contains compulsory subjects and indi-
cates such things as minimum number of hours per week and basic contents in
each subject. The elective part of the curriculum allows “for the existence of a
differentiated approach” (DES 2001: 19).
In reality, many problems remain. Despite expressed desires to decentralize
some education functions, there is still a high degree of centralization. The Ram-
522 Mark Turner
bouillet Agreement of 1999 even identifies the provision of education as a com-

mune responsibility; a matter not mentioned in the subsequent Constitutional
Framework of 2001. The MEST operates through its central office in the capital
and through its deconcentrated offices in five regional administrations. There
has been a heavy reliance on external advisers under UNMIK, while it is not
clear how Kosovan teachers are to learn and adopt new teaching methodologies
and knowledge. Over 30% of teachers already lack the necessary qualifications
for their profession (UNDP 2002b: 55). Talk of aspiring to European standards
and methods may be an unrealistic goal given the environment for education in
Kosovo. Teachers’ salaries are low and likely to remain that way, although they
consume 80% of the education budget (UNDP 2002b: 51). Many teachers seek
additional jobs or leave the profession, while the poor terms and conditions of
teacher employment discourage young adults from entering the profession. Edu-
cational managers with the required skills are scarce. Student drop-out rates are
problematic – at primary level the figure is 7% but at secondary it is up to 34%
(OECD 2001).
Supplies for schools are constrained by harsh budgetary conditions. There has
been a heavy reliance on foreign aid in the post-conflict years, but these levels
of support are unlikely to be maintained. Government management of finance
in Kosovo is in its infancy with numerous matters still to be resolved. Fiscal de-
centralization with accompanying functional decentralization of education to
municipalities is sometimes promoted as a way to bring about greater equity in
resource distribution and efficiency in its usage. However, the lack of good finan-
cial management systems and differences in administrative capacity at municipal
levels would make such a move a very risky venture. Decentralization has also
been identified as a way to utilize local initiative and energy, but central govern-
ment has been reluctant to devolve authority for managing education to the mu-
nicipalities (UNDP 2002b; OECD 2001).
A major problem in Kosovo’s evolving education system is a persistent one –
how to accommodate the different ethnic communities. Some groups are still fa-
voured over others (OECD 2001). In 2000-2001, Kosovo’s primary schools taught
303,590 students in four languages – Albanian (92.3% of students), Serbian (5.2%
of students), Bosnian (1.3 per cent of students) and Turkish (0.7% of students).
There is no integrated curriculum, as each community obtains its resources from
different systems and locations. Only 22 of the 541 primary schools in Kosovo
teach in more than one language and even then each language group is taught in
separate parts of the schools. Moving towards greater integration of the educa-
tion system is a declared priority for Kosovo, but may be difficult to obtain as
divisive ethnic politics are still prominent.
There have been concerted efforts to modernize Kosovo’s one university, the
University of Prishtina. Approximately 10% of the total Kosovan population in
the 19-25 years age group attend the university (UNDP 2002b: 56). All teaching
is in Albanian, but legally instruction could be given in other languages. Admin-
istrative and curricula changes have been undertaken, but, as has been found
in vocational education, the existing fields of specialization and courses do not
16 Education 523
necessarily represent the skills required in the Kosovan economy. There are no
private institutions of higher education.
H Northern Ireland
The educational experience of Northern Ireland contrasts greatly with the other
case studies. Firstly, Northern Ireland’s education system has been very little dis-
turbed throughout the long conflict. Although there have been 3,600 deaths and
30,000 casualties since the eruption of hostilities in 1968, this environment of
violence has not disrupted the delivery of education (IBE 2003). Schools have not
been destroyed, damaged, or closed down by the conflict, and there has been no
significant displacement of population. A second distinguishing feature of North-
ern Ireland’s education system is its level of resources. In the other cases, there
were resource constraints, often severe, which hampered efforts to reconstruct
or develop education. Northern Ireland has assets and recurrent spending typi-
cal of other Western European countries. There has been a pattern of sustained
investment at levels far higher than in the other case-study territories. Northern
Ireland is a high income country with a high cost education system. There are
schools, colleges of further education, teacher training institutions, schools for
specialist needs, and two universities. Indicative of the country’s high income
status is the Classroom 2000 initiative, in which each pupil is expected to have
access to a computer. Such aspiration is unthinkable in other post-conflict terri-
tories such as Bougainville, Mindanao, and Gagauzia, where finding a computer
in any classroom is unlikely.
The administrative complexity of education in Northern Ireland reflects the
system’s size, maturity, and degree of specialization, as well as its religious segre-
gation. Ten statutory bodies are required to run education in the province. The
overarching statutory body is the Department of Education, which is responsible
for central administration and dispensing an annual budget of about gbp 1.3 bil-
lion. The department follows the UK’s penchant for results-based management
and the measurement of performance through a Service Delivery Agreement that
contains the department’s objectives, targets, and actions to achieve the latter.
The idea of continuous improvement is incorporated into the Service Delivery
Agreement methodology.
Northern Ireland is divided into five Education and Library Boards (ELBs),
which operate as local education authorities. They ensure that there are sufficient
schools and colleges of further education to meet area needs and channel the
finance from the Department of Education to maintain the schools that fall un-
der ELB management. The amount of funding is determined by the Assessment
of Relative Need Exercise, an instrument which employs a range of indicators
but especially pupil population. The ELBs also have the tasks of ensuring that
the schools are equipped and maintained, monitoring school attendance, provid-
ing curriculum advice, and regulating the employment of young persons. An-
other significant statutory body is the Council for Catholic Maintained Schools
(CCMS). It is responsible for the employment of teachers in Catholic maintained
524 Mark Turner
schools and undertakes some advisory functions. All of its funds come from the
Department of Education.
In Northern Ireland, education is compulsory and free from ages four to six-
teen. It is also free for students opting to stay on until the age of eighteen years.
The schools of Northern Ireland increasingly operate under the decentralized
governance of boards of governors. Schools teach a common curriculum, but
there is pronounced segregation along religious lines – Catholicism or Protes-
tantism. The ELBs provide the funds to Controlled schools, which are typically
Protestant, and to Voluntary Maintained schools, which are typically Catholic. In
addition, there are Voluntary Non-Maintained schools, mainly selective gram-
mar schools which are usually Protestant. The selection for secondary school in
Northern Ireland is still determined by examination taken at the end of primary
school and known as the ‘11-plus’. It seems that this long-standing system of selec-
tion will change. A Review Body on Post-Primary Education was established by
Northern Ireland’s education minister in 2000 to make recommendations on the
future structure of post-primary education in Northern Ireland. The Report by
the Post-Primary Review Body (RPPRB, commonly known as the Burns Report)
was published in October 2001 and recommended scrapping the 11-plus and,
among numerous other suggestions, urged the creation of a ‘collegiate’ system
“where schools would retain their independence, but cooperate in 20 clusters
across Northern Ireland” (Woodward 2002). The support for ending the 11-plus
is strong, but there is steadfast opposition from at least two major loyalist parties
in the Northern Ireland parliament. Even among supporters of the 11-plus’s aboli-
tion, there is widespread disagreement about the future shape of the education
system. Major political decisions on this matter are still awaited.
Another significant development has been the growth of integrated education,
that is, schools which have no religious affiliation. In 1990, the Department of
Education was given the statutory duty to encourage, facilitate, and fully fund in-
tegrated education. At present only 5% of students in Northern Ireland attend in-
tegrated schools, a figure which emphasizes the embedded nature of segregation
in education. At the curriculum level there are also efforts to address negative ste-
reotypes of people from the ‘other’ group that are established early in childhood.
Various organizations have cooperated to produce a curriculum called Local and
Global Citizenship, which is expected to become a statutory part of the common
Northern Ireland curriculum in 2006 (IBE 2002). Finally, there has been modest
growth of schools utilizing Irish as the language of instruction. The Department
of Education has the duty to encourage, facilitate, and fund such schools. Accord-
ing to the department, there are now thirteen primary and one secondary school
of this type, and seven dual-language primary schools (DoE 2003).
16 Education 525
II Conclusions
A Socio-economic Development
The case study countries and regions vary widely, between high income North-
ern Ireland and low income Bougainville. The economic difference is enormous.
However, most of the countries have poor economic indicators, a fact which can
threaten the aspirations to develop a well-resourced modern education system.
The question arises about how such countries can make significant investments
in educational improvement while extending coverage. When GDP per capita is
low and in decline, budgets are stretched and investment capital in short supply.
Sometimes teachers may be unpaid for months and there is no money for school
supplies. In such circumstances, it is often impossible to make significant gains
in the pursuit of a high quality education system. By contrast, in high income
Northern Ireland there has been sustained investment and guaranteed payment
of recurrent costs to ensure that both coverage and quality of educational provi-
sion remain at the levels associated with advanced European economies.
B Levels of Disruption
The case studies reveal different degrees of disruption to the education systems.
In Bougainville, almost the entire educational infrastructure was destroyed, ne-
cessitating a considerable rebuilding programme. In other places, such as Kosovo
and Bosnia and Herzegovina, many schools were destroyed or damaged. In Ab-
khazia, about 70% of schools were burned or looted. Physical damage in Min-
danao mainly occurred in the 1970s, but population displacement still occurs
after major engagements between government forces and Muslim insurgents.
Macedonia has escaped major disruption to its education system as has Northern
Ireland where the education system continued to operate throughout the many
years of conflict. Where massive population displacement has occurred this has
created the subsequent problem of return and re-integration. In some instances,
this process is by no means complete. The differing degrees of disruption mean
that territories commence their efforts to rehabilitate and develop their educa-
tion systems from different starting points, and obviously some are more advan-
tageous than others. Where population displacement has occurred, educational
planning is conducted in conditions of high uncertainty about future demands
on the system.
C Contents of Peace Agreements

In general, peace agreements say little about education. They normally make ref-
erence to equity and human rights and give guarantees that education will be
accessible to all. In some instances, there is specific mention of respecting the
rights of all communities or minorities. But instructions about what the educa-
tion system should look like are absent from all but one peace agreement in these
526 Mark Turner
case studies. The task of providing education is usually allocated to a relevant

authority, and this delegation may involve no more than one line of text. A pos-
sible interpretation of such minimal treatment is that education is regarded as an
unimportant issue. This is plainly not the case in any of the territories. Education
is important to governments and citizens. The urgency to get peace agreements
signed means that there is a strong imperative to leave the details of service deliv-
ery until later. Functions such as education are allocated to post-conflict authori-
ties for them to determine policy and organization at a later date. In the case of
Northern Ireland, education continues its incremental development, although
the peace agreement has led to more radical proposals and wider debate about
the nature of educational change. The exception to this pattern is Mindanao. The
peace agreement for the autonomous region went into several pages of details
about the division of responsibilities for education among different levels of gov-
ernment, the structure of the education system, finance, the role of private edu-
cation and many more matters. The nature of the process which led to the final
agreement is perhaps why such an approach could be taken. In the protracted
negotiations, there were committees on specific fields, that fed recommendations
into the main committee. Also, many items were reiterations of the earlier peace
agreement. This process of attending to many of the details of educational ar-
rangements has the benefit of anticipating problems and dealing with them in
advance. Wider usage of such a technique will depend on the urgency of secur-
ing a peace agreement and the willingness of conflicting parties to negotiate and
make concessions.
D Language, Religion, and Segregation

A major concern in many of the case studies is the language of education. Often,
there is an imperative to guarantee the rights of minorities to receive education
in their own language. In some cases, language has been a central issue in the
conflict. For example, in Kosovo, where Serbian authorities closed the Albanian
system; in Macedonia, where Albanians wanted their language to be recognized
as an ‘official language’; or in Gagauzia, where tensions were raised by Moldovan
language policy. While there are provisions in Kosovo, Bosnia and Herzegovina,
Gagauzia, and Macedonia for minority language rights in education, they are not
always fully implemented. Such situations can reflect either a politics of obstruc-
tion or a lack of capacity in the system to cope with the terms of the agreements.
What is clear in post-conflict arrangements is continuity in ethnic or religious
segregation, despite rhetoric about integration. In Abkhazia, the parties have yet
to get as far as an agreement on language in education and a ‘clandestine’ system
of education for the Georgian minority operates. In Northern Ireland, despite
efforts to promote integrated schools, such non-sectarian institutions still only
account for 5% of school students. Ironically, in the most linguistically diverse
case study, Bougainville, language does not appear to have been a divisive issue.
Ples tok (the local language) continues to be used in the early grades where local
teachers are available, but in general it is English that is used in both primary and
16 Education 527
secondary levels. In Mindanao, the major linguistic concern of the Muslim popu-
lation was the use of Arabic in the madrasah system. But this was not disputed
and Arabic has been used for many years in these schools. Filipino and English
are used in the mainstream system, as they are keys to progress in education and
the job market although many people in the ARMM use local languages in their
everyday communication. Mindanao does demonstrate how language and reli-
gion often go together. In other cases, language and religion together demarcate
one group from another and lead to segregation of education. Northern Ireland
is in contrast to other case studies in this regard. While religion is a major de-
terminant of which school a student attends in Northern Ireland, the conflicting
parties are not separated by language. All speak English and almost all are taught
in English, despite official support for Irish-language schools.
E Education for Peace

The formal education system has the opportunity to make a major contribution
to the promotion of peace between communities in conflict. Through curricular
innovations it should be possible to encourage greater understanding of differ-
ent communities and ameliorate hostility between them. Thus, in Northern Ire-
land there is the Local and Global Citizenship Project, which aims to contribute
to consensus-building in a deeply divided society. The promotion of integrated
schools pursues similar aims. In Mindanao, the peace agreement states that the
formal education system is responsible for “inculcation of values of peaceful set-
tlement of disputes” (RA 9054 Art 14(2b)) and elsewhere recommends teaching
about “the cultures of the Muslims, Christians and tribal minorities” (RA 9054
Art 14(2d)). Additionally in Mindanao, Northern Ireland, and elsewhere, NGOs
have been active in promoting peace education on a less formal basis.
Even when officially promoted, education for peace is a difficult task. In North-
ern Ireland, there has been very slow progress in attracting students to integrated
schools. In Mindanao, stereotypes of other nationalities are deeply embedded
and sporadic violent encounters serve to re-confirm these beliefs. Unfortunately,
in some of the post-conflict societies, there appears to be little effort to mount
significant programmes of peace education. While the fighting may have stopped
there may be little happening in the education sector to promote greater under-
standing of different cultures. Anyway, children’s values and behaviour patterns
are largely formed within their families and communities. They will not change
overnight because of a curriculum change. There may be incremental gains in
peace education, but it is the wider community that must also change, as that
provides the all-important context for socialization.
F Curriculum Development
Curriculum development is one of the most important educational matters in
post-conflict societies. If the curriculum is not designed to the satisfaction of all
parties, it is likely to remain a focus of ethnic discontent. The challenge facing
528 Mark Turner
all curriculum designers is how to incorporate two leading objectives of ethnic

groups. The first is to guarantee, develop, and transmit knowledge and under-
standing of ethnic identity. This concerns history, culture, language, and religion.
The second is to provide the modern knowledge and skills required for the econ-
omy. For example, the New Kosovo Curriculum Framework takes great care to
identify elements of a core curriculum to satisfy the economic objective while
allowing for the expression of difference through an elective curriculum. Prob-
lems can emerge when early action is not taken to reform old curricula or when
the identity objective dominates. The former is seen in Bosnia, where nation-
alist ideologies have continued to influence curricula that perpetuate bias and
intolerance. The latter has occurred in Mindanao, where the peace agreement
has guaranteed the position of the madrasah schools for the Muslim population.
However, the curriculum there is religion-focused and provides few marketable
skills. Thus, the qualifications from madrasah schools are not recognized by the
education department and so severely limit the options of madrasah graduates
for further education or employment. A potential problem for curriculum devel-
opment is the existence of a gap between the statements in policy documents and
the reality of implementation. Capacity problems are particularly important. For
example, in Gagauzia or South Ossetia, the shortage of resources in the context of
low levels of economic development mean that curriculum objectives may be dif-
ficult to realize. Even in Macedonia, the financial unattractiveness of the teaching
profession raises questions about the education system’s capacity to deliver the
desired quality of education. Finally, curriculum development can be used as an
integrative device to promote better understanding between conflicting groups.
For example, in Northern Ireland, efforts to develop a statutory curriculum called
Local and Global Citizenship is hoped to lead to better understanding between
the two religious denominations. However, the injection of such initiatives into
school curricula can only work if there is some degree of support from the wider
ethnic community.
G Foreign Aid and Educational Finance

In many of the case studies, there has been considerable destruction, dislocation,
and loss of life. Infrastructure is damaged and systems have broken down. Re-
construction and development can only be accomplished with outside assistance.
The territories are frequently in poor economic shape and are located among
the ranks of the low income or lower middle income countries. Thus, in many
instances, there is great reliance on foreign aid. Organizations such as the World
Bank, European Union, UNDP, and many NGOs provide financial and technical
help. The relative size of this help can vary but in most cases has been vital for
rehabilitating the education system. The donors appear to be able to undertake
repairs and rebuilding works quickly and easily. They also contribute to re-de-
signing the educational system and ensuring that it produces quality outputs and
outcomes. This has proved a much more difficult task than physical infrastruc-
ture and one that is far from complete. Even new infrastructure creates a cost for
16 Education 529
recipients, for example, the recurrent costs of maintaining a school. There have
been variable degrees of dependence on the donors, but in all cases questions
arise as to the sustainability of the systems being created and of the aid itself.
Dire economic circumstances mean that the post-conflict countries are unable
to shoulder the full resource burden of educational development. Increased re-
current costs may be difficult to maintain and investment capital may disappear.
Whether this will lead to cynicism, anger, or simple disappointment among con-
stituent populations remains to be seen, but all scenarios are possible. Northern
Ireland is once again an exception as it does not receive foreign aid.
H Decentralization
In some case study territories, there are questions about further decentralization.
In general, there has been a central reluctance or slowness to decentralize edu-
cational responsibilities to lower levels such as municipalities or districts. There
appear to be two aspects to this retention of central control. The first is the per-
ceived need to consolidate new or revitalized institutions at the centre before giv-
ing out power to lower levels. The second is that administrative capacity at lower
levels may be lacking. Decentralization could then lead to increased inefficien-
cies in newly established or reinvigorated systems. In Bosnia and Herzegovina
where decentralization was introduced the overall system suffered from too rapid
design and implementation with some adverse consequences for education. By
comparison, the major decentralization of functions in the Philippines in 1991 did
not include education. The devolution of authority for education to the ARMM
is accompanied by various central controls. Northern Ireland appears to have de-
centralized authority successfully, first to local education authorities and then to
school governing boards. But while formal decentralization may not have taken
place in many of the case studies, there may be a de facto decentralization. Many
of the states or regions are weak in the sense that they lack the capacity to make
and implement the rules that determine people’s everyday behaviour. They also
have substantial rural populations, some of whom are difficult to reach. In such
circumstances, decisions about what goes on in educational institutions may be
less determined by central ministries than by local practice.
I The Private Sector

The peace agreements make few mentions of private education, with the excep-
tion of Mindanao. There are some references to private sector participation in
documentation produced by aid agencies such as the World Bank, and these
agencies are often champions of privatization, contracting out, and user-pays
services. However, their reluctance to promote private sector solutions seems to
reflect the perception of institutional fragility in many instances and a common
popular belief in the duty of the state to provide free education. Poor economic
conditions accompanied by widespread poverty also make the promotion of pri-
vate sector remedies both difficult and potentially destabilizing. There are a few
530 Mark Turner
private schools in Mindanao but they are the exceptions among the case studies
rather than the rule.
J Education and Power-sharing

The case studies in this chapter have clearly demonstrated that in all power-shar-
ing agreements the arrangements for education are perceived by all stakeholders
as important and as a result are often contested. Resolving these arrangements
to the satisfaction of the ethnic groups is often elusive. Two objectives determine
the educational claims of these parties. Firstly, there is the desire to defend or
consolidate identity. Secondly, there is the demand for equal access to economic
opportunities provided by education. If an ethnic group perceives that it is being
disadvantaged in its capacity to achieve these objectives then education contin-
ues to be a highly contested policy issue, a focal point for interethnic conflict.
The identity objective most often leads to the perpetuation or even reinforce-
ment of segregation in the education system. If resources are not seen to be eq-
uitably distributed between the separate groups, or that one group is exerting
undue influence over matters such as curriculum and language, then disaffec-
tion occurs. While there have been some efforts towards greater integration, the
dominant theme in power-sharing arrangements is for separation of the groups.
This reflects embedded distrust following from often lengthy histories of conflict.
Devices such as peace education and common curricula can assist in promoting
greater understanding, efficiency and equity in education but messages trans-
mitted in classrooms may be contradicted by more powerful influences in their
communities. The major requirement of educational reformers is the develop-
ment of political skills that can be employed in persuading different groups that
education policy is not a zero-sum game in which the protagonists see things as
one side’s loss and another’s gain. The realistic reformer should adopt a ‘satisfic-
ing’ model, that is, seeking solutions that are both satisfactory and sufficient to
the various parties (March and Simon 1993). There are possibilities in activities
such as curriculum development, teacher training, school equipment, and equal
access for different sides to select satisfactory policy alternatives. The idea of op-
timal solutions that can be achieved rapidly, often involving substantial interna-
tional support, is misguided. Educational development in post-conflict societies
is a long-term task and an important one in which incremental progress, rather
than great leaps, should be the aim of those making and implementing policy.
Chapter 17
Resolving Self-determination Disputes Using Complex
Power-sharing: The Role of Economic Policies
John Bradley
I Introduction
The treatment of the role of economic policies in the context of complex pow-
er-sharing systems is one of a number of ‘horizontal’ issues, the others include
education; policing, security, and military; administration of justice; human and
minority rights; and external relations and trans-border cooperation. In this set
of issues, the role of economic policies (and, perhaps, education) can usefully be
differentiated from the roles of the others in some important ways. At the most
basic level, disputes concerning policing, the administration of justice, the treat-
ment of human and minority rights, and the modalities of external relations are
often at the very centre of the initiation, conduct, and resolution of self-determi-
nation disputes. However, economic policies and education tend not to feature
near the top of any list of the fundamental underlying causes of self-determina-
tion conflicts, or on the ‘traditional’ list of issues that need to be addressed when
trying to resolve such conflicts.
It would be naive, of course, to claim that economic and educational inequities
and inequalities play absolutely no role in self-determination disputes or in their
resolution. A more reasonable position would be to acknowledge that, where they
do play such a role, it is more usually as a manifestation of the consequences of
other reprehensible actions, involving a more widespread discrimination against
specific ethnic or regional sub-groups, rather than as a primary cause. The denial
of opportunity, the denial of resources, the ‘unfair’ use/extraction of resources,
etc., are usually part of self-determination claims. But, as McGarry and O’Leary
have asserted in connection with the conflict in Northern Ireland:
The Northern Ireland conflict has been waged paramilitarily and politically between
two communities with different national identities, not between two aggregates of in-
dividuals mainly interested in promoting their economic well-being. Economic factors
have impinged upon the conflict, but they do not determine it. (1995a: 306).
532 John Bradley
Another way in which economic policies matter, in a secondary rather than in

a primary role, reflects the fact that, until the early 1990s, five of the eight case
study regions were integrated into the former Communist bloc. Three regions
– Kosovo, Macedonia, and Bosnia – were part of the former Yugoslav Federation
and its Serb-dominated precursor since shortly after the First World War. The
Yugoslav Federation then became part of the Communist bloc, albeit a semi-de-
tached member, after the Second World War, and remained so until its breakup
during the early 1990s. Two of the remaining five case study regions – Georgia
and Moldova – were constituent Republics of the USSR and remained so until
the breakup of the USSR and the formation of the looser Confederation of Inde-
pendent States (CIS) in 1992. All five of the these regions– to a greater or lesser
extent – had centrally planned economies, or at least economies that did not
possess many of the institutions that are considered necessary for a functioning
market economy.
But disputes over the transition from Communist central planning to capitalist
market institutions could hardly have been at the core of the self-determination
conflicts in any of these five ex-communist regions. The obvious difficulties in-
volved in making the transition to a market economy may have exacerbated an
already long list of political, ethnic, and socio-economic problems facing each
of these regions, but many other centrally planned economies have successfully
made this transition without descending into interethnic conflict. Nevertheless,
there are some compelling arguments that suggest reasons why some formerly
communist countries became more prone to internal conflict, and suffered from
official corruption and organized crime, when they liberalized their politics and
their economies (Olson 2000: xxv).
If the transition from an economic policy regime of central planning to one of
market-based institutions tends not to act as a primary driver of conflict, could
it be that stresses due to particular conditions of poverty may have generated, or
at least exacerbated, conflict in the case study regions? Seven of the eight regions
rank as poor, and in the case of Mindanao and Bougainville, are among the very
poorest in the world (World Bank 2003). But the standard of living in one case
study region – Northern Ireland – is very high, in that it broadly shares the living
standards of the United Kingdom, of which it is a region, and which itself ranks
among the group of very rich countries. Thus, prosperity is not a complete in-
sulation against internal self-determination conflicts. Nevertheless, other things
being equal, the fact that a country has a reasonably high standard of living ap-
pears likely to guide internal disputes away from violent conflict towards peace-
ful solutions. If a country has a very low standard of living – a situation that is
often accompanied by very wide inequalities of regional income levels – this may
Georgia and Moldova were part of the Soviet system of central planning, where both
production and prices were determined by the central planners. The former Yugosla-
via – while formally part of the Communist bloc - was not centrally planned, but had
an economy characterized by state ownership, labour-managed firms and controlled
prices.
17 Resolving Self-determination Disputes Using Complex Power-sharing: The role of Economic Policies 533
increase the probability of internal conflict. If such poor regions have different
ethnic composition, this can often tip them into conflicts of self-determination.
Does economic ideology, interpreted in a wide political-economic sense, play
any role in conflict? During the period of the Cold War, national liberation move-
ments and conflicts about self-determination have sometimes been closely asso-
ciated with clashes of political and economic ideologies. In the post-Communist
era, one does not, of course, have to accept all aspects of the “end of ideology”
thesis of Francis Fukuyama (Fukuyama 1992). Nevertheless, in the post-com-
munist era, the homogenizing forces of globalization place severe restrictions
on the menu of economic policy options that are realistically available to small
countries and regions. When conflict moves towards resolution, the first dawn-
ing realization of devolved or self-government is usually the grim realization of
just how powerless countries and regions can be in designing and implementing
well-meaning strategies to improve their economic welfare!
Much attention has been paid elsewhere in this project to the role of interna-
tional involvement or intervention in the process of negotiating power-sharing
solutions to self-determination conflicts, often in the form of financial or other
aid. But foreign direct investment, by far the dominant form of international in-
volvement in any modern economy, usually manifests itself only when the con-
flict is fully resolved, or when it is partially resolved and violence has reduced to
an ‘acceptable’ or ‘tolerable’ level. Host country experiences of inward investment
can take different forms: many such experiences are development-enhancing
while some are simply exploitative. But openness to the global economy, and in
particular to foreign private capital, is a crucial lever of growth, and post-conflict
economic policies are often directed at enhancing the ability to attract such in-
vestment, in a context of fierce competition from other countries and regions, in
order to augment their own domestic successes (Rodrik 1999).
The absence of conflict is almost always a necessary condition for economic
stabilization and recovery, as well as for the encouragement of an inward flow of
foreign capital. But it is hardly ever a sufficient condition alone. A wide range of
other issues must also be considered, such as a region’s geopolitical location, its
resource endowments, the quantity and quality of physical infrastructure and hu-
man capital, and the ability and willingness to implement appropriate economic
policies and plan rationally for a better future. But, as Olson suggests, tackling
such issues is not easy:
A major challenge of poor societies is not their lack of resources, but rather their diffi-
culty organizing large-scale activities, particularly governmental activities (Olson 2000:
xvii).
In its economic aspects, Fukuyama claims that “capitalism is a path toward economic
development that is potentially available to all countries … provided that (countries)
play by the rules of economic liberalism” (Fukuyama 1992: 103).
534 John Bradley
These are some of the themes that we take up in this chapter. To provide context,
and to justify the inclusion of economic policies as a ‘horizontal’ issue in the proj-
ect, we start in Section II with a review of what economic theory tells us about
policy-making in a context of conflict resolution, drawing mainly on the insights
of North (1990) and Olson (2000). In the foreword to Olson (2000), Leszek Bal-
cerowicz – previously Minister of Finance in post-Communist Poland – used the
illuminating phrase “moments of extraordinary politics” to describe the dramatic
efforts that are often needed to consolidate authority by appealing to encompass-
ing interests in an effort to solve conflict. The concept of ‘encompassing inter-
est’ in its economic interpretation is crucial to an understanding of the limited,
but still vital, role that can be played by economic policy-makers and economic
policies in mitigating the human costs of conflict. North’s framework of analysis
points to the crucial importance of institutions in promoting economic growth.
Olson’s extension and application of the framework shows how the legacy of the
era of communism has been particularly poisonous, and has served to impede
the emergence of encompassing socio-economic institutions in five of the eight
case study regions. Similar considerations can be shown to apply to the remain-
ing three ‘non-communist’ case study regions.
In Section III, we briefly describe the economic context of seven of the eight
case study regions, with a view to identifying the external and internal constraints
that condition the scope for economic policy formation. In two of the case stud-
ies (Bougainville and Mindanao), the self-determination dispute has taken place
within a region of a nation-state (Papua New Guinea and the Philippines, respec-
tively) that appears likely to continue to play a dominant role in the destiny of the
region experiencing the self-determination conflict. In the case of the remaining
five case studies (Kosovo, Macedonia, Bosnia, Georgia, and Moldova), the self-
determination disputes took place within a region of a nation-state whose exter-
nal economic context continues to be influenced by interactions with elements of
the wider federation of which it was formerly a part (Yugoslavia and the USSR),
or the relevant rump of that former federation (Serbia and Russia). However,
looking at these case study regions from the perspective of the global economy
suggests another classification. The future of the three West Balkan regions al-
most certainly lies with their ability to use their proximity to, and possible future
membership of, the European Union as a driving force for stability, cohesion, and
development. In the case of Georgia and Moldova, this role is likely to be played
by Russia, in conjunction with the EU. Finally, in the cases of Mindanao and Bou-
gainville, their main driving force for economic development will rest with their
ability to link with and benefit from the ASEAN (Association of South East Asian
Nations) economies in the Mindanao case and ASEAN plus Australia and New
Zealand in the Bougainville case. But such classifications and distinctions are not
dogmatic or precise, and are merely intended to clarify the context of, and scope
The case of Kosovo is somewhat anomalous, but it seems more logical to include it
here with the other two West Balkan states and to assume that it will follow them
either on the road to independence, or to a form of devolution.
for, economic policies in post-conflict power-sharing arrangements. In what fol-

lows, our main sources of socio-economic information and data are drawn from
the documented project case studies and from international agencies, assisted by
access to the material that have been helpfully collated and annotated in Boyle
(2002).
In Section IV, we examine the case of Northern Ireland. Northern Ireland is
different from the other seven regions in some very important ways. The level
of economic welfare was high at the start of the conflict in 1969, and remained
high for its duration. The conflict – although serious – was less intense than in
most of the other case study regions. The conflict is now effectively over, and
present difficulties concern mainly the operation of the post-conflict institutional
arrangements. The possible relevance of Northern Ireland to the other seven case
studies lies in the fact that the power-sharing arrangements now in place were ar-
ticulated well beyond the political institutional arrangements needed to stop the
conflict, or aspirational economic policies that were unlikely to progress in the
absence of adequate funding. These arrangements have been extended to cover
a wide range of encompassing socio-economic policies and institutions that at-
tempt to address the underlying causes of the conflict. While we must take the
very special situation of Northern Ireland into account, and in particular its ready
access to massive financial support from within the United Kingdom, we try to
use this case study to generalize our insights and draw lessons for the other case
study regions.
In Section V, we draw together the main strands of our study, covering the key
insights from economic theorizing about the nature of encompassing economic
policies and the likelihood that they can implemented in power-sharing arrange-
ments; the constraints placed on the case study regions by their intrinsic charac-
teristics and the particular nature of their internal conflict; and the insights that
we can obtain from development experience of some of the less-developed small
EU member states.
II Institutions, Economics, and Conflict
A Economic Theory and Economic Policy

In any consideration of how economic theory provides a guide to the design of
policies useful in terms of conflict resolution, it is worth suspending disbelief for
a moment to consider whether or not explicit policies are needed at all. This is
not quite so foolish as it might first appear, as there are two dominant schools of
thought within economics that propose radically different answers to this ques-
At the time of writing, the power-sharing institutions of the Belfast Agreement, in

particular the local legislative assembly at Stormont, are suspended. But there has
been no resumption of paramilitary violence, and there is a reasonable expectation
that the institutions will be restored in the near future.
536 John Bradley
tion. The outcome of these debates can have an important role in shaping policy
towards the provision of development aid funding.
Orthodox theory takes the view that all one has to do to promote growth and
convergence is to put in place policies that facilitate the free movement of goods
and the factors of production (i.e., labour and capital). In that case, orthodox the-
orists claim that factor incomes (wages and the returns on capital) will converge
to a common level across all regions. So, if all markets are competitive, any initial
regional disparities will eventually vanish and there is no need for specific policy
intervention. Indeed, such policies will run the risk of wasting scarce resources
and simply get in the way of the competitive market process!
One of the consequences of more recent advances in the study of spatial eco-
nomic processes is that the conditions required for such automatic convergence
to take place are increasingly seen as not holding in practice. New theoretical ap-
proaches focus attention on the importance of the initial level of regional physi-
cal infrastructure, local levels of human capital, or on the fact that regions which
start off at a structural disadvantage may never converge in any reasonable time
period. Such theories suggest that the removal of barriers to trade and factor
movements may actually lead to a relative deterioration rather than an improve-
ment of some regions (Fujita, Krugman, and Venables 1999).
As with much else in economics, there is no right or wrong answer here. A
crude erection of trade or other barriers, or an indiscriminate dependence on
external aid in order to ‘protect’ weak regions, can be shown to be damaging to
economic welfare. Equally, a blind belief in competition policy and the forces of
market liberalization can also be shown to be inadequate (Rodrik 1999). A bal-
ance needs to be struck between these polar approaches, with particular atten-
tion paid to the evolution of appropriate institutions.
B Economics and Institutions

Douglass North has suggested that “the present and the future are connected to
the past by the continuity of a society’s institutions” (North 1990: vii). A common
feature of all eight case studies is that conflict was accompanied by a complete
breakdown of the societal institutions, which further exacerbated violence. It is
impossible to assign causality here, but the breakdown of institutions is usually
the most important element of any self-determination dispute insofar as the col-
lateral damage to the economy is concerned.
North makes a distinction between what he refers to as the “rules” and the
“players.” The rules (or institutions) define the way the game is played. But the
players (or organizations) operating within a given set of rules determine the op-
portunities in a society. As organizations evolve, they alter the institutions. If
the underlying institutional framework reinforces incentives for organizations to
engage in productive activity, then this usually produces stable economic growth
and development. But if the institutional framework (or, indeed, an inability to
provide any such framework in a situation of conflict) favours activities that pro-
mote redistributive rather than productive activity, and creates monopolies rath-
er than competitive conditions, then this restricts growth rather than encourages
it.
Logic of this kind illuminates the debate between the political science approach
to conflict resolution and the role of economics, which has been summarized as
follows by McGarry and O’Leary in the context of the conflict in Northern Ire-
land:
Affirmative action, economic growth, anti-racketeering offensives, cross-border eco-

nomic cooperation, the promotion of more competitive enterprises, and European eco-
nomic integration … must be welcomed for their own sake, but it must always be re-
membered that at best they will fall short of what is required to resolve the conflict, and
at worst they may even deflect attention and energy from the crucial political measures
necessary to change the logic of the cruel game in which the participants are presently
trapped (1995a: 307).
Such a conclusion obliges one to question whether conventional economics – a

discipline founded on rational choice theory – has any role in conflict resolution.
Here, North’s work augments conventional economics and constructs a theory of
institutions by combining a theory of human behavior with a theory of the costs
of transactions, and led him to delineate three types of archetypical exchange
mechanisms that underpin all economic activity:
i. Personalized exchange, involving small-scale production and local trade;
ii. Impersonal exchange, in which the parties are constrained by kinship ties,
bonding, exchanging hostages, or merchant codes of conduct; and
iii. Impersonal exchange with third-party enforcement, of a type that is a criti-
cal underpinning of successful modern economies involved in complex con-
tracting necessary for modern economic growth.
History demonstrates the tendency to evolve sequentially from the first to the
third exchange mechanism over time, but some of the case studies describe situ-
ations of conflict where economies have imploded and exchange mechanisms
have reverted to earlier, more primitive, and often illegal forms. North concludes
that: “One cannot have the productivity of a modern high income society with
political anarchy” (1990: 35), and suggests a possible way in which the dilemma
identified above by McGarry and O’Leary can be addressed, without conceding
their assertion of the limited role of economics:
Political rules in place lead to economic rules, though the causality runs both ways. That
is, property rights and hence individual contracts are specified and enforced by political
decision-making, but the structure of economic interests will also influence the politi-
cal structure. (North 1990: 46).
We can interpret North’s concept of institutions, and his assertion that they are
the underlying determinant of the long-run performance of economies, as being
consistent with what McGarry and O’Leary referred to above as “political mea-
538 John Bradley
sures.” Any artificial distinction between politics and economics now fades, and
the need for a dual concern with both can be established. But this is only likely
to be valid if one moves away from an excessively narrow concept of ‘economic
policy’ as merely a random collection of disorganized and unfocused initiatives.
In the next section, we will illustrate the ‘systemic’ approach to economic policy-
making by reference to the EU use of Structural Funds to address the fundamen-
tal political-economic objective of ‘cohesion’.
Olson (2000) uses a related classification of markets into “self-enforcing” and
“socially contrived.” Without an appropriate institutional environment, a country
will be restricted to trades that are self-enforcing (equivalent to North’s catego-
ries (i) and (ii) above). But with appropriate institutions (legal system, political
order, etc.), all possible gains from trade can be realized (North’s category (iii)
above). Olson suggests two general conditions that are required to sustain a mar-
ket economy that is likely to generate economic success: secure and well-defined
property rights, as well as the absence of predation of any kind. He concludes:
Given the extraordinary gains available from adopting the advanced technologies avail-
able in the post-war world and the possibility of interacting with a reasonably successful
world economy, these two conditions, if fully met, are nonetheless sufficient to bring
prosperity to a society (2000: 197).
A Centralization and Regionalization

The degree to which economic policy autonomy differs between regions of a na-
tion-state tends to reflect the extent to which the state is organized along de-
centralized rather than centralized lines. For example, the United Kingdom is a
highly centralized state, and has normally permitted only very limited devolved
powers of political and economic governance to its regions. Northern Ireland
was the exception, and had the potential for some policy autonomy after the pas-
sage of the Government of Ireland Act in 1920, which partitioned the island of
Ireland and set up a separate parliament in Belfast with control over a wide range
of local matters. The referenda of the late 1990s on devolution for Scotland and
Wales further relaxed the degree of UK centralization. However, the overall level
of public expenditure in the regions of the UK, though not its detailed allocation
between programmes, is still decided at the centre in Westminster, subject to
the application throughout the UK of a common level of income support, health,
education, and other entitlements-based payments, where expenditures are ring-
fenced by the central government. In a federal political system, on the other hand,
a large degree of freedom in both political and economic governance is accorded
to the individual regions.
An essential part of any political resolution of conflict situations often involves
the establishment of a devolved administration to oversee the operation of public
policy and administration. A central issue concerns the degree of autonomy in
policy making that is granted to the regional authority, and the implications for
regional and national financing. A devolved system of governance tends to be
valued for a variety of reasons. It encourages an efficient allocation of national re-

sources; it fosters political participation and a sense of the democratic communi-
ty; and it helps to protect basic liberties and freedoms (Inman and Rubinfeld 1997:
44-45). A common principle in all types of devolved systems of governance is the
desirability of having the most decentralized structure capable of internalizing all
economic externalities, in the following sense. Central government is assigned
responsibility for services that are pure public goods (e.g., national defence, basic
research) or when there are inefficiencies arising from externalities across juris-
dictions (e.g., transport, communications, and energy infrastructure). Lower-tier
government is assigned responsibility for provision of ‘congestible’ services, i.e.,
ones that become congested as more households use the service (health, educa-
tion, police, sanitation), and may also fulfill a role in conflict resolution.
Ohmae (1996) has argued that regions of centralized states face a further di-
lemma as they participate in the global economy because the nation-state has
tended to become a somewhat unnatural, even dysfunctional, unit for organizing
human activity and managing economic endeavour in what, he argues, is becom-
ing a borderless world. He contends that a nation-state sometimes represents no
shared community of economic interests and may not define meaningful flows
of economic activity. He argues that the nation-state tends to overlook many of
the true linkages and synergies that exist among often disparate populations by
combining important measures of human activity at the wrong level of analysis.
In what has become an extensive debate, others have argued that the much dis-
cussed phenomenon of ‘globalization’ does not mean the hollowing-out or death
of the nation-state (for example, Anderson 1995), nor indeed a withering away of
the economic strength and policy autonomy of national governments (Hirst and
Thompson 1995). A more accurate picture would seem to be one that suggests
that:
[The nation-state’s] powers and roles are changing and it is interacting with a plethora
of other, different kinds of political institutions, organizations, associations and net-
works. Contemporary globalization is overlaying the mosaic of nation-states and na-
tional communities with other forms of political community and non-political market
relations (Anderson 1995: 103).
The important point is that increased regional autonomy and the continuing
importance of the nation-state are not contradictory, but are mutually-benefi-
cial phenomena. For example, Dunford and Hudson (1996) argue that strong re-
gional economies tend to be associated with strong national economies and that
an important factor is the degree of regional policy autonomy. The eight case
study regions have all displayed, to differing degrees and for different reasons,
elements of this wider and seemingly paradoxical phenomenon of increased re-
gionalism/nationalism during a period of increased globalization. It is likely that
these trends will continue, with an irreversible impact on the constitutional and
policy framework of nation states, and possibly the eventual reconstitution of
these states.
540 John Bradley
III Economic Issues in Regions of Conflict
A Introduction
In this section, we review the case studies and explore a series of issues related
to economic factors involved in situations of conflict. We have already seen that
there were some important implications for economic policy arising from the
breakup of the two encompassing states/economies of the USSR and the former
Yugoslavia into a series of smaller states/economies. The main consequence was
the complete destruction of the former system of communist central planning
– including the less rigid Yugoslav variant – and its gradual replacement during
the 1990s by a broadly market-based system built around a series of new institu-
tions and laws.
But the breakup of the two larger economic units – the USSR and the Former
Yugoslavia – occurred as a result of the breakup of these two federations due to
political and ethnic factors that had no simple or direct connection with econom-
ic matters. The emergence of Kosovo, Bosnia, and Macedonia from the former
Yugoslavia, and of Georgia and Moldova from the USSR led, or is in the process
of leading, to the creation of a series of smaller states whose economies have a
certain logic in terms of their economic independence and governance.
In the case of the three case study regions of the West Balkans, it is perhaps
probable that they will become full member states of the European Union after a
period of transition and reform, and perhaps on a time-scale similar to Romania
and Bulgaria (ESI 2002; 2003). Given the likelihood of eventual EU membership,
the long-term economic strategy for these three states will almost certainly in-
volve full integration into the Single European Market, adoption of the Euro as
their currency, and transitional investment aid to promote economic cohesion
prior to EU membership.
Georgia and Moldova are less likely to join the European Union as full mem-
bers, although the close relationship that exists between Romania and Moldova
suggests that this is not inconceivable. In the case of Georgia and Moldova, it is
their economic relationship with Russia that is likely to continue to dominate
economic policy-making, together with a growing trade relationship with an en-
larged European Union. Consequently, the stabilization and development of the
Russian economy will be a key driving force of the development of Georgia and
Moldova, in the context of deeper links with the economies of the EU.
For two of the remaining three regional case studies – Mindanao and Bougain-
ville – it is more difficult to make a strong case in favour of their emergence as
sovereign, independent states with viable self-sustaining economies likely to de-
velop and prosper in the absence of permanent dependence on outside aid. What
will be crucial for Mindanao and Bougainville is the nature of their new relation-
However, excessive fragmentation needs to be avoided, and, in the case of Serbia-

Montenegro, the EU has exerted pressure in favour of sub-regional integration as a
pre-condition for eventual EU membership.
ship with the states of which they are likely to remain part, the nature and level of
support that will be given to them from the central policy-making authorities, as
well as from the international community, and the viability and economic health
of the states of the Philippines and of Papua New Guinea.
The final case study – Northern Ireland – is unique and demands separate
treatment from the viewpoint of economic policy. Here, we have a region of the
United Kingdom with a reasonably prosperous, developed, and modern econo-
my, whose growth has certainly been damaged by the conflict that lasted from
1969 to the mid-1990s, but which enjoys all the benefits of the United Kingdom’s
ability to give long-term massive financial support to its poorer regions in the
form of an annual financial subvention.
While there is a logic to including Northern Ireland with the other regions as a
case study of self-determination conflict, in economic terms it has almost noth-
ing in common with the other seven case studies. However, the fact that complex
power-sharing institutions have been set up in Northern Ireland, and these in-
stitutions have strong economic policy aspects, is of relevance to the other case
studies. In none of the other regions has the design and operation of such insti-
tutions progressed as far, and been financed as well, as they have in Northern
Ireland, and for these reasons we treat it separately in Section IV below.
B Regions of the Former Yugoslavia

The three regions that make up the Western Balkans – Bosnia, Macedonia, and
Kosovo – form a logical group from the point of view of the design and execu-
tion of economic policies that are likely to be supportive of the political institu-
tion-building that is at the centre of conflict resolution. Each region was part of
the former Yugoslav Federation and broadly shared in the general level of de-
velopment as well as the Yugoslav form of communist economic policy based
on labour-managed firms rather than on formalized central planning. All three
regions experienced violent ethnic-based conflict, a situation that has been sta-
bilized to a considerable extent, particularly in the aftermath of political change
in Serbia. All three regions have received significant economic development aid,
If Northern Ireland were an independent state like Macedonia (which has a similar
population size: 1.95 million compared with 1.6 million in Northern Ireland), then
a hypothetical ‘state’ of Northern Ireland would have a public sector borrowing re-
quirement of some 30% of its GDP if it were required to finance its present level of
public sector spending through foreign borrowing rather than through an intra-UK
financial allocation.
It should be noted that one region of the former Yugoslavia – Slovenia – was more
prosperous than all the others, and contained much of the more modern, export-ori-
ented, manufacturing firms. Slovenia escaped from the federation in June 1991 – with
almost no violence, made very rapid economic progress, has a standard of living that
now equals that of Greece, and became a full EU member in 2004 along with seven
other former Communist countries.
542 John Bradley
mainly from the European Union. And all three regions have a strong expectation
that full EU membership will eventually be offered to them.
1 Bosnia
As pointed out in the case study in Chapter 7, the origins of the war in Bosnia lay
in the weakening of the communist and federal authority of Yugoslavia from the
mid-1980s. Disputes on the future shape of Yugoslavia eventually spread to an
internal ethnic conflict within Bosnia, and the consequences have been described
in the case study.
The power-sharing arrangements in Bosnia – based on the Dayton Peace Ac-
cords of 1995 – were designed in the context of a country that was not yet fully
sovereign, and one where a considerable degree of power was vested in interna-
tional institutions and individuals and could not be shared among domestic po-
litical actors (e.g., in the security, civilian, and financial areas). The main aspects
of economic policy-making take place in a two-tiered governance system that
involves the centralized institutions of Bosnia and two decentralized ‘entities’ –
the Federation of Bosnia and Herzegovina and the Republika Srpska. The central
institutions of Bosnia have competencies in such aspects as international rela-
tions, communications, and border control. In terms of economic policy, these
competencies include trade, customs, monetary, fiscal, and transport policies
(the ‘public goods’ referred to in the previous section). The weakness of the joint
institutions of Bosnia has resulted in financial and fiscal problems caused by an
inability to raise substantial resources of their own in the absence of payment to
the centre by the ‘entities’.
2 Kosovo
Kosovo was an autonomous province of the former Yugoslavia, and was among
the poorer of the Yugoslav regions. Its tortured relationship with the Yugoslav
Federation and with Serbia has been described in the case study in Chapter 8.
Indeed, its conflict is so recent and was so violent that it is probably premature to
speculate about the exact form of its long-term resolution.
The EU has had a major involvement in the region, and this initially took the
form of humanitarian assistance: emergency repairs to infrastructure, rebuilding
houses, restoring electricity and water supplies, and establishing a financial in-
frastructure. At the time of writing, this initial phase is coming to an end, and the
EU agencies are turning their attention to the more challenging and costly task
of developing Kosovo as a modern market economy that is capable of initiating
and sustaining economic growth. In Kosovo, the EU is centrally involved in the
work of the UN (UNMIK), and heads the pillar of the mission responsible for
Economic Development and Reconstruction.
There is general agreement that the following are the policy framework priori-
ties for the future of Kosovo. To avoid a permanent dependence on international
Commitment to Kosovo, the newsletter on European Union activity (www.euinkos-

ovo.org).
aid, there needs to be more progress towards budget sustainability, by broaden-

ing the tax base, reducing public sector employment, maintaining a reasonable
level of income, and recovering the costs of some public services. To encourage
the growth of private markets, further reforms are needed in trade policies and
customs administration, enactment and enforcement of laws to foster new busi-
ness formation, and efforts to resolve property ownership questions and devise an
efficient and effective privatization programme. Finally, there is a need for modern
labour legislation and the development of a social safety net targeted at the poor
in all ethnic groups.
There remain many policy questions that can only be addressed when the po-
litical status of Kosovo is eventually resolved. For example, if Kosovo reverts to
its previous status as a region of Serbia with devolved powers, then fiscal ar-
rangements – in particular, generation of viable public budgets by enforcement
of fiscal obligations and sharing of resources with the devolved local government
– transnational links, and the legal basis for ensuring equal economic opportuni-
ties, will need to be designed and implemented within a federal system. However,
if Kosovo gains independence, these issues will become the responsibility of an
internal power-sharing administration between the two ethnic groups. For the
present, these matters are mainly the responsibility of the international commu-
nity, operating through UN and EU representatives in the area.
3 Macedonia
The separation of Macedonia from the former Yugoslavia in September 1991 was
a relatively peaceful process. Through a combination of international and domes-
tic actors, together with a moderate approach by local political leaders, the coun-
try remained stable. After the 1999 Kosovo conflict, however, violence broke out
in early 2001 between Macedonian forces and a group that claimed to be fighting
for improved rights for ethnic Albanians in Macedonia. The parallels with North-
ern Ireland were noted, and the media referred to the possibility that the region
risked becoming a ‘Balkan Ulster.’10 In terms of the power-sharing institutional
arrangements set up under the Framework Agreement of 13 August 2001, with its
move away from a system of majoritarian democracy, there are further parallels
with the Belfast Agreement of Northern Ireland.
The Framework Agreement preserved the unitary character of the Macedo-
nian state, but developed local self-government to deal with a broad range of
socio-economic matters, such as the provision of public services, urban and rural
For a discussion of how “rule” by the international community works in practice, in

the case of Bosnia and Herzegovina, see Knaus and Martin 2003.
10 In Macedonia, the ethnic Albanians made up some 25% of the total population. In
Northern Ireland, the proportion of Catholics grew from some 30% in 1920 to about
45% today. So, both regions had a large minority population, and in both cases the
minority group appeared to be economically disadvantaged (in terms of educational
attainments, high quality jobs, etc.) – a matter that can easily be interpreted in terms
of overt discrimination.
544 John Bradley
planning, environmental protection, local enterprise development, culture, edu-

cation, social welfare, and health care. As in the cases of Bosnia and Kosovo, the
EU was the main provider of international finance, and promised over EUR 100
million linked to the confirmation of the Agreement by the Macedonian parlia-
ment. This was intended to finance repairs to damaged infrastructure and sup-
port public expenditure more generally.
4 Economic Future of the Western Balkans

The European Stability Initiative (ESI) has pointed out that, as the period of initial
reconstruction and stabilization draws to a close in the Western Balkans, there
is a looming crisis of social and economic dislocation (ESI 2002). The crisis is
emerging just as existing EU assistance is being scaled down, and in a context
where the countries of the region find themselves excluded from the EU enlarge-
ment process. The ESI suggests that the existing EU policy instruments – de-
signed with post-conflict reconstruction in mind – need to evolve into a genuine
and long-term commitment to address the region’s chronic economic and social
problems. A possible suggestion could be that the EU undertakes to include the
Western Balkans in its commitment to economic and social cohesion across Eu-
rope prior to opening negotiations with states in the region.
The term ‘cohesion’ in such a context first came into use in the late 1980s at
the time when major reforms and expansions of EU regional aid were being car-
ried out. As set out in Article 130a of the Treaty on European Union, there is an
explicit aim to promote “harmonious development” with a specific geographi-
cal dimension: by “reducing disparities between the levels of development of the
various regions and the backwardness of the least favoured regions” (Treaty on
European Union 1993). Thus, there is a recognition that wide disparities are intol-
erable in any community.
Progressive trade liberalization within Europe was always likely to entail sub-
stantial industrial disruption in the periphery, either defined as the member
states on the western and southern edge of the EU or as those sub-regions of
large member states that were located far from the main centres of population
and economic activity. One of the main reasons for the reform and enlargement
of EU regional policy (the so-called Structural Fund aid packages), was the in-
troduction of the Single Market in 1992. The political rationale came from the
fear that not all EU member states were likely to benefit equally from the Single
Market, whose purpose was to dismantle all remaining non-tariff barriers within
the Union. Initially, the less advanced economies of the southern and western pe-
riphery (mainly Greece, Portugal, Spain, and Ireland) were felt to be particularly
vulnerable unless they received development aid.
There are three main channels through which the development-enhancing
effects of EU Structural Fund aid operate. These investment programmes im-
prove the physical infrastructure of the economy; raise the level of human capital
(through enhancing the skills and education of the labour force); and directly as-
sist private sector performance by subsidizing investment. What is special about
the Structural Fund-type policies is their goals, i.e., to design and implement
policies with the explicit aim of transforming the underlying structure of the
beneficiary economies in order to prepare them for exposure to the competitive
forces being unleashed by the Single Market. These policies moved far beyond
a conventional stabilization role, being directed at the promotion of structural
change, faster long-term growth, and real convergence through mainly supply-
side processes.
How relevant are these policies to the West Balkan states? The first phase of
the transition of the former command economies of Central and Southeastern
Europe involved considerable disorganization and a very basic overhauling of
industrial and institutional capacity. Socio-economic mechanisms operating
during this phase entailed the creation of market-based institutional structures
accompanied by substantial re-allocation of labour between the public and pri-
vate sectors, as well as between manufacturing and market services. The initial
impacts of restructuring generate the well-known U-shaped pattern for income
and employment (Blanchard 1997). This process is not yet complete in the West-
ern Balkan region.
However, the processes that characterize the early years of transition in the
Western Balkans should not be taken as the pattern of behaviour for the future.
The second phase of transition is more likely to resemble the path followed in re-
cent decades by EU countries like Greece, Ireland, and Portugal, where the driv-
ing forces behind cohesion (or catch-up) include progressive trade integration,
foreign direct investment inflows, technology transfer, and EU-aided investment
programmes, mainly for the support of infrastructural and human-capital devel-
opment. However, in the real world, the two phases of transition are intertwined
and operate simultaneously.
What does one look for from an economic development strategy? It should
provide a coherent and comprehensive set of policies for economic and social
development for the foreseeable future, it must be consistent with the availabil-
ity of resources, and its various parts must be well integrated. More formally, it
must have four elements. First, it must set explicit targets and ensure that they
carry no ethnic bias. Prosperous developed countries can enjoy the luxury of set-
ting a diverse range of socio-economic targets, since they tend to have adequate
resources. Developing countries, on the other hand, need to focus on the over-
riding target of accelerating growth of income per head. Given appropriate care
in influencing income distribution within the country, other desirable socio-eco-
nomic and cross-community objectives tend to come into line as a direct conse-
quence of income growth (e.g., facilitating improvements in social conditions in
the health and education areas, improved environmental care, etc.).
Second, a development strategy must clearly identify constraints. For any tran-
sition country, particularly if there are violent conflicts, constraints are only too
plentiful. The country may be located in an unfavourable geographic-economic
area, close to zones of conflict and/or of deteriorating economic performance.11
11 One notes the fact that the very detailed IMF country reports are completely silent
on the issue of self-determination conflicts in progress, or recently concluded, in
546 John Bradley
In almost all such economies, there is an inadequate set of institutional struc-

tures. The initial level of development is usually low, putting constraints on the
availability of resources (i.e., policy-makers face financial constraints from the
balance of payments and in funding public sector borrowing requirements). It
will probably have an unfavourable initial economic structure, with a very large
agriculture sector, an unsuitable configuration of industrial sectors, and a very
under-developed market services sector. The level of human capital as well as the
level of technology tends to be low, placing constraints on feasible development
strategies. The need to reduce exploitation of non-renewable resources places
further constraints on growth.
Third, it must identify suitable policy instruments. The availability of suitable
social and economic policy tools is as important for the success of any develop-
ment plan as the clear statement of objectives. These usually include increased
investment, fiscal incentives aimed at stimulating the private sector, and appro-
priate wage-setting institutions. In particular, public investment must focus on
infrastructure and human capital, with increased investment in health, educa-
tion, and research, while stressing the need to improve telecommunication and
transport infrastructures as well as building information technology competenc-
es. Both foreign and indigenous firms require incentives, designed to make the
region more attractive to inward investment, since this will be the main facilita-
tor of technology transfer and the re-orientation of manufacturing towards fast-
growing export markets. Finally, the main threat to restructuring will be unrest in
labour markets. Any evolution of wage rates that disregards the need to maintain
international cost competitiveness will choke off growth. Consequently, econom-
ic development strategies can benefit from forms of social partnership among
the three main economic institutional players – government, trades unions, and
employers - aimed at producing an orderly evolution of wage costs that preserves
cost competitiveness.
C Nations of the Former USSR

Whereas the nations of Central and Eastern Europe (CEE) were only incorpo-
rated into the Communist economic system after the Second World War, the
constituent Republics of the USSR (with the exception of the three Baltic states,
incorporated after the Second World War) had organized their economies under
a system of centralized planning since shortly after the First World War. If one
examines the performance of the former communist countries, one sees that the
states of the CEE area made a very rapid transition, involving an initial collapse of
their economies by up to 50% from their 1989 size, followed by a rapid restructur-
ing and periods of fast growth. As a result, eight of these states will become full
the territories about which they are formulating economic policy advice. In the case
of reports on countries like Georgia, the Philippines, and Papua New Guinea, such
denial is unsettling!
members of the EU in June 2004.12 However, with the exception of Slovenia, their
standard of living remains low, ranging from about 50% of the EU average (the
Czech Republic) to about 40% (Poland).13
The two least-developed second-wave EU-candidate CEE states – Bulgaria and
Romania – most resemble the states of the CIS that emerged from the collapse of
the USSR. The economies of Bulgaria and Romania between the two World Wars
were less developed than those of the first wave of EU candidate states, so their
level of physical and social capital was also much lower when they liberalized in
the early 1990s. Their standard of living today is below 30% of the EU average and
they have been deemed to be unprepared for full EU membership until at least
2007. Moldova is not even an EU candidate state. Even if Georgia and Moldova
had been free of internal self-determination conflicts, they would be facing – to-
gether with Belarus and the Ukraine – formidable development challenges. Their
internal conflicts have simply added to already difficult problems of economic
restructuring by diverting energy and resources away from the economic chal-
lenges so that political problems can be solved.
1 Georgia
The self-determination conflicts in Georgia have occurred in Abkhazia and South
Osettia, and are described in the case study by Tsikhelashvili and Ubilava. Ab-
khazia was a reasonably well-developed and prosperous region of Georgia, with
a good level of infrastructure that included Georgia’s largest electricity power sta-
tion, important road and rail links, rich agricultural land, and mineral resources.
South Ossetia, on the other hand, was less economically significant, with little
by way of industrialization, and a considerable amount of income generation –
much of it illegal – arising from an important transit route through the Caucasus
mountain range along the central part of Georgia’s northern border with North
Ossetia/Alanya within the Russian Federation.
In the case of Abkhazia, one of the three Working Groups set up in Novem-
ber 1997 under the Geneva Process dealt with economic and social questions.
Mutual understanding was reached on a range of issues: foreign economic ties,
customs, energy, transport and communications, etc. For example, Russia is the
main consumer of electricity from the Inguri power station and CIS troops guard
the complex, and a series of trade treaties between Abkhazia and CIS republics
has facilitated its self-sustainability. But the conclusions reached by Tsikhelash-
vili and Ubilavaare pessimistic:
The effective independence of the Abkhazia breakaway region, with full sovereignty
over domestic and in part, external affairs against weakly developed ties with Georgia
proper makes it clear that for now all is about monopolizing power to the largest ex-
12 The eight CEE states who became EU members in 2004 are: Estonia, Latvia, Lithua-
nia, Poland, the Czech Republic, Slovakia, Hungary, and Slovenia.
13 For data on standards of living, as measured by gross domestic product per capita,
see Economic Commission for Europe 2003: Statistical Appendix.
548 John Bradley
tent possible in Sukhumi rather than sharing it with Tbilisi. Therefore, given the same
momentum and format of the settlement process, hardly any, even rudimentary joint
arrangements are in sight.
The Georgian-South Ossetian peace process is organized through a Joint Control

Commission (JCC) between Georgia, North and South Ossetia, and Russia, with
the participation of the OSCE, UNHCR, and the EU, and the JCC has a Working
Group on Economic Issues. A major economic policy problem is that Georgia is
unable to finance even modest allotments for the Tskhinvali regional rehabilita-
tion projects that are annually earmarked by the state budget. One of the major
drawbacks for effective progress towards a final resolution is the lack of economic
back-up. The EU has funded projects aimed at regional reconstruction and the
construction of interrelated and interdependent infrastructure between South
Ossetia and Georgia, as an initial stage in a process that may eventually trans-
form Georgia-South Ossetia economic relations into well-integrated and coher-
ent paths.
2 Moldova
In the case study in Chapter 10, Järve describes Moldova as a paradox. After the
collapse of the USSR and the communist system of governance, the post-Soviet
Republic of Moldova suffered from the classic problems of feeble state institu-
tions that were buffeted by ethno-nationalist conflicts. But Moldova is exception-
al in that it legislated to create an autonomous territorial unit – Gagauzia – and
codified power-sharing between the centre (Chisinau) and the autonomous re-
gion. The origins of the Gagauzia self-determination dispute arose from the de-
sire of the Popular Front of Moldova party to make the Romanian-like language
of Moldova into the only state language, and to promote a merger between the
states of Romania and Moldova. These moves were opposed by the two regions
with high concentrations of non-Moldovans – Transdniestra and Gagauzia. In
1990, Gagauz separatists unilaterally declared a breakaway republic, fearing a re-
surgence of ethnic Romanian nationalism within Moldova as it moved towards
independence. A state of emergency was declared after the separatists held un-
authorized elections to an independent parliament. However, in 1994, the newly
independent state of Moldova adopted a constitution that granted the Gagauz
special autonomous status, meeting some separatist demands. Since 1995, the
Gagauz have had their own parliament and adopted their own laws, though the
Moldovan government retains the right to approve decisions made by the Gaga-
uz parliament. This power-sharing arrangement is unique in Central and Eastern
Europe and the former USSR.
The aspect of interest to this chapter concerns the economic implications of
the power-sharing arrangement. The preamble to the law stresses, inter alia, the
aim to ensure “economic self-reliance.” The vertical institutional arrangements
have three levels: the state (capital Chisinau); the county level (twelve regions,
including Gagauzia); and the local level. Within Gagauzia, the People’s Assembly
has competences in the economic areas of education, housing and urban plan-
ning, health, and local budgetary matters, including taxation, the economy, la-
bour relations, etc. These competences are exercised by an Executive Committee,
and matters such as the budget, having been approved by the People’s Assembly,
are executed. However, these competences are not exercised in a vacuum, and
there are restrictions on the actions of the Gagauzia Assembly and Executive.
The Gagauzia local budget must include economic matters fixed by legislation by
the Republic of Moldova. In addition, the Gagauzia budget must conform to the
laws of the Republic of Moldova in the form of “fixed payments out of all forms
of taxes and payments.” These economic restrictions, and taxation in particular,
have proved to be troublesome for the relations between Chisinau and Gagauzia
and have led to complaints about the lack of support for the region from the
central government and to a lack of resources needed to implement local policy
decisions.
The Gagauz Assembly and Executive lay claim to all sources of revenue col-
lected on their territory, plus the revenue from customs and excise taxes col-
lected (by the centre) for goods actually consumed in Gagauzia. The Moldovan
government reject this approach, even though it appears to be well-founded in
the 1994 law! An approach based on the so-called ‘Barnett formula’ in the United
Kingdom has been suggested, where the centre allocates financial resources in
the form of a lump sum (or block grant), based on a determination of the ‘needs’
of each region. However, no agreement has been reached, and the poor economic
situation of Gagauzia has become a further source of tension.
In broad terms, the assignment of tax bases to sub-national governments in
Moldova is in accordance with general public finance principles (de Melo 1999).
Mobile tax bases are assigned to the central government budget, and immobile
tax bases are assigned to sub-national budgets. The main tax bases assigned to
sub-national governments are personal income, property, and land taxes. Sub-
national governments are also allowed to collect non-tax revenues in the form of
royalties and fees for the use and exploitation of natural resources, user charges
for goods and services provided locally, and land fees and duties. The central gov-
ernment collects excises and foreign trade taxes. The revenues of the corporate
income tax and VAT are shared between the central and sub-national govern-
ments on a derivative basis.
Two features of the Moldovan form of fiscal federalism are noteworthy. First,
despite the large ratio of own tax revenues to total revenues, the sub-national
governments are not free to set tax rates, which restricts their policy-making au-
tonomy. Second, the revenue-sharing system tends to deepen vertical imbalances
between central and local budgets. As a result, expenditure needs and revenue-
mobilization capacity are not equalized among the sub-national jurisdictions on
a per capita basis, and regional income inequalities are perpetuated. Over time,
the perpetuation of regional differences in income can accumulate into regional
differences in stocks of financial, infrastructural and human capital. ‘Catch up’
would then require ‘affirmative action’ of a kind that would compensate disad-
vantaged regions for past fiscal neglect. This kind of action is almost certain to
550 John Bradley
provoke strong political opposition in the wealthier ‘donor’ regions, particularly

when the general economic climate is static or deteriorating.
The de Melo (1999) IMF study concludes that the devolution of expenditure
functions to sub-national governments should be matched by a sensible assign-
ment of revenue sources across government levels, without depriving the centre
of revenue sources and tax instruments, which it is best equipped to use more
efficiently (in the sense of operating services that are pure public goods (e.g.,
national defence, basic research) or when there are inefficiencies arising from
externalities across jurisdictions (e.g., transport, communications, and energy
infrastructure).
But complex arguments about the best way to distribute tax revenues between
the various levels of sub-national government in Moldova are almost irrelevant
when the real problem is that the ability of the Moldovan economy as a whole to
generate tax revenues of any kind falls far short of supporting the level of public
expenditure that is required to implement the necessary economic restructur-
ing. The rise in the level of poverty over the transition period has been caused
by both internal and external factors. The dissolution of the Soviet Union, the
accompanying disruption of trade and payments, and the price shock after the
liberalization of the price of imported energy, all served to damage the economy.
The conflict with Transdniestra in 1992, and the continuing regional instability in
Gagauzia exacerbated matters further. This made it very difficult for the Moldo-
van government to implement its poverty reduction strategy, based on sustain-
able and inclusive growth, human development policies emphasizing increased
access to basic services, and social protection policies targeting those most in
need (IMF 2000).
D Mindanao
The previous five case study regions can be examined within a wider European
– or more specifically, a European Union – context, as can the case of Northern
Ireland to be discussed in the next section. But the two remaining case studies
– Mindanao and Bougainville – are remote from Europe.
Mindanao, a large island in the Philippines archipelago, falls within the context
of the group of ten ASEAN (Association of South East Asian Nations) economies
that includes Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philip-
pines, Singapore, Thailand, and Vietnam. Mindanao provides the only case study
among the eight where the ethnic basis for the demand for self determination – in
this case, the differences between the Christian Filipino and Muslim population –
arose in a context where there had previously been overt and serious discrimina-
tion (against the Muslim people). At the time of independence, the Muslim polit-
ical leaders were too remote from the centre of power in Manila to be influential.
The education system was biased against Muslim Filipinos, and low educational
standards restricted Muslim access to high quality administrative and industrial
jobs. This led to a spiral of deprivation that fuelled paramilitary violence.
Although it is less developed than other regions of the Philippines, the large
island of Mindanao is recognized as having considerable potential, but this po-
tential could only be exploited if there was peace and political stability. Unless
and until the violence ends, it is difficult to explore this case from an economic
policy point of view. Any political solution to the self-determination dispute of
the Muslim population is almost certainly to be within the context of a degree of
regional political autonomy conceded by Manila. The economic initiatives that
will have follow after the political solution are likely to be similar to those already
in operation in Northern Ireland or Moldova.
E Bougainville
The self-determination demands of Bougainville were primarily related to issues
of ethnic identity. But the economic prospects of self-government or autonomous
status were enhanced by the presence on the island of one of the world’s largest
copper and gold mines at Panguna – operated by a subsidiary of the international
mining conglomerate, Rio Tinto Zinc Ltd. Of course, the revenues from the mine
were also important to the economic viability of Papua New Guinea itself. Early
attempts in 1975 by the national government and the Bougainville Interim Pro-
vincial Government to reach agreement on an equitable sharing of the revenues
between the national and local authorities failed.
Disputes over the mine – relating to rents, compensation for land, environ-
mental damage, and perceived discrimination against local workers – came to
the surface again during the second secession attempt in 1989–2001. This led to
the closure of the mine in May 1989, and that closure sent a shock through the
already fragile local economy that caused further deterioration in the economy.
During the accompanying violence, much of the island’s infrastructure was de-
stroyed or damaged.
It is difficult to say anything profound about the socio-economic dilemma that
faces the island of Bougainville, with its population of 200,000 people. Bougain-
ville finds itself an unwilling part of the state of Papua New Guinea simply because
of what was probably an unthinking, unreflecting, and callously indifferent act of
the imperialist powers of the early twentieth century. Perhaps if the island had
been joined with the Solomon Islands, the present ethnic violence between the
centre (Papua New Guinea) and Bougainville might have been merely a peaceful
dispute between local administrations about how the rents from the now closed
copper and gold mine should be shared between the islands, and there might be
a greater willingness to make a common case to the outside world seeking devel-
opment aid.
IV Economic Policy and Power-sharing: Northern Ireland

The details of conflict resolution in Northern Ireland have been described in the
case study by O’Leary in Chapter 4. Here we are concerned with the economic
policy aspects of the Belfast Agreement, the document that sets out the new in-
552 John Bradley
stitutional arrangements within Northern Ireland (subsequently referred to as

the North), and between Northern Ireland and the Republic of Ireland (subse-
quently referred to as the South).
At the time of writing, the Assembly set up by the Belfast Agreement has been
suspended, precipitated mainly by the slow pace of decommissioning of weapons
by the IRA, the paramilitary arm of the Republican movement. But this did not
lead to a collapse of the peace process. Many of the benefits that came with the
Belfast Agreement are still operating, due both to the generally impartial behav-
iour of the public sector agencies (including the civil service and the police force),
as well as the very harmonious relationships that exist between the British and
Irish governments (under Strand 2 of the Agreement, as we shall see).
A Economic Governance in Northern Ireland: Strand 1

The central aspect of the Belfast Agreement is Strand One, dealing with the polit-
ical governance of the North. Northern Ireland is a peripheral regional economy
within the United Kingdom that also happens to be a region that is part of the
island of Ireland. Although it enjoyed some policy autonomy in fiscal and public
expenditure areas under the Stormont parliament prior to 1972, today the North-
ern authorities have only limited discretion on regional economic policy. Mon-
etary policy in the North, and almost all elements of fiscal policy, are set by UK
norms, although there remains some discretion in the area of regional industrial
and labour market policies.
The favoured position of the Unionist parties is for the North to continue as a
constitutional part of the UK, with minimal formal provision for any North-South
political and economic policy structures. The Belfast Agreement established a
Northern Assembly and, discounting for the present the important North-South
implementation bodies of Strand 2, the Northern economy will continue to func-
tion as a region of the UK in much the same way as do Scotland and Wales. This
means a perpetuation of the situation where the standard of living that the North
experiences is underwritten to a very large degree by the British taxpayer, and
where the size of the public sector, and the role played by financial transfers and
subsidies, is likely to continue a situation of economic dependency into the me-
dium term.
From a Unionist perspective, being a region of the UK carries with it an auto-
matic guarantee of parity in the quality of public services. Having started from
a lower base, and having experienced exceptional economic difficulties during
the conflict, Unionists feel it quite natural that the North should require a pe-
riod of ‘catch-up’ during which it will be entitled to attract a greater per-capita
subvention than more prosperous UK regions. Furthermore, they argue that this
dependence on transfers from Britain acts as a bulwark against any possibility of
a ‘united Ireland’, given the inability and/or unwillingness of southern Irish tax-
payers to take over the present subvention, and thus sustain the standard of living
in the North at its current level.
The above regional, peripheral framework has implications for the economic
governance of the North, with policy-making centralized in London and little in
the way of real local policy autonomy. But in the global economy, economic logic
commonly cuts across political/national boundaries while leaving these bound-
aries intact. Political realities, as well as geographical proximity to the Republic
of Ireland, have presented the North with what might be termed a ‘Southern’
problem, the resolution of which will require creative and innovative movement
in both the spheres of political and economic governance (Bradley 2001). Neither
independence nor incorporation into the UK appear to offer an optimum eco-
nomic solution for the North. The former seems quite infeasible for the foresee-
able future, in light of the heavy dependence on financial transfers from London
that the South would be unable, or unwilling, to take over (Bradley 1996). The
latter is quite feasible (indeed, it represents something approximating the status
quo), but it leaves political problems unsettled, has led to a Mezzogiorno situa-
tion of chronic economic dependency and fails to normalize relationships with
the South.
Because of its history, any exploration of the North’s economic relationship
with the South is bound to be contentious. A comparison with the situation of
other EU regional economies provides an alternative, perhaps less contentious,
framework that offers the potential for new insights into the future economic de-
velopment of the North. However, as noted in Dunford and Hudson (1996), there
had been little or no earlier examination of the consequences of placing the North
fully in the EU regional economic context, or government policy documents that
were explicitly influenced by policy developments elsewhere in Europe.
Dunford and Hudson (1996) compared the North with the political and eco-
nomic governance of four different, but more successful, European regions - Jut-
land in Denmark, Rhone-Alpes in France, Saarland in Germany, and Abruzzo in
Italy. They concluded that the most active regional governments are to be found
in the most economically successful regions, and that their ability to exercise a
high degree of pro-activity is predicated upon their location within nation-states
characterized by decentralized systems of governance. Successful regions tend to
be characterized by distinctive forms of local regulation and governance. They
also have systems of governance that embrace enabling and facilitating institu-
tions within the local state and civil society, as well as bridging the permeable
boundaries between them and adjoining regions and states. Part of the problem
of less successful regions (such as the North) is that they are locked into institu-
tional structures that were relevant to an earlier phase of successful economic
and political development, but which now constitute a barrier to moving onto a
new development trajectory.
What is revealing in EU regional comparisons is that inter-regional coopera-
tion does not necessarily require complete, or even extensive, harmonization of
economic policies. Rather, it thrives where policy differences are fully understood
and are made more transparent against the background of removal of non-tariff
barriers to trade through the implementation of the Single Market. Such findings
suggest that mutually-beneficial North-South as well as East-West cooperation
554 John Bradley
could be built along similar lines if contentious political issues could be resolved.
Indeed, the Belfast Agreement points exactly to this conclusion and it is to that
we now turn.
B North-South Cooperation and the Belfast Agreement: Strand 2

The cornerstone of Strand Two of the Belfast Agreement is the North-South
Ministerial Council whose remit was described as follows:
[T]o bring together those with executive responsibilities in Northern Ireland and the
Republic of Ireland, to develop consultation, cooperation and action within the island
of Ireland – including through implementation on an all-island and cross-border basis
– on matters of mutual interest within the competence of the Administrations, North
and South.
This Council was designed to meet in different formats: in plenary format twice a
year, and in specific sectoral formats on a regular and frequent basis. The activi-
ties of the Council includes the exchange of information; discussions and consul-
tation on areas of cooperation; the use of best endeavours to reach agreement on
the adoption of common policies in areas where there is a mutual cross-border
and all-island benefit; to take decisions by agreement on policies for implemen-
tation separately in each jurisdiction in relevant meaningful areas; and to take
decisions by agreement on policies and action at an all-island and cross-border
level to be implemented by other bodies to be established.
During the transitional period between the elections to the Northern Ireland
Assembly and the transfer of power to it, it was intended that representatives
of the Northern Ireland transitional administration and the Irish government,
operating as the North-South Ministerial Council, would undertake a work pro-
gramme, in consultation with the British government, with a view to identifying
and agreeing areas where cooperation and implementation for mutual benefit
could take place. In an annex of Strand Two in the Belfast Agreement, twelve
possible areas where either new implementation bodies or the use of existing
institutions and co-operative arrangements could be used are listed. As part of its
work programme, the Council was to identify and agree at least six matters for
cooperation and implementation in each of the following categories:
a) Matters where existing bodies would be the appropriate mechanisms for co-
operation in each separate jurisdiction;
b) Matters where the cooperation would take place through agreed implemen-
tation bodies on a cross-border or all-island level. The implementation bod-
ies would have a clear operational remit and would implement all-island and
cross-border basis policies agreed in the Council.
The Council was also to consider the EU dimension of relevant matters, including
the implementation of EU policies and programmes and proposals under con-
sideration in the EU framework. The views of the Council were also to be taken
into account and represented appropriately at relevant EU meetings. Finally, the

new Northern Ireland Assembly and the Southern Oireachtas (parliament) were
to consider developing a joint parliamentary forum and consideration was to be
given to the establishment of an independent consultative forum appointed by
the two administrations, representative of civil society, comprising the social
partners and other with expertise in social, cultural, economic, and other issues.
The political sensitivities involved in drawing up the list of possible areas for
North-South cooperation were clearly reflected in the highly technical and tight-
ly drawn nature of the suggested functions. In particular, the list seemed to reflect
political compromises made between parties in the negotiations leading up to
the Belfast Agreement. This is seen most clearly, for example, by the absence of
significant economic and industrial matters, in particular issues such as the pro-
motion of industrial development and the attraction of inward investment. It is
interesting to note that the business community – North and South – has argued
that the Belfast Agreement did not go far enough in facilitating the economic and
business initiatives that they have been promoting on a cross-border or all-island
basis (IBEC/CBI 1998). At least in this context, the business community appears
to be ahead of politicians in seeing the economic logic of more extensive North-
South economic cooperation, in a situation where the political willingness for
such ‘deep’ cooperation is absent on the Unionist side (Bradley 2001).
Following intensive negotiations, an agreement was reached on 18 December
1998 on matters of North-South cooperation and cross-border policy implemen-
tation (Trimble and Mallon 1998). The Trimble-Mallon statement of 18 Decem-
ber listed six North-South implementation bodies: inland waterways, food safety,
trade and business development, special EU programmes, language, and aqua-
culture and marine matters. Finally, the Trimble-Mallon statement contained an
initial list of six matters for North-South cooperation through the mechanism of
existing bodies in each separate jurisdiction. These included non-controversial
aspects of transport, agriculture, education, health, the environment, and tour-
ism.
A better understanding of the outcome of the Trimble-Mallon talks can be ob-
tained from an earlier statement of Mr Trimble where he set out three principles
that his party intended to apply to the selection and design of the cross-border
implementation bodies (Trimble 1998). First, any new body should have demon-
strable advantages both for Northern Ireland and for the Republic of Ireland. Sec-
ond, Northern Ireland’s ability to develop a vibrant and competitive economy on
a sound basis should not be impaired. Third, Northern Ireland’s identity should
not be submerged in a new all-Ireland identity. The Belfast Agreement enshrined
the principle that the people in Northern Ireland consent to the continued ex-
istence of Northern Ireland within the UK, and others, including nationalists,
endorsed the legitimacy of that choice. While the logic of the first point – the
necessity of mutual benefit – is obvious, the defensive logic of the second and
third points – access to local policy instruments and regional identity – stands
in contrast to the latest thinking on the dynamics of regional economic develop-
ment where national and regional governments have a series of vital roles to play
556 John Bradley
in promoting economic growth and development (Porter 1998). So, the cross-
border institutions set up under the Belfast Agreement, however beneficial, leave
many unresolved issues and appear to have had only a weak impact on easing
inter-community tensions.
V Conclusions
Poor countries, and in particular regions with self-determination conflicts, often
have inadequate financial resources to address their developmental problems. If
left to their own devices, such countries run the risk of drifting further into pov-
erty, and thus further threatening internal socio-economic cohesiveness. Poverty
and conflicts of self-determination tend to form a vicious circle that can easily
spiral out of control. In the immediate aftermath of conflict, regional, national,
and international economic policies must focus on the restoration of stability
and the repairing of war-damaged infrastructure. But the major contribution of
economic policy tends to come in the longer term.
When preparing national and regional strategies for investment in the post-
conflict case study regions, we have suggested that it may be useful to reflect on
the pattern that emerged in the EU structural aid programmes. These are usually
classified into three broad economic categories: support for basic infrastructure
(i.e., roads, telecommunications, etc.); support for human resources (training, re-
training, etc.); and support for productive structures (investment and market-
ing subsidies, etc.). Such programmes provide examples of how economic policy
within the EU is shifting from one appropriate for independent and individualis-
tic states to that of region-states fully integrated into an encompassing European
economy. The willingness to go down this road was conditioned by domestic
policy failures of the past.
Few people believe any longer that impacts emanating from national discre-
tionary demand management and ‘fine tuning’ policies provide durable support
for success in the long run. At best, policies designed to stabilize the economy
will provide the basis for subsequent sustained growth. Stabilizing at a low level
of welfare, exacerbated perhaps by regional inequalities, is never an enduring so-
lution.
Krugman (1997) has described the issues that will be central over the next de-
cade to the management of many of the small open economies described in the
case studies. For example, he suggests that industrial revitalization will be led by
an initial clustering of similar industries supported by local suppliers of special-
ized inputs subject to economies of scale. These clusters are likely to generate a
local labour market for skilled workers, which further facilitates the growth of the
cluster. Appropriate human resource policies are crucial at this stage. Spillovers
of information will further encourage growth in the initial sectoral cluster and
provide the basis for additional clustering effects, often in traditional areas that
benefit from new technologies (e.g., food processing, textiles, clothing). To facili-
tate this stage, the improvements in physical infrastructure and in the produc-
tive environment are crucial. Finally, a consensual process of social partnership
needs to be put in place to ensure that there are as few losers as possible in the
economic restructuring that accompanies such a virtuous circle, with the result
that growth is less likely to be choked off by industrial unrest if the social partners
fight over their respective shares of added value.
More generally, and almost without exception, the small states of the case stud-
ies will have to carve a path between liberalism and statism, and will need to
evolve towards indirect forms of economic control (Katzenstein 1985). What will
characterize the economic and political experience of small states and what will
set them apart from the large industrial countries will be the ‘premise’ of their
planning efforts: namely, adaptation to external market forces. Detailed compre-
hensive sectoral planning efforts will be increasingly inapplicable, simply because
of the openness and vulnerability of their economies to external pressures. Their
main challenge will be that of selecting the planning devices that are in harmony
with their social objectives. Hence, the rationale for state intervention depends
on the ability or otherwise of market forces to yield results consistent with these
social objectives. Because of their lack of autonomy, their strategies must be flex-
ible, reactive, and incremental. They cannot oppose change by shifting its costs
to others. Neither can they ignore change if they wish to prosper. Their economic
policies are seldom the root cause of conflicts of self-determination, and neither
are they the only solution. But they can be an important part of the longer term
solution.
Chapter 18
Policing Territories Previously Subject to Civil War and
Ethnic Violence
Philip Towle
I Introduction
In unsettled communities plagued by ethnic tensions, the establishment of an
efficient and impartial police force is crucial to attempts to restore harmony. By
definition, a peaceful, settled community trusts its police forces to enforce the
law equitably and to bring criminals to justice. A deeply divided community nat-
urally produces a police force distrusted by one or more ethnic or social group.
Such a force, in turn, reinforces the divisions in society, leading to a vicious circle
of suspicion and tension. A major objective of recent peace settlements after civil
conflict has been to break this circle by establishing a police force that has the
confidence and support of all parts of the community.
Despite the importance of policing, much less has been written about the
problems of policing ethnically divided communities than about peacekeeping
with military forces. This is partly because armed forces ensure that they publi-
cize their activities and partly because efforts by the police to mediate fairly are
undramatic and may take years to have an impact. Military peacekeepers are like
the ambulance that rushes to the scene of an accident to save the lives of those
involved, whereas police forces resemble the family doctor who has responsibil-
ity for day-to-day health care – the former is visible and spectacular but it is the
latter who has the greatest impact on people’s lives.
Monitoring, advising, and training local police forces to reduce sources of con-
flict has long been an aspect of peacekeeping operations. More recently, in East
Timor, Haiti, and Kosovo, the peacekeepers actually took over executive policing
(Hansen 2002). It is not, however, with the progress of these operations that this
chapter is primarily concerned but with measures to reform or create indigenous
police forces after internal conflicts.
For efforts to publicize peacekeeping activities, see Clifford and Wilton 2002. Note
the virtual absence of references to police work in this collection.
560 Philip Towle
In a civil war, the police force normally either collapses and disperses or takes
one side in the conflict. In a guerrilla or terrorist conflict, the police force is iden-
tified with the government both symbolically and actually. It is the key agent in
anti-insurgent activities, hence the emphasis placed on its efficiency by theorists
and practitioners of anti-guerrilla operations. In many ethnic conflicts, the police
force itself will have been involved in widespread abuse of human rights, further
alienating minority communities. On its side, the police force is likely to develop
bitter hostility to those who attack its members and, because they are normally
the first people on the scene after an attack, against those who kill or maim civil-
ians during the course of the conflict.
For all these reasons, most recent settlements following complex ethnic and
religious disputes place great stress on the constitution and make-up of the police
forces, because the negotiators know that such forces’ future behaviour will quite
largely determine relations between the communities. Signs of bias amongst the
police will re-awaken antagonisms. Conversely, if the police can show that all will
be treated equally before the law, confidence in the organs of the state may gradu-
ally be re-established. The negotiators of peace settlements in Northern Ireland,
Kosovo, Bosnia, Montenegro, and elsewhere have tried to increase the propor-
tion of ethnic minorities in the police force. In many cases, international organi-
zations have played a constructive role by monitoring and improving the quality
and the skills of these forces, helping to establish police academies and seconding
officers from other police forces to work with the reconstituted police.
II Case Studies
A Northern Ireland
The Northern Irish case illustrates the difficulty of increasing the proportion of
ethnic minorities in an existing police force while simultaneously maintaining
the efficacy of the force and the trust of the majority community. Despite the loss
of 300 police officers in 30 years of ethnic violence since 1969, the Royal Ulster
Constabulary had survived and borne the brunt of much of the rioting and civil
unrest. It was drawn almost entirely from the Protestant community, not least
because the IRA killed, or threatened to kill, any Catholics who ventured to join.
The Northern Ireland peace settlement led to the establishment of an Indepen-
dent Commission under Christopher Patten to consider reforming the constabu-
lary. This recommended the creation of a Policing Board to oversee the force,
made up of elected representatives from the various communities, and suggested
that the police should recruit equal numbers from the Catholic and non-Catholic
populations over the next decade, so that 30% of the force would be of Catholic
origin within ten years.
For example, the account of the behaviour of the Royal Ulster Constabulary in Hast-
ings 2002: 33 ff.
18 Policing Territories Previously Subject to Civil War and Ethnic Violence 561
The British government believed it had to take more account of feelings within
the Protestant majority in Northern Ireland, the morale and loyalty of the Royal
Ulster Constabulary, and the advice of the British security services. As a conse-
quence, they tried to implement the proposals in modified form, although some
of their modifications were withdrawn under pressure from the Catholic com-
munity. Patten wanted the force’s name changed to the Northern Ireland Police
Service, while the title chosen was ‘The Police Service of Northern Ireland (in-
corporating the Royal Ulster Constabulary)’ (PSNI). The government reduced the
period when the police would have to recruit equally from the two communities
from ten years to three. The government also insisted that the Union Jack fly on
government buildings on specified days – Patten would have removed the flag
and the Queen’s portrait from police stations. For both sides, these were deeply
emotional issues: the RUC believed that their sacrifices had been ignored by the
Patten Commission, while the Catholics often maintained that the new Police
Service was just the old RUC under a thin disguise. As a result, the IRA used the
changes to justify its refusal to ‘decommission’ its weapons and this has periodi-
cally led to the suspension of the whole Agreement.
The power of the Policing Board was less than Patten had proposed, so that
the Secretary of State for Northern Ireland oversees the Board, and the Chief
Constable of the PSNI can refuse requests from the Board if he should choose
to do so. On the other hand, the Board has significantly more power than its
predecessor, the Police Authority for Northern Ireland. Of the nineteen mem-
bers of the Board, ten are from the National Assembly. The other nine positions
are advertised and the applicants interviewed. All members are appointed by the
Secretary of State to reflect the various parties and communities in the province.
The Chief Constable answers to the Board on any significant aspect of policy,
although he can make decisions that he believes are necessary to maintain law
and order. After consultation with the Chief Constable, the Board sets targets for
police performance and monitors progress towards these. It monitors police ac-
tions in line with the Human Rights Act of 1998, issues a Code of Conduct for all
officers, and appoints visitors to make certain that those detained by the police
are properly treated. The Policing Board appoints Independent Community Ob-
servers to monitor the recruitment of police officers and to make sure that they
are treated fairly. Anyone can volunteer to become an observer, except current
and former police officers and those with some criminal convictions. District Po-
licing Partnerships have also been established by each District Council to consult
with local people and ensure they are satisfied with policing (Northern Ireland
Policing Board).
B Macedonia
In contrast to the situation in Northern Ireland, the Albanian minority in Mace-
donia had traditionally joined the Macedonian police forces and army, albeit in
small numbers, and often despite feeling that they were treated like second-class
citizens in these bodies. A brief civil war broke out in February 2001 between
562 Philip Towle
government forces and Albanian rebels, ending with the compromise Framework
or Ohrid Agreement negotiated between the communities on 13 August 2001.
The Agreement set out measures to increase confidence between the communi-
ties. These included better representation of the different ethnic groups in the
civil service, police, and armed forces. The Ohrid Agreement allowed local police
commanders to be appointed from a list drawn up by the Interior Ministry. At
the same time, to try to make the force more representative of the various com-
munities, 1,000 new police officers were to be recruited from the “communities
not in the majority” by 2004. The new recruits would be stationed within their
local communities and the changes would increase the percentage of Albanians
within the whole force from about 3-23%. By June 2003, 632 cadets had gradu-
ated from the new nine-month Basic Training Programme, 84% of whom were
from ethnic minorities (International Criminal Investigative Training Assistance
Programme 2003).
Even before the outbreak of the civil war, the US International Criminal Inves-
tigative Training Program (ICITAP) had been assisting the Macedonian police. In
May 2000, ICITAP concluded an assessment of the police and subsequently began
to assist the Macedonian National Police Academy with its training programmes,
particularly by establishing a unit to raise professional standards. ICITAP is also
reviewing policing in a number of Macedonian towns to assess the possibility of
introducing community policing and, possibly, to encourage the donation of rel-
evant equipment. ICITAP is also working with the Macedonian authorities to try
to reduce the incidence of human trafficking through the country (International
Criminal Investigative Training Assistance Programme 2003).
In August 2001, the OSCE secretariat convened an informal working group
in Vienna of the European Union Police Unit, ICITAP, and OSCE representa-
tives. This carried out a number of investigations in Macedonia and eventually
established an OSCE Police Development Unit to assist the Macedonian govern-
ment with its police reforms and training. ICITAP and the PDU work together
to encourage this activity. The PDU has assigned police trainers to the former
crisis areas to work with the local police, mayors, municipal leaders, and citizen
advisory groups. Reports from these groups suggest that citizens have increasing
confidence in the ability of the police to deal with crime.
Neither in Northern Ireland nor in Macedonia was an effort made to specifi-
cally integrate the insurgents themselves into the police force, merely to encour-
age representatives from the minority community to become police officers. Yet
the danger of former insurgents reverting to violence either as criminals or as
members of a new violent political organization is ever-present after prolonged
conflicts. They have, after all, often known nothing but violence. They owe their
status in the community to the violence they have used. Some may eventually be
sickened by the violence and killing but many would miss the feeling of impor-
tance and the sense of belonging to a pivotal group which may disappear when
a settlement is negotiated. On the other hand, to integrate them into the police
For an account by one IRA defector see Collins 1997.

force itself implies a degree of trust and reconciliation which, for obvious rea-
sons, is often difficult to achieve.
C Mindanao
In Mindanao, in the Southern Philippines, by contrast with the two previous
cases, the insurgents of the Moro National Liberation Front (MNLF) were sup-
posed to be integrated into the Philippines’ National Police (PNP) and the armed
forces by the settlement of September 1996. Quite quickly, 1,500 former guerrillas
were incorporated into the police and 5,250 into the armed forces, as part of the
number the agreement had stipulated. Those former insurgents not incorporated
were to be re-trained in a “special socio-economic, cultural and educational pro-
gramme” so that they could improve their chances of earning a livelihood and
not feel a need to revert to violence. The combined police force was to maintain
order and investigate crimes in the autonomous Moro region. Given the ambi-
tious scope of the whole enterprise and the difficulty of former enemies working
together peacefully, however, it was not surprising that there were some subse-
quent complaints of discrimination in the police and armed forces against the
new recruits.
D Bougainville
After nine years of conflict with Papua New Guinea, Bougainville gained a some-
what greater degree of autonomy in policing matters than the Moro region had
gained from the Philippines. Under the Bougainville Peace Agreement of 30 Au-
gust 2001, Bougainville could establish its own police force and the role of the
national or Papua New Guinea Police would subsequently be restricted to en-
forcing national criminal laws. Even this work could be taken over by the Bou-
gainville Police, although the national government would audit the new force’s
compliance with national laws. The new force and its commander would come
under the control of a minister of the autonomous Bougainville government, but
the commander would be appointed (and could be removed) by an independent
Bougainville commission and would hold a title below that of the Commissioner
of the National Papua New Guinea Force. The Bougainville police were to wear
national (PNG) emblems and to develop standards compatible with the national
force. Under the Agreement, the national government would also help fund the
Bougainville police, and consideration would be given to the wishes of police
serving in the national force who wanted to transfer to the new Bougainville
service. Despite the good intentions behind these proposals, however, financial
problems and local hostility meant that little progress was made and Bougain-
ville was described in June 2003 as “starting from a zero basis in infrastructure
terms in most areas except Buka.” New long and medium term plans were then
drawn up for policing, the courts, and the justice system. These look forward to
the rapid establishment of a Bougainville Transitional Police Service (BTPS) tai-
lored to suit the special circumstances in the area. Former combatants, who were
564 Philip Towle
members of the JBLEB, are to become special constables until 100 additional full-
time officers have been trained. The aim is to increase the number of full-time
police to some 200, to be made up of the 68 existing personnel, 30 experienced
police from other parts of the country, and the 100 newly trained officers. In the
meantime, communities will have to be reassured about the entry into their areas
of the full-time officers who will be deployed across the country. The officers’
role will be to support and supervise the 396 existing community police officers
and, if finances permit, to expand their numbers. Communities will be involved
in the selection of the new police officers. Bougainville will need extra funds for
these projects – many community police officers have resigned in recent months
because they have not been paid (Bougainville Administration Officers’ Working
Group 2003).
Full-time and community police are to work closely with chiefs and other tra-
ditional leaders. In rural areas, full-time police are to be directly involved only
with serious criminal cases and with issues that cannot be sorted out by the chiefs
or village courts. Community policing will also be adapted so that it can be ap-
plied to urban areas. To emphasize the break with traditional police methods, a
new badge and uniform are planned, and the new title ‘Bougainville Transitional
Police Service’ is also intended to symbolize the changes. The officers will not
carry weapons in the normal course of their duties.
If the long years of ethnic strife have undermined policing in Bougainville, they
have also weakened or destroyed the court and justice systems. The National
Court has not sat in Bougainville since 2000, partly because of anxiety amongst
the court personnel about security and partly because of funding shortages. The
result has been that even serious cases, such as rape, have not been tried, and
standards of human rights have not been maintained. All district court houses
have been destroyed and funds are short for their replacement and for the nec-
essary equipment. The intention is to have court houses in Buka, Arawa, and
Buin, providing the local people prove amenable to the plan. Village courts are
evidently more popular and many have been re-established, though funding is
again restricted and allowances have not been paid since late 2002. The Council
of Elders can play a major role in the law and justice system and the adminis-
tration is seeking funds for their training (Bougainville Administration Officers’
Working Group 2003).
Land disputes cause serious problems in many areas. They should not nor-
mally go to the Land Court until they have been dealt with by a land mediator, but
no mediators have been trained since the conflict began. Bougainville is seeking
help from Australia and New Zealand to help re-establish this system. The prison
or lock-up system has also broken down. Lock-ups are needed in Buka, Arawa,
and Buin, but there is local resistance. The plan is for the Hutjena facility in Buka
for people on remand or serving short sentences to be replaced, and for other
lock-ups to be re-established when local communities agree.
E Kosovo
In Kosovo, following the conflict between the Serb and Albanian inhabitants of
the province, a wholly new police force was established under the aegis of the
United Nations Mission in the province (UNMIK) and the OSCE. The two na-
tional communities were encouraged under the Interim Agreement to volunteer
for the new police force. The recruits were then to be sent to the Kosovo Police
School, and only after they had successfully completed the course would they be
offered employment in the police force. The School was thus given a vital role in
the construction of an effective force that could restore the rule of law and end
the revenge killings that were so frequent immediately after the war. In Febru-
ary 2000, a Professional Standards Unit was established to investigate any com-
plaints made against the police by the public. During the course of that year,
twenty-one officers were reprimanded, eighty-three were suspended without pay,
and seventy-two were sacked.
It was posited that previous political activities would specifically not act as
a bar to joining the police. Although the Agreement did not encourage former
insurgents to volunteer, many did so. By the end of 2002, 39% of officers were
former KLA members. That they took a risk was clearly demonstrated by the
murder of Ekrem Rexha outside his home on 8 May 2000. Rexha was a former
commander of the Kosovo Liberation Army and one of the first to join the new
police force. The Agreement allowed for a force of up to 3,000 communal police
officers, although this could be increased or decreased as the need arose and as
the Criminal Justice Administration advized. Recruitment from local communi-
ties was vital if trust was to be re-established. In Zubin Potok, for example, 90%
of the community were Serbs, with a deep suspicion of the police force. Until of-
ficers were recruited from the local community, it proved impossible to create the
necessary confidence. These and other police officers were not armed normally,
although a rifle was to be maintained for every fifteen police officers. The com-
munal police commander had to gain permission from the Interior Ministry to
use the weapons except in cases of self-defence. When the peace settlement was
reached, the Ministry Interior Police of the Yugoslav state were still operating
within Kosovo. They were to be withdrawn within a year and the combatants,
including the Kosovo Liberation Army (KLA), were to surrender their heavy ar-
maments and some small arms. Subsequently, the Yugoslav police could enter
Kosovo in hot pursuit of serious criminals, but they were to notify the authori-
ties in Kosovo within an hour of doing so. The Republic of Serbia’s Border Police
would continue to exercise authority at the international border crossings around
Kosovo but they were to recruit from the Kosovo population and they were also
obliged to go for training at the Kosovo police academy.
The UNMIK Police Commissioner, Christopher Albiston, reported in 2002
that the number of serious crimes had remained constant. Of these, however,
the number typical of a society recovering from conflict fell significantly. The
exceptions included the destruction of one of the coaches in the so-called Nis
Express – a crime that killed nineteen people and injured twelve more. When it
566 Philip Towle
was blown up by a mine, the coach was taking Serbs back to the graves of their
ancestors in an area now dominated by Muslims. Forensic evidence enabled a
suspect to be found, but he subsequently escaped from a US base. UNMIK Police
made a major effort to protect witnesses so that this and other serious crimes
could be investigated. They formed a Close Protection Unit to defend those who
might be vulnerable to assassination. In May 2001, UNMIK also established a
Police Aviation Support Unit (with a helicopter) to overcome some of the difficul-
ties of covering the rugged Serbian terrain. Initially, the International Criminal
Tribunal (ICTY) was responsible for searching for the bodies of those who had
been murdered during the ethnic violence. However, UNMIK set up a Missing
Persons Unit (MPU), and by the end of 2000 this had role had been taken over
from ICTY. The MPU had handled 3,915 cases, of which 3,251 relate to Albanian
victims, 414 to Serbians, and 59 to Roma. By the end of 2002, about 2,000 bodies
had been discovered but not all identified (UNMIK 2002).
Like Macedonia, Kosovo lies on a major route for smuggling to and from West-
ern Europe. In particular, the smuggling of drugs, women, and cigarettes causes
major problems. Interdiction is particularly difficult because of the length of the
border, the rough countryside, and the ferocity and heavy armaments of some of
the gangs involved. UNMIK made a major effort to collect the weapons that had
been spread out during the conflict. An amnesty brought in 524 rifles, 116 pistols,
30 anti-tank weapons, and 25 missiles. In the month following, 414 rifles and 3
anti-tank weapons were again seized by the police, demonstrating that only a
small proportion had actually been surrendered.
In February 2003, the UN Secretary-General Kofi Annan was still concerned
about the level of crime as well as the violence within the Kosovo Albanian com-
munity and towards ethnic Serbs. He recalled that the ethnic leaders in Kosovo
often challenged the authority of the UN’s Special Representative there and asked
for additional powers, but “they first need solid accomplishments for the ben-
efit of all communities in the areas for which they do have responsibility under
the Constitutional Framework.” Similarly, the Secretary-General’s Representative
insisted that local leaders must “take a stand against crime … Refrain from ex-
tremist statements and … call on the public to cooperate with the police and the
courts. Tacit tolerance for crime and corruption must stop.” Plainly, there is some
way to go before Kosovo meets the required standards (UNMIK 2003).
F Bosnia and Herzegovina

After the bitter war in Bosnia and Herzegovina between Serbs, Muslims, and
Croats, the negotiators of the 1995 Dayton Agreement set a very board frame-
work for dealing with the most urgent issues, such as military disengagement.
They set in motion a number of measures for re-establishing law and order, and
for building an effective police force. Under Annex 11 of the Dayton Accord, the
For a graphic account of the spread of bodies through the forests, see Lloyd
1999: 1 ff.
parties were supposed to provide “a safe and secure environment for all persons
in their respective jurisdictions, by maintaining civilian law enforcement agen-
cies operating in accordance with internationally recognized standards.” To assist
them, the UN was charged with training the police and monitoring their fairness
and professionalism. As part of the UN Mission to Bosnia and Herzegovina (UN-
MIBH), it established an International Police Task Force (IPTF) numbering more
than 1,500 for this purpose. Its members were to accompany local police forces
and ensure that they carried out their work fairly and responsibly. Resistance from
some of the ethnic groups was initially often fierce, as Carl Bildt, the European
High Representative, later recalled (Bildt 1998: 354). Moreover, at the beginning
of 2001, only 3% of the police came from ethnic minorities, compared with a tar-
get of 28%. Nevertheless, by December 2002, when the UN Mission in Bosnia and
Herzegovina (UNMIBH) was dissolved, it was considered that great progress had
been made. Kofi Annan said that it had been the most extensive police reform
and restructuring effort ever made by the UN and that Bosnia and Herzegovina
now had a police force “fit for Europe” (UNMIBH 2002c). One indication of this
was the growing number of refugees who dared to return to their homes: 67,000
in 2000, 92,000 in 2001, and even greater numbers in 2002. The number of regu-
lar police had by then been reduced from 40,000 during the war years to 18,000.
159 police officers had been sacked for wartime and other offences. 1,165 cadets
had been trained or were under training, including over 400 women. 13,000 cas-
es of alleged human rights abuses by police had been investigated. The UN had
also made a start with establishing non-political Commissioners to insulate them
from political influence, and by June 2000 had begun to deploy a multi-ethnic
State Border Service to combat smuggling and other crimes. This was estimated
to have increased state revenues by some 20%. The UN had brokered an agree-
ment between Bosnia and Herzegovina, Croatia, and Serbia-Montenegro to set
up a Committee of Ministers and a Regional Task Force (RTF) to combat illegal
migration, terrorism, and organized crime. Subsequently, the UN negotiated an
agreement with Hungary to join the RTF, further enhancing its efficiency. It also
encouraged the establishment of a National Bureau of Interpol in Sarajevo.
Assembling international groups of policemen, such as the IPTF, which served
in Bosnia and Herzegovina, is often more difficult than assembling military forc-
es for peacekeeping. Armed forces are usually only intermittently employed in
warfare and their governments may be pleased to second them to peacekeeping
forces to gain prestige, to give them further experience, particularly of working
with more professional forces, and to remove them from the national budget.
But police forces constantly have to deal with criminal activities and thus have
no ‘spare capacity’ in the same way. Moreover, because of the steep pyramid-
shaped hierarchy of military forces, there is a large pool of former officers who
have retired, not because of their incompetence but for family and other reasons.
They can be called upon by the OSCE or the UN for peacekeeping activities.
For information on policing in Bosnia and Herzegovina, see UNMIBH 2002a;

2002b.
568 Philip Towle
Because of the different career structure of the police, there is a far smaller pool
of competent and reliable retired police officers. On the other hand, the shortage
of police officers makes it essential that, for an organization the size of the IPTF,
recruits have to come from a larger group of countries than is usually the case
with peacekeeping forces. In October 2001, the more than 1,400 members of the
IPTF included officers from forty-three countries including Argentina, Austria,
Bangladesh, Bulgaria, Kenya, and Switzerland (UNMIBH 2002d).
G Moldova
The extent of external involvement in the settlements in Kosovo and Bosnia-
Herzegovina was unusual. Following the tension in Moldova between the cen-
tral authorities and the inhabitants of Gagauzia, the ensuing peace settlement
allowed no such involvement and concentrated mainly on the economic and cul-
tural grievances of the inhabitants of Gagauzia. By granting them a measure of
autonomy, they were enabled to pursue links with Turkey, Russia, Italy, and other
countries, but not to establish a semi-autonomous police force. Nevertheless,
the Law on the Special Legal Status of Gagauzia passed by the Moldovan parlia-
ment was careful to delineate the way the police forces were to be controlled. The
Chief of the Department of Internal Affairs in Gagauzia was to be appointed by
the Moldovan Minister of Internal Affairs on the recommendation of the elected
Governor of Gagauzia with the agreement of the People’s Assembly. The Chief
was to appoint the Commissars of the police commissariat, but the chief of the
municipal police was to be appointed by the Governor of Gagauzia following a
proposal by the Chief of the Department of Internal Affairs of Gagauzia. Such an
apparently cumbersome and balanced scheme was, no doubt, intended to abate
criticisms and fears of the power of the Moldovan state.
III Conclusions
Thus, there is a spectrum of ways in which complex power-sharing arrangements
deal with police forces after ethnic conflicts and no pattern can be prescribed for
all cases, not least because arrangements acceptable in one country could exac-
erbate the violence in others. In some cases, such as Moldova, the negotiators
have felt that carefully-tailored constitutional arrangements may be sufficient to
provide the minority communities with confidence in the arrangements. The Pat-
ten Commission wanted to transform the Royal Ulster Constabulary in Northern
Ireland by re-naming it, changing the way it was controlled, and encouraging
the recruitment of Catholics. Although its proposals were not all accepted by
the British government, those implemented represented a substantial change
that has become more obvious as the peace process advances. In Macedonia, the
peace settlement increased the proportion of ethnic Albanians in the police and
particularly encouraged such recruits to carry out their duties in areas where Al-
banians predominated. More ambitiously, the settlement with the MNLF in the
Phillipines specifically encouraged former insurgents to join the police and army.
In Bougainville, funding shortages and the sensitivities of local communities

about seeing a formal system of policing introduced and prisons and law courts
re-established have led to heavy reliance on councils of elders and community
policing and courts. In Kosovo, a new force was established that, unlike its prede-
cessor, was representative of the ethnic makeup of the region. The province con-
tinues to suffer from high levels of crime, but progress has been made with the
establishment of a local force. In many of these cases, international institutions,
such as the UN or the OSCE, laid down standards and monitored the various
forces to ensure that the agreed-upon standards were upheld. In Bougainville,
neighbouring countries, including Australia and New Zealand, have had a major
role in providing financial assistance.
Each complex power-sharing agreement has to be tailored to the situation, but
all show the importance of establishing police forces that are or can grow to be
trusted by the various communities after ethnic and religious violence. Each case
may have some lesson(s) to offer future cases of complex power-sharing arrange-
ments. In the last decade, UN peacekeeping operations have been intensively
studied in armed forces, research institutes, and university departments to avoid
previous mistakes – but nothing like as much effort has been expended on the
ways in which police forces have been reformed or re-created in divided states
during the same period, yet it is the policing that may in fact be more important
to long-term peace and stability. These policing activities demonstrate the role
that international institutions can play in monitoring police actions and encour-
aging fair behaviour towards the different ethnic communities. The education
and training of police forces are clearly vital, and the international community
has played a major role in the education of many of the newly-established or
reformed police forces. It will be years before it becomes clear how far the new
or reformed police forces have lived up to the international community’s hopes
and the expectations and wishes of the local populations. Much will depend on
the national leaders’ insistence that the police treat all communities fairly, protect
them against attack, and investigate their complaints against their neighbours or
the police themselves. A respected and efficient police force is the basis of a set-
tled, ordered community: only by the establishment of such a force can progress
be made on breaking the vicious circle of distrust and violence between ethnic
communities.
Have we addressed the critical importance of policing in complex power-shar-
ing? In many ways, it is one of the more important and visible day-to-day inter-
faces of such arrangements with the population(s) concerned. It is thus a critical
issue and, while it may seem largely unregarded, it is actually a vital step in mak-
ing such power-sharing arrangements work. Is this a step that should ideally be
undertaken very early in the process as it can foster much-needed confidence in
such processes and the future? Or is it that because it is potentially so very impor-
tant (and often important symbolically also) that it is actually such a contentious
area that it may be a subject that the parties may find it hard to reach agreement
on how to participate?
Chapter 19
The Military Dimension of Security Sector
Governance in Complex Power-sharing Arrangements
Paul Cornish
I Introduction
Do complex power-sharing arrangements offer the means to address and re-
solve the underlying causes of self-determination conflicts, and might such ar-
rangements offer inclusive and durable futures, within which political and so-
cial stability can reasonably be expected? It is clear that the interest in complex
power-sharing has both practical and normative dimensions. In this respect, ex-
pectations of complex power-sharing are similar to those found in another area
of contemporary political science concerned with conflict resolution – security
sector governance.
Security sector governance (SSG) seeks to eliminate armed conflict by de-
veloping policies specifically focused on the reorganization, rehabilitation, and
governance of organized armed forces (whether formal or informal) and other
elements of the security sector, such as armed police and border guards. One
recent and authoritative study of peace operations and post-conflict reconstruc-
tion described effective reform of the security sector as the “sine qua non of post-
conflict reconstruction” (Dahrendorf 2003: 13). Lakhdar Brahimi, head of the UN
mission in Iraq makes a similar point: “When you speak about creating a national
army you are talking about the first development project … because without it
you are not going to have anything else that lasts” (Financial Times 2003). In its
more ambitious guises, however, security sector governance is also characteristi-
cally normative. More than merely addressing local causes of tension and dispute
and removing the means to fight, SSG sets out to improve civil-military relations
and, in short, to improve society by privileging means other than armed force in
the settlement of disputes and disagreements.
The purpose of this chapter is to test the compatibility of two sets of ideas:
complex power-sharing (CPS) and SSG. Both are central features of the debate
on post-conflict reconstruction and state (re)formation, and the literature in each
case is extensive. Orthodox academic and policy-oriented analysis generally ac-
cepts that post-conflict reconstruction programmes must be comprehensive, ad-
572 Paul Cornish
dressing both power-sharing and security sector issues, as well as issues concern-
ing social, judicial, and economic reform (CSIS 2002). Yet in the process, the
complex (and, occasionally, dysfunctional) relationship between power-sharing
and reform of the security sector has largely been overlooked. One notable ex-
ception is Davis and Pereira’s collection of essays on the role of irregular armed
forces in state formation, which argues trenchantly that where state formation
and post-conflict reconstruction are concerned, a broad definition of the secu-
rity sector is essential in order to comprehend the destructive and constructive
potential of irregular armed forces. Irregular armed forces, of course, are a par-
ticular challenge to power-sharing arrangements in that they are a forceful, if not
violent, expression of the cultural, ethnic, and civil society differences that must
be bridged. At the same time, irregular armed forces might also be expected to
be most resistant to security sector governance initiatives, which might be seen
as a threat to their power and influence (Davis and Pereira 2003). More typical of
the literature is Rees’ study of security sector reform in post-conflict transitional
administrations, which describes the challenge as “a unique and complex under-
taking”, but merely hints at the possibility that many of the difficulties confront-
ing the United Nations Transitional Administration in East Timor were caused
by the inability to co-ordinate security sector reform and power-sharing activities
(Rees 2002).
To ask whether security sector governance initiatives can function and be
effective in the context of complex power-sharing arrangements is more than
merely an academic inquiry. The scope and utility of both CPS and SSG must, to a
considerable degree, be reflected in the vitality of the relationship between them.
From the outset of any attempt to settle a self-determination dispute through
power-sharing, it would be reasonable to suppose that military (both regular and
irregular), paramilitary, and police agencies must in some way be drawn into the
process. But there are practical and conceptual difficulties here. How can power-
sharing arrangements manage such agencies when they might, typically, be more
concerned with power maximization and physical and economic dominance? By
what argument and by what means can advocates of CPS bring military and po-
lice forces into the process, to ensure that a reformed system of governance based
upon power-sharing is matched by a reformed security sector that is designed to
be compatible with power-sharing? Conversely, security sector governance ini-
tiatives are characteristically Weberian and centralist, in that they emphasize the
indivisibility of key state security and coercive functions (police, military, border
guards, customs, etc.). In that case, how can such initiatives manage the deep
cultural and ethnic differences that might be expected in post-conflict situations,
particularly where regular and irregular armed forces remain significant actors,
other than through reliance on a carefully crafted power-sharing arrangement,
such as consociation between different ethnic groups?
Following a discussion of security sector governance, this chapter draws gen-
eral and specific lessons from eight case studies in complex power-sharing: Bou-
gainville; Bosnia and Herzegovina; Gagauzia; Kosovo; Macedonia; Mindanao;
Northern Ireland; and South Ossetia.
19 The Military Dimension of Security Sector Governance in Complex Power-sharing Arrangements 573
II Security Sector Governance

Interest in the politico-military character and stability of economically undevel-
oped and/or democratically deficient states has, since the 1950s at least, become
something of a preoccupation for policy-makers and political sociologists alike
(Huntington 1957; Finer 1988). Since the end of the Cold War, in the context of the
progressive blurring of the boundaries between so-called ‘hard’ and ‘soft’ secu-
rity, much of this interest has been re-energized by the ‘security sector’ policy and
research agenda. While the decades-long pedigree of much of this thinking ought
generally to be acknowledged, it is nevertheless clear that the 1990s brought two
new dynamics to the fore. The first was a shift within the development assis-
tance community from interest in (reluctant or otherwise), to engagement with
the security sector. Clare Short, formerly UK Secretary of State for International
Development, noted how: “Development organizations have in the past tended
to shy away from the issue of security sector reform” (Short 1999: 2).
The second dynamic was the associated realization that the security sector is
a complex, multi-level phenomenon, management of which requires a complex,
inter-disciplinary, and inter-departmental response. The concepts that frame dis-
cussion of the security sector are sharply contested, and the new technical lan-
guage that has been developed is too often imprecise or overlapping in meaning:
‘security sector reform’; ‘security sector transformation’; ‘security sector recon-
struction’; ‘defence transformation’; and ‘security sector governance’. This chapter
uses the latter expression, on the basis that it appears in the literature to incorpo-
rate and to have superseded the alternatives (Dahrendorf 2003: 36).
But what is the ‘security sector’, when does it become a matter of interest, and
what is meant by its ‘governance?’ Definitions vary greatly, within and between
the research and policy sectors. Cooper and Pugh argue that the cardinal feature
of the post-Cold War security sector debate is the wide (or perhaps widening)
recognition that the security sector constitutes “something broader than solely
the military” (Cooper and Pugh 2002: 5). Adopting this approach (itself drawn
from the work of security sector analyst Nicole Ball (Ball 2001)), the security sec-
tor can be said to incorporate the following actors and organizations:
• Security forces (military, paramilitary, police);
• Agencies of government and parliament responsible for oversight of these
forces;
• Informal security forces;
• The judiciary and correction system;
• Private security firms; and
• Civil society (Cooper and Pugh 2002: 5).
Others, however, have been more sceptical that a consensus definition has
emerged. Hendrickson and Karkoszka argue that “there is not a universally ac-
ceptable definition of the security sector” (Hendrickson and Karkoszka 2002: 5),
choosing to describe the security sector more functionally:
574 Paul Cornish
• Forces authorized to use force: armed forces; police; paramilitary forces; gen-
darmeries; intelligence services (including both military and civilian agen-
cies); secret services; coast guards; border guards; customs authorities; and
reserve and local security units (civil defence forces, national guards, presi-
dential guards, militias, etc.).
• Security management and oversight bodies: presidential and prime ministe-
rial offices; national security advisory bodies; legislature and legislative se-
lect committees; ministries of defence, internal affairs, foreign affairs; cus-
tomary and traditional authorities; financial management bodies (finance
ministries, budget offices, financial audit and planning units); and civil so-
ciety organizations (civilian review boards, public complaints commissions,
etc.).
• Justice and law enforcement institutions: judiciary; justice ministries; pris-
ons; criminal investigation and prosecution services; human rights commis-
sions and ombudsmen; correctional services; and customary and traditional
justice systems.
The security sector largely comprises these three categories. But there is also a
fourth category – non-statutory security forces – which can have a “significant
influence of economic and political governance” (Hendrickson and Karkoszka
2002: 5), such as guerrilla armies and political party militias.
Perhaps the most comprehensive definition, incorporating key agencies as well
as key functions, is that provided by the UK Department for International Devel-
opment (DFID). DFID has a strong reputation internationally for exploring the
policy dimensions of the security/development overlap, and has produced policy
accordingly. DFID’s comprehensive definition of the security sector is as follows:
Core security actors:

Armed forces; police; paramilitary forces; gendarmeries; presidential guards; intelli-
gence and security services (both military and civilian); coast guards; border guards;
customs authorities; reserve of local security units (civil defence forces, national guards,
militias).
Security management and oversight bodies:

The Executive; national security advisory bodies; legislature and legislative select com-
mittees; ministries of defence, internal affairs, foreign affairs; customary and traditional
authorities; financial management bodies (finance ministries, budget offices, financial
audit & planning units); and civil society organizations (civilian review boards and pub-
lic complaints commissions).
Justice and law enforcement institutions:

Judiciary; justice ministries; prisons; criminal investigation and prosecution services;
human rights commissions and ombudsmen; customary and traditional justice sys-
tems.
Non-statutory security forces:
Liberation armies; guerrilla armies; private body-guard units; private security compa-
nies; political party militias.
(DFID 2002: 7).
Some differences of opinion remain over the relative merits of a ‘narrow’ defini-
tion of the security sector (i.e., one which focuses largely on military actors and
issues), as opposed to the ‘broad’ definitions of the sort outlined above (Hen-
drickson and Karkoszka 2002: 4; Costa and Medeiros 2002: 26, 28, 29, 41). There
is, however, widespread consensus among security sector analysts that broad
definitions are both more useful and more accurate. While accepting that prefer-
ence, this chapter is nevertheless focused more narrowly on the reform of the
military dimension in complex power-sharing arrangements.
Having established what comprises the security sector, the next step is to ask
when it is appropriate to seek to modify a given security sector: before, during,
or after a conflict. The notion of security sector modification as prophylaxis (i.e.,
initiatives that take place before deterioration into tension and conflict – armed
or otherwise) is practically and conceptually incoherent; the crisis that might ra-
tionalize the modification of the security sector has not yet taken place, and as a
result the necessary processes and political will may not be available. Empirically
at least, a better case could be made for security sector modification during a con-
flict. But it seems unlikely that elements of the security sector, while engaged in
what they see to be their primary function (and perhaps even seeking their own
survival), would accept or even be interested in ideas that would be perceived to
involve the redistribution, reduction, or removal of the authority and armed ca-
pacity of security sector actors. There is some evidence of disarmament, demobi-
lization, and reintegration (DDR) activities that have taken place during conflict,
but this seems sufficiently counter-intuitive to be regarded as the exception that
proves the rule (Hagman and Neilsen 2003: 3).
As the cycle of conflict progresses, and eventually tends towards stabilization,
reconciliation, and reconstruction, so the opportunity for intervention in, and
modification of, the security sector becomes more apparent, simply because it
is not until this point that the environment has eased sufficiently to allow the
concentration of all those concerned to focus beyond the immediate or very
short-term crisis they have been confronting. But this is also the point at which
the scale of the challenge of intervening in the security sector becomes apparent
– any intervention will involve several agencies and disciplines, engaging with a
variety of security sector actors, at various levels. In other words, it can be said
that whenever the long-awaited moment for security sector intervention arrives,
the demand is instant and total, and the response must be complex and multi-
level.
For a useful account of the DDR experience in Angola in the early 1990s, see Anstee
2003: 476-478, 496. For DDR in Bosnia and Herzegovina, see Heinemann-Grüder
and Pietz 2003.
576 Paul Cornish
Security sector governance can be said to incorporate three types of activity:

stabilization, oversight, and, finally, development of the security sector. Stabiliza-
tion is characteristically a practical activity, and is clearly the necessary precondi-
tion for oversight and development. But it is important not to regard security sec-
tor governance as a linear activity that begins with a set of easily comprehensible
practical measures and gradually becomes more progressive and norm-driven as
the market can bear. They might not take place simultaneously, but practical and
normative aspects of security sector governance are all essential aspects of the
overall process.
A Security Sector Stabilization

The working assumptions of security sector stabilization are that the target se-
curity sector is sufficiently discrete and can be reoriented into a more benign
form, rather than be replaced in its entirety. The simple goal of security sector
stabilization is to make the established security sector more efficient, stable, and
predictable, improving the prospects for longer-term projects – economic, de-
velopmental, military, industrial, commercial, diplomatic. Stabilization activities
are characteristically practical, and often low-level, addressing primarily military
structures (assumed to be the best-trained, and most effective, users of violence),
but including paramilitary forces and police forces if necessary and appropriate.
Typically, attention will be given to DDR activities, to levels and standards of
training, and to equipment procurement and standardization. Stability of anoth-
er sort might also be pursued, in terms of improving the efficiency and oversight
of security sector budgeting (including off-budget military expenditure).
For all their significance in the broader project of security sector governance,
stabilization activities are short-term, and deal mainly with symptoms. If mishan-
dled, activities of this sort can be counter-productive and damaging to long-term
goals. For the recipient, a tightly-focused, practical approach to security sector
change can appear to be a thinly disguised attempt to impose western or An-
glo-Saxon models and procedures (Chuter 2000: 14), rather than the preliminary
phase of a longer-term, more enlightened and constructive engagement. Security
sector stabilization must be perceived to be a means to an end (rather than the
construction of a new military elite enjoying a privileged relationship with the
donor) and must be expected to bring about change, the result of which will be its
subordination to civilian authorities. The pledge of usd 10 million by the US De-
partment of Defense in 2000 to ‘re-professionalize’ the Nigerian army prompted
the following comment:
The programme, run by the private security company Military Professional Resources
Incorporated (MPRI), includes initiatives appropriate to the broader conception of re-
form, such as developing a civilian oversight structure. But it also has the stamp of
an old-fashioned military aid package in its provision of military equipment. At the
On the subject of off-budget military expenditure, see Hendrickson and Ball 2002.
extreme, a concern with the importance of security as a precondition for development

(and thus the need for reform of the security forces to achieve this) can offer an oppor-
tunity for those who wish to reintroduce Cold War conceptions of the broader mod-
ernizing potential of the armed forces. This approach needs to be guarded against, as
focusing exclusively on the role of military forces as agents of, but not targets for, mod-
ernization risks elevating them to precisely the kind of pre-eminent social and political
role that security-sector reform aims to prevent .
(Cooper and Pugh 2002: 20).
B Security Sector Oversight

Security sector oversight initiatives introduce a more normative dimension.
Broader in scope than the stabilization activities, oversight programmes are
characteristically politico-military, and begin to address the norms, standards,
and prospects for civil-military relations in the target security sector. As well as
continuing with DDR activities, oversight programmes may include the profes-
sionalization of armed forces, as seen in the US support to the African Centre
for Security Studies, and in the UK advice and assistance to the governments of
Rwanda and Bosnia and Herzegovina in the creation of military/security sec-
tor staff colleges. Oversight activities have been most advanced in Central and
Eastern Europe, involving the development of democratic institutions and the
insistence on democratic control of armed forces as part of a wider process of
political, societal, and economic transition (or ‘westernization’ into the NATO
and EU fold). The essence of this approach has been summarized as:
Redefining the respective roles of local security organizations and institutions in post-
conflict states, and training them to be professional, transparent and accountable insti-
tutions that are responsive to the needs of the society and its citizens
(International IDEA 2002: 3).
Once again, however, oversight projects can create adverse impressions if mis-
handled. Attempts to professionalize local armed forces, for example, might be
perceived to be attempts to impose culturally inappropriate ‘Sandhurst’ (UK) or
‘West Point’ (US) models of military professionalization and civil-military rela-
tions. A more serious objection might be that oversight activities run the risk of
raising expectations, through the use of expansive, ambitious, and normative lan-
guage, which are not or cannot be met at the practical level. DDR programmes,
for example, if not carefully grounded, might result in demobilized but deeply
disaffected former soldiers unable to obtain gainful, non-security-related em-
ployment.
C Security Sector Development

Security sector development is the most ambitious and explicitly normative
set of activities, as well as being the area of policy where the goals of social and
578 Paul Cornish
economic developmentalists are most evident. It is at this point in the broader

programme of security sector governance that connections are made between
the target security sector and “wider issues, including the political economy of
conflicts and the ways in which the economic and security policies of developed
states can perpetuate local war economies” (Cooper and Pugh 2002: 57). Simply
put, security sector development connotes a shift in thinking and practice from
the injunction to ‘do no harm’ (through stabilization activities, DDR, retraining
etc.), to ‘do good’, by encouraging a stabilized and supervised security sector to
become a net contributor to broader human and societal development. The goal
of security sector development is to achieve structural change and improvements
that will benefit the state in the long-term. Typical security sector development
activities include the following:
• Civilian control over the military;
• International humanitarian law and human rights training for members of
the military;
• Parliamentary oversight of the security sector, and accountability and trans-
parency;
• Promotion of civil society oversight of the security sector; and
• Support for demobilization and reintegration of ex-combatants in post-con-
flict situations (Short 1999: 5).
An important aspect of security sector development is that it should see the initia-
tive in security sector governance shifting from external to indigenous agencies.
A good example of transition was the Mandela government’s stress on security
sector change after the 1994 election in South Africa, where: “The process ben-
efited from the provision of external funding and advisors, but it was primarily
generated from within the country” (Smith 2001: 10). As responsibility is assumed
locally, the demand for external intervention and direct donor involvement be-
gins to diminish, but other forms of constructive international engagement will
be encouraged. As it becomes embedded locally, security sector development can
also lead to the development of a regional or even global consciousness, in the
form of “a multilevel and multifunctional strategy that addresses not just the do-
mestic environment of the state but gives equal weight to change at the regional
or global level” (Cooper and Pugh 2002: 23).
As security sector governance approaches its zenith, and expectations rise, the
costs of failure can also be at their highest. From the perspective of an intervening
or donor government, the costs in financial and political terms could still seem
very high, while the rewards and indicators of success appear increasingly mar-
ginal, or based on assessments so long-term and conditional as to be outside the
usual political cycle. Locally, as the reform agenda reaches its most mature and
complex state, the demand for trained and reliable officials and specialists can far
exceed supply. The experience in Iraq in 2003, and in so many other cases, shows
that when regimes or states disintegrate for whatever reason, key institutions, as
well as the physical and cultural infrastructure, can all cease to function. Recon-
struction in such circumstances involves not only durable conflict resolution and
post-conflict settlement – daunting enough tasks in their own right – but also the
rebuilding of state institutions and social networks. The task of ‘cleansing’ from
society all elements of a tainted former regime can simply be counter-productive,
even if feasible. In a comment strongly resonant of the ironies of post-conflict
Iraq in 2003, International IDEA note that “the training of security officials in
a post-conflict society all too often means transforming former oppressors into
guarantors of human rights and security for individual citizens” (International
IDEA 2002: 5).
Having summarized the goals and modalities of security sector governance,
focusing particularly on activities and initiatives concerning military actors, the
next step is to extract key lessons and insights from the Resolving Self-determina-
tion Disputes Using Complex Power-sharing project’s eight case studies.
III Case Studies
A Bosnia and Herzegovina

When conflict began in April 1992, it became heavily militarized almost immedi-
ately. External actors were also quickly brought into the conflict, as the main mili-
tary actors appealed to external sponsor states. The military actors also quickly
took on a very high degree of political authority. For example, Serb forces con-
trolled some 70% of Bosnian territory within months. During the 1993–1994 war,
the Croatian HDZ sought to establish Herceg-Bosna in western Herzegovina as
part of Croatia. It was also the actions of one military organization in particular
that brought the international community to act: the Srebrenica atrocity by the
Bosnian VRS, backed by the Yugoslav army.
Thereafter, however, the various models for stabilization failed to take full and
proper account of the central role of the militaries. The February–March 1992
Cutileiro peace plan marginalized the militaries as agents of change by insisting
on a solution based on borders and territory, and did not even address the issues
of control of the police and the army. The 1993 Vance-Owen peace plan was based
on the newly established lines of control, and in that regard effectively accepted
the results of ethnic cleansing projects carried out by the militaries. So, rath-
er than marginalize the militaries, the Vance-Owen plan arguably encouraged
them to continue with their atrocities, as it was this behaviour that clearly set the
agenda – witness the increased activity by Croat forces and the Croat/Bosniak
advances in summer 1995.
The active militaries, having been the de facto key agents for change in the
region, then met with complex marginalization in the form of the Dayton Peace
Accords (DPA) of 1995. In the first place, the DPA were not based on the defeat
on any of the three parties – decision had not been reached through force of
arms. But the DPA also sought to avoid settling grievances between the militaries
through politics and diplomacy; the militaries seemed to be regarded as untouch-
For details of the case studies refer to Part 2.

580 Paul Cornish
able. Under the new constitutional arrangements for Bosnia and Herzegovina,
issues relating to defence were decentralized to the level where cross-communal
dialogue and negotiation was not necessary. In other words, such was the per-
ceived strength and implacability of the local military forces, that they were left as
entity, rather than federal matters. Unlike other federal arrangements, no central
or joint competency in military matters was sought or established. The federal
authority not only declined to establish its rights in the area of military force
(often understood as the key attribute of sovereignty), it permitted, and possibly
even encouraged, rivals in this field. In part, this can be explained by the financial
and fiscal weaknesses at the centre. The federal government would not have been
able to afford to run reorganized and integrated armed forces even if it had been
given the opportunity to do so. In the resulting vacuum, both entity armed forces
maintained close relations with their ‘mother states’; it was not until March 2002
that the Yugoslav army ended financial aid to the army of Republika Srpska.
Reluctance to engage with local military forces was compounded when the
federal authority ‘borrowed’ its military capability from outside. Under Annex
I of the DPA, foreign forces (NATO’s Implementation Force (IFOR) and subse-
quently the Stabilization Force (SFOR)), were given the task of supervising the
disengagement of military forces in Bosnia, by establishing a ceasefire along the
Inter-Entity Boundary Line (IEBL), seeking the withdrawal of neighbouring and
foreign forces, and ‘stabilizing’ the region through their physical presence and
through close supervision of local military forces. The main focus of IFOR and
SFOR activities was to secure the status quo rather than look forward and modify
or improve the political landscape. These forces also had an abiding reluctance
to support or participate in civilian aspects of the DPA, choosing to limit their
activities to the contents of the DPA military mandate.
The DPA was a weak agreement with little real authority vested at the centre.
Local military forces, having been closely involved in every aspect of the crisis
and conflict in the region and having become heavily politicized, were neither
silenced by defeat nor discharged honourably by an armistice. Instead, they were
pushed to the periphery of the new federal constitutional arrangements, and
finally were substituted by foreign military forces. To compound the problem,
these external forces self-consciously saw themselves not as agents of change (as
local military forces had been), but of stability, and avoided the process of politics
as far as possible. As a result of all this, the case study notes that joint security
initiatives such as the creation of a joint army command or a joint border serv-
ice “were met with fierce resistance by the dominant political forces of Croats
and Serbs,” and “the parallel power structures of the three nationalist parties re-
mained largely undisturbed” (see Ch 7: 240). It is not only that an ambitious se-
curity sector governance agenda was not pursued from the outset in Bosnia and
Herzegovina, but that it was carefully and deliberately avoided. To date, the focus
of the DPA has been to prevent future conflict, through intensive and expensive
initiatives to stabilize the security sector. But stabilization has been seen as an
end in itself, and has been pursued at the expense of establishing durable security
arrangements firmly embedded locally. That said, there are some indications that
the Ashdown administration is developing a longer-term strategy, with initiatives

to centralize customs and intelligence services and to improve integration of the
military. Nevertheless, for the short and medium-term, a large proportion of the
Bosnian population fear a new war when the international military presence is
finally withdrawn. Even if war did not break out immediately, the fragile fed-
eral political system would certainly face great difficulty without its international
sponsors, and might be unable to manage considerable population movements
that may take place. At that point, local military forces who not only have a large
stake in the political development of the region, but who have also shown them-
selves especially interested in the movement of populations, might well take up
arms once again, on their own terms.
B Bougainville
The Bougainville self-determination dispute has been long-running and complex,
with issues of ethnic identity compounded by competition for resources (copper
and gold). Protagonists in the dispute and conflict have been the government of
Papua New Guinea (PNG) and its associates in Bougainville, notably the Bou-
gainville Resistance Force (BRF). PNG consistently rejected Bougainville’s claims
for independence, was especially adamant that national sovereignty required se-
curity and police forces to be controlled centrally, and could not contemplate
devolving authority in these matters to Bougainville as autonomous functions.
Opposed were the Bougainville Revolutionary Army (BRA) and its associated
civilian administration, the Bougainville Interim Government (BIG). With its un-
compromising demands for secession, the BRA began as an informal militia but
steadily expanded its recruiting base to include older and more educated people,
together with some serving and former army officers. In terms of manpower and
military capability, the armed forces on both sides were not high level. By the
mid-1990s, about 800 PNG Defence Force (PNGDF) troops and about 150 police
riot squad members were deployed in Bougainville. At that time, BRA had about
2,000 members, armed with about 500 modern weapons and 2,000–3,000 vin-
tage or homemade weapons. The BRF, numbering about 1,500, were armed by the
PNGDF. At various points, there has been strong international involvement in
the settlement of the conflict, including the UN, the Commonwealth Secretariat,
various NGOs, the South Pacific Peace Keeping Force, and even the private Sand-
line security organization.
On neither side, however, were armed forces particularly well-organized. Both
the PNGDF and the Royal PNG Constabulary (RPNGC) had considerable au-
tonomy, often working in their own interests and against official policy. For its
part, BRA was not a well-organized and coherent force but a loose collection of
many small, armed groups with local loyalties. It was BRA’s loose organization
that made it unable to exploit the surprise withdrawal of all PNG security forces
in March 1990 by taking control of Bougainville. On both sides, the looseness
of organization and/or loyalties also meant that progress towards a negotiated
settlement became difficult. Lack of clear and unequivocal direction probably
582 Paul Cornish
contributed to human rights abuses committed by both sides, practices which

can only have reinforced ethnic loyalties and deepened hostility.
The Bougainville Peace Agreement was signed at Arawa on 30 August 2001.
The agreement established two lists of powers and functions, one for the PNG
national government and another for an eventual autonomous Bougainville gov-
ernment. The national government list included those powers and functions that,
“consistent with national sovereignty,” the national government would hold “in
relation to [PNG] as a whole, including Bougainville.” The first item on the na-
tional government list was defence. Elsewhere, the Arawa agreement made clear
that the national government would exercise its defence responsibility in Bou-
gainville, albeit following consultation, and that the PNGDF would be considered
“a national organization which recruits throughout [PNG].” The agreement also
included an elaborate, three-stage Weapons Disposal Plan, which welcomed “the
way in which the [BRA] and the [BRF] have come together.” The plan went be-
yond weapons disposal to include DDR activities, and involved not only the com-
batant elements but also local chiefs and elders, and the UN observer mission as
arbiter. Former combatants were instructed not to attempt to re-arm, while the
national government committed itself not to redeploy defence or police units in
demobilized areas. The Plan made an explicit connection between weapons dis-
posal and reconciliation, with the parties undertaking “to cooperate in promoting
reconciliation among ex-combatants and in the wider community, and restora-
tion of civil authority in Bougainville.” The commanders of the various groups
agreed to disband their organizations on completion of the Plan.
The Bougainville peace process is still underway, and important aspects of the
self-determination dispute are yet to be resolved. Nevertheless, in terms both of
language and process, the Peace Agreement exemplifies the pursuit of an ambi-
tious security sector governance agenda (stabilization, oversight, and develop-
ment) in the context of a power-sharing arrangement. The Peace Agreement
places combatant factions at the centre of the process and achieves the notable
feat not only of disarming the combatant groups (mainly BRA and BRF) but also
of achieving their agreement to the centralization of military power by the PNG
national government.
C Gagauzia
The recent origins of the Gagauzia dispute lie in the 19 August 1990 independence
declaration of the Gagauz Soviet Socialist Republic. The new entity aspired to be
independent of Moldova but part of the Soviet Union. Gagauzian administra-
tive institutions were duly created and the Moldovan Supreme Soviet responded
on 26 October by declaring a state of emergency in southern Moldova, and by
moving 40,000 armed volunteers into the area. Gagauzians appealed to Moscow
for military assistance, and local political movements begin to set up paramili-
tary units. Assistance was offered by Transdniestria, itself in a bitter and violent
dispute with Moldova. A very major conflict looked to be in prospect, but So-
viet authorities persuaded the volunteer forces to withdraw from confrontation.
Armed clashes did take place but they were sporadic and did not escalate, unlike
in Transdniestria. Following the end of the Transdniestria conflict, negotiations
began on political autonomy for Gagauzia in the autumn of 1992.
On 23 December 1994, a power-sharing arrangement was agreed by the
Moldovan parliament, establishing an autonomous region of Gagauzia. The Law
on the Special Legal Status of Gagauzia (Gagauz Yeri) stressed the independent
and unitary status of the Republic of Moldova and confirmed the precedence of
the Moldovan Constitution. Within those parameters, however, Gagauzia was
accorded a high level of autonomy, with the Gagauzian People’s Assembly au-
thorized to legislate on almost all aspects of public policy, other than defence and
foreign affairs.
The Gagauzian/Moldovan security sector does not figure prominently in the
power-sharing arrangement. Article 12, listing the competences of the Gagauzian
People’s Assembly, establishes a rudimentary relationship with the parliament of
the Republic of Moldova in the matter of declaring and administering a state of
emergency on the territory of Gagauzia (Article 12(3)(h)). Article 17 commits the
Executive Committee of Gagauzia (the region’s permanent executive authority,
appointed by the People’s Assembly) to ensuring the “protection of legality and
public order.” The most sophisticated security sector arrangements are found in
Articles 23 and 24. Article 23 establishes that the chief of Gagauzia’s Department
of National Security will be appointed (and dismissed) by the Moldovan Minister
of National Security, on the recommendation of the Governor of Gagauzia and
with the agreement of the Gagauzian People’s Assembly. Article 23 also provides
for the Gagauzian national security chief to be an ex officio member of Moldova’s
Board of the Ministry of National Security. Article 24 makes similar provision for
the appointment of the chief of Gagauzia’s Department of Internal Affairs. The
commander of ‘interior forces’ (‘Carabineer troops’) based in Gagauzia is, simi-
larly, to be appointed and dismissed by the Moldovan Minister of Internal Affairs
at the recommendation of the Governor of Gagauzia.
The case study (Järve) acknowledges major shortcomings in the Gagauzia/
Moldova power-sharing arrangement. The relationship has been uneasy and has
at times threatened to degenerate into large-scale conflict. There have, howev-
er, been some more auspicious moments, such as the political confrontation in
February–March 2002, which saw threats and some violence, but no casualties.
Violence is a largely latent phenomenon, but it is nevertheless curious that the
governance of the Gagauzian/Moldovan security sector has received so little ex-
plicit attention. Three explanations might be offered. First, it would appear that
the mismanagement of, or abuse by, the security sector was not a key component
of the dispute, the resolution of which could therefore be sought by constitu-
tional, economic, and ethno-national devices. Little attention was paid to the role
of organized military forces, simply because they had not been significant actors
in the dispute, which was more a complex constitutional matter than a struggle
for territory or resources. Second, at key moments in the dispute, the external in-
volvement of the Council of Europe, the CSCE/OSCE, the Soviets/Russians, and
others has been crucial. External involvement has not, however, been military,
584 Paul Cornish
with the exception of Soviet Interior Ministry troops in 1990. Most participation
in the conflict, whether internal or external actors, has been civil rather than mili-
tary, and this might have contributed to a process in which the security sector was
neither part of the problem, nor part of the solution. Finally, the existence of the
parallel conflict in Transdniestria must be borne in mind. That conflict was a seri-
ous shock to the fledgling Moldovan government, and seems to have persuaded
Moldova to seek peaceful resolution of the Gagauzia dispute if possible. Gagauz-
ians were certainly inspired by the Trandniestrian secession, but they were also
influenced by Moldova’s response and consequently positioned themselves care-
fully. It was clear that the weak Moldovan government was severely stretched by
the Transdniestrian conflict. The Gagauzians benefited by, effectively, playing a
military hand by proxy, and by doing so ensured that it was largely unnecessary
for the power-sharing agreement to address security sector governance.
D Kosovo
The Kosovo case is complex and far-reaching in terms of its origins, the number
and variety of actors, the intensity and scope of the violence, the close relation-
ship between political and military actors, and the consideration given to the
principles of security sector governance in the settlement of the dispute. The case
study describes the Kosovo dispute as a shift from a struggle for identity and
power within an autonomous province of a wider federal structure to a “genu-
ine self-determination conflict aiming towards secession” (see Ch 8: 244). Dur-
ing 1998 and 1999, parties to the conflict included paramilitary/terrorist groups
(principally the Kosovo Liberation Army (KLA)), Serb volunteers, police and
army units from the Federal Republic of Yugoslavia, and the armed forces of
NATO and other intervening countries. As elsewhere in the Balkans, the Kosovo
crisis was also distinguished by the massive, deliberate, and brutal involvement
of the civilian population in the conflict. When the decisive round of hostilities
broke out in 1999, Kosovo was subjected to rapid and large-scale ethnic cleans-
ing, with over half of its 1.7 million ethnic Albanians being moved out of the
territory within weeks, and with many thousands of others becoming internally
displaced. By June 1999, following intervention by NATO, ethnic cleansing had
been stopped and Kosovo was temporarily under the control of the KLA. Repri-
sal attacks then began against Serbs and Roma, in spite of the presence of NATO
and, eventually, UN civil administrators. The outcome is a complex but uneasy
peace, where power has not been so much shared as held in trust by the UN:
“Kosovo has become, in fact, a protectorate, which is not dissimilar from the
condition of Class A mandates of the League of Nations, which were sovereign
but not independent” (Groom and Taylor 2000: 303).
The Kosovo conflict has been marked by a very high degree of external inter-
est and involvement, ranging from neighbouring states (and some more distant,
particularly the United States) to major civil and military international organiza-
tions. The CSCE/OSCE has had a long-standing interest in the Kosovo dispute,
which it saw not only as challenging the organization’s advocacy of the human
dimension of security and stability, but also as a potential source of instability and
even open conflict on a regional scale. The OSCE assumed an unusually active
role in the form of the 2,000-strong Kosovo Verification Mission (KVM), estab-
lished in October 1998 to monitor compliance with UN Security Council Resolu-
tion (UNSCR) 1199. NATO also became a prominent actor in the dispute during
1998. Activity increased under the auspices of NATO’s Partnership for Peace, in-
volving Albania and Macedonia. Both countries consented to NATO air exercises
in their airspace in June, and to NATO ground force exercises in August and
September. By the end of the year, NATO had positioned itself as the guarantor of
the mediation and conflict resolution efforts of the UN Security Council and the
OSCE. NATO’s 13 October ultimatum to Belgrade, in the form of an activation
order for limited air strikes and a phased air campaign, was unprecedented in the
alliance’s history. The deployment of the KVM also necessitated the organization
of a 1,500-strong ‘extraction force’ deployed to Macedonia. The United States,
which had been making military threats against Serbia over Kosovo since 1992,
viewed NATO increasingly as a useful device for the politico-military coercion
of Belgrade.
Serb military and police activity over the Christmas period, followed by the Ra-
cak massacre in January 1999, brought the October 1998 Holbrooke Agreement
to the point of collapse. The Agreement, between Serbian President Slobodan
Milosevic and US Special Envoy Richard Holbrooke acting on behalf of the Con-
tact Group, was a complex arrangement by which Milosevic would cease military
operations in Kosovo and withdraw those forces that had moved into Kosovo
during the preceding seven months. The Kosovo Verification Mission – a team of
unarmed, civilian monitors from the OSCE – would be deployed to monitor Serb
compliance. The Holbrooke Agreement offered short-term stability and even, in
the longer term, the means by which the Serb government could accept settle-
ment of the Kosovo issue. But the Agreement was overtaken by events; the first
six months of 1999 saw armed intervention by NATO and other forces, as well as
the development of arrangements for the administration of Kosovo. As far as the
security sector and its governance are concerned, negotiations and agreements
made during the first half of 1999 provide an example of an attempt to agree to
an ambitious and sophisticated security sector governance agenda (in the form
of the Rambouillet Accords), revisited in a slightly less sophisticated form in the
Belgrade/NATO Military Technical Agreement and UNSCR 1244 of June 1999.
Chapter 7 of the Rambouillet Accords is the most detailed part of the provi-
sional agreement. The Accords allowed for a wide range of security sector sta-
bilization and oversight measures. A ceasefire was to come into force immedi-
ately upon signature of the Accords, and any violations of it would, importantly,
be reported to international observers and would not be admitted as justifying
a retaliatory use of force. An amnesty would be offered to former combatants,
other than those accused of violating international humanitarian law. Regular
police and security forces would be withdrawn and redeployed according to a
detailed programme. Air defence systems in Kosovo would be dismantled, and
heavy weapons would be placed under international supervision. All forces
586 Paul Cornish
would be expected to commit themselves to demilitarization programs and to

the renunciation of violence. Paramilitary and irregular forces in Kosovo, on the
other hand, were simply to be excluded as personae non grata, incompatible with
the settlement process and the resolution of the conflict. This would have become
something of a sticking point for the KLA, which considered itself neither a para-
military nor a terrorist force.
Under the Rambouillet Accords, the parties would have agreed to the deploy-
ment by NATO of a Kosovo Force (KFOR) of about 28,000 troops, with Rus-
sian and other contingents, under the political control of NATO’s North Atlantic
Council. A UN Security Council Chapter VII enforcement mandate would be
sought to legitimize the deployment. KFOR would be authorized to use military
force as necessary to ensure compliance with the Accords and, unlike the Dayton
Peace Accords in Bosnia and Herzegovina, would be available to support civil-
ian implementation by the OSCE and others. Although the KFOR military com-
mander would have final authority in the interpretation of military aspects of the
Accords, that authority would, if only to a limited extent, be supervised by a Joint
Implementation Commission (JIC). The purpose of the JIC would have been to
consider complaints and grievances and would comprise Yugoslav military com-
manders, Yugoslav and Kosovo Serb officials, representatives of Kosovo, and rep-
resentatives of the military and civilian implementation missions.
However, Chapter 7 of the Rambouillet Accords also had an annex containing a
proposed status of forces agreement for deployed NATO forces. Belgrade claimed
that this document represented an abuse of Yugoslav sovereignty by NATO and,
on that basis, the Rambouillet Accords and the follow-on Paris negotiations were
rejected.
The situation in Kosovo worsened rapidly in the early months of 1999, even
as the Rambouillet talks were taking place. The OSCE abandoned its monitor-
ing mission on 20 March and air strikes began against Yugoslavia within days.
The intense air campaign (Operation Allied Force) lasted until early June, when
Milosevic accepted a G-8 plan for the withdrawal of all Serb forces from Kosovo.
A Military Technical Agreement (MTA) was signed on 9 June between NATO’s
operational commander in the area and Serb generals providing for the with-
drawal of Serb forces and the entry into Kosovo of NATO-led ground forces (to
be known as KFOR, as proposed at Rambouillet). The following day, UNSCR
1244 was adopted for the establishment of the administration of Kosovo.
The MTA established a wide range of security sector stabilization measures.
KFOR was authorized to use military force as necessary to ensure compliance
with the MTA and “to contribute to a secure environment for the international
civil implementation presence, and other international organizations, agencies
and non-governmental organizations” (Art I: 4.b). Air safety and ground safety
zones beyond Kosovo’s borders were agreed, from which Yugoslav forces would
be excluded as well as from Kosovo itself. Under the threat of further NATO
air strikes, Yugoslav forces – broadly defined to include regular forces, paramili-
taries, armed civilians, border police, intelligence services, riot police, etc. – were
committed to an immediate ceasefire and to a detailed, eleven-day withdrawal
plan, removing mines and other obstacles as they left. KFOR would have assumed
control of Kosovo’s borders and airspace. The MTA also allowed for a JIC, and
would have made the KFOR commander the final authority for interpretation of
the MTA “and the security aspects of the peace settlement it supports” (Art V).
UNSCR 1244 of 10 June 1999 began by attempting a rather more balanced ap-
proach to the Kosovo conflict and its perpetrators, by condemning “all acts of
violence against the Kosovo population as well as all terrorist acts by any party.”
The resolution called for Yugoslavia to end violence and repression in Kosovo and
to withdraw its military, paramilitary, and police forces (Art 3). International civil
and security ‘presences’ were to be deployed to Kosovo under the auspices of the
UN. The international civil presence was to be headed by a Special Representa-
tive of the Secretary-General, and was instructed to “co-ordinate closely” with
the “security presence” (i.e., KFOR) (Art 6). Article 9 of the resolution lists the
tasks of the security presence, including: maintain the ceasefire and ensure the
withdrawal of forces from Kosovo; establish a secure environment within which
the ‘civil presence’ can operate and to enable the return of refugees and displaced
people; to ensure public safety and order “until the international civil presence
can take responsibility for this task”; and to support and co-ordinate closely with
the work of the international civil presence. Articles 9 and 15 also explicitly call
for the demilitarization of the KLA and “other armed Kosovo Albanian groups.”
The Kosovo dispute required a complex settlement that continues to be fragile
and to require constant maintenance. Whatever the strengths and weaknesses
of the settlement, from the narrow perspective of security sector governance,
and the even narrower concern with the role of the military in such initiatives,
the Kosovo settlement is noteworthy in a number of ways. First, and perhaps of
most significance in the longer term, the KLA had been a very significant com-
batant organization and had temporarily administered much of Kosovo before
the full deployment of KFOR. But the KLA subsequently found itself treated as
part of the Kosovo ‘problem’, rather than an element of the solution – the UN in
particular had long regarded the KLA as a terrorist organization and little more.
This is not to offer a defence of the KLA, simply to observe that organizations
such as this do not often go into oblivion voluntarily. The settlement provided
for the replacement of the KLA by a new, lightly armed, multiethnic civil policing
organization, the Kosovo Protection Corps (KPC). This arrangement, however,
convinced few, least of all former KLA members, who continue to see the KPC
as the kernel of an army of Kosovo, which will be expanded when KFOR departs.
The Serb population of Kosovo take a similar view of the KPC, and for that reason
are reported to have made arrangements to revive civil volunteer guard forces in
the event of a post-KFOR crisis. Small arms and light weapons are in abundant
supply in Kosovo; a resource that can easily be exploited by both sides when nec-
essary. Second, the Kosovo settlement is built largely on a denial of that which the
population of Kosovo most desires: independence from the Federal Republic of
Yugoslavia (FRY). Economic and geostrategic arguments can be advanced against
the separation of Kosovo from FRY, but it remains the case that “substantial au-
tonomy within the FRY” (UNSCR 1244 Art 10) was precisely not what many in
588 Paul Cornish
Kosovo wanted, and was not something they would, in the words of the UNSCR,
“enjoy.” Third, the Kosovo settlement sees the United Nations acting as a form
of trust or mandate power – critics would say ‘colonial’ – which self-evidently
denies the inhabitants of Kosovo control over their own institutions (at least for
the present) and does not require them to face the difficult challenges of finding
the motivation and the means to resolve their own disputes. Finally, the heavily-
involved international community has indulged in its own form of power-shar-
ing, with the ‘security’ and the ‘civil’ presences cooperating with each other but
not bound closely in an overarching constitutional arrangement of some sort.
What this amounts to is the curious spectacle of western powers, with their very
developed models of civil-military relations, acting in Kosovo as if the civil and
military could and should be separated ideologically and functionally. With the
notable exception of the reform of the police service (see Ch 18), therefore, at-
tempts to indigenize the governance of Kosovo’s security sector have been largely
unproductive. And where Kosovo’s various military bodies and groups are con-
cerned, the reform and governance agenda has in effect been avoided altogether.
E Macedonia
The long-acclaimed success of Macedonian stability and conflict prevention was
marred by the outbreak of fighting between Macedonian security forces and the
ethnic Albanian National Liberation Army (NLA) in 2001. Most ethnic Albanians
in Macedonia wanted the state of Macedonia to remain united and were not anx-
ious to press a claim for self-determination, yet were sympathetic to the NLA
for championing the ethnic Albanian cause within Macedonia. The Macedonian
government, however, viewed things very differently. For the government, the
dispute was a territorial struggle, with the NLA fighting for the absorption of eth-
nic Albanian regions of Macedonia into ‘Greater Albania’. Accordingly, the Mace-
donian government responded militarily to the NLA as a terrorist or insurgent
organization, not as an armed body representing an aggrieved ethnic community.
The NLA agreed a ceasefire but the Macedonian government then refused to
discuss the rights of ethnic Albanians and began a full-scale military offensive in
late March 2001. The small and ill-equipped armed forces of Macedonia faced a
difficult struggle against the NLA, and certain practices only served to alienate
ethnic Albanian populations still further. As the conflict continued into the sum-
mer of 2001, both communities became progressively more radicalized.
International organizations were closely involved in the crisis, not least be-
cause the Macedonian government blamed authorities in Kosovo for lax border
controls that allowed NLA fighters to cross into Macedonia. The international
perspective was that the Macedonian government had reacted inappropriately
and had adopted a confrontational position from which it could not hope to re-
solve the ethnic problems that lay at the heart of the dispute. After several days
For the argument that post-conflict transitional administrations should be unified

(civil-military), see Einsiedel 2002: 7.
of extremely violent fighting, a power-sharing Framework Agreement was signed

in Skopje on 13 August 2001 by the main ethnic Macedonian and ethnic Alba-
nian parties, with the encouragement of the United States, EU, NATO, and the
OSCE.
The Framework Agreement offers very little with which to inform analysis of
the role of security sector governance in a complex power-sharing arrangement.
Although the NLA had been kept informed by NATO it had not been directly
involved in the negotiation of the Framework Agreement. As a result, the Frame-
work Agreement allows for no formal role for the NLA, insisting starkly that
“Macedonia’s sovereignty and territorial integrity, and the unitary character of
the State are inviolable and must be preserved. There are no territorial solutions
to ethnic issues” (para 1.2). The contribution expected of the NLA to the agree-
ment and to the post-conflict reconstruction process is, quite simply, to elimi-
nate itself as an actor by agreeing to “complete cessation of hostilities,” “complete
voluntary disarmament,” and “complete voluntary disbandment” (para 2.1). The
Framework Agreement makes what might be a vague acknowledgement of DDR
when it refers to “rehabilitation and reconstruction in areas affected by the fight-
ing” (para 8.3) but otherwise the only concession to security sector governance
lies in the commitment to increase the representation of minority communities
in the military and other public services (para 5.1). Minority representation in
the armed forces was a serious issue – the development of parallel ethnic cul-
tures in Macedonia meant that very few ethnic Albanians had been recruited
into the Macedonian military and police forces – but its inclusion in the Frame-
work Agreement is the barest evidence of a security sector governance agenda at
work.
On 21 August 2001, the NLA indicated their consent to the agreement and
agreed to surrender their weapons to NATO. In Operation Essential Harvest (27
August–26 September) NATO collected and destroyed some 3,875 NLA weap-
ons. NLA involvement went no further. Former NLA members were not formally
integrated into the Macedonian army, for example. Some left for Kosovo, others
continued a sporadic struggle. In effect, the NLA had brought an ethnic griev-
ance to the surface but was not closely involved in the resolution of that griev-
ance, and so announced its own disbandment on 27 September 2001.
The case study (Daftary and Friedman) concludes that the power-sharing ar-
rangement in Macedonia is neither a success nor a failure, but incomplete. As to
the military dimension of security sector governance in this work in progress,
beyond practical stabilization measures and weapons removal, the Framework
Agreement provides little evidence that the principles of security sector govern-
ance had been given much thought. Unusually, one of the protagonists in the
dispute, the NLA, was substantially excluded from the negotiation and imple-
mentation of the Framework Agreement. The NLA leadership accepted this fate,
but many former fighters did not and formed the Albanian National Army, pledg-
ing to reject the Framework Agreement and to renew the struggle for a ‘Greater
Albania’. This sense of an unresolved conflict is reinforced by the knowledge that
ethnic Macedonian hardliners and paramilitaries are also ‘keeping their powder
590 Paul Cornish
dry.’ In other words, in Macedonia there is a fight yet to be finished, according to

some. The Macedonian settlement also featured intense international involve-
ment, in both the negotiation and the implementation of the Framework Agree-
ment. Whatever the motives and merits of this involvement, in some respects it
exacerbated rather than resolved the crisis. Many ethnic Macedonians were and
are convinced that the international community instinctively favours the ethnic
Albanians and accuse NATO of seeking the creation of an Albanian protectorate
in Macedonia. For these critics, NATO’s weapon collection programme was in-
sincere, gathering only a tiny proportion of an NLA weapon stockpile that could
contain as many as 60,000 weapons. As in other cases, however, the main prob-
lem with relying on external political influence and military power in the Mac-
edonian settlement is that when international interest and involvement come to
an end, unresolved local grievances may re-emerge with renewed vigour.
F Mindanao
Conflict in and over Mindanao is centuries-old and exceptionally complex, with
confrontation occurring between Muslims and Christians, Philippine govern-
ment troops and Muslims, private armies and either Christians or Muslims, and
private armies and government troops. The peace agreement between the Philip-
pine government and the Moro National Liberation Front (MNLF) was signed
in September 1996. The first step of the agreement was the establishment of a
Special Zone of Peace and Development (SZOPAD). The agreement also con-
tained extensive DDR arrangements. Under the first phase of the peace agree-
ment, some 1,750 MNLF members would be integrated with the Philippine Na-
tional Police (PNP) and a further 5,750 with the Armed Forces of the Philippines
(AFP). Other MNLF members would be absorbed into ‘socio-economic, cultural,
and educational’ programmes, and would be considered a priority when recruit-
ing for SZOPAD development projects. In addition, a Special Regional Security
Force would be established as a regional command of the PNP in order to em-
ploy former MNLF fighters. The broader relationship between the Philippines
national government and that of the new Autonomous Region of Muslim Min-
danao (ARMM) is also set out in the agreement. National defence and security
was retained as a national function and would not be devolved to the ARMM;
the defence and security of the ARMM would continue to be the responsibility
of the central government. With this in mind, an AFP regional command has
been established in the ARMM, while remaining under the control of the AFP
and central government. The ARMM regional governor – a member ex officio
of the Philippines National Security Council – may request deployment of the
AFP regional command in the event of invasion, rebellion, or general lawlessness.
Tellingly, however, central government is not empowered to decide unilaterally to
send the AFP into the ARMM.
The successful implementation of the peace agreement was undermined by
the reluctance and/or inability of the Philippine government to deal effectively
with organizations other than the MNLF, notably the Moro Islamic Liberation
Front (MILF) (which had earlier broken away from the MNLF) and the extremist
Abu Sayyaf terrorist organization. The MILF had withdrawn from negotiations
at an early stage when they perceived that the central government was princi-
pally interested in negotiating with the MNLF only, and the MILF had therefore
not endorsed the peace agreement. After the peace agreement, the MILF con-
tinued to consolidate its military strength and remained a concern for the AFP,
which culminated in war between the AFP and MILF that broke out once again in
Mindanao in early 2000. A peace agreement with MILF is yet to be negotiated.
Economic and social development were regarded as the great hope for stabil-
ity in Mindanao. After the 1996 agreement, it was discovered that over 60% of
MNLF personnel were unemployed, with their communities in a wretched eco-
nomic condition. A massive development assistance programme was geared to
assisting former MNLF fighters, their families, and communities, in the process
ensuring that social and economic suffering would not trigger further conflict.
However, by focusing closely on the needs of former MNLF fighters, other fac-
tions with a legitimate claim to assistance felt marginalized, with the result that
the aid programme threatened to prevent, rather than encourage reconciliation.
Overall, the benefits of the development programmes have yet to be seen by
the majority of Mindanao’s population, and the much-needed reconciliation be-
tween the opposing factions has yet to take hold. Disaffection is widespread, to
the extent that the Mindanao crisis cannot yet be said to be resolved. As the case
study notes, “Underlying conditions which give rise to conflict still persist and are
unlikely to diminish significantly in the short-term:” (Ch 6: 191). As far as security
sector governance is concerned, the Mindanao case demonstrates the need for a
close relationship between stability and economic and social development. But
the Mindanao case also proves the dangers of excluding a key actor (the MILF)
from the process, and of moving too quickly to the development phase without
first having stabilized the situation by involving all armed actors, and without
having established effective and consensual oversight programmes.
G Northern Ireland
The recent conflict in and over Northern Ireland stems from the outbreak of inter-
community violence in the late 1960s – known euphemistically as the ‘Troubles’
– and the subsequent large-scale involvement of Britain’s armed forces in polic-
ing and counter-insurgency operations. The epicentre of the conflict was North-
ern Ireland itself, although violent acts were also carried out on the UK mainland
and against British military bases in Europe. Occasionally, loyalist paramilitary
forces made incursions into Ireland.
The British government always perceived and portrayed the conflict as an in-
ternal security matter, and was resistant to external involvement (other than from
the government of Ireland). From 1972 until 1985, the British government pre-
sented itself as a neutral arbiter between rival groups of Northern Irish Catholics
and Protestants, justifying the British Army’s involvement as an armed polic-
ing accompaniment to a neutral intervention. Unlike many other self-determi-
592 Paul Cornish
nation disputes around the world, Britain’s armed forces were generally closely
restrained throughout the course of the conflict. There were, nevertheless, some
abuses of authority and military strength. Furthermore, certain security policies,
such as internment without trial and Britain’s perceived support for loyalist inter-
ests and armed factions, were counter-productive and fuelled support for the re-
publican paramilitary groups such as the Irish Republican Army (IRA). For this, if
for no other reason, Britain cannot be said to have been objectively neutral in the
conflict. Being closely involved in the conflict, with decisive negotiating powers
in the resolution of the conflict, and with a practical role in the reconstruction
phase (in terms of demilitarization), Britain’s position at the centre of every as-
pect of the Northern Ireland conflict is unique, shared by neither the government
of Ireland, the political parties of Northern Ireland, nor the paramilitary organi-
zations (republican and loyalist).
On 10 April 1998, settlement of the Northern Ireland conflict was reached in the
form of an agreement known respectively as the ‘Belfast’, the ‘Good Friday’, or the
‘British-Irish’ Agreement. Five years later, full implementation of the Agreement
is still awaited. Nevertheless, important indications of the role of security sector
governance in the Agreement can be discerned. Paramilitary organizations were
represented in the negotiations by political parties: the IRA by Sinn Féin, and the
Ulster Defence Association and Ulster Volunteer Force by the Progressive Un-
ionist Party and Ulster Democratic Party, respectively. This was a significant act
of security sector stabilization through bargaining; the case study notes that the
governments of Ireland and Britain effectively “recognized the paramilitaries that
have organized ceasefires as political agencies” (Ch 4: 100). Just as important,
those paramilitary organizations abiding by the Agreement were given a special
status and were not required to formally surrender themselves to what would
have been perceived as their former opponents.
The Agreement was, in essence, a political framework that sought to provide
opportunities to seek peaceful resolution of disputes and avoid armed conflict. At
the strategic level, the Agreement proposed that the internal security functions
of the state – policing and the courts – would be retained in Westminster by the
British government, subject to review. The Agreement made special provision for
the reconstruction of the police force. As the case study observes: “Democratic
consociation cannot exist where those of military age in one community are al-
most the sole recruitment pool for policing all of those in another community”
(Ch 4: 92). As far as the military side of the security sector was concerned, no
military or paramilitary organizations were formally involved in negotiating the
Agreement. In several respects, however, the Agreement was explicitly military;
in terms of demilitarization activities (i.e., removal of watchtowers and border
posts, reduction of troop numbers), release of paramilitary and military prison-
ers, and weapons decommissioning.
Weapons decommissioning was at the heart of the Agreement. Of all the secu-
rity sector reconstruction and governance initiatives contained in the Agreement,
only weapons decommissioning was a necessary precondition for the construc-
tion of a new political culture and institutions in Northern Ireland. In the latter
years of the conflict, the issue of decommissioning saw a number of different

approaches being advocated. In the mid-1990s, the British government insisted
on prior decommissioning before paramilitary organizations or their representa-
tives could join the negotiations. The Mitchell Commission, however, argued for
decommissioning to take place in parallel with negotiations. From May 1997, the
approach of the new Labour government in Britain was to negotiate a renewed
ceasefire by the IRA in exchange for placing the decommissioning issue into a
separate negotiating forum. The Labour government’s approach was met with the
vehement opposition of many unionists, for whom decommissioning was a vital
feature of any agreement and its implementation and who regarded the British
government’s approach as the appeasement of terrorists.
The SSG components of the British-Irish Agreement are not especially promi-
nent (in the military sphere at least; the police sphere is much more developed
and comprehensively treated), but at the same time they reveal an extraordinarily
sophisticated approach to the role of the security sector in the settlement of dis-
putes. In military terms, the Agreement reduces largely to a discussion of DDR
matters, and further still to a protracted debate over just one initiative among
many actual and possible initiatives – weapons decommissioning. The govern-
ance of the respective state armed forces (British and Irish) is not addressed, al-
though specific cases of abuse and misuse of authority are to be pursued. The
British Army, having been closely involved in every aspect of the conflict since
its inception in the late 1960s, is largely confined to a set of demilitarization ac-
tivities. In other words, the British Army’s principal contribution to the settle-
ment is to withdraw. The Agreement calls for a variety of confidence-building
measures, some from governments and others from paramilitary organizations.
The principal contribution of the paramilitary bodies is agreement to, and main-
tenance of a ceasefire; the precondition for all security sector stabilization, over-
sight, and development initiatives. Generally, ceasefires have been maintained.
Where ceasefires have been broken and atrocities carried out, these have largely
been the responsibility of breakaway groups. The key point in this, however, is
that those paramilitary organizations that have maintained a ceasefire as a confi-
dence-building contribution to the negotiation and eventual implementation of
the Agreement have done so without having been directly involved in the nego-
tiations. Paramilitary organizations could only contribute indirectly to the nego-
tiations through a political party, and only then if they agreed to and maintained
a ceasefire. In other words, the Agreement is a political enabling framework
rather than an explicit attempt to reconstruct the security sector governance of
Northern Ireland as part of a complex, new, constitutional arrangement. Like all
such settlements, the British-Irish Agreement requires a degree of security sector
stabilization before it can proceed. The notable achievement of the Agreement,
however, is that rather than put the more normative and contentious aspects of
security sector governance to one side (oversight and development initiatives), it
imposes normative standards of SSG on the key participants as a precondition for
participation in the negotiations.
594 Paul Cornish
H South Ossetia
The dispute between Georgia and the people of the South Ossetia region begin
in November 1989 with armed clashes between South Ossetian militia (aided by
Russian volunteers) and local Georgians. Sporadic violence lasted until December
1990, when a somewhat larger-scale military confrontation began. The conflict
saw about 1,000 deaths and the internal displacement of about 5,000 Ossetians
and Georgians, and lasted until a ceasefire in July 1992. The conflict was a limited
affair throughout, even though local populations took part in street clashes and
mercenary forces from Russia and North Caucasus became involved. On neither
side were the armed forces particularly well organized, with few if any attempts to
achieve a decisive build-up of troops and equipment. Ossetian and Georgian par-
tisan movements either failed or declined to organize themselves and establish a
protected enclave from which to continue the struggle after the ceasefire.
The power-sharing arrangement between the Ossetians and Georgians was al-
ways more of a process than a single, formally negotiated document. In all this
informality there were, nevertheless, some indications of a systematic effort to
address security sector stabilization and oversight issues. The April 1996 Memo-
randum on Measures to Provide Security and Strengthen Mutual Trust Between
Sides in the Georgian-Ossetian Conflict begins by congratulating the parties on
the achievement of basic security sector stabilization, “stating with pleasure that
… military actions were terminated in the zone of conflict.” In Article 1, the par-
ties commit themselves to “refuse to use force or the threat of force, to exert po-
litical, economic or any other form of pressure on each other.” Article 4 offers an
amnesty to combatants, other than those suspected of military crimes or crimes
against civilians. Article 6 refers to the gradual demilitarization of the zone of
conflict, and the deployment of peacekeeping forces into the area. Peacekeeping
forces and activities will themselves be subject to an eventual “phased reduction”
and will generally be managed by a Joint Control Commission (JCC) (Art 7).
The relative ease with which a power-sharing process was established has sev-
eral explanations. First, there had been a long record of peaceful coexistence be-
tween Ossetians and Georgians. The lack of “deep-seated rancour” (Ch 11: 351)
between the two sides, contributed to the low scale and intensity of the con-
flict, reduced the chances of reprisal and escalation, and meant that recovery and
reconstruction projects could begin more quickly and with more effect than in
other, deeper conflicts. The absence of intense antagonism also made possible the
replacement, after the conflict, of both sides’ radical leaders with more moderate
figures.
The Soviet Union and then the Russian Federation had been involved all along,
even to the extent in the early months of the conflict of using airpower against
Georgian villages. As conflict turned into settlement negotiations, Russia’s main
contribution came to be that of a regional “custodian of stability .” (Ch 11: 359).
Russian troops deployed in 1990 later became the core of the Joint Peace-keep-
ing Force (JPKF). In 1992, Georgian and Ossetian elements were incorporated
into the JPKF. But the Russian contingent continued to dominate the JPKF in the
absence of sufficient volunteers and military equipment from Georgia and South
Ossetia. Russian military preponderance in the JPKF managed to create an envi-
ronment of stability within which a settlement could be negotiated, essentially by
making it strategically and diplomatically possible for the two major actors in the
dispute to continue to have a minor role in military terms. Russian involvement
was also central to the establishment of the Joint Control Commission (JCC) in
June 1992, involving Georgia, South Ossetia, Russia, and the CSCE/OSCE, and
comprising three working groups in military, economic, and repatriation affairs.
The final explanation for the relative ease of the settlement process was that
western governments were barely interested in the dispute from the beginning,
as it was not considered to be on the same scale as the crisis in the Balkans, and
which, in any case, took place in Russia’s ‘near abroad’. International civil organi-
zations proved vital, however, particularly the CSCE/OSCE, which monitored
the JPKF and provided support to the JCC and its three secretariats. Gradually,
more organizations became involved, including the EU, UNHCR, and UNDP, in-
troducing programmes such as weapons amnesties in exchange for community
development projects.
The South Ossetian settlement is unusual, particularly as far as the security
sector is concerned. A power-sharing process or arrangement was agreed infor-
mally in order to settle a crisis that had seen violence but that had not escalated
into full-scale war. Certain security sector issues were addressed in the settle-
ment process, but the process was not explicitly driven by the need to improve
the governance of the security sector. The conflict between the two main parties
had never been as irreversible as in many other crises, and the military prepon-
derance of the Russians further demilitarized the negotiation process. External
actors became involved but those that did were characteristically non-military,
and as such did not detract from the civil negotiation process. South Ossetia
demonstrates that the right mix of internal and external forces can produce an
agreed, civil settlement to a dispute (even where violence is involved) without
a formal agreement. But the South Ossetia case study also carries a warning;
informal agreement on the regulation of security, military, and economic issues
might be effective, but it is also unconstitutional and legally challengeable with-
out democratic endorsement. In the end, there can be no substitute for a complex
and formal settlement.
IV Conclusion
Kriesberg defines power-sharing as “a governance system that provides for pro-
portional representation of diverse groups in policy making and administrative
institutions, particularly in the police and military services” (Kriesberg 2003:
120). The principal conclusion of this chapter, however, is that this may be a par-
ticularly difficult prescription to fill. There exists a fundamental tension between
complex power-sharing arrangements, which seek to dilute centralized and ex-
clusive control over aspects of governance, and security sector governance initia-
tives, especially after violent conflict, which seek to reduce the proliferation of
596 Paul Cornish
control over security structures to diverse actors. In stable, conflict-free countries

generally, and especially in post-conflict societies that aspire to such stability, the
control of armed forces simply cannot be treated in the same way as, say, control
over transport infrastructure, education, or hospitals. It follows that in post-con-
flict situations, and depending on local and historical circumstances, certain ele-
ments of security sector governance must be incorporated into a power-sharing
arrangement if the centralization of organization and control of the armed forces
is to be achieved. Failure to incorporate normative security sector governance
into the power-sharing apparatus will ensure that those inclined to undermine or
spoil the power-sharing arrangement will have the means and thereby the incen-
tive to do so. Put another way, without the achievement of credible and durable
security sector governance, complex power-sharing can appear most vulnerable
to those with the greatest capacity to wreck.
The eight cases reviewed here – several of which remain ‘works in progress’
– all offer very different experiences and lessons, and taken together illustrate the
difficulty of incorporating complex power-sharing and security sector govern-
ance effectively. As far as the military dimension of security sector governance is
concerned, the case studies show enormous variation in the scale of the conflict,
the handling of military actors and issues, the level of external involvement, the
military component of any settlement agreement, and the quality and value of
DDR initiatives. Each case study offers a different insight into the role of secu-
rity sector governance in complex power-sharing arrangements. The Mindanao
study shows that SSG – however elaborate – cannot achieve its potential without
parallel economic and social development. In the Gagauzia and Macedonia cases,
SSG issues were the least relevant in the settlement. In South Ossetia, agree-
ment was the easiest to be reached, but was also the least formal (and perhaps,
therefore, least durable) settlement. Considering the military dimension of SSG,
Kosovo was the most ambitious project, but a disappointment in terms of the
failure to indigenize SSG. Bosnia and Herzegovina emerge as the least successful
and most volatile experiment in integrating complex power-sharing and security
sector governance; the widespread fear that armed conflict will break out again
following the departure of the international presence indicates a basic lack of
confidence and durability. Northern Ireland was perhaps the cleverest process
among those reviewed, in that normative standards of security sector governance
were imposed on key participants as preconditions for participation in negotia-
tions. Bougainville was the most successful overall; combatant factions were not
only disarmed and fully involved in negotiations, they also agreed to the centrali-
zation of military power by the national government of Papua New Guinea.
The relationship between complex power-sharing and security sector govern-
ance is intricate and often self-contradictory, but no less necessary. The task of
complex power-sharing is to build common structures without seeking to ho-
mogenize; differences in experience, ethnicity, culture, and civil society are all
to be acknowledged and put to use in the effort of state-building. To ignore the
security sector in that effort would be to overlook not only what is often the
clearest expression of difference but also potentially the most destructive. For
its part, the central insight of security sector governance is that military, police,
paramilitary, and other armed forces and services must be, not the target of post-
conflict reforms in civil society and governance, but an intimate and inescapable
component and even agent of that reform. These studies in complex power-shar-
ing support the contention that security sector governance should be approached
comprehensively, involving all armed services and forces (Dahrendorf 2003: 13),
and that security sector governance is a collection of activities (stabilization,
oversight, and development), not all of which will be appropriate in all circum-
stances at all times.
Complex power sharing can benefit from well-conceived security sector gov-
ernance initiatives in three ways. First, comprehensive security sector govern-
ance can prevent break-out from one element of the security sector that has been
incorporated in power-sharing arrangements (for example, the army) to another
that has not, or which is at a less advanced point in post-conflict reform (for
example, a paramilitary gendarmerie). Second, a comprehensive approach to the
security sector makes a long-term differentiation between different components
of the security sector more possible: “armies exercise power, but the police must
exercise authority.” (Costa and Medeiros 2002: 28). An undifferentiated security
sector, in which military, police, and gendarmerie forces are all publicly perceived
to be, and are in practical terms, interchangeable, is one in which the very dif-
ferent tasks of establishing civilian control over the military, on the one hand,
and the police, on the other, can only be made more difficult. Finally, insisting
upon a working relationship between complex power-sharing and security sec-
tor governance also offers an important device by which potentially destructive
armed forces (particularly the military) can be kept engaged in the process of
post-conflict stabilization and reconstruction. As reconstruction begins a stark
choice presents itself; the military (who were, after all, a large part of the prob-
lem) can be either excluded or involved. Simple exclusion, however, can often
lead back towards the vulnerable status quo ante or, at best, merely short-term
stability, while exhausted military forces rest and recuperate. Inclusion of the
military can be either passive (in the sense that the military holds the ring, pro-
viding a modicum of stability while governance and power-sharing initiatives are
devised and initiated), or active and normative, whereby military bodies ‘buy in’
to the settlement process and reinforce it. Active involvement of the military will
not easily be achieved. But, as the Northern Ireland and Bougainville experiences
demonstrate, it is not impossible.
Chapter 20
Transforming Justice, Reclaiming the Rule of Law:
Legal Transition in Complex Power-sharing Agreements
Angela Hegarty
I Introduction
Central to any political transformation are the values, institutions, and practices
of justice. They provide the legal means of guaranteeing the rights and equality
promised in power-sharing agreements and are, in an ideal form, the independent
bulwarks against repression and abuse of power that drive so many conflicts. But,
in conflicts throughout the world, the institutions of justice have failed this impor-
tant test, and thus if the transition out of conflict into a power-sharing democracy
is to be successfully made, those institutions must themselves be transformed.
They must move away from rule by majority towards the democratic rule of law.
II Definitions and Themes
A Defining ‘Administration of Justice’

The term ‘administration of justice’ covers a multitude of functions and institu-
tions, so arriving at a definition of the term is in itself challenging. For example,
the teaching of the ‘administration of justice’ usually involves a discussion of the
legal system, its purpose, and its institutions. Thus one includes a description of
the courts, the judiciary, and the laws that create them. But legal systems vary
depending upon tradition, culture, and legal models, so the best place to start
looking for a definition is the relevant constitution. Most constitutions create
institutions to administer justice or lay down parameters for their operation, and
most modern constitutions refer in some fashion to fundamental rights. Those
that do not are still bound – in theory anyway – by international norms and by
commonly held ideas about justice. In the UK, for example, where the constitu-
tion is ‘unwritten’ and un-entrenched, the incorporation of the European Con-
vention on Human Rights into domestic law has led to formidable change in the
judicial ethos and in the standards and practice of the administration of justice
(Starmer 2003: 23). Most systems of justice often tend to incorporate notions of
600 Angela Hegarty
democracy; for example, the practice of due process, which owe as much to sev-
eral centuries of political norm building as to any legal stricture. In some aspects,
therefore, the political culture dictates the legal one.
International practice and standards may also provide insights into what may
constitute the ‘administration of justice’. For example, the UN General Assembly
has adopted a number of resolutions on human rights in the administration of
justice (UN 7 December 1987; 11 December 2001) and has also set up, under the
Sub-Commission on the Promotion and Protection of Human Rights, a sessional
working group on the administration of justice. This group looks at issues such
as the use of the death penalty; prisons standards; summary, arbitrary, and extra
judicial executions (UN 2001); improvement and efficiency of national judicial
institutions for the protection of human rights; transitional justice and mecha-
nisms of truth and reconciliation; rules of evidence with regard to rape and sexual
assault; and current trends in international criminal justice (UN 2002).
Thus, any discussion of the administration of justice ought to include a consid-
eration of the legal system and its components – such as the courts, the judiciary,
and the criminal justice system – but should also encompass an exploration of
the laws and values that create and inform that system.
For the purposes of this chapter, the consideration of the ‘administration of
justice’ will be primarily focused upon the constitutions or constitutional frame-
works laid down in the case studies and at the institutions and reforms created
as a consequence. In this chapter, a survey is made of the various indications of
change in the administration of justice, such as the creation of a new constitu-
tion and a constitutional court and the institution of any other new courts or the
restructuring of the courts system. Other indicators include changes to courts,
the appointment of new judges, the creation of a judicial commission to consider
those appointments, as well as other new legal institutions.
B Themes
The themes that emerge from this diverse group of case studies are those pri-
marily associated with the role of justice in transition and the importance of the
‘bedding down’ of universal legal values, which are vital in the delivery of human
rights, stability, and equality. In a number of the case studies, the use of law by
the previous government or regime was a central aspect of the conflict, as the ap-
propriation of law in defence of the political status quo contributed to repression
and the undermining of the rule of law. The task, therefore, for the new arrange-
ments for the administration of justice is to restore those values undermined by
conflict and to become the guarantor against such repression that the preceding
arrangements were not.
20 Transforming Justice, Reclaiming the Rule of Law: Legal Transition in Complex Power-sharing Agreements 601
III A Survey of the Case Studies
A Some General Comments

In cases as diverse as those surveyed, it is advisable to be somewhat cautious
about identifying common trends or features. The cases cover an array of evolv-
ing political situations and cultures, and there are also differences as to the
level at which the peace agreements operate and the extent to which they have
been implemented. Nonetheless, all of the agreements provide in some way for
changes in some aspect of the administration of justice. Some of these changes
envisaged are extensive and some are very limited, but the fact that all of the
agreements contemplate change in the administration of justice demonstrates its
relevance to the political processes being negotiated. Because these agreements
are designed to address complex problems, they envisage complicated solutions.
This complexity is frequently reflected in the planning and implementation of
the new administration of justice arrangements. As some of the agreements only
sketch out in broad terms what those changes should be and a number are not yet
properly implemented, it is unwise to make final judgments about the meaning
and effect of those changes. The implementation of the agreements is still evolv-
ing and in their implementation they also display changes, as they are re-negoti-
ated and refined. For example, in Northern Ireland the details of the changes to
policing envisaged by the agreement were left to a Commission, whose detailed
report (Independent Commission on Policing in Northern Ireland, 1999) became
the subject of an ongoing political battle, the outcomes of which are still being
negotiated and disputed. In another example, Bougainville, a blueprint for con-
stitution-making – including the jurisdiction of the courts – was laid down in the
peace agreement, but the drafting was left to a constitutional convention.
In all of the agreements, the arrangements made or envisaged are complicated
by the necessary trade-offs that occur during political deal-making and by the in-
evitable checks and balances that are a feature of complex power-sharing agree-
ments. These tend to produce a proliferation of structures, which sometimes
overlap and may be unclear. For example, there remains some confusion over
the apparently coinciding jurisdiction of the Human Rights Chamber and other
higher courts in Bosnia (Council of Europe 2001). This exemplifies the problems
created by the reform and overlaying of previous systems with different legal re-
gimes and cultures. Further, “[t]he discontinuity of the territorial structure be-
queathed by the Dayton Peace Accords is compounded by Bosnia’s mixed legisla-
tive inheritance. The statute books contain a multitude of outdated, overlapping
and inconsistent laws from the pre-war, wartime and post-war periods. They are
applied (or not) by courts which are too numerous, too expensive, too inefficient,
and too vulnerable to political influence.” (International Crisis Group 2002b: 1).
Bougainville Peace Agreement 2001 Ch 3.

Bougainville Peace Agreement 2001 Ch 3(14-21).
602 Angela Hegarty
Furthermore, that these are self-determination conflicts, where the role and
sovereignty of law is both an aspect and a symptom of the conflict, means that the
nature of the changes to the legal system reflect the extent to which autonomy has
been achieved in the agreements. For example, the Bougainville peace agreement
allows for the development of indigenous Bougainvillean courts, but specifies
that these will always be ultimately subject to the national Papua New Guinea
courts. This is mirrored in the Mindanao agreement, which allows for the en-
hancement of existing Shari’a courts and the development of indigenous tribal
courts, but both are ultimately subject to the Philippine Supreme Court. In other
cases, the courts are more-or-less independent within the autonomous political
unit, e.g., Kosovo (CEELI 2002b: 2-3).
Thus, the complexity of the problems, the diversity of the parties involved, the
layering of the authorities, and the degrees of autonomy offered under the agree-
ments all create complex arrangements that are still evolving.
B Key Issues
1 Constitution Making
A key element of any new political settlement is constitution making, providing
both the legal stability required for the new arrangements and institutions to
take hold and an opportunity to articulate the vision of the form of the agreed
democratic arrangements. Constitutions are therefore both ‘mission statements’
and political blueprints. They are also often monuments to political compromise,
and in the translation from agreement to law new problems can emerge. Some
of the institutions created by this constitution making may be the ‘price of the
deal’ or may be safeguards for the political and legal arrangements created by the
agreement; as such their implementation may be resisted. Any changes in legal
sovereignty envisaged or laid down in the agreement may have to be given further
legal effect and that may well be done through new constitutional provisions, the
negotiation of which may create further difficulties. Insulating the new arrange-
ments against change may be another concern. It is sometimes the case that, hav-
ing arrived at an agreement, some parties seek to prevent its re-negotiation and
thus seek to entrench the new institutions and governing principles. However, a
number of the agreements permit further change in the constitutional status of
the region (e.g., Northern Ireland) or later full or limited secession by parts of
the region from the newly created autonomy (e.g., Gagauzia).
Some agreements are more detailed than others in all of these respects. In Bou-
gainville, a detailed blueprint for constitution-making was laid down in the agree-
Bougainville Peace Agreement 2001 Ch 8(128; Ch 13(276).

Bougainville Peace Agreement 2001 Ch 13(283); (285); (289-291).
RA 9054, 2001 Art VIII 19.
Northern Ireland Act 1998 Sec 1.
The Law On The Special Legal Status Of Gagauzia 1994 Art 1.
ment, and in Bosnia the agreement had an interim constitution attached to it.
In Kosovo, which also had an interim constitution appended to the agreement, a
further ‘constitutional framework’ was produced in 2001. In other places, such as
Northern Ireland (Harvey 2003: 1013-1014) and Macedonia,10 only limited consti-
tutional changes were contemplated, but the manner in which these changes took
place led to further negotiation difficulties. The failure to implement agreed consti-
tutional changes raises further problems, not least an erosion of confidence in the
deal brokered and a lessening of general confidence in the peace-making process.
An interesting issue is the interplay between the general principles or ethos
that the agreement embodies and the existing constitutional model that it seeks
to modify or replace. In Northern Ireland, for example, there was an inevitable
friction between the output of the constitutional negotiations (the Belfast Agree-
ment) and the pre-existing British constitutional model, which places a premium
on parliamentary sovereignty. This was further complicated by the manner cho-
sen to embed the agreement politically, which was by referendum throughout the
whole of Ireland. Referenda have constitutional legal status in the Irish Republic,
but not in the UK and thus not in Northern Ireland. This meant that transforma-
tion of the agreement into law took place through the UK parliament drafting and
enacting the Northern Ireland Act 1998, in a process that was fractious and conten-
tious and over which many arguments persist. (Mageean and O’Brien 1999: 1524-
1537) In some of the Eastern European cases, there are continuing problems with
the fulfillment of the agreements; for example, Gagauzia, where the Tribunal of
Gagauzia has not been set up, thus delaying and perhaps negating one of the legal
changes envisaged in the agreement. In Macedonia, a new constitution was en-
acted in November 2001, which contradicted the ‘ethnically neutral’ ethos agreed
in the August 2001 Framework Agreements (see Ch 9). This ‘ethnically neutral’
approach appears to have been an acknowledgement that one of the difficulties
with the pre-existing legal environment was its partisan ethos, a problem that runs
throughout many of the case studies.
Significant changes to existing constitutional rules or the creation of a wholly
new constitutional culture may affect how well the new arrangements take hold.
If the new values embodied in the agreements are significantly different to the
previous order, there may be difficulties in the ‘bedding down’ of the arrange-
ments. New values, which draw upon standard international norms of democ-
racy and justice, may frame the agreement, in the form of a preamble, or form
part of its structure – in the form of references to various international treaties,
such as the European Convention on Human Rights. This may create problems
domestically: if there is no local ownership of the new constitutional culture, the
new institutions that stem from it may founder. This may be a particularly acute
concern where the peace-making process is largely managed by international ac-
Bougainville Peace Agreement 2001 Ch 3(14-21).

Rambouillet Interim Agreement for Peace and Self-Government in Kosovo 1999
Ch 1.
10 Framework Agreement Annex A.
604 Angela Hegarty
tors. In some of the cases, the agreement provides a draft constitution or articu-
lates certain constitutional principles.11 In some cases, a wholesale re-invention of
the constitution was necessary, because of the nature and extent of the conflict,
e.g. Kosovo.12 In others, amendments were made to the pre-existing arrange-
ments to take into account the cultural differences that the agreement attempts
to resolve, e.g., Mindanao. In one or two cases, these changes were fairly minimal,
e.g., Northern Ireland (Harvey 2003: 1013-1014).
One way around some of these difficulties may be to invoke and to engage inter-
national standards in the making of new constitutional arrangements. A number
of the agreements expressly refer to international and regional treaties, such as
the UN Charter, the UN Covenants and the European Convention on Human
Rights (ECHR) – which is widely referred to in Bosnia,13 Kosovo,14 and Northern
Ireland,15 and given formal status. However, it seems that there was little recourse
to international legal standards when designing the institutions. The real prob-
lem has been the failure to employ the numerous relevant ‘soft law’ standards in
creating and developing the institutions. No doubt this is partly because they are
so numerous and therefore difficult to assimilate into one coherent set of bench-
marks, particularly by non-specialists who are negotiating agreements quickly
and under considerable pressure. This may be one of the reasons for the more
widespread use of the ‘rule of law’ rubric, as it provides a shorter, more coherent,
and comprehensible set of benchmarks. The ‘rule of law’ is a somewhat nebulous,
jurisprudential concept, which makes it harder to quantify and implement. Fur-
ther, it does not come replete with a ready-made enforcement mechanism and
case-law, unlike the international human rights standards.
2 New Legal Institutions

Most of the agreements envisage some new legal institutions in one form or an-
other. Even in places like Northern Ireland and Georgia, where the emphasis was
upon the reform of existing institutions, new legal entities were created or pro-
posed along with new laws to govern them.16 Legal institutions can range from
legislatures and formal courts to commissions, public inquiries, quangos, judicial
authorities, ombudsmen, etc. It is from the initial agreement that these institu-
11 Dayton Peace Agreement Annex 4; Bougainville Peace Agreement 2001 Ch 3 Art 21;
Northern Ireland Peace Agreement 1998 Ch 2, which specifies principles and pro-
vides draft clauses for the legislatures.
12 Rambouillet Interim Agreement for Peace and Self-Government in Kosovo Ch 1; A
Constitutional Framework for Provisional Self-Government in Kosovo’.
13 Dayton Peace Agreement Annex 6.
14 A Constitutional Framework for Provisional Self-Government in Kosovo’, Ch 3.2(b).
15 Good Friday Agreement (Northern Ireland Peace Agreement) 10 April 1998 Strand 1
para 26(a); ‘Rights Safeguards and Equality of Opportunity’, para 4.
16 For example, the Northern Ireland Human Rights Commission, Northern Ireland
Act 1998 Sec 68-70; Equality Commission for Northern Ireland, Northern Ireland
Act 1998 Sec 73-74.
tions flow – either they are built into the agreement itself or they are created un-
der it, as illustrated in Tables 1-2.
Table 1 Key Legal Developments
Measure New con- Const Other Reform of New Judicial

stitution? court? New old system judges ap- Commis-
Case courts? pointed? sion?
NI Partial No No limited No No
NI Act (see
In Re Rob-
inson)
Ko UN Yes Yes ? yes Yes (JKJP-
– in Ch1 of PC)
1999 Ac-
cord;
2001 Con-
stitutional
Framework
BiH Dayton + Yes Yes – in partial Some Yes IJC
Annex 4 Art VI progress – ongoing (2001)
Geo Amended. Yes, pre- Yes Yes + re- ?
exist certific
ABK has
form of
constitu-
tion
South Os-
setia mem-
orandum
1996
Mol Amended Yes Basic Law, Implied by Council
Gaugauzia Art 20.; Art 20 of Magis-
Basic Law Tribunal of trates
1998 Gaugauzia
Legal (not yet set Ministry of
status con- up). Justice
tested.
Mac 1991 Yes – pre- 1996 reyes Yes Rep. Judi-
amended exist structuring cial Com-
by 2001 mission
FW agree-
ments
606 Angela Hegarty
Boug 2001 Subject Permitted, Provided

Agreement to PNG but subject for
lays down supreme to PNG
principles court courts;
and mech- possibil-
anism for ity of ad-
develop- ditional
ment jurisdiction
Min 1996 Peace Remains New Of some
Agreement the Phil- Shari’ah existing
lippine Appellate Shari’ah
Supreme & other courts
court courts;
tribal
courts
Table 2 Other Justice Changes

Measure Ref to Int. Positive New ar- Past ad- Other
Standards? action mea- rangements dressed in
Case sures in jus- for criminal legal form?
tice system? justice?
NI Yes PSNI 50-50 Changes toPrisoner re- Human
recruitment policing &lease; Other- rights Com-
(time limited) Criminal wise limited mission
Justice (Corrie in-
vestigation)
Ko UN 1999 Implied? Policing; Refugees; Police Om-
– yes; Criminal Jus- prisoner re- budsman
2001 much tice (eg Com- lease; agree-
more specific missions) ment to be
bound by int.
humanitarian
law and ICTY
BiH Yes. Annex 2. Memo of Un- yes partially International
ECHR sover- derstanding Judges on
eign over all with specific Constitution-
other laws measures al Court
to increase
minorities in
judiciary; but
no veto rights
in courts
Geo Brief Promised in
ABK
brief
Mol
Mac Amendments Limited/im- policing Refugees; Public At-
include gen plied amnesty law torney role
ref. emlarged
Boug No. Yes policing, Provisions
criminal law for amnesty
etc. and weapons
disposal
Min Yes – for Policing and Integration of Shari’ah Pub-
Philippine arrest powers former com- lic Assistance
Supreme Ct. batants Office (Legal
Aid)
Note: The details in each category are illustrative, not exhaustive. Where a category is
left blank, there was insufficient information to complete it accurately. Where a ‘?’ is
entered, the information was unclear.
Few of the agreements specify the full details of the courts. Only in those agree-
ments that had constitutions appended and created constitutional courts are
there detailed descriptions of the new institutions, e.g., Bosnia. In the other cas-
es, this was left to subsequent legislation, enabled in many cases by the agree-
ment.17 The cases provided for a range of changes – from an almost completely
new courts system (Kosovo);18 a combination of new courts and partial reform of
existing courts (Bosnia,19 Macedonia); new specialized courts or legal tribunals
(Macedonia,20 Bougainville);21 to limited reforms (Moldova).22 In one case, that
of Northern Ireland, there was no reform of the courts structure at all, although
certain aspects of the ethos and procedure of courts are to be modified (Criminal
Justice System Review Group 2000: paras 8.41-8.65).
Most of these new or reformed institutions are permanent, although some
contain transitional measures – such as the presence of international judges on
the Bosnian Constitutional Court.23 Some legal institutions of limited life span
were created, in order to implement parts of the agreement, with the intention
17 For example, Bougainville Peace Agreement 2001 Ch 13, Art 276.

18 A Constitutional Framework for Provisional Self-Government in Kosovo Ch 9.4.3.
19 For example, Decision Enacting the Law re-amending the Law on Court of Bosnia
and Herzegovina, Office of the High Representative, 24 January 2003.
20 Framework Agreement Annex A Arts 104; 109.
21 Bougainville Peace Agreement 2001, Chs 12; 13.
22 Republic of Moldova, Parliamentary Resolution 152-XIII on Approving the Concept
of Judicial and Legal Reform in the Republic of Moldova (1994).
23 Dayton Peace Agreement Art VI(1).
608 Angela Hegarty
that these will be wound up once the implementation task is carried out (e.g., the
Decommissioning Body and the Sentence Review Body in Northern Ireland).24
Given the autonomous nature of some of the political entities created, an inter-
esting question is the nature of the relationship of the courts inside the area (‘lo-
cal courts’) governed by the agreement with national or transnational (‘external
courts’). For example, in Mindanao, administration of justice is a matter reserved
to the national government, but the creation of new Shari’a and tribal courts is
devolved to the Mindanao region.25 However, the Shari’a and tribal courts only
have jurisdiction over the Muslim and tribal communities and are subject to the
oversight of the Philippine Supreme Court.26 In Bougainville, the agreement al-
lows for the development of Bougainvillean courts27 but these too will be subject
to the jurisdiction of the Papua New Guinea ‘national courts’, although the agree-
ment allows for the limited development of additional jurisdiction.28
Because of the necessity of a robust and independent judiciary, the question
of how judges and law officers are appointed is a vital concern. In a number of
the cases, commissions or investigatory bodies have been set up to recommend
changes to the administration of justice (e.g., Kosovo).29 In other cases, there have
been attempts to increase the representativeness of the judiciary by introducing
positive action measures. A number of the agreements contain implied quotas,
specifying that courts shall be made up of a certain number of representatives
from one region and a certain number from another region, community or eth-
nicity (e.g., Macedonia,30 Mindanao).31 In Bosnia, a Memorandum of Understand-
ing was signed in 2001 in an attempt to specify particular measures aimed at in-
creasing the representation of minorities in the judiciary.32 In Northern Ireland,
however, despite long-standing concerns about the performance of the judges
and the identification of the judiciary with Unionism, little has been done to ad-
dress this issue, although a judicial appointments commission has been created
to advise upon the appointment of judges.33 One significant change has been the
removal of the requirement to make a declaration of loyalty to the Crown for
24 Northern Ireland Arms Decommissioning Act 1998; Northern Ireland (Sentences)

Act 1998.
25 Mindanao Peace Agreement 1996 Ch III(A).
26 RA 9054 Arts VII; VIII.
27 Bougainville Peace Agreement 2001 Ch 8(128); Ch 13(276).
28 Bougainville Peace Agreement 2001 Ch 13(283); (285); (289-291).
29 A Constitutional Framework for Provisional Self-Government in Kosovo Ch 11.1(b).
30 Framework Agreement of Ch 4; Annex A Art 109(2).
31 Mindanao Peace Agreement 1996 Ch III(A).
32 Memorandum of Understanding on Appointment Procedures Concerning the Judi-
ciary of Bosnia and Herzegovina between the Independent Judicial Commission and
the Ministries of Justice of Republika Srpska and the Federation, 4 July 2001.
33 Justice (Northern Ireland) Act 2002 Sec 2-8.
those becoming QCs (senior barristers),34 and that change was in part prompted
by a court case taken by two prominent Nationalist barristers (In Re Treacy &
McDonald 2000).
Reform and re-training of the judiciary is another important concern. In some
places, such as Kosovo, the legal infrastructure and the judiciary largely disap-
peared during the conflict, necessitating the creation of a new, indigenous one.
In other examples, there is continuing re-training and re-accreditation of judges,
prosecutors and other legal officers, e.g., in Georgia, where new judges are also
being appointed (Council of Europe 1999a: 20). In Bosnia, existing judges and
prosecutors have had to re-apply and undergo an evaluation (OHR 2002a). In
most of the cases, there is an emphasis on re-training and reform of the judiciary
and the legal process, with a number of large scale projects focused on this, e.g.,
Central and East European Law Initiative (CEELI). Only in Northern Ireland has
there been no attempt to re-train or re-accredit the judiciary, beyond the UK-
wide training in preparation for the Human Rights Act 1998.
Consequently, much re-training of judges and lawyers is being conducted,
particularly across Eastern Europe. This involves training in international law
standards and particularly in the ECHR. A number of the programmes focus on
UN ‘soft law’ standards, but questions remains over the effectiveness of the train-
ing. These problems are most acute in countries where the independence of the
judiciary is not assured, where there are funding problems, e.g., in Macedonia
(CEELI 2002a: 7), and where there is poor political will for the implementation
of legal reforms, e.g., Bosnia, where the HRSG intervened and imposed laws to
speed up judicial reform (OHR 2002b). There is also evidence of resistance to
new ideas about law and legal practice and widespread corruption in the public
services, including the judiciary (Mertus 1999: 23). As the International Crisis
Group observed about Bosnia: “millions of dollars have been spent since 1996
by an assortment of international agencies to promote the rule of law in Bosnia,
including hefty salaries for over 200 foreign legal experts who have worked to
improve the performance of Bosnia’s 1,200 judges and prosecutors. In compari-
son to the sums expended, the results achieved have been pitiful” (International
Crisis Group 2002b: 10).
3 Changes to Law and Criminal Justice

Because repressive or draconian criminal justice systems and emergency law have
contributed to many of the problems the agreements seek to address, an impor-
tant area of change can be the new criminal justice arrangements and the aboli-
tion of emergency law. However, the latter is now less likely to occur given the cur-
rent international political climate and as the major global powers seek to increase
anti-terrorism powers and thus further erode civil liberties. In Northern Ireland,
for example, despite references in the Good Friday agreement to the need to ‘nor-
malize’ criminal justice, long-standing ‘emergency’ legislation has been replaced
with permanent anti-terrorist legislation applicable to the whole of the UK, with
34 Justice (Northern Ireland) Act 2002 Sec 19.

610 Angela Hegarty
special sections retained for Northern Ireland.35 There is a continuing wrangle over
the proposed reforms to policing and criminal justice, with further amendments
to the policing legislation. Therefore, it can be seen that a significant element of
any new arrangement will be changes to policing, criminal justice, and emergency
laws. In Bougainville, for example, the agreement envisages that the autonomous
government can declare and initiate a state of emergency in Bougainville.36
4 Dealing with the Past

One of the features of internal conflict and civil war in the modern era has been
the attempt by various institutions, including the United Nations, to address the
issue of gross human rights violations that have occurred during those conflicts.
Many of those conflicts, including some of those encompassed in the case stud-
ies, fall outside the legal definition of ‘war’ and thus outside the ambit of the
international laws of war. Whilst some of the Geneva standards may apply to
conflicts that are not technically wars and attempts have been made to generate
international humanitarian law standards that apply to internal conflicts, there
have been problems in finding a legal framework for addressing the many human
rights violations that conflicts generate. A central problem has been the fact that
human rights treaties are enforceable against governments or states and not usu-
ally against third parties or non-state actors. Therefore, where there is a civil war
or other form of ethnic and largely internal conflict with rights violations carried
out by paramilitary organizations, it is difficult to apply those treaties. It is also
practically difficult to apply those standards to non-state actors who may be act-
ing at the behest of the government or another kin state, however covertly. Thus
achieving accountability for violations of rights and upholding the rule of law can
be highly problematic. Notwithstanding these problems, dealing with the past
and with the abuses committed by prior regimes has become a necessary element
of resolving conflict and creating political agreement.
There are many ways in which the law can respond to and seek to address these
violations and by which individuals and states can be made accountable for such
violations. It can be done either by a criminal or a civil process: former dictators
and generals can be put on trial for their crimes, or a form of public investigation
can be carried out. The format of the criminal trial is familiar – from the Israeli
trial of Adolf Eichmann to the Argentinean trials of the former junta. Latterly,
the model has been employed to address the gross violations arising from the
conflicts in Rwanda and the Balkans. However, successor trials are relatively rare
(Teitel 2000: Ch 2), and although the criminal process may demonstrate the shift
to liberal democratic values, there are dilemmas about the application of due
process. Trying former dictators and their collaborators may be politically attrac-
tive but it gives rise to complex legal problems, and the conduct of such trials may
actually undermine the new legal ethos that they seek to establish. This is clearly
exhibited in some of the problems experienced in a number of the case studies,
35 Terrorism Act 2000.

36 Bougainville Peace Agreement 2001 Ch 12.
for example, Bosnia. There are other legal models available and in use – such as
public investigation. This is a less obvious, but pervasive, model, of which two
principal variations can be identified – the truth commission and the public in-
quiry. Public inquiries can be seen as a British export, and tend to be highly legal
and are usually narrowly focused, whereas truth commissions are more usually
quasi-legal with a wider ranging remit (Hegarty 2003).
Since the early 1970s, truth commissions have developed as ways of addressing
patterns of human rights abuses (Hayner 1994). Truth commissions are usually,
but not exclusively, addressed at abuses committed by the state or by previous
regimes exercising state power and usually arise during or as a result of political
transition. They are often a way of reconciling the new version of the conflict
with the old, although they are sometimes perceived as ‘victor’s justice’. The use
of the term ‘truth commission’ implies that ‘the truth’ has been denied or covered
up and that the purpose of the commission is to find and establish that truth.
Truth commissions have been heard all over the world,37 including South Af-
rica,38 Guatemala,39 Chile,40Argentina,41 and El Salvador.42 Each truth commission
is a separate and distinctive process, but they are all essentially flexible products
of political transition, often perceived as a necessary part of that transition. Pub-
lic inquiries, on the other hand, are a creation of the common law, principally
in the UK, although they have been exported to other common law countries.
They are often very formal, highly legal, and, although they are intended to be
inquisitorial, they can become highly adversarial. In Northern Ireland, the public
inquiry model is currently being used to investigate the events of Bloody Sunday,
where thirteen civilians were shot dead by the British Army during a civil rights
demonstration on 30 January 1972.
Some, but not all, of the cases surveyed have made attempts to address the
issue of the past. Table 3 provides examples of some of the key measures, which
include the incorporation of international human rights standards in the agree-
ments, changes to policing and criminal justice, the release of prisoners, the in-
tegration of former combatants into the police forces, and the setting up of insti-
tutions to address the violations of the past and to provide accountability in the
future.
37 For example, truth commissions are also planned or in process in Bosnia and Sierra
Leone.
38 Truth And Reconciliation Commission of South Africa Report (1998), Cape Town,
South Africa.
39 Guatemala: Memory of Silence: Report of the Commission for Historical Clarifica-
tion (Conclusions and Recommendations) (1999), Guatemala City, Commission for
Historical Clarification.
40 Report of the Chilean National Commission on Truth and Reconciliation (1993).
41 Nunca Mas: Report of the Argentine National Commission on the Disappeared
(1986).
42 From Madness to Hope – The 12 Year War in El Salvador: Report of the Commission
on the Truth for El Salvador, (1993), UN Doc S/25500, United Nations, New York.
612 Angela Hegarty
Table 3 Addressing the Past through Legal Change – Some Examples from the
Case Studies
Measure Is there ref- Positive ac- New ar- Past ad- Other
erence to in- tion meas- rangements dressed in
Case ternational ures in jus- for criminal legal form?
standards? tice system? justice?
Northern Yes – ECHR Police: 50-50 Changes to Prisoner re- e.g., Human
Ireland recruitment policing and lease; other- Rights Com-
(time limited)criminal jus- wise limited mission
tice
Kosovo Yes – UN Implied Policing; Refugees; e.g., Police
1999; 2001 – guaranteed Criminal prisoner Ombudsman
(much more places on justice (e.g., release;
specific) courts for commis- agreement to
ethnic groups sions) be bound by
international
humanitar-
ian law and
International
Criminal
Tribunal on
the former
Yugoslavia
Bosnia and Yes (Annex Memoran- Yes Partial International
Herzegovina 2) dum of Un- Judges on
ECHR sover- derstanding Constitution-
eign over all with specific al Court
other laws measures
to increase
minorities in
judiciary; but
no veto rights
in courts
Georgia Brief Promised in
Abkhazia
Macedonia Amendments Limited and Policing Measures Public At-
include gen- implied to deal with torney role
eral reference refugees; enlarged
amnesty law
Bougainville No Yes – chang- Provisions
es to polic- for amnesty
ing, criminal and weapons
law, etc. disposal
Mindanao Yes – for Policing and Integration of Shari’ah Pub-
Philippine arrest powers former com- lic Assistance
Supreme batants into Office (legal
Court police forces aid)
The extent to which any of the agreements engage in a discourse about the res-
toration of the rule of law or the acknowledgement of past wrongs or abuses is
limited. Despite the increasing incidence of truth processes in transitional societ-
ies, few of the agreements require the creation of mechanisms to deal with such
past wrongs, such as a truth commission, although in some of the cases there are
parallel legal processes aimed at providing some form of accountability, e.g., the
International Criminal Tribunal for the Former Yugoslavia. Some of the agree-
ments provide for amnesties or the release of prisoners, e.g., Northern Ireland,
but many of these measures are not politically agreed or possible at the point at
which the agreement is made.
Another form of accountability arises when considering the issue of whether
those who have been involved in the previous regime should be allowed to par-
ticipate in government, law, policing, or the civil service. In Eastern Europe, the
process of lustration has been used to remove many such individuals from public
positions and offices. Lustration is “a mode of accountability that bypasses the
criminal law by removing or disqualifying whole categories of people from gov-
ernment jobs” (Cohen 2001: 231). The extent to which this is contemplated by
agreements is important. Most of the agreements refer in some fashion to the
need for reconciliation after the conflict, or at least to provide mechanisms to
dispose of arms (e.g., Kosovo),43 to release prisoners (e.g., Northern Ireland),44
to provide for the return of refugees (e.g., Macedonia),45 or to integrate former
combatants into new policing arrangements (e.g., Mindanao).46 However, lustra-
tion is not referred to in detail in most of the agreements and, whilst it has taken
place to some extent in some of the cases, in others it has not. This may have
consequences for the stability of the agreement. In Northern Ireland, for exam-
ple, where there was no provision for lustration but former paramilitary combat-
ants were specifically barred from the police, many of those associated with the
former police service and judiciary remain in office.
The issue of how victims of violence associated with the conflict are dealt with
is important, but almost entirely overlooked. Few of the agreements do anything
other than pay lip service to victims. Few provide directly for reparations, pub-
lic hearings, land reform or reassignment, and in those that have done so (e.g.,
Bosnia)47 implementation of the guarantees (of the right to return, for example)
has been poor (Bell 2000: 253-254). The notable silence of the agreements on
practical measures to help victims of the conflict may be because such arrange-
ments can only follow a stable political settlement and thus may not be part of
the agreement. Nonetheless, it is a salutary lesson about the political muscle that
43 Rambouillet Agreement Ch 7 Arts III-VII; XVI.

44 Northern Ireland (Sentences) Act 1998.
45 Framework Agreement Annex C.
46 Mindanao Peace Agreement 1996 Ch II(19).
47 The 1995 Dayton Peace Agreement includes explicit recognition of the right of refu-
gees and displaced persons to return to their homes and repossess their properties in
Annex 7.
614 Angela Hegarty
victims and the groups that represent them have in the hotbed of political nego-
tiations.
5 Using Law to Implement Change

Law is a primary tool by which political change can be driven, although it is not
the only one. Thus, key to the success of any agreement will be the manner in
which the law implements and safeguards it. An independent judiciary and an
active civil society are crucial in achieving this. In respect of the administration of
justice provisions, there are wide variations in the extent to which there are spe-
cific implementation provisions for the administration of justice. In some cases,
there are no formal implementation provisions, but informal ‘side-deals’, ‘letters
of comfort’, or annexes that relate to implementation, e.g., Kosovo.48
In some cases, implementation of the agreement is left to particular bodies,
which may either be specifically created under the agreement49 or subsequently
set up,50 and there is usually implementing legislation supplementing (or in some
cases, amending) the agreement. Finally, some agreements provide for periodic
reviews,51 and, in some cases (e.g., Northern Ireland), there are controversies or
disagreements about implementation in relation to the administration of justice.
Perhaps as important as the implementation mechanism is the set of values
or principles that the agreement purports to incorporate into the newly created
or amended justice system. An examination of those values can indicate much
about the nature of the justice system the agreement seeks to create.
IV Law in Transition
A The Role of Law

Law is the main method by which policy change is instituted and by which new
institutions that are a product of that policy change are established. Thus, in
Northern Ireland, the Good Friday Agreement is a political deal between the
various political parties and the two governments that was ratified politically
by an all-Ireland referendum. Technically, the agreement is comprised of two
agreements – one agreed between the political parties and then the international
agreement between the two sovereign governments. However, in order for it to
48 Rambouillet Agreement Ch 5.
49 The Bougainville Peace Agreement 2001 creates a Constitutional Commission to
propose a constitution (Ch3 Art 14-17) and a Constituent Assembly to debate and
adopt a constitution.
50 For example, Macedonia (Ch 8.1 of the Framework Agreement of 2001 requires the
presentation of specified draft constitutional amendments to the Macedonian Par-
liamentary Assembly within forty-five days); Northern Ireland (Ch 2 and Annexes A
and B of the Good Friday Peace Agreement 1998 contain draft legislation).
51 For example, Northern Ireland (Ch 1, sections 5-8 of the 1998 Peace Agreement,
which provide for emergency and four year reviews).
be implemented it had to be transformed into law, primarily in the form of the

Northern Ireland Act 1998. This was supplemented by legislation in the Oireach-
tas (the Irish parliament), further legislation in the Westminster parliament, and
by bilateral international treaties between the two governments. Law was there-
fore the necessary vehicle by which political agreement was implemented.
Law has another important role in political transition, that of accountability
and the re-assertion of the rule of law. The rule of law is seen as an intrinsic
element of good governance (Mertus 1999), providing the stability and account-
ability necessary for such governance to flourish. As Neil Kritz observes: “The
rule of law does not simply provide yet one more vehicle by which government
can wield and abuse its awesome power; to the contrary, it establishes principles
that constrain the power of government, oblige it to conduct itself according to a
series of prescribed and publicly known rules” (Kritz 1996: 588).
Yet the practical experience for many is very different. Law may deliver the
accountability and truth sought by the victims of human rights violations, but
it is also often the tool employed by states to avoid or deny responsibility. Thus,
the law is a paradox – how can it be the primary means by which governments
are held to account, when it has also provided the convenient vehicle through
which those same governments have validated abuses and criminalized their op-
ponents? For those who have hitherto opposed the state and for others who are
the victims of state wrong-doing, the law is part of the state and is thus identified
with its ethos and culture. In Northern Ireland, for example, the law has been that
of the British state, draped with all the symbols and signs of the colonial power.
That is why the re-shaping of law and state and the development of new legal
processes are such central aspects of political transition (Teitel 2000: 66-67).
Without this change, the law remains the property of the state that was alien in
culture and political legitimacy and which may have carried out human rights vi-
olations. Without such change, the state remains deeply distrusted by significant
sections of society. Substantive change also signals a willingness to acknowledge
the faults of the past and the abuses they spawned. However, with the advent of
such change, the state risks losing control of the law and thus of a primary means
by which its policies and behaviour can be validated. For newly created states
or political entities, the assumption of law and the means of making it (through
fully or partially independent legislatures) is a key achievement. However, those
communities or ethnicities that supported the previous incarnation of the state
may regard the pre-existing legal system as sacrosanct and so may view any at-
tempt to reform or transform it as deeply suspect. This explains why, in Northern
Ireland for example, change to the legal and criminal justice systems remains so
controversial.
One way to resolve this paradox is to see the law’s role as facilitating both ex-
position and validation. The methodology of the legal process may allow access
to information that might otherwise be unavailable – the nature of many types of
legal process is suited to such an investigatory approach. Secondly, and despite its
limitations, the law is one of the very few mechanisms by which the state may be
made accountable for its actions by individuals. Thirdly, the law provides a means
616 Angela Hegarty
of ‘drawing a line’ under difficult or controversial events and thereby stamping

a final authority upon a version of those events. In such an acknowledgement
there is the minimum form of accountability that is necessary for the rule of law
to prevail. It is for this reason that many of those who have been abused by legal
processes still seek legal redress, arguing that the wrong done in the shape of the
law may only be undone by the law. It is also for this reason that the state chooses
to make use of legal processes – law provides the official version it requires.
Since the end of the Second World War, there has been a marked growth in
the priority given to accounting for human rights violations generally. Despite
this, genocidal conflicts and gross violations of human rights persist (Bassoui-
ni 1996: 10). As Eide (1983) argues, the development of human rights has three
stages – idealization, positivization, and realization. The evolution of the inter-
national human rights machinery is currently somewhere between stages two
and three. There has been an increase in the acceptance of international human
rights norms, as evidenced by the increasing accession by states to international
human rights treaties, the proliferation of treaties on specialized areas, and the
expansion of ‘soft law’. But the challenge is no longer to persuade states to sign up
to human rights declarations, but rather to ensure that they adhere to them.
Governments are generally more anxious than they previously were to sign hu-
man rights treaties, but many continue to commit serious human rights abuses
despite these apparent commitments. This gap between rhetoric and practice has
led to a greater focus upon accountability and investigation, particularly in politi-
cal transition. If a state agrees to be bound by international human rights treaties,
then those treaties create a normative standard against which that state may be
judged. These international standards have been utilized across the agreements
to provide benchmarks of core normative values. The importance of the promo-
tion of human rights in the democratization of states has become increasingly
important, as the upheavals in Southern Africa and Eastern Europe demonstrate.
The attempts to resolve self-determination disputes, such as those arising from
the break up of the Soviet Union, have permitted the wider disclosure of vio-
lations committed by the prior regimes. The political agreements arrived at in
resolving some of these conflicts have also involved, as part of the condition of
the deal, an accounting for the violations committed by the prior regime. The
emergence of evidence of such gross violations and the perceived need for some
formal legal response to them have led to the creation of both ad hoc mechanisms
(e.g., the International Criminal Tribunal for the former Yugoslavia, International
Criminal Tribunal For Rwanda) and permanent structures (such as the Inter-
national Criminal Court) to address the violations. These institutions are them-
selves subject to criticism of their effectiveness and their reach (Akhavan 1997;
Morris and Scharf 1998), but their formation can be seen as another step towards
the full ‘realization’ of rights.
B Sources and Standards

Law therefore provides both normative standards by which change can be judged
and a mechanism by which change can be driven. It is useful therefore to reflect
on the legal ideas, values, and benchmarks that may influence new systems and
institutions of justice and by which those new systems may in turn be judged.
1 International Standards
There are a plethora of international standards and guidelines for the administra-
tion of justice. In the modern era, these have tended to be driven by concerns
about human rights and by adherence to international human rights standards.
In addition to global and regional human rights treaty law (e.g., the UN Cov-
enant on Civil and Political Rights; the jurisprudence of the UN Human Rights
Committee; the European Convention on Human Rights and the jurisprudence
of the European Court of Human Rights), there are many basic standards that
have quasi-legal status (known as ‘soft law’). They largely emanate from the UN,
but also come from the Council of Europe and other organizations.
The list of relevant UN principles alone runs to several pages, but amongst the
most important are:
• The UN Basic Principles on the Role of Lawyers;
• The UN Guidelines on the Role of Prosecutors;
• The UN Basic Principles on the Independence of the Judiciary;
• The UN Code of Conduct for Law Enforcement Officials (and the Commen-
tary thereto);
• The UN Body of Principles for the Protection of All Persons under Any Form
of Detention or Imprisonment;
• The UN Principles on the Effective Prevention and Investigation of Extra-
legal, Arbitrary and Summary Executions;
• The UN Minimum Rules for the Administration of Juvenile Justice; and
• The UN Victims Declaration.
There are other types of standards, less overtly concerned with human rights.
Useful sources of this nature are documents such as the United Nations Frame-
work for Strengthening of the Rule Of Law, which incorporates a number of
concepts and benchmarks, for example, a strong overarching constitution, an
independent judiciary, limits on derogations from international law and human
rights standards, non-discrimination, national human rights institutions, and “a
strong legal framework, under the Constitution, which protects human rights
and democracy, and which provides for effective redress” (United Nations 1994:
para 3).
The OSCE has also promulgated relevant standards on human rights, justice,
and the rule of law. These numerous documents include affirmations that de-
mocracy is part of the rule of law; that everyone is equal before the law, the im-
portance of an independent judiciary, and the relevance of international human
rights norms to the rule of law (OSCE 1990); the importance of accountability
618 Angela Hegarty
and state action compliant with the law (OSCE 1991); and the centrality of due
process and fair trial rights (OSCE 1989; 1990; 1991). Like the OSCE, the EU has
increasingly adopted the language of human rights and the rule of law, inserting
the concepts overtly into the Treaty of Rome via the Amsterdam Treaty and by
linking them specifically to its conditions for accession to the EU and to democ-
ratization and human rights (EU 1991).
There are many other legal and human rights considerations relevant to the
case studies, such as the accommodation of minority rights and protections for
linguistic and ethnic groups within the legal system. Such protections may form
part of the agreement reached in the settlement of the political dispute, but unless
they form an inherent part of the legal system, through, for example, guaranteed
places or quotas for minority groups in the judiciary, they may have little effect
in practice. In Northern Ireland, a time-limited form of quota was implemented
in the recruitment of police officers to the new police service, guaranteeing 50/50
recruitment of Protestants and Catholics.52 Interestingly, there was no compara-
ble mechanism proposed for the courts or the legal system, and nor was there any
attempt to provide a similar guarantee of recruitment to the police for women or
ethnic minorities, despite the acknowledgement by the Commission on Policing
that both these latter groups were similarly under-represented in the police force
(Independent Commission on Policing in Northern Ireland 1999: 14.5).
Whilst it is clear that many of the ‘soft law’ standards are used when training or
educating personnel involved in the administration of justice, there appears, ini-
tially at least, to be little employment of these or other standards when designing
the system. Indeed, in Northern Ireland, the UK government specifically rejected
the proposal by the Northern Ireland Human Rights Commission and a range of
human rights NGOs that the new police service created under the Good Friday
Agreement should be required to have recourse to such standards (O’Rawe 2003:
1049). Even if such standards are not used in the design of the systems, they do
provide useful benchmarks by which progress towards change can be measured,
yet they do not appear to have been utilized to any great extent in this regard.
1 The Rule of Law

The rule of law is an old-fashioned concept that has been undergoing something
of a revival. Carothers provides a useful description:
The rule of law can be defined as a system in which the laws are public knowledge, are
clear in meaning, and apply equally to everyone. They enshrine and uphold the political
and civil liberties that have gained status as universal human rights over the last half-
century. In particular, anyone accused of a crime has the right to a fair, prompt hear-
ing and is presumed innocent until proved guilty. The central institutions of the legal
system, including courts, prosecutors, and police, are reasonably fair, competent, and
efficient. Judges are impartial and independent, not subject to political influence or ma-
nipulation. Perhaps most important, the government is embedded in a comprehensive
52 Police (Northern Ireland) Act 2000 Sec 46.

legal framework, its officials accept that the law will be applied to their own conduct,
and the government seeks to be law-abiding (Carothers 1998: 2).
It should of course be observed that the concept of the rule of law embodies
western liberal notions about justice and fairness that cannot be transferred au-
tomatically to other cultures, which may not interpret or apply the concept in the
same way.
The principle of the rule of law was most famously articulated by A. V. Dicey,
who argued that every act of government or its agents had to be done in accord-
ance with and subject to the strictures of the law. This was an elaboration of a
principle that had long been identified by legal theorists and philosophers, but
Dicey went further and described a number of characteristics of the rule of law.
Amongst those characteristics were that all action was subject to the supervision
of parliament, particularly that power which was discretionary, or delegated by
parliament; and that an important bulwark against abuse of power by the state
were the courts. Dicey saw the judiciary as having a crucial role in defending
the rule of law in the face of the state (Dicey 1985: 175–186). Thus, in any recon-
struction or re-invention of a nation, a robust and independent judiciary is key,
particularly in places where the allegation made is that the pre-existing judiciary
was simply a tool of the old regime. The concept of due process applies across the
legal system, but is most developed in respect of the criminal law. The 1959 New
Delhi Conference on the Rule of Law devoted some time to the specifics of the
criminal process and the rule of law. Among the concepts identified as central to
the concept of the rule of law were the presumption of innocence, the need to
ensure that arrest powers and detention before trial are exercisable within strict
limits, and the principle that the duty of the prosecution is not to secure a con-
viction – rather it is fairly to place the relevant evidence before the court (Marsh
1959: 8-18).
The theory of the rule of law has been criticized, principally by the critical
legal studies movement, as unattainable, unrealistic, and ultimately detrimental
to democracy (Horwitz 1977; Tushnet 1983; 1988). It has been re-formulated as a
highly formalistic conceptualization, which tends to contract it towards a narrow
‘rule by law’ definition, which would mean, for example, that slavery, provided it
conformed with the law, would be compatible with this theory of the rule of law
(Raz 1977: 211). In the modern era, the concept has been sidelined and subsumed
to some extent by modern international human rights standards. The primacy of
the contemporary discourse about rights meant that the concept of the rule of law
is one less visited by legal commentators. However, with the tumultuous events
of the past few decades and the needs of reconstruction and ‘democratization’
in places as diverse as the Balkans, Central and Southern America, and Africa,
the ideas embodied in the concept of the rule of law have been re-popularized.
Increasingly, it is being linked to delivering due process and core human rights
norms as part of a bundle of standards regarded as necessary for democratization
and political stability, what Juan Méndez calls, “the fundamental character of the
new order to be established, an order based on the rule of law and on respect for
620 Angela Hegarty
the dignity and worth of each human person” (Méndez 1997: 1). Thus the liberal
definition of the concept (Dworkin 1986; Allan 2001) provides a useful rubric of
what values a legal system ought to embody, and perhaps also offers a clearer
blueprint for politicians and political scientists than that of the often confusing
and technical legal language of rights.
There are essentially two elements to the rule of law – no one is above the law
and everyone is equal before the law. From these flow the workings of the rule
of law, or the benchmarks by which we can judge whether a society conforms to
the rule of law. These benchmarks, which have emerged from numerous interna-
tional bodies, such as the EU, the OSCE, and the Council of Europe, as well as the
UN, produce a form of modern consensus around what is meant by the rule of
law. Most current definitions of ‘good governance’ include adherence to the rule
of law (e.g., EU 1991: para 5) and therefore it can be assumed that the following
benchmarks are also necessary for that to exist. The benchmarks are:
• equality before the law;
• accountability under the law;
• the application of due process;
• fair trial;
• an independent judiciary;
• the incorporation of core international human rights;
• democratic elections;
• anti-corruption measures;
• a strong civil society.
This re-popularization of the rule of law is aided by the fact that the key ideas
that it enshrines are also concepts embedded in the western idea of democracy
– such as fairness, openness, equality, and due process. It may also have suited
the political ideologies and agendas of some of those countries involved in these
‘democratization’ processes. Confronted by the great differences between the
pre-existing legal systems of the eastern bloc and others from the common law
model predominant in the USA, UK, and Australia, the rubric created by the rule
of law provided a common benchmark by which new systems could be developed
or old ones reformed. The warnings about legal colonialism apart, it may be that
the modern formulation of the rule of law provides an accessible and culturally
familiar kind of normative standard by which new institutions and the laws and
the practices that emerge from them can be judged.
However, the rule of law projects can also be seen as exercises in legal colonial-
ism and the imposition of capitalist values on transitional societies, at a point at
which they are weakest to resist them. “In post-colonial times ‘democracy, hu-
man rights, and good governance’ and ‘the Rule of Law’ are exported as part of a
market driven ideology” (Twining 2000: 43). Capitalism is keener to export legal
standards that allow markets to flourish than it is to export the standards of liv-
ing deemed essential in western liberal democracies. The criticism can be made
that the rule of law’s focus on the creation of democratic and legal institutions
conveniently neglects the abject poverty in which many of those who are the
focus of this export exist (Faundez 2000; Wetlaufer 1999). It can be said that the
imposition of western liberal ideas about the rule of law on emerging democra-
cies has more to do with ensuring that the sort of society that emerges is one
that will support market economies and globalization than with concerns about
due process and human rights. Such concerns can be ameliorated if the linkage
with rights is not just about ‘first generation’ civil and political rights, but also
embraces economic, social, and cultural rights, which may be more important to
the citizens of these emerging democracies. The reluctance of the great western
powers to make this link to so-called ‘third generation rights’ is illustrative of the
problem. The dangers of this approach are obvious. As Julie Mertus remarks:
[T]he process of legal transplants by which politically strong states such as the United
States persuade weaker states to adopt U.S.-style laws and institutions and the proc-
esses of international law making (through which transnational bodies determine the
content and impact of international law) run the danger of violating democratic norms.
Weaker states may feel forced to agree to adopt U.S.-style laws, and the process by
which those laws are adopted may not be open to public scrutiny. Bosnia-Herzegovina
provides one extreme example of this phenomenon. American advisers have played an
incredibly forceful ‘behind the scenes’ role in the formulation and adoption of U.S.-style
criminal and civil laws by the Bosnian entities (Mertus 1999: 17).
It certainly seems that, in a number of the case studies, the ‘great powers’ have
imposed their own values and institutional ideals on the new arrangements. Yet
in one case, that has noticeably not been the case. In Northern Ireland, there has
been very little change to the core administration of justice institutions. There are
no new courts, few new legal entities, and ongoing struggles over the legal powers
and level of funding of those that have been created. There are some pre-existing
entities (e.g., the Police Ombudsman) that are affected by the new arrangements.
It may be said that this is because in Northern Ireland– unlike Kosovo or Bosnia
– the pre-existing legal system at least operated some semblance of the rule of
law. Because of this, it already looked much like the kind of system the ‘great pow-
ers’ sought to create or impose in many of the other cases.
This was made simpler by the fact that the legal system in Ireland, the kin-
state, did not look much different either, with the one notable exception of the
possession of an entrenched written constitution. This is largely because the kin-
state is a former colony of the other state party in the conflict and because it is a
relatively recently liberated former colony that is very much part of the western
liberal democratic consensus. That no entrenched written constitution was cre-
ated for Northern Ireland as part of the agreement speaks volumes about where
the real power in the negotiations lay. The UK’s refusal to allow a variation in its
constitutional model has created continuing problems in the implementation of
the agreement, not just in the areas of policing and security, but was also dem-
onstrated in the recent (May 2003) unilateral decision by the UK government to
suspend elections in Northern Ireland. Legal sovereignty remains in the hands of
the UK, which means that it can suspend the democratic institutions by passing
622 Angela Hegarty
legislation through the UK parliament, even when, as in this case, the other gov-
ernment involved and most of the political parties in Northern Ireland disagreed
with its decision.
There are other problems with the rule of law as a concept and as a standard.
One real criticism that can be made is that it can be used as a method of obscur-
ing repression. It is possible to have systems that appear to conform to the rule
of law, but in reality flout it, particularly those systems where national security
is employed by the state as a shield to deflect the investigation and the discovery
of activities that flout the rule of law. In Northern Ireland, for example, the state
maintained that the rule of law ran throughout the conflict, when in reality it was
engaging in very nefarious activities (Stevens 2003: para 1.3). The appearance that
all was well, despite the criticisms of many international human rights NGOs,
was maintained by the illusion of due process institutions such as the courts.
Thus a failure to undertake a root-and-branch reform of the pre-existing system
led to problems in creating the democratic accountable society the agreement
strives to bring about.
The problem of transposing alien legal cultures is better addressed by learn-
ing from the mistakes of the past. A bottom-up approach, which agrees rather
than imposes new institutions and which spends time working out their struc-
tures and ethos through a thorough application of UN standards may address this
problem. Combined with an understanding that rights and standards have to be
enforced before they can be real, such an approach has a better chance of estab-
lishing long-term democratization. It may not suit the agenda of the ‘great pow-
ers’ to take such an approach, as it may produce outcomes that are anti-capitalist
or fundamentalist, or which shine a light on the role of those ‘great powers’ in the
conflict, but that is the price of a real commitment to democracy. The problem of
corruption, endemic in some of the cases, may be a symptom of a failure to take
such a grassroots approach, which would allow indigenous populations to decide
and develop their own systems and institutions in a way that is compatible with
international standards.
V Conclusion
The rule of law is often misunderstood. Lawyers use it as shorthand, with the
consequence that the term is not always properly understood by non-lawyers.
The most common misapprehension is that it simply means ‘rule by law’. The dif-
ference between the two is usefully explained thus:
In a rule-by-law system, power is tightly held by a small social elite that uses the legal
system to protect and consolidate its privileges and power. The voiceless majority has
no tools to challenge this monopoly on the self-interested use of law. In a rule-of law
system, by contrast, power is fragmented and dispersed among rival social groups and
organized interests, none being powerful enough to work its will by intimidation or
force (Holmes 2002: 90).
Whilst the re-emergence of the rule of law may owe much to the attempt to
advance western capitalist ideology, the modern concept is now linked to the
protection of fundamental rights. The rule of law may concentrate more on the
protection of certain types of rights – such as due process, fair trial, and free
speech – but that is due in large part to the nature of the concept, which focuses
upon systems and processes of justice. The concept of human rights is, however,
a much wider, deeper concept, only some aspects of which are guaranteed by the
full implementation of the rule of law.
Implementing change in the administration of justice is complicated by the
fact that the issues to which law and justice relate are often major factors in the
conflict and continuing areas of dispute. In a number of the cases, it might be
thought that changes to the administration of justice were settled as a part of the
agreement. Christine Bell, for example, in her comparison of Bosnia and North-
ern Ireland argues that in both cases rights were conceded by one side in the
conflict as the price of tying the other into the settlement (Bell 2000: 159). Yet
the problems in both cases have been complicated by ongoing controversies over
rights (Campbell, Ní Aoláin, and Harvey, 2003: 342).
Technically, rights cannot be conceded: they are universal and inherent. Per-
haps one of the errors of political negotiations, however tempting, has been to
treat rights as if they could be conceded, as if they were bargaining counters in the
gift of one side, to be handed over to the other parties at a price. This has an ulti-
mately detrimental effect on peace negotiations, as the illusion of concession may
be created where one does not exist. In Northern Ireland, for example, rights are
still being contested, demonstrated by the continual re-negotiation of changes to
policing and criminal justice.53 If these rights were capable of being conceded and
had in fact been conceded, then no contestation would be taking place.
The difficulties in bringing about change are exacerbated by the fact that not
only are the new or reformed courts emanations and creatures of the agreements,
they are also often the channels through which controversies are directed and
settled. This raises issues about how the agreements themselves are enforced.
The matter of how well self-determination is delivered via the agreement may be
for the courts to decide, but if those institutions remain largely unchanged from
before then the trust of the general public in the verdicts of those institutions may
be insecure. If the institutions of justice are unrepresentative of the population
or if they fail to act independently and uphold both the rule of law and protect
fundamental rights, then they fail to deliver the bulwark against corruption, re-
pression, and discrimination promised by the agreements.
All the case studies demonstrate the need for the reassertion of the rule of law
and due process after conflict. In the majority of the cases, there is some official
acknowledgement that the rule of law has been disrupted, at least, in the past,
and that has contributed to the conflict. Even in Northern Ireland, where the
state has generally been in denial about its role in the conflict, there is no doubt in
the mind of most independent observers that the rule of law had been damaged.
53 Joint Declaration by the British and Irish Governments, April 2003, paras 20-24.
624 Angela Hegarty
In many discourses about the connections between conflict and the rule of law,
a primary theme is the role of law as an instrument of repression. In such a situ-
ation, a partisan legal system is used by the powerful to oppress, persecute, and
criminalize minorities, a prominent aspect of which is lack of accountability for
the agents of the state who commit what would otherwise be criminal offences.
This raises once again the paradox of the rule of law: the law as an instrument
of abuse, repression, and conflict versus the law as the means of accountability
and political transition. It is only a visibly transformed legal system that can re-
solve that paradox, through a justice system that in its values, ethos, and practices
guarantees and delivers the rule of law. A deal which does not secure the rule of
law and which makes elements of it negotiable may work in the interim but is
unlikely to survive. The principles and ideals that underpin and guarantee legal
transformation cannot be subject to negotiation. It is only through the rehabilita-
tion of the rule of law that the peace sought in the agreements can take hold.
Despite the geographical and cultural diversity of the cases, the one thing they
all exhibit is the need for a political stability that recognizes and incorporates
diversity, defends human rights, and that does so in a manner that promotes due
process and the rule of law. The imposition of US-style institutions in some of
the Balkan cases should lead one to distinguish between models and principles.
Principles provide the physics that govern the erection of the institutions, models
provide architectural blueprints that only permit the erection of certain kinds of
institutions. Courts and other legal institutions should be allowed to take any
shape provided they meet the requirements of the rule of law: their form should
not be unnecessarily prescribed by any one kind of model, particularly where that
model springs from a Western cultural experience of those principles. If there has
been one particular flaw in the administration of justice projects, it has been to
assume that only Western-style institutions can produce the rule of law. Indeed,
perhaps one should deliberately seek to avoid importing particular models and
prefer to design indigenous ones, which will satisfy the twin aims of embedding
an understanding of the rule of law and creating institutions that incorporate and
defend it in a culturally appropriate way. Whilst the agreements that create the
new political and legal orders contemplated in the case studies are reflections of
the trade-offs and safeguards inherent in negotiations, they should also seek to
bring accountability and fairness where previously there was none.
Therefore, all the cases demonstrate the need to re-imagine the rule of law in
a way that links it to the protection of the full range of human rights, including
economic, social, and cultural rights. A courts system that excludes from judicial
office certain minorities or a constitution that does not prohibit discrimination
on ethnic grounds can no longer satisfy the strictures of the rule of law. The rule
of law, thus re-imagined, can only be grown in the native soil of the transitional
state: it cannot be transplanted from the greenhouses of the west. It will only
provide the stability and good governance sought if it is a concept embedded in
every aspect of society and in its new institutions. Simply imposing it as a stan-
dard from ‘on high’, without building in an understanding and a feel for what it
means in practice is insufficient.
As Mary Robinson put it:
[T]oo little attention has been given to building effective institutions. Human Rights
cannot be realized in the absence of effective and accountable institutions. Where
courts are corrupt, over-burdened and inefficient, basic civil rights will be violated.
Where social ministries are under-resourced, disempowered or lack qualified staff, ba-
sic rights to adequate health care, education and housing will remain unfulfilled. Insti-
tution building and reform is neither easy nor particularly newsworthy – it is, however,
essential (Robinson 2003: 7).
A Resolutions and Regulations

Conference for Security and Co-operation in Europe (1991). “Report of the Third Confer-
ence On The Human Dimension Of The CSCE Moscow 10 September - 15 October
1991” CSCE.
Conference for Security and Co-operation in Europe (1990). “Document of the Second
Conference On The Human Dimension Of The CSCE Copenhagen 5 June–29 July
1990” CSCE.
Conference for Security and Co-operation in Europe (1989). “Concluding Document of
Vienna Conference” CSCE.
EU (1999). Regulations 975/99 and 976/99 of 29th April 1999.
EU (1991). European Council Resolution of 28 November 1991 on Human Rights, Democ-
racy and Development.
EU Council Regulation (EC) 975/1999 of 29 April 1999, Official Journal L 120, 8/5/1999, 1.
EU Council Regulation (EC) 976/1999 of 29 April 1999, Official Journal L 120, 8/5/1999, 8.
Organization For Security And Cooperation In Europe (OSCE) (1994). “Budapest Docu-
ment: Towards A Genuine Partnership In A New Era” OSCE.
UN HRC, (2001). “Creation Of A Pre-Sessional Working Group On The Administration
Of Justice Of The Sub-Commission On The Promotion And Protection Of Human
Rights”, Commission on Human Rights Decision 2001/106.
UN General Assembly Resolution A/RES/42/143 (7 December 1987).
UN General Assembly Resolution A/RES/56/161 (11 December 2001).
UN Basic Principles on the Role of Lawyers, Eighth United Nations Congress on the Pre-
vention of Crime and the Treatment of Offenders, Havana, 27 August to 7 September
1990, UN Doc. A/CONF.144/28/Rev.1 at 118 (1990).
UN Guidelines on the Role of Prosecutors, Eighth United Nations Congress on the Pre-
vention of Crime and the Treatment of Offenders, Havana, 27 August to 7 September
1990, UN Doc. A/CONF.144/28/Rev.1 at 189 (1990).
UN Basic Principles on the Independence of the Judiciary, Seventh United Nations Con-
gress on the Prevention of Crime and the Treatment of Offenders, Milan, 26 August
to 6 September 1985, UN Doc. A/CONF.121/22/Rev.1 at 59 (1985).
UN Code of Conduct for Law Enforcement Officials, GA Res 34/169, Annex, 34 UN GAOR
Supp (No. 46) at 186, UN Doc. A/34/46 (1979).
626 Angela Hegarty
UN Body of Principles for the Protection of All Persons under Any Form of Detention or
Imprisonment, GA Res 43/173, Annex, 43 UN GAOR Supp (No. 49) at 298, UN Doc.
A/43/49 (1988).
UN Principles on the Effective Prevention and Investigation of Extra-Legal, Arbitrary and
Summary Executions, ESC res. 1989/65, Annex, 1989 UN ESCOR Supp (No. 1) at 52,
UN Doc. E/1989/89 (1989).
UN United Nations Standard Minimum Rules for the Administration of Juvenile Justice
(“The Beijing Rules”), GA Res. 40/33, Annex, 40 UN GAOR Supp (No. 53) at 207, UN
Doc. A/40/53 (1985).
UN Declaration of Basic Principles of Justice for Victims of Crime and Abuse of Power, GA
40/34, Annex, 40 UN GAOR Supp (No. 53) at 214, UN Doc. A/40/53 (1985).
B Cases
In Re Treacy & McDonald (2000). (Northern Ireland H.Ct 2075), available at http://www.
courtsni.gov.uk/en-gb/judicial+decisions/judgments/j_j_kerf3196.htm (24 Novem-
ber 2003)
Chapter 21
The Role of Human and Minority Rights in Complex
Power-sharing
I Introduction
A fundamental concern for the well-being of the various individuals and com-
munities who have been involved in or affected by ethnic conflict underscores all
complex power-sharing agreements. At issue in such arrangements is the recog-
nition and accommodation of distinct and potentially competing interests, en-
titlements, and beliefs. The obvious political objective is the restoration and pres-
ervation of peace, order, and good government within states, and, by extension,
also stability between states. The central premise behind complex power-sharing
is that ethnic groups who are recognized by and accommodated within the state
are far less likely to challenge its authority or to threaten its territorial integrity.
Recognition and accommodation may take different forms: legal or political.
Legal guarantees are embodied in treaties, in constitutions, and in laws; political
guarantees are embodied in the structure of the state, in political parties, and in
electoral systems. In practice, of course, the two forms often overlap, and argu-
ably function best when they are mutually reinforcing. For example, constitu-
tional protection is dependent upon the judicial system, which is in turn embed-
ded within the larger political structure of the state. Without a fair and impartial
judiciary whose judgements are respected by political actors and institutions,
constitutional provisions are worth little more than the paper they are written on.
Nevertheless, there is a qualitative difference between legal and political provi-
sions. Political provisions are disproportionately utilitarian and reflect changing
calculations of interest and efficacy. In other words, they are reasonable subjects
for negotiation and are usually intended to encourage a political culture of ac-
commodation. In contrast, legal provisions are overwhelmingly deontological.
They are designed to recognize fundamental normative entitlements and to de-
liberately put these beyond the reach of political compromise. Human and mi-
nority rights fall into the latter category.
Curiously, the role of legal provisions – including human and minority rights
– is often ignored, or at least downplayed, within the existing literature on ethnic
628 Jennifer Jackson-Preece
conflict regulation (Sisk 1996: 66). Instead, both consociational and integrationist
perspectives are disproportionately, if not exclusively, concerned with political
provisions (e.g., proportional representation and consensus rules in executive,
legislative, and administrative decision-making) and state structures (e.g., federal
and electoral systems). The explanation for this gap is not immediately obvious.
Legal practices, including human and minority rights, are increasingly identified
as fundamental prerequisites for good governance by both state and non-state
actors and feature prominently in complex power-sharing agreements. It may
well be that this oversight is a consequence of a methodological predisposition
towards social science positivism, which tends to avoid normative claims and
categories to the detriment of empirical analysis in this area. This chapter aims
to redress the balance somewhat by examining the implications of a rights-based
response to ethnic conflict as disclosed in the various complex power-sharing
agreements studied in this project.
II The Idea of Rights
A Rights as Normative Entitlements

The rights discourse is an almost ubiquitous feature of contemporary politics.
States, international and nongovernmental organizations, individuals, and groups
all increasingly speak the language of rights. This tendency is apparent in the his-
tory of the self-determination disputes studied here: the ‘troubles’ in Northern
Ireland are generally perceived to originate in the failure of a prior movement for
civil rights symbolized by the human rights atrocity known as ‘Bloody Sunday’;
the break-up of Yugoslavia and its political consequences are graphically revealed
in the documentation of gross human and minority rights abuses catalogued by
international and nongovernmental organizations; similar, albeit less sustained,
reporting defines the self-determination disputes in Gagauzia, Transdniestria,
South Ossetia, Abkhazia, Bougainville, and Mindanao in accordance with inter-
national standards of human and minority rights.
In part, such developments are a consequence of the growing recognition of a
variety of rights at both the domestic and the international level, but they are also
much more than that. The salience of the rights discourse, and indeed the domes-
tic and international standards it invokes, is an important reminder that politics
is normative as well as instrumental. International agreements, state structures,
public policies, and so forth entail moral issues – questions of good and bad, right
and wrong (Jackson 2000: 38). To claim a right transforms what might initially
be viewed as a mere expediency into a moral imperative: rights are normative
entitlements. In other words, the power of the rights discourse originates in its
normative content and corresponding moral authority. That is why the Bougain-
ville Freedom Movement compiled two volumes on the human rights abuses
committed against the people of Bougainville, and why the Mindanao National
Liberation Front claim that the ethnic nationalities of Mindanao have been vic-
tims of ethnic cleansing, and why the South Ossetian leadership has accused the
21 The Role of Human and Minority Rights in Complex Power-sharing 629
Georgian government of discriminating against the non-Georgian population.

Examples like these may be found in all the self-determination conflicts included
in this project.
B Main Elements of Rights

Despite its widespread appeal, the idea of rights is far from obvious. Rather, there
is often a great deal of confusion between the use of ‘right’ as an adjective (the
description of proper conduct) versus the use of ‘right’ as a noun (something
one has) (Vincent 1986: 8). The idea of a right as a normative entitlement is the
‘stock-in-trade’ of lawyers and thus is usually a feature of legal or quasi-legal (e.g.,
non-binding international resolutions and documents) provisions. A right in this
sense consists of five main elements:
• a right holder (the subject of the right) has,
• a claim to some substance (the object of a right),
• which he or she might assert, or demand, or enjoy, or enforce (exercising a
right),
• against some individual or group (the bearer of the correlative duty),
• citing in support of his or her claim some particular ground (the justification
of a right) (Vincent 1986: 8).
For example, as an individual (the right holder), I may claim a right to property
(the object of the right), which I assert in the form of ownership and exclusive
enjoyment (the exercising of the right) against other potential property holders,
be these individuals, companies, or indeed the state (the bearers of the correlative
duty) on the grounds that I should be able to enjoy the proceeds of my own labour
(the justification of the right).
C Classifying Rights
Although all rights possess these five main elements, they are far from identical.
Accordingly, rights theorists have often found it expedient to classify rights ac-
cording to their right holder (individual versus group), their substance (civil and
political, economic and social, or cultural) and their correlative duty (positive or
negative).
1 Right Holders: Individuals versus Groups

Rights may be held either by individuals or by groups. Individual rights are de-
signed to preserve and protect the autonomy of persons as individuals. Group
rights are designed to preserve and protect the individual’s propensity for com-
munal attachments and associations. Individual rights are held by individuals
either as humans or as members of specific groups (e.g., citizens of a particular
state or members of minority communities). The rights included in the Universal
Declaration on Human Rights and the UN covenants are all individual rights, with
the exception of the right to self-determination of peoples (Donnelly 2003: 21).
This generalization holds true even for those minority rights that have been rec-
ognized in Article 27 of the International Covenant on Civil and Political Rights
and in the Universal Declaration on the Rights of Persons Belonging to National
or Ethnic, Religious or Linguistic Minorities. Group rights are held by corporate
entities. In international law, the classic group rights are the right of sovereignty
held by states and the right of self-determination held by peoples.
Within the academic literature on rights, there is currently a great deal of dis-
cussion about the relative merits of individual versus collective formulations, par-
ticularly as these would apply to sub-state groups and their members (Lyons and
Mayall: 2003). This discussion may be theoretically interesting, but it often seems
to ignore the practical purpose of human and minority rights guarantees. Both
individual and group rights are necessary to give full legal recognition to the fact
that we exist as social beings and not as atomistic individuals. It is precisely this
aspect of human nature that gives rise to families, tribes, ethnicities, and states
– and which creates those social circumstances in which conflicts of interest,
entitlement, and belief may occur. Self-determination disputes generally contain
elements of all three conflicts, which may explain their apparent intractability.
Practically speaking, rights provide a moral framework in which to understand
and respond to these conflicts; to recognize only one category of right holder
risks ignoring a key element of the problem.
The difficulty of balancing between rival claimants (individuals versus groups)
remains a crucial issue. But in practice, there is a reasonably well-established hi-
erarchy within the various standard-setting documents that is designed to do
precisely this. Where conflicts of rights occur between individuals and groups,
moral primacy is usually accorded to the individual as the ultimate source of col-
lective legitimacy. In other words, the normative value assigned to groups is de-
rivative of the more fundamental normative value vested in individuals. Accord-
ingly, groups must respect the autonomy of their individual members, including
those who freely choose to dissent from majority opinions or ways of life. This
rationale is apparent in the complex power-sharing agreements, which in most
cases specifically require those institutions of self-government that are represen-
tative of minority communities to abide by national or international individual
rights guarantees.
2 Substance: Civil and Political, Economic and Social, or Cultural

During the Cold War, there was intense debate about the relationship between,
and indeed the relative priority of, distinct substantive entitlements. Civil and
political rights were championed by Western liberal states. These rights include
those guarantees associated with liberal democracy: the right to life, liberty, se-
curity of the person, privacy, and property; freedom of thought, conscience, and
religion; freedom of assembly and association; the right to a fair trial; freedom
from slavery, torture, and arbitrary arrest; and the right to free elections, univer-
sal suffrage, and participation in public affairs. Such provisions may be found in
the Universal Declaration on Human Rights (1948), the International Covenant
on Civil and Political Rights (1966), and in various regional human rights instru-
ments, including the European Convention on Human Rights (1950), the Ameri-
can Convention on Human Rights (1969), and the African Charter on Human
and People’s Rights (‘Banjul Charter’) (1981).
Economic and social rights were championed by Soviet bloc states. These
rights include those guarantees associated with socialism and the welfare state:
the right to work and a living wage; the right to form and join trade unions; the
right to adequate health care and housing; the right to education; the right to so-
cial security; and the right to paid holidays. Such provisions may be found in the
International Covenant on Economic, Social and Cultural Rights (1966), various
conventions sponsored by the International Labor Organization (ILO), and in re-
gional instruments such as the European Social Charter (1961) and the Additional
Protocol to the American Convention on Human Rights in the Area of Economic,
Social and Cultural Rights (‘Protocol of San Salvador’) (1988).
Cultural rights were and remain the rallying point for national or ethnic, re-
ligious or linguistic minorities, and indigenous peoples who are unable to form
their own independent states for whatever reason(s). Cultural rights include
those rights directed at the preservation of and participation in cultural commu-
nities. The substance of cultural rights may overlap with both civil and political
rights and even economic and social rights. However, the correlative duties they
create may be distinguished by the emphasis on policies and actions of specific
relevance to culture and identity, especially (although not exclusively) regarding
minorities and indigenous peoples. The most important provisions concerning
cultural rights as these apply to minority groups may be found in Article 27 of
the International Covenant on Civil and Political Rights, in the UN Declaration
on the Rights of Persons Belonging to National or Ethnic, Religious or Linguistic
Minorities (1992), ILO Convention 169 Concerning Indigenous and Tribal Peo-
ples (1989), and in regional instruments such as the Council of Europe (COE)’s
Framework Convention for the Protection of National Minorities (1995), the Eu-
ropean Charter for Regional or Minority Languages (1992), and the Copenhagen
Document of the Organization for Security and Co-operation in Europe (OSCE)
(1990).
Today there is widespread agreement that all rights – whether civil and politi-
cal, economic and social, or cultural – are indivisible, interdependent, and inter-
related (Donnelly 2003: 22). Accordingly, provisions from each of these catego-
ries of rights may be found in a variety of combinations within the power-sharing
agreements encountered in this project. Nevertheless, as will be discussed in
greater detail later in the chapter, cultural rights would appear to be especially
significant within power-sharing arrangements – presumably because such ar-
rangements are by definition disproportionately concerned with ethno-cultural
minority communities.
3 Correlative Duties: Positive or Negative

Rights may also be distinguished by the kinds of correlative duties they create.
Some rights create negative duties, which require restraint on the part of actors
other than the right holder, as with freedom of speech or assembly. Other rights
create positive duties, which require particular actions on the part of actors other
than the right holder. In other words, they are designed to promote specific prac-
tices that benefit the right holder. For example, the right to education is usually
exercised through access to educational institutions (schools, universities, etc.).
Such institutions cannot be created or maintained by the right holder acting on
their own; they require a collective effort and may necessitate state involvement
(in funding, etc.).
It is often claimed that negative duties tend to correspond with civil and po-
litical rights, and positive duties tend to correspond with economic, social, and
cultural rights. In practice, however, this supposed relationship frequently breaks
down. For example, as the various power-sharing agreements demonstrate, the
(negative) right to life may in fact require positive policing measures; similarly,
the (positive) right to housing may require a (negative) withdrawal of interference
to permit the right holders to return to their homes. Thus, it is empirically more
accurate to analyze each right/duty relationship within the particular context in
which it arises.
4 Operationalizing Rights
Because all rights imply duties, they are by definition embedded in social rela-
tionships. A right can only be claimed against someone other than the right hold-
er; therefore an atomistic individual in principle cannot possess rights. For this
reason, it has been suggested that Robinson Crusoe had no rights until he met
Man Friday (Benn and Peters 1964: 111). This point may appear to be unnecessar-
ily abstract, but in fact it has important practical implications: it reminds us that
all rights operate through social interaction (Donnelly 2003: 21). In other words,
in order to become actualized, rights require rules that are incorporated into ev-
eryday social and political life. For example, my right to property is meaningless
unless the judicial and police systems are prepared to enforce it against other
would-be claimants. This reality underscores the fundamental connection be-
tween legal guarantees and political practices. It also becomes the litmus test for
compliance with human and minority rights obligations as defined in the com-
plex power-sharing agreements and, indeed, elsewhere.
III Rights and Complex Power-sharing
A Rights Response to Ethnic Conflict

A rights response to ethnic conflict will include legal and quasi-legal provisions
that identify specific right holders and accord them substantive guarantees that
address their particular circumstances. A series of corresponding duties will thus
also be created and compliance procedures (judicial or political, domestic or in-
ternational) may be identified to ensure these duties are fulfilled. Such provisions
may be located at either the domestic or the international level, or a combina-
tion of the two. They may be found in different kinds of texts – constitutions,
domestic legislation, power-sharing agreements, international treaties, and de-
claratory statements. Thus, many of the power-sharing agreements incorporate

explicit human and minority rights provisions while also referring to existing in-
ternational standards (e.g., the Northern Ireland Peace Agreement, the Dayton
Agreement, the Constitutional Framework for Provisional Self-Government in
Kosovo, and the Bougainville Peace Agreement) or constitutional rights (e.g., the
Framework Agreement for Macedonia, the Law on the Special Status of Gagauz,
and the Mindanao Peace Agreement). This multi-level layering of human and mi-
nority rights affords further confirmation of the complexity of the power-sharing
arrangements studied in this project.
The different modes of establishing human and minority rights within the vari-
ous power-sharing agreements reflect the distinction between complex consocia-
tion (Northern Ireland, Bosnia, and Macedonia) and complex autonomy (Bougain-
ville, Mindanao, Moldova, and possibly Kosovo) as defined by Brendan O’Leary in
his conceptual chapter (see Ch 3). In cases of complex consociation, both human
and minority rights tend to function as forms of personal autonomy; the rights
they recognize apply to all relevant individuals regardless of their geographic loca-
tion within the state. For example, the cultural guarantees included in the 2001
Framework Agreement for Macedonia are incorporated as general amendments to
the Constitution of Macedonia and thus apply equally to all citizens of Macedonia
without distinction. In cases of complex territorial autonomy, while human rights
remain personal – and are often associated with national constitutional provisions
– minority rights are more likely to be a provision of the territorial autonomy ar-
rangements. For example, the cultural guarantees included in the Mindanao Peace
Agreement are specific to the agreement and so apply only within the Special Zone
of Peace and Development (SZOPAD) created by it. Where consociations are com-
bined with territorial autonomy (as in Northern Ireland and Bosnia and Herze-
govina), different levels of rights are often recognized, such that universal forms
of personal autonomy that apply throughout the state are supplemented by ter-
ritorially limited guarantees within the autonomous areas. For example, the 1998
Northern Ireland Peace Agreement specifically affirms the cultural guarantees of
the Framework Convention for National Minorities and the European Charter for
Minority Languages, which are applicable throughout the United Kingdom. In ad-
dition, the Northern Ireland Agreement itself recognizes supplemental cultural
guarantees specific to Northern Ireland and affirms the authority of the Northern
Ireland Assembly to draft its own human rights legislation.
As normative entitlements, all human and minority rights are intended to cre-
ate a lasting series of rules and relationships. Once accepted, they are no longer
legitimate subjects for political negotiation or compromise, and instead are de-
liberately placed beyond the usual reach of political actors. Generally speaking,
rights may only be modified or derogated from in exceptional circumstances and
according to stipulated procedures. In this sense, rights trump political interests.
In both complex consociations and complex autonomies, the aim of a rights-
based response is to recognize those normative entitlements that are considered
fundamental for the well-being of the various individuals and communities who
have been involved in or affected by the self-determination dispute. Several cat-
egories of would-be right holders can be discerned: individuals as humans, indi-

viduals as citizens, individuals as members of ethnic communities, the civic or
state community, and the various ethnic communities within its jurisdiction.
While distinct, these categories nevertheless overlap and so create a web of
rights and obligations that crosscut civic and ethnic divisions. For example, an
individual resident within the territory of the state may at one and the same time
possess rights as a human, as a citizen or member of the state/civic community,
and as a member of an ethnic community. As a result of these distinct categories
of rights, the civic or state community, the various ethnic communities within its
jurisdiction, and indeed the individuals resident within the state will all acquire
duties towards one another (Table 1).
Table 1 Overlapping and Cross-cutting Rights and Obligations

Human Citizen- Rights Duties Duties Duties Duties Duties
Rights ship of Per- to Hu- to Citi- to the to Own to Other
Rights sons mans zens State Ethnic Ethnic
Belong- Group Groups
ing to and its and
Ethnic Mem- their
Groups bers Mem-
bers
Member X X X X X X X X
Ethnic
Group
One
Member X X X X X X X X
Ethnic
Group
Two
Statei X X X
Ethnic X X X X X
Group
One
Ethnic X X X X X
Group
Two
i The state or civic community and the ethnic communities within its jurisdiction are
considered here to be distinct categories. Accordingly, all ethnic groups are by defini-
tion ‘other’ to the state.
This web of cross-cutting rights and obligations is especially important in cases
of wide-ranging autonomy where individuals or minorities may perceive them-
selves to be threatened by the newly empowered autonomous community. By

recognizing a variety of right holders, the human and minority rights provisions
aim to prevent autonomy from giving rise to ‘little tyrannies’. Individual rights
protect individual freedom from unwarranted intervention by both the civic and
ethnic communities as well as other individuals – thus providing an individual
dissenter with recourse against the ethnic group to which he belongs as well as
public authorities at both the national and sub-national level. Group rights for
ethnic communities protect cultural identities from unwarranted intervention
by both civic and other ethnic communities and their individual members – thus
providing any minorities that might exist within the autonomous area with re-
course against the dominant ethnic group there, and also against the autonomous
governmental authority itself.
The 2001 Constitutional Framework for Provisional Self-Government in Koso-
vo provides a useful example of how cross-cutting rights and obligations func-
tion within circumstances of wide-ranging autonomy. This agreement recognizes
Kosovo as an “undivided territory” under “interim international administration”
and provides for institutions of self-government at both the local or municipal and
Kosovo or regional level. Human rights and fundamental freedoms are granted
to all persons in Kosovo without discrimination on any ground (Art 3.1). Ethnic
communities and their members are granted a wide range of cultural guarantees
(including provisions for language, religion, and education) that are designed to
preserve and promote their distinct identities (Art 4.4). At the same time, the
provisional institutions of self-government are required to ensure both human
rights and fundamental freedoms (Art 3.2) and the rights of communities and
their members (Art 4.5).
Substantively, a web of overlapping and cross-cutting rights and obligations
may be the most important conflict-regulating outcome of human and minor-
ity rights provisions. Analytically, it may also represent an important distinction
between the complex power-sharing approach and both the consociational and
the integrationist perspectives. Whereas consociational strategies are designed
to produce separation and integrationist strategies are designed to produce in-
clusion, a rights strategy that recognizes human, citizen, and minority right hold-
ers – which is typical of the complex power-sharing agreements included in this
study – does both.
B ‘Minorities by Force’ and ‘Minorities by Will’

A rights response to ethnic conflict ought properly to begin with a careful con-
sideration of the particular circumstances of the communities involved in the
self-determination dispute. Without such an understanding, a rights response
may create inappropriate normative entitlements and thus produce ineffective or
even counter-productive results. Since, for the most part, such conflicts involve
an identifiable majority and an identifiable minority, the normative relationship
between these groups is particularly significant. Such an understanding is of
course complicated by the fact that minority/majority relations may shift accord-
ing to their territorial context. So, for example, the relationship between Serbia
and Kosovo (Serbian majority versus Albanian Kosovo minority) is normatively
distinct from the relationship within Kosovo (Kosovo Albanian majority versus
Serbian and Roma minorities). Thus, in all cases, relationships must be carefully
contextualized.
The normative relationship between a minority and a majority may take differ-
ent forms, the precise content of which will change according to the particulari-
ties of each case. Nevertheless, we can usefully distinguish between two contrast-
ing types of relationships: “minorities by force” and “minorities by will” (Laponce
1960: 12-13; Jackson-Preece 1998: 24-29). ‘Minorities by force’ want to be inte-
grated into the larger society, but are prevented from doing so by the dominant
group. In such cases, a distinct normative status is primarily a result of major-
ity hostility; such minorities do not want to be treated differently, instead they
want to be treated equally. The African-American or Black minority in the south-
ern United States prior to the 1960s is a classic example of a ‘minority by force’,
and accordingly their struggle was for civil rights equal to those of other (white)
Americans. In contrast, ‘minorities by will’ want recognition as a distinct group
either within or apart from their present state. In such cases, a distinct norma-
tive status is subjectively desired by the minority rather than objectively imposed
by the majority. The Quebecois minority within Canada is a classic example of a
‘minority by will’; their ongoing struggle is to become masters of their own politi-
cal life (‘maitre chez nous’), defined in terms of either a special status within the
Canadian federation or outright political independence from Canada.
The problems experienced by these two types of minorities are qualitatively
different and consequently so is the rights response required to address them.
‘Minorities by force’ tend to suffer from unsolicited efforts by the majority to
perpetuate their separation within the larger society. The appropriate rights re-
sponse to a ‘minority by force’ is equal rights (economic and social, as well as civil
and political) and non-discrimination guarantees. In contrast, ‘minorities by will’
tend to suffer from unsolicited majority efforts to deny or erode their distinctive-
ness. The appropriate response to a ‘minority by will’ is special guarantees over-
and-above the usual equal rights of citizens that are intended to preserve and
promote their distinct identity and culture.
It would be wholly inappropriate to respond to a ‘minority by force’ with provi-
sions designed to perpetuate minority distinctions, since it is unsolicited separa-
tion and discrimination by the majority that such minorities seek to overcome.
By the same token, it would be similarly inappropriate to assume that ‘minorities
by will’ require only equal rights and protection against discrimination, as such
policies do not address their desire for recognition as a distinct community. In
sum, to be effective a rights response must recognize and respond to the circum-
stances of minorities in a way that is sympathetic to their subjective needs and
aspirations.
These points may appear obvious, yet for many years international practice has
not distinguished between the normative circumstances of ‘minorities by force’
and ‘minorities by will’. During the Cold War, international actors deliberately
ignored ‘minority by will’ claims for special rights over-and-above those of equal
citizenship and non-discrimination because it was feared that such guarantees
might become pretexts for secessionist demands. Accordingly, most of the leading
international rights texts adopted during this period gave considerable emphasis
to equal rights and anti-discrimination provisions but said little or nothing about
the special rights ‘minorities by will’ assert are necessary to preserve and promote
their distinct identities and ways of life. Article 27 of the International Covenant
on Civil and Political Rights is a notable exception to this generalization, but it
nevertheless gives state signatories the freedom to determine whether or not eth-
nic groups in their jurisdictions constitute minorities. Predictably enough, many
states that possess minorities, particularly the more contentious ‘minorities by
will’, have effectively avoided their political obligations by redefining these groups
under another rubric, be it ‘immigrant’, ‘migrant’, or whatever.
The complex power-sharing agreements analyzed in this project address ‘mi-
nority by will’ circumstances. The ethnic groups identified in these agreements
want to preserve their distinct identities and cultures, and the power-sharing
arrangements are an attempt to provide the legal and political recognition that
will allow them to do precisely that. Accordingly, the substantive rights content
of these agreements tends to be disproportionately concerned with issues relat-
ing to the preservation of minority cultural distinctiveness. This tendency is in
keeping with international developments since 1989, which have placed a much
greater emphasis on cultural rights. It does not imply, however, that civil and
political or economic and social rights are irrelevant to complex power-sharing.
Far from it, as these rights are generally considered crucial both for the preserva-
tion of the existing civic or state community and for the protection of dissident or
non-conformist members of minority communities they also appear in most of
the power-sharing agreements. Finally, in those cases where the self-determina-
tion dispute has resulted in refugees or displaced persons, the right of return and
other normative entitlements relevant to such individuals are also addressed.
C Cultural Rights
Cultural rights are central to minority rights provisions because they are intended
to create circumstances that will permit the survival and development of distinct
ethnic communities within states. The basic statement of this principle is con-
tained in Article 27 of the International Covenant on Civil and Political Rights,
which stipulates that:
[I]n those states in which ethnic, religious or linguistic minorities exist, persons belong-
ing to such minorities shall not be denied the right, in community with other members
of their group, to enjoy their own culture, to profess and practice their own religion, or
to use their own language.
The Declaration on the Rights of Persons Belonging to National or Ethnic, Reli-

gious or Linguistic Minorities supplements these basic provisions with additional
rights to participate in relevant national and regional decisions, to establish and

maintain associations, and to have contact both within and across internation-
al frontiers. It also recognizes specific state measures designed to give effect to
these various normative entitlements: states are required to protect the existence
of minorities within their territories and to encourage conditions for the pro-
motion of their distinct identities through policies aimed at improving minority
language rights, increasing minority participation in the governmental process,
and creating a greater public awareness for the minority’s history, culture, and as-
pirations. Such provisions constitute a global standard of minority entitlements.
In their domestic practice, states may improve upon this minimum international
standard, but they should not go beneath it.
Taken together, the substantive content of the various complex power-sharing
agreements is very much in keeping with recent international developments in
the area of minority rights. Emphasis is clearly placed upon the legitimate inter-
est of minorities to preserve and develop their distinct national, ethnic, linguistic,
or religious identities and cultures within their existing states. Provisions for lan-
guage rights, improved minority participation in the decision-making process,
and a greater recognition and respect for minority cultures and traditions as per
the United Nations recommendations for state policy feature prominently. At
the same time, the complex power-sharing agreements also contain innovative
formulations designed to protect minority cultural and historic sites, recognize
traditional legal practices and provide a more flexible response to self-determina-
tion claims than has hitherto been the norm in international law and practice.
1 Identity
The recognition of minority identity is usually a fundamental demand of ‘minor-
ity by will’ communities. Thus, the preamble of the Law on the Special Legal
Status of Gagauz identifies the “aim of satisfying the national needs and preserv-
ing the identity of the Gagauzes”. Similarly, Article 4(a) of the Bougainville Peace
Agreement defines the “objectives of autonomy” to include the “expression and
development of the Bougainville identity.” Statements like these may be found
in every one of the substantive complex power-sharing agreements included in
this project. The Constitutional Framework for Provisional Self-Government
in Kosovo, the Framework Agreement for Macedonia, and the Mindanao Final
Agreement also include specific provisions for the development of identity. For
example, Article 48(3) of the Constitution of Macedonia, pursuant to the con-
stitutional amendments stipulated in the Framework Agreement, declares that
“members of communities have the right to establish institutions for culture, art,
I have been unable to locate any power-sharing agreement for Abkhazia. An OSCE-
led meeting to discuss power-sharing in Abkhazia took place in September 2001,
but did not result in an agreement. Human and minority rights provisions are not
included in the 1996 South Ossetia Memorandum or the Transdniestria Memo-
randum. However, these documents are arguably better understood as preliminary
peace agreements rather than full power-sharing agreements.
science and education, as well as for scholarly and other associations for the ex-
pression, fostering and development of their identity.” Article 7.1 of the Frame-
work Agreement further extends this general guarantee, such that “local authori-
ties will be free to place on front of local public buildings emblems marking the
identity of the community.”
2 Language
Language rights are increasingly recognized within international organizations as
crucial for the protection of minorities. For example, in Europe, a dedicated Char-
ter for Regional or Minority Languages has been developed under the auspices
of the Council of Europe. In keeping with this trend, language provisions may
be found in the Northern Ireland Peace Agreement, the Constitutional Frame-
work for Provisional Self-Government in Kosovo, the Framework Agreement for
Macedonia, the Law on the Special Legal Status of Gagauz, and the Mindanao
Final Agreement. Such provisions may take the form of language rights assigned
to the minority or of language requirements assigned to the state. For example,
Article 4.4 of the Constitutional Framework for Provisional Self-Government in
Kosovo grants “communities and their members” the right to “use their language
and alphabets freely, including before the courts, agencies and other bodies in
Kosovo,” “receive education in their own language,” and “enjoy access to informa-
tion in their own language.” In contrast, Article 4 of the chapter on Economic,
Social and Cultural Rights in the Northern Ireland Peace Agreement does not
recognize right holders as such, but instead requires the British government to
take “resolute action to promote the [Irish] language” by encouraging the “use of
the language in public and in private life,” removing restrictions that would “dis-
courage or work against the maintenance or development of the language,” mak-
ing provisions for “liasing with the Irish language community,” placing a statu-
tory duty on the Department of Education to “facilitate Irish medium education,”
explore the scope for achieving more widespread availability of Irish language
broadcasting; and provide financial support for Irish language film and television
production.
3 Religion
Religion, like language, is another defining feature for many minority groups. Ac-
cordingly, guarantees for religious minorities within states have featured promi-
nently in international treaties for many centuries, far predating the emergence
of a universal rights discourse. The rights of minority religions and their mem-
bers are recognized in several of the complex power-sharing agreements. As one
might expect, these all involve conflicts in which religion is a key characteristic
distinguishing the minority from the majority. For example, the Mindanao Fi-
nal Agreement stipulates that “Muslim culture, mores, customs and traditions …
shall be preserved through the regular and special schools in the Autonomous
Region” and provides government funding for traditional madrasah schools.
Likewise, Article 19(4) of the Constitution of Macedonia, pursuant to the consti-
tutional amendments stipulated in the Framework Agreement, gives the “Islamic
Religious Community in Macedonia, the Catholic Church, and other Religious

Communities and groups” the freedom to “establish schools and other social and
charitable institutions.” Finally, Article 4.4 of the Constitutional Framework for
Provisional Self-Government in Kosovo recognizes the right of “all communi-
ties and their members” to “operate religious institutions” and “preserve religious
sites.” It is something of a curiosity that neither the Dayton Peace Agreement nor
the Northern Ireland Peace Agreement contains explicit provisions for minority
religions – although both do provide general individual guarantees for freedom
of religion or belief.
4 Traditional Law and Custom

The recognition of traditional law and custom is an interesting feature of the
Mindanao Final Agreement, the Bougainville Peace Agreement, and the Inter-
im Agreement for Peace and Self-Government in Kosovo (which preceded the
Constitutional Framework for the Provision of Self-Government). The Mindan-
ao Final Agreement provides for an Islamic Banking Unit within the Philippines
Central Bank to be staffed by qualified Islamic banking experts (Ch D). It also
empowers the regional legislative assembly to establish Shari’a Courts (Ch E).
The Bougainville Peace Agreement incorporates provisions for the integration of
traditional law and custom into the judicial system. Article 129(a) recognizes the
“aspirations of the Bougainvilleans for the integration of custom and introduced
law.” Article 128 establishes a:
[c]ommission to examine and report on the issues that would be involved in giving the
autonomous Bougainville Government power to make laws permitting courts or Coun-
cils of Elders to require clan-groups to which persons convicted of criminal offences
belong to meet customary, non-custodial obligations.
Article VII(4) of the Interim Agreement for Kosovo gives “national communi-
ties acting through their democratically elected institutions” the right to “protect
national traditions on family law” by “establishing their own rules with respect to
inheritance, family and matrimonial relations, tutorship and adoption.” Require-
ments of this kind are not included in either Article 27 of the International Cove-
nant on Civil and Political Rights or the UN Declaration on the Rights of Persons
Belonging to Minorities. However, Article 8 of ILO Convention 169 on Indig-
enous and Tribal Populations requires states to take indigenous and tribal custom
and customary law into account when applying national laws and regulations to
the peoples concerned. In the Bougainville case, the similarity with Convention
169 may be more than coincidental, since the Bougainville Freedom Movement
appears to identify itself with the global indigenous peoples movement.
5 Cultural and Historic Sites

The protection of minority cultural and historic sites is a distinctive characteristic
of the Dayton Peace Agreement, the Constitutional Framework for Provision of
Self-Government in Kosovo, and the Framework Agreement for Macedonia. An-
nex 8 of the Dayton Peace Agreement is explicitly dedicated to the protection of

national monuments of this kind. Article 4.4 of the Constitutional Framework for
Provision of Self-Government in Kosovo grants communities and their members
the right to “preserve sites of religious, historical or cultural importance of the
Community.” Similarly, Article 56(2) of the Constitution of Macedonia pursuant
to the constitutional amendments stipulated in the Framework Agreement guar-
antees the “protection, promotion and enhancement of the historical and artistic
heritage of Macedonia and all communities in Macedonia and the treasures of
which it is composed, regardless of their legal status.” The destruction of cultural
and historic sites was a salient feature of the ethnic cleansing experienced in the
former Yugoslavia, which may explain why the explicit protection of such sites is
thus far unique to this region.
6 Self-determination
All the cultural rights discussed thus far are compatible with the sovereignty and
territorial integrity of existing states. That is hardly surprising given that the cen-
tral premise behind recent international activity in this area is that minorities
who are recognized and supported by the state are far less likely to challenge its
authority or threaten its territorial integrity. International efforts to protect mi-
norities aim to prevent ethnic conflict by encouraging domestic circumstances in
which the language, culture, religion, and way of life of all ethno-cultural com-
munities can be preserved and promoted within existing borders; they are not
intended to further minority claims for self-determination. Indeed, the leading
international statements on minority rights, including the UN Declaration on
the Rights of Persons Belonging to National Minorities, the Council of Europe
Framework Convention for the Protection of National Minorities, and the Or-
ganization for Security and Co-operation in Europe’s Copenhagen Document
all specifically affirm the sovereignty, territorial integrity, and political indepen-
dence of existing states. This practice is also apparent in several of the complex
power-sharing agreements, which confirm the right of sovereignty held by exist-
ing states, presumably with a view to precluding putative claims to self-deter-
mination in the form of outright political independence. Such is the case with
respect to Bosnia and the Dayton Peace Agreement, Serbia-Montenegro and the
Constitutional Framework for Provision of Self-Government in Kosovo, Mace-
donia and the Framework Agreement, and the Philippines and the Mindanao
Final Agreement.
It is therefore rather remarkable that three of the agreements specifically rec-
ognize a minority right to self-determination understood as secession, albeit only
under specified circumstances. Article 1(4) of the Law on the Special Legal Status
of Gagauzia affirms that in the event of a change of status of the Republic of Mol-
dova, the “people of Gagauz shall have the right of external self-determination.”
Similarly, Article 1(i) of the Chapter on Constitutional Issues in the Northern
Ireland Peace Agreement recognizes the “legitimacy of whatever choice is freely
exercised by a majority of the people of Northern Ireland with regard to its (po-
litical) status.” Article 1(ii) further adds that “it is for the people of the island of
Ireland alone … to exercise their right of self-determination on the basis of con-

sent.” Even more explicitly, Article 2 of the introduction to the Bougainville Peace
Agreement provides for the “right … for a referendum among Bougainvilleans
on Bougainville’s future political status,” including “separate independence for
Bougainville.” These statements represent a considerable departure from previ-
ous international practice in the area of minority rights and self-determination.
Since 1945, international law and practice have generally interpreted self-deter-
mination as a right to form separate states. As a result, in order to preclude wide-
spread claims for secession, the right to self-determination has been limited to a
very narrow range of right holders: (1) mandated territories, trust territories, and
other territories considered to be self-governing according to Chapter XI of the
UN Charter (most recently Palau in 1994, which was the last remaining United
Nations trust territory); (2) distinct jurisdictions subject to carence de souveraine-
té (e.g., Bangladesh); (3) territories in which self-determination is agreed through
democratic negotiation or plebiscite (e.g., Slovakia and the Czech Republic); (4)
highest level constituent units of a federal state in the process of dissolution or
break-up according to the principle of uti posseditis iuris (e.g., Slovenia, Croa-
tia, Bosnia and Herzegovina, Macedonia, and Serbia-Montenegro); (5) and for-
merly independent entities reasserting their independence with at least the tacit
consent of the established state in which incorporation was either illegal or of
questionable legality (e.g., Latvia, Lithuania, and Estonia) (Kingsbury 1992: 487).
Accordingly, provisions for self-determination in the form of independent state-
hood are not ordinarily included in minority rights agreements and instead are
much more likely to be explicitly denied. For similar reasons, collective minority
right holders are not usually recognized. Instead, minority provisions tend to be
formulated in terms of individual right holders, who may exercise their rights in
community with other members.
Recently, however, it has been suggested that new developments directed at
national minorities and indigenous peoples may be modifying the substance of
self-determination to include internal arrangements for autonomy or self-gov-
ernment that fall short of separate statehood. For example, Section IV(35) of the
Copenhagen Document contains a recommendation for “appropriate local or au-
tonomous administrations corresponding to the specific historical and territorial
circumstances of … minorities.” The draft national minorities protocol to the Eu-
ropean Convention on Human Rights as proposed by the Parliamentary Assem-
bly of the Council of Europe in 1993 would have gone much further in recogniz-
ing a minority right to appropriate local or autonomous authorities or a special
status matching their specific historical and territorial situation. Such substantive
measures would in effect expand the category of self-determination right holders
to accommodate sub-state groups and their claims for recognition.
It should by now be evident that the provision of internal autonomy for sub-
state groups is a primary objective of complex power-sharing. Thus, the practice
of complex power-sharing would seem to substantiate the emergence of a collec-
tive right to internal self-determination for ‘minorities by will’. The recognition of
collective right holders is another defining feature of the complex power-sharing
agreements, which adds further credence to this point of view. A quick glance at
the substantive provisions for identity, language, and religion confirms this. No-
tice the references to “the people of Northern Ireland,” “national communities”
in Kosovo, “entities” in Bosnia, “the Gagauz People” in Moldova, “Bougainville-
ans,” “the Bangsamoro People” in Mindanao, and so forth. Moreover, many of the
rights are formulated in terms of “communities” or “communities and their mem-
bers” rather than the usual “individuals … in community with other members”
phraseology commonly employed by the United Nations, the Council of Europe,
and the Organization for Security and Co-operation in Europe.
In sum, the cultural rights provisions of the complex power-sharing agree-
ments are fundamentally concerned with the recognition of groups per se and
their collective normative entitlements, as well as those of their individual mem-
bers. Arguably the justification for these cultural rights is an understanding of
self-determination, defined both as an internal as well as an external practice. In
this respect, complex power-sharing is a novel response to self-determination
disputes precisely because it recognizes rather than denies the self-determina-
tion claims of sub-state groups while still preserving the principle of territorial
integrity upon which the post-1945 international order has been constructed.
These developments would seem to offer a new perspective on the traditional
understanding of the relationship between sovereignty, self-determination, and
democracy. John Stuart Mill famously remarked in his 1861 treatise On Represen-
tative Government that “free institutions are next to impossible in a country made
up of different nationalities” (Mill 1865: 392-393). For more than a century, that
presumption has mitigated against the recognition of ‘minority by will’ claims for
a distinct normative status appropriate to their cultural circumstances and politi-
cal aspirations. Should the practice of complex power-sharing prove successful,
Mill’s argument may no longer be so persuasive.
D Civil and Political Rights

Civil and political guarantees are usually associated with the ideal of equal citi-
zenship in a democratic system of government. Insofar as divided societies are
concerned, such provisions are intended to counter Mill’s claim about the viabil-
ity of democratic institutions by fostering a common civic identity amongst the
population of the state, regardless of any ethnic or cultural differences that may
exist. Civil and political rights therefore define the normative relationship be-
tween the state and the individual in terms of citizenship or residency rather than
membership of a particular ethnic community. Unfortunately, for these reasons,
a crude interpretation of civil and political liberties has often become the pretext
for assimilationist policies designed to reduce, deny, or erode the cultural distinc-
tiveness of ‘minorities by will’ – which is precisely why such groups usually seek
additional cultural guarantees. Despite these abuses, civil and political rights are
nevertheless intended to constrain the power of the state and ensure that its insti-
tutions and officials are limited and accountable rather than oppressive and dic-
tatorial. This objective is beneficial for all members of society regardless of their
ethnic attachments. At the same time, because civil and political liberties create
reciprocal rights and duties between all the individuals within the state, they also
afford important protection to dissident members of ethnic groups. Just as the
state must respect the liberty of individuals, so too must associated individuals,
including ethnic communities.
Complex power-sharing is arguably intended to transform autonomous mi-
norities into cooperative minorities, consequently civil and political rights are
incorporated into all the substantive agreements. This incorporation, however,
may take different forms. The Northern Ireland Peace Agreement and the Dayton
Peace Agreement include specific provisions for civil and political liberties, as
well as specific references to relevant international human rights texts such as the
European Convention on Human Rights and the Framework Convention on Na-
tional Minorities. The Transdniestria Memorandum and the Bougainville Peace
Agreement include general references to both international human rights stan-
dards and domestic constitutions, but no specific statements of civil and political
rights per se. The Framework Agreement for Macedonia, the Law on the Special
Legal Status of Gagauzia, and the Mindanao Final Agreement only include gen-
eral references to domestic constitutions. Whether or not these different forms
of incorporation have any effect upon the relative success or failure of democratic
government is worthy of further consideration.
E Economic and Social Rights

Economic and social rights are usually associated with the ideals of equality of
opportunity and basic human needs. As with civil and political rights, these pro-
visions are also intended to preserve and promote an overarching civic or state
community, but their normative focus is directed at economic rather than politi-
cal relationships. In other words, they are designed to ensure that all citizens or
residents of the state have equal access to economic opportunities and social wel-
fare benefits. While economic and social rights do feature in all of the substantive
power-sharing agreements, only the principle of non-discrimination is univer-
sally affirmed – perhaps because the allegation of discrimination on the part of
the dominant or majority group is often a significant ‘minority by will’ grievance,
even if it does not define their normative circumstances. Needless to say, the
resolution of self-determination disputes is not facilitated by the existence, real
or imagined, of social and economic inequalities. These considerations make pro-
visions for non-discrimination an essential ingredient of complex power-sharing.
In other respects, however, there is tremendous variety in both the scope and
content of economic and social rights across the cases.
One of the striking features of this variation – and perhaps also its explanation
– is the extent to which economic and social provisions reflect considerations
specific to the self-determination disputes. For example, in the Northern Ireland
Peace Agreement, the right to “freely choose one’s place of residence” and to be
“free from sectarian harassment” is recognized (Art 1 of the chapter on Rights,
Safeguards and Equality of Opportunity). These provisions are arguably a direct
response to the experience of segregated Protestant and Catholic residential

areas, which is a feature of the sectarian conflict. Similarly, Article 2(ii) of the
chapter on Economic, Social and Cultural issues requires the British government
to “review the national security aspects of the present fair employment legisla-
tion at the earliest possible time” – the salient point here being that “nation-
al security” has been viewed as a pretext for Protestant discrimination against
Catholics. Other examples of case-specific economic and social provisions may
be found in the Mindanao Final Agreement and the Bougainville Peace Agree-
ment. Since claims to economic exploitation featured in both disputes, each of
these agreements includes special provisions intended to address this perceived
inequality. In the Mindanao Final Agreement, special attention is given to the
problem of economic under-development within the autonomous zone and to
equal opportunities for its inhabitants within the Philippines national police and
the civil service. In the Bougainville Peace Agreement, the right of autonomy for
Bougainville includes the “power to decide on foreign investment applications
for Bougainville” (Art 53). This provision is arguably a direct response to allega-
tions of “plundering” and “unfair distribution” of wealth and opportunities asso-
ciated with mining activities by international companies like Rio Tinto Zinc Ltd.
– which was a significant causal factor behind the Bougainville independence
movement (Regan 2003: 6-7).
F Refugees and Displaced Persons

A further category of rights aimed at refugees and displaced persons is also in-
cluded in those complex power-sharing agreements that address cases in which
extensive population displacement occurred as a result of ethnic conflict: Bosnia,
Kosovo, and Macedonia. Refugees and displaced persons are in such situations
the logical consequence of a prior failure to respect the rights of minority groups
such that their continued existence within their traditional territories is no longer
possible. Moreover, that failure is threefold: a failure to respect their equal rights
as citizens; a failure to respect their special rights as minorities; and a failure to
respect their universal human rights.
Since the end of the Second World War, international law and practice has
developed a series of normative entitlements designed to address the distinc-
tive circumstances of refugees. Refugees thus possess the right not to be refused
admission by states unless such refusal is dictated by national security or public
order (sometimes referred to as the principle of non-refoulement); the right of
asylum (subject to the domestic procedures of the state concerned); and the right
of international protection by the Office of the United Nations High Commis-
sioner for Refugees.
In practice, however, the number of potential right holders is limited by the in-
ternationally accepted definition of a refugee. According to the Geneva Conven-
tion on Refugees (1951) and its additional protocol (1967), a refugee is someone
who:
[O]wing to a well-founded fear of being persecuted for reasons of race, religion, na-
tionality, membership of a particular social group or political opinion, is outside the
country of his nationality and is unwilling to avail himself of the protection of that
country or who, not having a nationality and being outside the country of his former
habitual residence as a result of such events, is unable or owing to such fear unwilling
to return to it.
This definition excludes those individuals who have been displaced for bona fide
reasons of fear within the boundaries of their state, but have not crossed an in-
ternational frontier. Thus, under the current international system, internally dis-
placed populations have far fewer normative entitlements than internationally
displaced populations. This normative inequality is justified according to realist
criteria – displaced populations who cross international frontiers are considered
to be greater threats to international peace and stability than those who remain
within. Needless to say, such a distinction is highly questionable in terms of hu-
man and minority rights criteria. Indeed, populations that remain within states
characterized by ethnic conflict may be much more likely to have these rights
infringed than those who leave.
Recently, realist and humanitarian priorities have combined to focus much
greater attention on the issues of prevention and solution of refugee crises in
those states where they exist or are deemed likely to arise (for instance, because
of a prior history of ethnic conflict). Ultimately, the problem of refugees can only
be overcome by eliminating the circumstances that create refugees in the first
place, in other words, by ensuring domestic practices of good governance and
compliance with international standards on human and minority rights. This
conclusion immediately brings us back to the role of complex power-sharing as
a way of resolving self-determination disputes and in so doing both solving and
preventing refugee crises.
Within all those complex power-sharing agreements that address refugee
problems, the right of return features prominently. However, the number of pos-
sible right holders has the potential for significant variation as is seen between
the Dayton Peace Agreement and the Constitutional Framework for Provision
of Self-Government in Kosovo, on the one hand, and the Final Agreement for
Macedonia, on the other. Whereas the former recognize “all persons” (Bosnia) or
“all refugees and displaced persons” (Kosovo), the Final Agreement for Macedo-
nia restricts this right to “citizens and legal residences of Macedonia.” Presumably
this more-qualified interpretation is designed to prevent claims of asylum (from
Kosovar Albanians) being misrepresented as claims of return. Nevertheless, it
opens up the possibility that the laws of citizenship and residence may be ma-
nipulated or contrived by the Macedonian authorities in order to obstruct the
otherwise legitimate return of individuals they consider undesirable.
Supplemental to the general right of return, other provisions are also recog-
nized with a view to giving practical effect to this basic provision. Such provisions
are most numerous in the Dayton Peace Agreement, which dedicates Annex 7
to refugees and displaced persons. This annex not only guarantees refugees the
right “freely to return to their homes of origin” but also the “right to have restored
to them the property of which they were deprived since 1991” (Art I(1)), the abil-
ity to “return in safety” “without risk of harassment, intimidation, persecution or
discrimination” (Art I(2)), immediate access to international humanitarian orga-
nizations and monitors (Art I(3)), and their choice of destination (Art I(4)). At
the same time, the Dayton Peace Agreement also outlines the responsibilities of
the Republic of Bosnia and Herzegovina, the Federation of Bosnia and Herze-
govina and the Republika Srpska towards returning refugees and displaced per-
sons. These responsibilities include the repeal of discriminatory legislation and
administrative practices, the prevention and prompt suppression of incitement
against returning refugees and displaced persons, and the prosecution, dismissal,
or transfer of public officials responsible for serious violations of these rights and
the rights of persons belonging to ethnic or minority groups (Art I(3)). Similar,
albeit less elaborate, requirements for government action are also contained in
Article 3.4 of the Constitutional Framework for Provision of Self-Government
in Kosovo and in Article 3 of Annex C to the Framework Agreement for Mace-
donia.
G Special Provisions
It is worth noting that some of the power-sharing agreements contain special
rights provisions directly associated with the conflict resolution process. For ex-
ample, Article 12 of the Chapter on Rights, Safeguards and Equality of Opportuni-
ty in the Northern Ireland Peace Agreement recognizes that “victims of violence”
related to the self-determination conflict have the “right to remember as well as
to contribute to a changed society.” Article 13 goes on to affirm the importance of
the reconciliation process to the “promotion of a culture of tolerance” conducive
to the respect for human and minority rights. Similarly, rights of amnesty and
pardon for those directly involved in the self-determination disputes are incor-
porated into both the Bougainville Peace Agreement and the Interim Agreement
for Peace and Self-Government in Kosovo (which preceded the Constitutional
Framework for the Provision of Self-Government). Article 331 of the Bougainville
Peace Agreement grants a general amnesty and pardon (as agreed in the previ-
ous Lincoln Agreement) “for all persons involved in conflict-related activities”
or “convicted of offences arising out of conflict-related activities.” A comparable
provision may be found in Article II(10) of the Interim Agreement on Kosovo,
which grants a right of release to all “persons held in connection with the con-
flict” as well as all “abducted persons or other persons held without charge.”
H Enforcement Measures
The recognition of human and minority rights standards is, of course, only the
first step in implementing a rights response to ethnic conflict. In order to be ef-
fective as an instrument of conflict prevention, these rules must be incorporated
into everyday social and political life. For that to happen, three essential condi-
tions need to be satisfied:
1) right holders must have some means of making complaints against alleged
infractions;
2) specific institutions must exist to receive and investigate complaints, and
make and enforce authoritative judgements on their merits; and
3) the courts, agencies, governmental institutions, and other public institu-
tions must be prepared to comply with these standards and any judgements
relating to them.
It is, however, important to stress that while these conditions may be necessary
for the effectiveness of human and minority rights, they are not sufficient to guar-
antee full compliance with these norms. This point will be discussed more fully in
the Assessing Compliance section.
All of the substantive complex power-sharing agreements included in this
project contain provisions that are designed to ensure compliance on these basic
points. In each case, the judicial system is recognized as having competence to
address issues arising from the agreements, including allegations of non-compli-
ance with human and minority rights on the part of public authorities. In order
to ensure that these judicial mechanisms will be fair and impartial towards all
individuals and communities, the agreements regarding Bosnia, Kosovo, Mace-
donia, and Mindanao incorporate requirements for minority representation on
the national courts of final appeal.
In addition, several of the power-sharing agreements create dedicated instru-
ments for the enforcement of human and minority rights. For example, both the
Constitutional Framework for Provision of Self-Government in Kosovo and the
Framework Agreement for Macedonia provide for an ombudsperson to receive
and investigate complaints against public bodies, monitor, and take preventa-
tive action. Thus, “natural and legal persons in Kosovo” have the right to make
complaints to the ombudsperson concerning “human rights violations or actions
constituting abuse of authority by any public authority in Kosovo” (Ch 10.1). The
Kosovo ombudsperson is expected to “give particular priority” to allegations of
“especially severe or systematic violations,” discrimination including “discrimina-
tion against communities and their members,” and “allegations of violations of
rights of communities and their members” (Ch 10.3). Similarly, Article 77 of the
Constitution of Macedonia, pursuant to the constitutional amendments stipu-
lated in the Framework Agreement, creates a public attorney (somewhat confus-
ingly referred to as an “ombudsperson” in the Framework Agreement) to protect
the constitutional and legal rights of citizens and, in particular, to safeguard the
principles of non-discrimination and equitable representation of communities.
Similarly, both the Dayton Peace Agreement for Bosnia and the Northern Ire-
land Peace Agreement create human rights commissions to perform a role of
investigation and monitoring. In Bosnia, this includes both an ombudsperson (to
be appointed by the OSCE in consultation with the various domestic entities) as
well as a Human Rights Chamber (with a set proportion of members appointed
by each of the entities and the Committee of Ministers of the Council of Europe).
In this arrangement, the ombudsperson has an investigatory and monitoring
role, while the Human Rights Chamber provides authoritative judgements. The
Northern Ireland Human Rights Commission can also investigate, monitor and
advise on infringements, and may bring proceedings to UK courts or provide as-
sistance to individuals doing so.
The Kosovo, Macedonia, and Northern Ireland agreements also provide for
consultative bodies with a remit that includes responsibility for human and mi-
nority rights. In Kosovo, the Committee on Rights and Interests of Communities
within the Kosovo Assembly is intended to comment on proposed laws with a
view to ensuring that “community rights and interests are adequately addressed”
and may also propose legislation of its own with a view to addressing the “con-
cerns of communities” (Art 9.1.14). In Macedonia, the Committee on Inter-Com-
munity Relations within the Macedonian Assembly has a similar function. It is
intended to resolve disputes regarding the representation of communities and
their interests. In a slightly different approach, Article 10 of the Chapter on New
Institutes in the Northern Ireland Peace Agreement “envisages” the creation of
a Joint Committee of representatives of “the two Human Rights Commissions,
North and South” as a “forum for consideration of human rights issues in the
island of Ireland.”
All the enforcement measures thus far discussed are located within the do-
mestic governmental structure of the states concerned. However, some complex
power-sharing agreements also incorporate – either directly or by implication –
international instruments. The Northern Ireland and Bosnian agreements make
specific reference to the enforcement provisions of the European Convention on
Human Rights. Likewise, the Kosovo and Macedonian agreements acknowledge
the importance of OSCE monitoring instruments. Moreover, by implication, all
of the signatories to the European Convention on Human Rights, the Frame-
work Convention on National Minorities, and the European Charter on Regional
or Minority Languages are subject to the enforcement instruments included in
these international standard-setting documents. For the ECHR, this enforcement
involves the Court of Human Rights at Strasbourg, for the Convention and Char-
ter it includes a combination of state self-reporting and international monitoring.
Similarly, all members of the OSCE are subject to the monitoring activities of the
High Commissioner on National Minorities and the Office for Democratic In-
stitutions and Human Rights. The location of international human and minority
rights and enforcement mechanisms applicable to each self-determination dis-
pute is summarized in the appendix.
I Assessing Compliance
Assessing the relative compliance of specific cases with human and minority
rights as outlined in the relevant standard-setting agreements is not an easy task.
The quantity and quality of empirical evidence varies enormously from one self-
determination dispute to another. At one extreme, there is no specific human
and minority rights information on Transdniestria and Bougainville; at the other

extreme, the documentation on Bosnia is so immense as to be approaching infor-
mation overload. These unavoidable evidentiary constraints are compounded by
the need to compare a variety of sources (internal versus external, governmental
versus nongovernmental) in order to obtain a balanced perspective. Neverthe-
less, based on the available evidence, it is possible to construct a taxonomy of
compliance (Table 2). A summary of the compliance indicators used to make this
assessment is included in the appendix.
Table 2 State Compliance with Human and Minority Rights

Full compli- Formal but not Partial compli- Non-compli- Insufficient in-
ance fully substan- ance ance formation for
tive assessment
N. Ireland Bosnia Kosovo Transdniestria
Macedonia Mindanao Bougainville
Gagauzia
1 Full Compliance
Full compliance denotes circumstances in which human and minority rights
standards are incorporated into everyday social and political life (e.g., Canada,
Finland, Denmark, Sweden, etc.). Full compliance would include the following:
• International and domestic, governmental and nongovernmental indicators
confirm that all courts, agencies, governmental institutions, and other pub-
lic institutions fully conform to these standards and any judgements relating
to them.
• Legislation and public policies are consistently formulated with reference to
human and minority rights norms.
• Remedial action recommended by the courts or other advisory bodies is
quickly and effectively implemented.
It is perhaps only to be expected that none of the self-determination disputes

included in this study are currently examples of full compliance as in all cases
the power-sharing arrangements are of relatively recent origin. As a result, the
rules they recognize (including human and minority rights standards) are as yet
unlikely to have become embedded in social and political practices.
2 Formal but Not Fully Substantive Compliance

Formal but not fully substantive compliance describes those circumstances in
which all legal requirements in the area of human and minority rights are ad-
hered to but there is nevertheless some reluctance on the part of certain public
Abkhazia and South Ossetia are not listed here because they arguably do not possess
substantive power-sharing agreements, and the preliminary peace agreements that
do exist fail to include human and minority rights provisions.
institutions or actors to review or amend policies or practices that may not be

fully in keeping with these principles. Thus, while the British government has
demonstrated a strong commitment to human and minority rights principles (as
disclosed in the recent domestication of the European Convention on Human
Rights), some public institutions within Northern Ireland continue to operate
in ways that threaten to undermine this commitment. The 2002 Report of the
Northern Ireland Human Rights Commission, for example, expresses concern
over the possible collusion between security forces and paramilitaries. Similarly,
the 2001 Human Rights Watch Report draws attention to what may be a practice
of impunity for police abuse. All local authorities in Northern Ireland must bring
their practices in line with human and minority rights norms before a condition
of full compliance can be achieved.
3 Partial Compliance
In circumstances of partial compliance, there is some evidence to suggest that
public institutions and actors are making efforts to comply with human and mi-
nority rights standards, but the practical effects of this effort remain both limited
and variable. The current state of affairs in Bosnia is representative of such situ-
ations. According to the Annual Report of the Bosnian Human Rights Chamber
(2000), the implementation of its judgements by local authorities was improving,
but remained uneven across the country. Similarly, recent reports by both the
Human Rights Coordination Centre within the Office of the High Representative
(2001) and the United Nations High Commissioner for Refugees (2003) note a
significant increase in the rate of refugee return, but suggest that sustainability
is problematic due to return-related violence. Such mistreatment of refugees is
subject to regional variability and tends to be much worse in the Republika Srp-
ska. The documentation for Macedonia and Gagauzia presents a similarly mixed
picture. While by 2002 most of the 170,000 displaced persons within Macedonia
had returned home (UNHCR 2002), human rights in Macedonia remain at best
“patchy” (Human Rights Watch 2001; UNHCR 2002) and ill-treatment of ethnic
minorities by public authorities persists (Amnesty International 2003). Likewise,
in Gagauzia progress has been made in the area of refugees, but institutional
support for returnees is nevertheless inadequate (UNHCR 2002) and Moldovan
support for Gagauz autonomy could be improved (Council of Europe 2000).
4 Non-compliance
Non-compliance describes a general failure on the part of public institutions and
actors to respect human and minority rights and a correspondingly high inci-
dence of violations. There is only limited information available on Mindanao,
but what does exist is suggestive of non-compliance. The December 2000 Re-
port of the Philippines Task Force for Detainees notes a substantial increase in
human rights violations in Mindanao. The 2003 Amnesty International Report
describes escalating violence between government forces and the Moro Libera-
tion Front that threatens the safety of the civilian population. Much more data
exists for Kosovo, all of it highly critical of the human and minority rights situ-
ation. The 2002 Report of the Human Rights Ombudsperson for Kosovo argues
that the United Nations Interim Administration Mission in Kosovo (UNMIK)
is not structured according to democratic principles, does not function in ac-
cordance with the rule of law, and does not respect human rights norms. It notes
that ethnic minorities remain isolated in ghettos and subject to violent attack,
and describes refugee returns as a “distant goal.” In a similar vein, the 2003 United
Nations High Commissioner on Human Rights characterizes the local treatment
of ethnic minorities as “unacceptable” and questions the quality of the local ju-
diciary. In sum, human and minority rights are largely an abstraction in both
Mindanao and Kosovo.
J Predicting Success or Failure

A final typological distinction may explain why some self-determination disputes
are more receptive to a power-sharing resolution than others. Among ‘minori-
ties by will’, two additional types may be distinguished: minorities who, while
refusing assimilation, are prepared to cooperate with the state; and minorities
who reject cooperation because their objective is outright political independence
(Laponce 1960: 12). Cooperative minorities are fully satisfied with a special legal
status within the state (e.g., the Swedes in Finland or the Danes in Germany);
separatist minorities refuse to accept any legal status short of outright political
independence (e.g., Slovenes and Croats in Yugoslavia during 1991). Between the
cooperative and the separatist minorities, we may add an intermediary (albeit
less clear-cut) type: the autonomous minority who would prefer independence
but may be prepared to cooperate in exchange for self-government (e.g., the Que-
becois in Canada).
Complex power-sharing agreements are arguably intended for this intermedi-
ary category – the autonomous minority – with a view to encouraging a politics of
accommodation that might eventually create a fully cooperative situation. When
they are inappropriately applied to avowedly separatist minorities, the results are
unlikely to be successful. Internal self-determination is still a poor consolation
prize for those minorities who aspire to full political independence. This argu-
ably explains the non-compliance evident in Kosovo and Mindanao, and perhaps
also the partial compliance apparent in Bosnia. By all accounts, the Kosovo Al-
banians, the Moros of Mindanao, and the Serbs of Bosnia do not want to remain
minorities within their present states, and arguably it is only the persistent threat
of force that keeps them in their current position. In other words, these self-de-
termination disputes remain unresolved, notwithstanding the complex power-
sharing agreements intended to ameliorate them.
IV Summary and Conclusion

Human and minority rights play an important role in complex power-sharing
agreements. They complement the political arrangements for the accommoda-
tion of divergent interests with the legal recognition of distinct normative entitle-
ments. These entitlements identify a variety of right holders – humans, citizens,

members of ethnic communities, the civic or state community, and the various
ethnic communities within its jurisdiction – and provide each of them with sub-
stantive guarantees that address their particular circumstances. While distinct,
these categories are nevertheless also overlapping and so create a web of rights
and obligations that crosscut civic and ethnic divisions. Such provisions establish
a lasting set of rules and relationships in which competing normative claims may
be articulated and resolved.
The efficacy of a rights response to ethnic conflict is fundamentally rooted in its
normative content precisely because the language of such disputes is overwhelm-
ingly a normative discourse: sovereignty, self-determination, freedom, exploita-
tion, discrimination, justice and injustice, loyalty and disloyalty are all normative
concepts. The parties to such disputes consider themselves to be morally as well
as politically aggrieved. Indeed, it is their moral claims that infuse their political
aspirations with passion and authority such that people are prepared to fight and
die for them. In other words, these conflicts are embedded in moral principles
and therefore any response to them must itself include a clear moral message.
For these reasons, complex power-sharing agreements must be widely seen and
understood as just settlements. The human and minority rights discourse has es-
tablished itself as a globally recognized source of moral authority. Consequently,
human and minority rights provisions are a highly effective way of ensuring that
complex power-sharing arrangements include an immediately identifiable moral
content so that justice can indeed be seen to be done.
Needless to say, where justice is not seen to be done, complex power-sharing
is unlikely to be successful. Such an outcome is probable where complex power-
sharing is inappropriately applied to separatist minorities. Minorities who are
convinced that outright political independence is their fundamental right are
likely to view complex power-sharing as yet another subterfuge by those who
would exploit and subjugate them – this may explain the continuing tension be-
tween the people of Kosovo and the United Nations Interim Administration Mis-
sion, the continued fighting between the Moro Islamic Liberation Front and the
Philippine government in Mindanao, and the distinctly uncooperative behavior
of the Republika Srpska within Bosnia Herzegovina.
This conclusion points to a fundamental shortcoming in international law and
practice as applied to minorities. Self-determination remains a highly indetermi-
nate right. Both its substance and its would-be right holders are deeply contested.
The extension of this right to incorporate internal arrangements for self-govern-
ment – which is apparent in the practice of complex power-sharing – may be
a notable improvement, but it is nevertheless insufficient for resolving disputes
that involve avowedly separatist minorities, of which there are many (examples
include Kosovo, Mindanao, Republika Srpska, Ossetia, Abkhazia, Ngorno-Kara-
bakh, Chechnya, and Palestine, to name only a few). Here, John Stuart Mill’s com-
ments on representative government remain prescient:
Where the sentiment of nationality exists in any force, there is a prima facie case for
uniting all the members of the nationality under the same government, a government
to themselves apart. This is merely saying that the question of government ought to be
decided by the governed. One hardly knows what any division of the human race should
be free to do, if not to determine with which of the various collective bodies of human
beings they choose to associate themselves
(Mill 1865: 392-393).
There is, in theory at least, a way out of this conundrum consistent with interna-
tional law and practice: extend the external right of self-determination to a some-
what larger group of right holders by encouraging boundary revision through
consent. Manipulating borders to fit peoples is extremely difficult. Consequently,
in most situations, such an extension is likely to rearrange rather than remove the
problem of minorities. Nevertheless, it may facilitate a more productive minor-
ity/majority configuration and thereby improve the likelihood of successful pow-
er-sharing. In such cases, the consensual separatist option merits consideration.
Unfortunately, however, past experience would suggest that such consent will
not be forthcoming in many, if not most, self-determination disputes involving
avowedly separatist minorities. Sadly, in such cases, the possibility for violence is
likely to remain much greater than the possibility for a Czechoslovak type ‘velvet
divorce’.
I Primary Sources
A Case-specific Documentation
1 Abkhazia
Not Available
2 Bosnia Herzegovina
General Framework Agreement for Peace in Bosnia (‘The Dayton Agreement’) (1995)
Office of the Human Rights Ombudsperson Report (Bosnia and Herzegovina) (1999)
Human Rights Chamber Annual Report (Bosnia and Herzegovina) (2000)
Office of the High Representative Human Rights Priorities (Bosnia and Herzegovina)
(2002)
Human Rights Coordination Centre within the Office of the High Representative Quar-
terly Reports (2001)
3 Bougainville
The Bougainville Peace Agreement (2001)
4 Gagauzia
The Law on the Special Legal Status of Gagauz (1994)
5 Kosovo
Constitutional Framework for Provisional Self-Government in Kosovo (2001)
Interim Agreement for Peace and Self-Government in Kosovo (1999)
Kosovo Human Rights Ombudsperson Report (2002)
6 Macedonia
Framework Agreement (for Macedonia) (2001)
7 Mindanao
Mindanao Peace Agreement (1996)
8 Northern Ireland
The Northern Ireland Peace Agreement (1998)
Report of the Northern Ireland Human Rights Commission (2002)
9 South Ossetia
Memorandum on Measures to Provide Security and Strengthen Mutual Trust Between
Sides in the Georgian-Ossetian Conflict (1996)
10 Transdniestria
Memorandum on the Bases for Normalization of Relations Between the Republic of
Moldova and Transdniestria (1997)
B International Organization Reports

1 Council of Europe
United Kingdom, State Report On the Framework Convention for the Protection of Na-
tional Minorities (1999)
Council of Europe Advisory Committee On the Framework Convention for the Protection
of National Minorities, Opinion on United Kingdom (2001)
Moldova, State Report On the Framework Convention for the Protection of National Mi-
norities (2000)
Serbia-Montenegro, State Report On the Framework Convention for the Protection of
National Minorities (2002)
2 Organization for Security and Cooperation in Europe

Organization on Security and Co-operation in Europe/United High Commissioner on
Refugees, Assessment on Ethnic Minorities (Kosovo) (2002)
Organization for Security and Co-operation in Europe, Annual Activity Report (1999-
2001)
3 United Nations
United Nations High Commissioner for Refugees, Global Operations Profiles (2002)
United Nations High Commissioner for Human Rights, Field Presence Reports (2003)
C Nongovernmental Organization Reports

Amnesty International, Reports (2000-2003)
Human Rights Watch, World Report (2001)
Minorities At Risk, Assessment (2001)
Task Force for Detainees of the Philippines, Report (2000)
Un-represented Nations and Peoples Organization, Report (2002)
D Human and Minority Rights Texts

1 Global
Universal Declaration on Human Rights (1948)
International Covenant on Civil and Political Rights (1966)
International Covenant on Economic, Social and Cultural Rights (1966)
International Labor Organization Convention 169 on Indigenous and Tribal Populations
(1989)
Declaration on the Rights of Persons Belonging to National or Ethnic, Religious or Lin-
guistic Minorities (1992)
2 Regional
European Convention on Human Rights (1948)
European Social Charter (1961)
American Convention on Human Rights (1969)
African Charter on Human and People’s Rights (‘Banjul Charter’) (1981).
Additional Protocol to the American Convention on Human Rights in the Area of Eco-
nomic, Social and Cultural Rights (‘Protocol of San Salvador’) (1988).
Organization for Security and Co-operation in Europe’s Copenhagen Document (1990)
European Charter for Regional or Minority Languages (1993)
Framework Convention for the Protection of National Minorities (1995)
3 Refugee Texts
Statute of the Office of the United Nations High Commissioner for Refugees (1950)
Convention Relating to the Status of Refugees (1951)
Protocol Relating to the Status of Refugees (1967)
United Nations Declaration on Territorial Asylum (1960)
II Appendix: Data Summary Tables
A List of Abbreviations
ICCPR: International Covenant on Civil and Political Rights
ICCPR – OP1: International Covenant on Civil and Political Rights
Optional Protocol One
CERD: Convention on the Elimination of All Forms of Racial

Discrimination
ICESCR: International Covenant on Economic, Social and Cul-
tural Rights
Coe: Council of Europe
IDP: Internally Displaced Person
HCNM: High Commissioner on National Minorities
HRCC: Human Rights Coordination Centre within Office of
the High Representative, Bosnia
ECHR: European Convention on Human Rights
Framework Convention: Framework Convention for the Protection of Nation-
al Minorities
HRW: Human Rights Watch
IAC: Interim Administrative Council for Kosovo
SFOR: Stabilization Force for Bosnia
KFOR: Implementation Force for Kosovo
N/A: Not Applicable
ORC: Office of Return and Communities, within SRSG Of-
fice, Kosovo
OSCE: Organization for Security and Co-operation in Eu-
rope
RS: Republika Srpska
SRSG: Special Representative of the UN Secretary-General
(Kosovo)
MAR: Minorities At Risk Project
MILF: Moro Islamic Liberation Front
Minority Language Charter: European Charter for Regional or Minority Languages
UNCR: United Nations Convention on Refugees
UNHCHR: United Nations High Commissioner for Human
Rights
UNHCR: United Nations High Commissioner for Refugees
UNMIK: United Nations Interim Administration Mission in
Kosovo
UNPO: Un-represented Nations and Peoples Organization
658
Table 1 Location of Human and Minority Rights Applicable to Case Studies

Case studies Power-sharing Agreement Constitution OSCE ECHR Coe Coe Minor- ICCPR ICCPR-OP1 ICESCR UNCR CERD
Framework ity Language
Convention Charter
Northern Ireland, 1998 Northern Ireland Peace Proposed Bill of Rights for Yes 1950# 1995# 2000# 1976 No 1976 1954 1969
United Kingdom Agreement Northern Ireland

Kosovo, 2001 Constitutional Framework Constitutional Framework /9 Yes# No*# No*# No*# 2001# 2001# 2001 2001# 2001#
Serbia-Montenegro and 1999 Interim Agreement
Bosnia and Herzegovina 1995 Dayton Agreement Constitution of Bosnia Yes# 2002# 2000# No# 1993# 1995# 1992 1993# 1993#
Macedonia 2001 Framework Agreement Constitution of Macedonia (in- Yes# 1995 1996 1996 1994 1994 1994 1994# 1994
cluding amendments specified
in Framework agreeements)
Abkhazia, N/A N/A Yes 1999 2000 No 1994 1994 1994 1999* 1999
Georgia
South Ossetia, 1996 South Ossetia Memo- N/A Yes# 1999 2000 No 1994 1994 1994 1999* 1999
Georgia randum
Gagauzia, 1994 Law on the Special Legal Constitution of Moldova Yes 1995 1995 2002 1993 No 1993 2002 1993
Moldova Status of Gagauz
Transdniestria, 1997 Transdniestria Memo- Constitution of Moldova Yes# 1995* 1995 2002 1993 No 1993 2002 1993
Moldova randum
Bougainville, Bougainville Peace Agree- Constitution of PNG; Proposed N/A N/A N/A N/A No No No 1986* 1982*
PNG ment 2001 Bougainville Constitution
Mindanao, Mindanao Final Agreement Proposed Regional Admin. N/A (a) N/A N/A 1986 1989 1974 1981 1967
Philippines 1996 Code; National Constitution N/A
• # indicates that the human and minority rights instrument • Moldova: Moldova has placed a reservation on its acces-
was specifically referred to in the power-sharing agree- sion to the ECHR, which stipulates that the government of
ment. Moldova cannot ensure compliance within the territory of
• Kosovo: Serbia-Montenegro is not a signatory to the ECHR, Transdniestria until the conflict there has been resolved.
the Framework Convention on Minorities, or the European • Georgia: Georgia has placed a reservation on its accession
Charter for Minority Languages; Kosovo by definition can- to the UNRC, which stipulates that the convention only ap-
not be a signatory at present because it is not an independ- plies in those territories where the government of Moldova
ent state. However, the UNMIK documents specifically exercises jurisdiction pending the “full restoration of the
refer to the provisions of the ECHR, Framework Conven- territorial integrity of Georgia.”
tion and Charter for Regional and Minority Languages and • Bougainville: The Bougainville Peace Agreement does not
indicate that these shall be “directly applicable in Kosovo as specifically cite any particular international text but refer-
part of the constitutional framework.” ence is made to “internationally accepted standards of good
governance, including protection of human rights.”
21 The Role of Human and Minority Rights in Complex Power-sharing
659
660
Table 2 Minority Rights Provisions Included in Power-Sharing Agreements

Agreements Right-Holders identified Non-discrimi- Autonomy External Identity Language Religion Education Refugees Special provisions
in text nation self-determi-
nation
N. Ireland, People of N. Ireland; Com- Yes Yes Yes Yes Yes No Yes No Sectarian Harassment; Victims
1998 Northern Ireland Peace munities; Individuals Irish and (language of Violence; Freedom to Choose
Agreement Ulster Scots education) Place of Residence
Kosovo, Individuals; National Commu- Yes Yes No – pre- Yes Yes Yes Yes Yes Amnesty and Release of
Interim Agreement for Self- nities; Members of National cluded Albanian, (language Combatants; Preservation of
Government 1999, Constitu- Communities Serbian, education) Cultural Sites
tional Framework 2001 Other
Bosnia, Individuals; Entities Yes Yes No – pre- Yes No No No Yes Preservation of National Monu-
1995 Dayton Agreement cluded (of refugees) ments
Macedonia, Individuals; Citizens; Religious Yes Yes No – pre- Yes Yes Yes Yes Yes Preservation of Cultural and
2001 Framework Agreement Communities*; Members of cluded Macedonian (language Artistic Heritage
Communities and others education)
spoken by
20% of
population
Abkhazia, N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
Not Applicable*
South Ossetia, No No N/A No No No No No No No
1996 South Ossetia Memo-
randum*
Gagauzia, Individuals; Citizens; Gagauz Yes Yes Yes Yes Yes No No No No
1994 Law on the Special Legal People Moldovan,
Status of Gagauz Gagauz,
Russian
Transdniestria, 1997 Transdni- Individuals Yes No No No No No No No
estria Memorandum
Bougainville, Bougainville People of Bougainville; Yes Yes Yes Yes No No No No Integration of Tribal Customs;
Peace Agreement 2001 Individuals Amnesty and Pardon of Com-
batants; Approval of Foreign
Investment
Mindanao, Mindanao Final Muslims; Christians; Cultural Yes Yes No – Yes Yes Yes Yes No Shari’ah courts; Islamic Banking
Agreement 1996 Communities; Bangsamoro precluded Filipino Eng- (religious Law; Representation in Police
People lish Arabic education, and Civil Service
especially
Madrasa)
* Three religious communities are singled out for recogni- • Neither human nor minority rights provisions are included
tion in the 2001 Framework Agreement: the “Macedonian in the 1996 South Ossetia Memorandum, which is arguably
Orthodox Church”; the “Islamic Religious Community”; a preliminary peace agreement rather than a full power-
and the “Catholic Church.” sharing agreement.
• I have been unable to locate any power-sharing agreement
for Abkhazia. An OSCE led meeting to discuss power shar-
ing in Abkhazia took place in September 2001 but did not
result in an agreement.
661
Table 3 Summary of Human and Minority Rights Compliance Indicators 662
Cases Specific Domestic Specific International OSCE Reports Coe Framework UNHCR UNHCHR HRW MAR Amnesty International
Mechanisms Mechanisms Convention Advisory Global Operations Field Presence Reports 2001 World Report 2001 Risk Assessment Reports
Opinions Profiles 2002 and Re- 2003
turning Refugees IDPs
N. Ireland 2002 Report of the See Coe Framework N/A 2001: N/A N/A Problems: “prospects are dim 2002:
N. I. Human Rights Reviews Under-representation concern re: Anti-Ter- that violence will Amnesty welcomes
Commission of Catholics in police rorism Act; govt. end”; keys to conflict House of Lords ruling
Problems: force; segregation in stalling on inquiries “no longer lie with that the Northern
possible collusion schools and housing into murder of human signatories” Ireland Human Rights
between paramilitar- estates; minority rights lawyers; Commission can
ies and security forces: language education impunity for police intervene in cases
paramilitary violence; and broadcasting abuse before the Northern
sectarian intimida- Progress: Ireland Courts;
tion; police use of human rights reforms Concern re: failure to
plastic bullets investigate security
Progress: force collusion with
consultation on Bill loyalist death squads
of Rights; 2001 Joint during “The Troubles”
H.R. Commission
established
Kosovo Ombudsperson See UNHCR Report OSCE/UNHCR N/A Problems: Problems: Problems: Increased tension 2002:
Report 2002: 2002; Assessment on Tension and inter- Unacceptable local human rights often an between KFOR and impunity for interna-
UNMIK not struc- UNHCHR Report 2003 Ethnic Minorities 2002 ethnic violence treatment of ethnic “abstraction”; mis- local authorities tional community in
tured according to Progress: continue; refugee minorities; limited treatment of ethnic Widespread anti-Serb, Kosovo violates rule
democratic principles, Gradual improvement returns modest return of minority minorities anti-Roma sentiment of law
does not function in security; creation because of security refugees; 2001:
in accordance with of ORC situation Progress: SRSG Executive orders
rule of law and does Problems: Progress: Quality of local of detention violate
not respect human small scale return IAC adopts Joint judiciary ECHR;
rights norms; ethnic of refugees; limited Framework for the
minorities isolated in freedom of move- Return of Serbs
ghettos and subject to ment; pervasive
violent attack; refugee discrimination
returns a “distant goal”
Bosnia and Office of the Human HRCC Quarterly See HRCC Report Not Available Problems: Problems: antiter- Problems: media Situation improving 2003: time to end
Herze- Rights Ombud- Report 06/01: housing and employ- rorism and human harassment but still volatile impunity for disap-
govina sperson: Rate of refugee ment shortages rights; citizenship Progress: pearances during war
no Annual Report return increasing but hinder return; return- criteria; S C abuse of return of refugees; 2002:SFOR detention
since 1999; no news- sustainability prob- ees face harassment power; mistreatment domestic compliance violates human rights
letters since 2000; lematic due to return and discrimination of returnees in RS with human rights 2001: violence against
Human Rights Cham- related violence; this minorities in Repub-
ber Annual Report quarter 117 reports lika Srpska must stop
2000: of incidents directed
6,675 individual ap- at ethnic minorities;
plications received to ethnic violence twice
date; implementation as high in SR
of decisions improv- OHR Human Rights
ing but uneven across Priorities 2002:
country; case content Sustainabil-
concerns property of ity of refugee return;
returning refugees, discrimination in
ill treatment on education; reform of
arrest and detention, police and judiciary;
treatment of Muslim capacity building of
minorities in SR domestic institutions;
minorities
HR Ombudsman
Annual Report: none
released since 1999
Macedonia No available infor- See OSCE; UNHCR; ODIHR : 2002 parlia- Not available 2002 Report: Problems: Problems: Some risk of conflict 2003: police ill-treat-
mation UNHCHR mentary elections more than 150,000 of judicial corruption; “patchy” human rights but reduced by in- ment and torture of
conducted “largely in the 170,000 persons treatment of opposi- record; police brutal- creased commitment ethnic minorities
accordance with OSCE displaced during 2001 tion parties; ity; concern for ethnic to democracy 2002: threat to free-
principles” have returned Progress: human minorities dom of expression;
HCNM: none since rights education practice of impunity
1999 for National Liberation
Army violates rule
of law
663
664
Abkhazia N/A N/A 2001 Annual OSCE Not available Problems: N/A Poor human rights Conflict unresolved; Not directly applicable
Activity Report: prospects for return Mission completed record; religious low-level violence; to conflict
“peace process at bleak because of 2002 minorities harassed refugee problems
virtual standstill” due continued conflict and
to increased conflict low security
South N/A N/A 2001 Annual OSCE Not available Problems: N/A As above Conflict unresolved Not directly applicable
Ossetia Activity Report: no significant but stable to conflict
“some positive signs progress on return of
in Georgian-Ossetian refugees and IDPs
conflict” including
some “grassroots
rapprochement” in
ethnically divided
villages
2002 Expert Meeting
on Ossetian Conflict:
“positive dynamics
of settlement process
have been signifi-
cantly upset in recent
months”
Gagauzia 2001: Peoples See OSCE No documents avail- -better support for For Moldova as a N/A Not available Low Not directly applicable
Assembly of Gagau- able since 1999; no Gagauz language whole Risk of conflict due to to conflict
zia demands greater mention in 2001 OSCE -add provision to Problems: inadequate autonomy agreement
representation in Activity Report constitution re: institutional capacities
Moldovan parliament Gagauz autonomy Progress: Directorate
and an end to dis- for Refugees estab-
parities in economic lished
distribution
Transdni- N/A See OSCE No specific docu- -no specific references For Georgia as a N/A Not available Declining risk of Not directly applicable
estria ments available since whole conflict; continued to conflict
1999; Problems: inadequate Slav demands for
2001 Annual OSCE Ac- institutional capacities greater autonomy
tivity Report: Moldova Progress: Directorate
Mission encouraged for Refugees estab-
return to negotiation lished
process after break of
6 months
Bougain- March 2002 PNG Peace Monitoring OSCE N/A N/A No relevant infor- N/A Not available 2001 N/A 2001: notes autonomy
ville parliament gave Group to withdraw UNPO Report 2002 mation Last review in 2000 agreement achieved
legal effect to peace in June 2003 so that Progress: constitu- prior to autonomy and implementation
agreement island can elect an tional commission agreement is pending
autonomous govern- established
ment; 01/03 Contain- Problems: large
ment programme number of decom-
progressing – hopeful missioned arms stolen
will retrieve arms
stolen in 2002
Mindanao No available infor- N/A N/A N/A No relevant infor- N/A No report since 1993 2001 N/A 2003:violence
mation Task Force for De- mation 2000 Review: violence between MILF and
tainees of Philippines continues government forces in
12/00 notes increase Mindanao escalating;
in human rights civilian safety at risk
violations
665
Chapter 22
‘Bridges’ in Self-determination Disputes?
External Relations of Sub-national Entities
and Minority Groups
Francesco Palermo
I Introduction
Borders divide. Especially in the context of the nation-state, the primary histori-
cal function of borders is that of forming barriers (Hutchman 2000; Mantovani
2000; Anderson, O’Dowd, and Wilson 2002). But borders also function as fil-
ters, as “discriminating mediators between two or more political-institutional
or economic systems,” creating economic advantages and disadvantages. A third
perspective, the concept of the “open border,” stresses the function of contact
(instead of separation) between two or more political-institutional systems or
social-economic subsystems, allowing synergies between border areas (Ratti
1991: 57).
In the context of a peaceful Western Europe after the Second World War,
border areas began to express two needs. It was increasingly perceived that the
greater integration between states and the consequent dilution of national sov-
ereignty transformed these areas from mere peripheries of the nation-states they
belonged to into “new centres in the periphery” (Toniatti 1997: 23). They were now
integrated and more efficient areas from economic, cultural, and social perspec-
tives, performing a sort of bridging role between different systems. Additionally,
the necessity of joint management of common problems between neighbouring
areas belonging to different states became more evident and of increasing impor-
tance. This is particularly the case for environmental issues (Macrory and Turner
2002: 59), but also applies to problems affecting the fields of economy (commerce,
trade and tourism), transport, the cultural sphere, and many others.
More recently, cross-border cooperation (CBC) activities have also emerged
between areas where ethnic, linguistic, or cultural groups live – sometimes con-
stituting a minority within their own state but the majority on the other side of
the border (for example, between the German-Czech, German-Danish, Italian-
Slovenian, and Austrian-Italian borders) (Council of Europe 1995; Strassoldo and
Delli Zotti 1982). Consequently, the external relations of sub-national entities and
groups began to draw political attention, especially as some cooperation projects
668 Francesco Palermo
were (and still are) seen as a tool for the creation of ethnically homogeneous
regions, in order to create a federation of small and ethnocentric areas (Wolff
2003). Projects of this kind have slowly integrated and, practically, even replaced,
at the sub-national level, the protecting role of nation (kin-)states where they
exist.
On the other hand, the overall tendency in governance has been the search
for more efficiency, which can be best achieved by more flexible and results-ori-
ented cooperation between (already existing) entities. This does not only explain
phenomena like cooperative federalism (and regionalism) but also supranational
forms of cooperation, for example, the European Community, which has proved
to be a guarantee for states with respect to fears for their own sovereignty. In this
new and less ‘ideological’ context, the phenomenon of legally regulated activities
outside the local jurisdiction can be a useful indicator for the institutional evolu-
tion of autonomy and minority-protection systems.
This chapter first explores the legal nature of external relations of sub-national
entities (and of CBC in particular) from the perspective of international and EU
law, as well as in the comparative constitutional law of federal and regional states,
in order to underline the rules that generally apply to all forms of external activities
of non-state entities. In the second part, I analyze the forms of external relations
provided in the power-sharing agreements of the case studies, including both in-
ternational relations tout court and CBC activities. Finally, some conclusions are
drawn on the present and future role of this kind of arrangement for resolving self-
determination disputes. It is argued that only mature power-sharing systems are
able to effectively (and cooperatively, which in this context is the same) manage
the powerful instrument of non-state external relations. Thus, it will be stressed
that most of the cases analyzed are still lacking the very pre-condition for a suc-
cessful implementation of external relations as an instrument for resolving ethnic
conflicts: cooperation established and guaranteed by law. However, the margins
for an improvement in this respect are probably greatest among all power-sharing
mechanisms.
II The Legal Nature of Sub-national External Relations
A Conceptual Context of the Phenomenon

1 External Relations of Sub-national Entities and Cross-border Cooperation
The concept of the border as a ‘barrier’ is increasingly regarded as anachronistic,
with great interest being devoted to overcoming such barriers by means of cross-
border initiatives and, more generally, of legally relevant activities beyond the
territorial jurisdiction of sub-state entities. Nevertheless, this phenomenon has
only recently been studied by legal scholars. While international lawyers have ex-
In this chapter, the terms ‘sub-national’, ‘sub-state’, and ‘non-state’ will be used as
synonymous, all indicating public-law entities subordinated to the authority of a pre-
vailing constitution, which guarantees their autonomy.
22 ‘Bridges’ in Self-determination Disputes? External Relations of Sub-national Entities and Minority Groups 669
plored it for some years from their perspective, until recently constitutional law-
yers usually did not pay too much attention to it, focusing mainly on the issue of
sovereignty instead. This approach led to a general presumption of sub-national
external activities – where allowed – as exceptional, whereas their prohibition
was considered the rule. However, the present legal situation seems to challenge
this assumption. On the basis of the numerous multilateral and bilateral agree-
ments concluded in the last two decades, as well as by means of constitutional
reforms (or progressive judicial interpretation of the constitutional provisions),
external relations of sub-national entities are now to be seen no longer as ‘patho-
logic’ (i.e., tolerated exceptions), but more and more as ‘physiologic’ (i.e., a natu-
ral phenomenon).
Given the apparent decreasing importance of space and physical distance (Or-
tino 2002), it becomes natural that almost all policies adopted within a territory
may have an impact on other territories subject to a different jurisdiction. The
legal regulation of external activities is a way to rationalize the external effects of
the policies and to provide limits to them.
External activities in some fields can be completely independent from borders.
This is the case, for example, of some economic policies that can be carried out
jointly by non neighbouring entities, as in the case of the so called “Four Engines
for Europe” – a strategic alliance in some crucial areas between economically
strong European regions (Baden-Württemberg, Lombardy, Catalonia, Rhone-
Alpes) (Raich 1995). However, it is quite evident that most of the sub-state ex-
ternal activities have to do with the joint management of border issues and are
thus performed by neighbouring entities (CBC). This is the case of many envi-
ronmental policies, traffic, territorial planning, etc., as well as of policies directly
or indirectly affecting minority groups living on a border and having an interest
in establishing special relations with the other side of the border. Thus, greater
attention shall be paid to the opportunities provided by public law-based external
(and CBC) relations of sub-national units or even specific groups of people (such
as minorities).
2 … Involving Ethnic Disputes

Since the end of the Cold War, the claim of self-determination has increasingly
been used as a justification for demanding independence or secession. Rapid
secularization, democratization, and urbanization often have a strong impact on
ethnic sentiments, which frequently serve as a substitute for religion (Bonazzi
and Dunne 1995; Kymlicka 1995). Despite the continuing changes in international
law, which seem to put an end to the concept of absolute state-sovereignty, as well
as to the related principle of non-interference, there is great resistance towards
changes to borders as a basis for settling ethnic disputes, as these borders are still
seen as important guarantees for stability. Additionally, since ethnic groups do
not always concentrate in a compact settlement area, means other than secession
or independence have to be found in order to solve conflicts resulting from ethnic
diversity within the state.
Contrary to the situation during the Cold War, where genuine autonomy of ter-
ritories and groups was sometimes seen as incompatible with state sovereignty, it
now not only represents an attractive alternative to secession but is also perfectly
in line with a pattern of general decentralization and the creation of multilevel gov-
ernance-structures by integration. These developments give more room for differ-
entiation and for activities organized at a regional and sub-state level, both within
the state as well as across its borders, as is well demonstrated by the (lobbying)
activities of various European regions in Brussels, their engagement in economic
development projects in other countries, and the formation of so-called Euro-re-
gions across state-borders. In this environment, CBC activities constitute a flex-
ible, necessary and important addition to the internal decentralization of states.
Where minority issues are involved, intra-ethnic relations between communi-
ties can also be carried out without altering borders. For this reason, an increas-
ing number of ethnic conflict settlement agreements more-or-less explicitly men-
tion the establishment of territorial or group relations with states or sub-national
units outside the state border. This instrument is seen as an efficient alternative
to self-determination claims by minority populations.
Due to its experimental nature as a ‘third way’ between secessionist threats and
overall centralization, minority-related external activities were originally devel-
oped in the penumbra of positive law. This also explains the immense variability
of the forms of external and CBC-activities (Beyerlin 1988). However, after devel-
oping in a sort of legal ‘grey zone’, the need emerged to make those activities more
visible (as a guarantee for states) and legally more secure and safeguarded (as a
guarantee for cooperating minority groups).
B Legal Foundations in International and European Law

After the immediate economic reconstruction that followed the Second World
War (and extended until the 1960s), and with the beginning of European integra-
tion, CBC between territorial units increased significantly in Western and Cen-
tral Europe, for example, along Germany’s western and northern borders. Imme-
diately after the fall of the Berlin Wall, territorial (mostly regional and local) CBC
developed on both sides of the former Iron Curtain, between German, Austrian,
and Italian entities, on one side, and Polish, Czech, and Slovenian ones, on the
other. Due to the lack of public law instruments allowing sub-national entities
to perform their functions across their borders or boundaries, the legal instru-
ments for this kind of cooperation were mainly forms of private law (associations,
foundations, companies, working groups, etc.). For this reason, states decided to
give these spontaneous external relations some legal recognition under public
law. In Europe, the legal framework was found in the Council of Europe, under
whose legal and political umbrella cross-border activities between sub-national
territorial entities have increased enormously in the last decades. In 1980, the
Recommendation 470 was adopted by the Consultative Assembly (now Parliamen-

tary Assembly) in 1966, followed by Resolution 8 regarding cooperation at the mu-
member states of the Council of Europe agreed upon an Outline Convention on

cross-border cooperation between territorial communities or authorities (Ma-
drid Outline Convention, MOC) that aims “to facilitate and foster cross-border
cooperation between territorial communities or authorities ... and to promote
the conclusion of any agreements and arrangements that may prove necessary
for this purpose with due regard to the different constitutional provisions of each
Party” (MOC Art 1).
The MOC established an international legal basis for existing activities and for
the further development of this new form of cooperation. Nevertheless, as is typi-
cal for international law mechanisms, the MOC only provides a programmatic
context and a minimum standard that shall be common to all the contracting
parts. In the case of the MOC, the minimum standard is represented by the op-
portunity for the local communities or authorities to cooperate “in particular by
agreements in the administrative field” (point 3 of the preamble), having regard to
the different constitutional provisions of each state (MOC Arts 1, 2.1, 3.1, 4). The
most relevant consequence of the MOC, therefore, is that it brings CBC into the
domestic legal system of the contracting states, transforming it from an activity
at best ‘tolerated’ into an explicitly ‘legal’ one, the promotion of which the states
have agreed upon.
On 20 October 1995, a notable Additional Protocol to the MOC was opened for
signature. This Protocol aims at strengthening the MOC by clarifying the legal
nature of CBC agreements and Euro-regions. It expressly recognizes, under certain
conditions, the right of territorial entities to conclude CBC agreements, the validity
in domestic law of the acts and decisions made in the framework of a cross-border
cooperation agreement, and the legal corporate capacity (“legal personality”) of
any cooperation body established under such an agreement (Additional Protocol
Art 5). Thus, the Additional Protocol opens new opportunities for internationally
regulated CBC, but its importance should not be overstated, as it needs ratification
by the involved states and additional implementation measures.
However, the Additional Protocol to the MOC marks the clear intention of the
contracting states to make the more and more numerous forms of CBC in Eu-
rope more visible and to explicitly recognize them under a legal framework. As
cross-border organisms, the so-called Euro-regions are the institutionalization of
spontaneous and informal CBC activities, and thus are the most developed link
nicipal level in border areas, adopted by the Committee of Ministers in 1974 (Council
of Europe 1997; Ricq 1996).
Signed in Madrid on 21 May 1980 (after five years of work on the drafts) and entered
into force (after four ratifications) on 22 December 1981 (ETS 106).
Open for signature in Strasbourg on 9 November 1995 by the states that have signed
the Madrid Outline Convention, and entered into force on 1 December 1998 (ETS
159).
See also Additional Protocol 2 to the MOC (ETS 169), 1998, which establishes a legal
basis for cooperation also between non-adjacent territorial authorities.
in the chain of evolution. They first appeared in three regions of Europe that are
particularly sensitive to issues related to traffic and/or environment: Scandinavia
(Nordic Council), along the river Rhine (Regio Basilensis and Euregio) and in
the Alps (Arge-Alp, Alpe Adria) (Bajtay 1996). An increasing number of “formal
structure(s) established by municipalities or regions for the purpose of CBC with
the participation of economic and social partners” (Gabbe 1995: 3) were estab-
lished. These structures do not represent a new level of government but rather
a horizontal and flexible link of governance for persons, enterprises, and public
bodies on both sides of the border by offering practical and psychological advan-
tages and respecting the common interests. There is a great variety in the insti-
tutional and administrative structures of the cross-border entities and these have
direct impacts on the possibility of effectively influencing policies related to the
territory, the environment, the cross-border traffic, or the economic cooperation
of its members.
The more recent attention paid to CBC as a viable instrument in minority is-
sues is demonstrated by the role conferred on CBC in the Council of Europe’s
1995 Framework Convention for the Protection of National Minorities (FCNM).
In particular, Article 17 FCNM explicitly requires that the contracting states “un-
dertake not to interfere with the right of persons belonging to national minori-
ties to establish and maintain free and peaceful contacts across frontiers with
persons lawfully staying in other states, in particular those with whom they share
an ethnic, cultural, linguistic or religious identity, or a common cultural heritage.”
Moreover, the states “undertake not to interfere with the right of persons belong-
ing to national minorities to participate in the activities of non-governmental
organizations, both at the national and international levels” (FCNM Art 17). This
provision clearly envisages a general obligation that the states undertake to allow
minority groups to engage in CBC activities in order to foster their ties with a for-
eign state. The clear distinction between external relations as a whole (generally
authorized by Article 17 FCNM) being specified in procedural rules of domestic
law and specific international treaties that minorities (or their representative in-
stitutions or territorial governments, or the central governments) may conclude
for the same purpose emerges from Article 18 FCNM. According to this pro-
vision, “the Parties shall endeavour to conclude, where necessary, bilateral and
multilateral agreements with other States, in particular neighbouring States, in
order to ensure the protection of persons belonging to the national minorities
concerned.”
A consequence of this is the increasing number of bilateral treaties concluded
by states for the protection of minority groups and for establishing permanent
relationships across the respective borders, particularly in Central and Eastern
Europe (Lantschner and Medda-Windischer 2003). Many of the overall frame-
works provided by the treaties find a more-or-less direct response in national
See the information on Euroregions provided by the Council of Europe at http://

www.coe.int/T/E/Legal_Affairs/Local_and_regional_Democracy/Transfrontier_co-
operation/Euroregions/
legislation. An interesting example is provided by Hungary, where the 1993 law

on the protection of minorities granted to members of minorities the “right to
motherland” (Art 3.4) and the right to establish “international relations with the
motherland” (Art 19).
Regardless of the increasing attention to framework legislation provided by
international law for external relations of minorities (and their territories), the
detailed procedures to exercise this emerging right need to be laid down in the
domestic legal system. In other words, the foundations for CBC provided by in-
ternational law are not sufficient in themselves as a legal base for concrete mea-
sures to be taken as cross-border activities.
This is also true for EC law, at least from a formal perspective. EC law cannot
directly intervene in this field because the internal organization and distribu-
tion of powers between the various levels of governance is still vested with the
individual member states (the so called ‘principle of institutional autonomy’).
Nevertheless, the regional policy of the EC, particularly by means of financial
incentives and support to promote CBC activities, plays a crucial role (albeit in-
directly) in impelling the states to improve the powers of their regions to connect
themselves with regions of other states (European Commission 1992; European
Parliament 1996). The most relevant instrument here is the INTERREG initia-
tive, which since 1990 has provided funds for common initiatives between border
regions within and outside the EU. The INTERREG funds promote CBC as a
means for the economic development of border regions in order to overcome
mainly economic problems of regional markets at the peripheries of the EU. Of
great importance in terms of awareness – and confidence-building – is the in-
volvement of the states themselves, who, by participation of the regional and lo-
cal entities, have to provide cooperation programmes (of five years duration) in
order to qualify for the funds. The total amount of INTERREG-IIA (1994–1999)
was around ECU 4 billion. Currently, the INTERREG initiative has entered its
third phase, with a constantly increasing budget.
The initiative is open to any region situated on borders, within the EU or with
non-member states. There are special programmes for the Eastern borders of
the EU (Phare-CBC Programme) and for cooperation with the successor states
of the former Soviet Union (Tacis-CBC Programme).10 Another significant Euro-
pean initiative for the promotion of CBC is the LACE project, aimed at inform-
The first financial support of CBC activities had been included in Article 10 ERDF
(European Regional Development Funds) in 1989.
See Communication from the Commission to the Member States of 28 April 2000
Laying Down Guidelines for a Community Initiative Concerning Trans-European
Cooperation Intended to Encourage Harmonious and Balanced Development of the
European Territory-Interreg III (OJ C 143, 23/05/2000, 6).
The specific aims of the EC’s structural policies are no longer a pre-condition. Local
and private bodies can also apply.
10 See Special Report 16/2000 of the Commission on Tendering Procedures for Service
Contracts under the Phare and Tacis Programmes, OJ C 350, 06/12/2000, 1.
ing and increasing the exchange of knowledge-related CBC between European

border regions.11 Yet other European initiatives grant financial aid to enterprises
operating across borders. In order to consolidate the legal foundations in EU law
for CBC activities, the Commission is working on the elaboration of a specific
legal instrument of secondary community law for making CBC possible directly
under community law.
In summary, the phenomenon of external and cross-border relations, with its
typical bottom-up approach, is increasingly determined and shaped by EC law
(and policies) even though still only in an indirect way, i.e., by means of promot-
ing economic cooperation, especially through generous financial support. Nev-
ertheless, the legal foundations of cooperation across borders by means of public
law instruments remains with the single (and often different) domestic constitu-
tional systems of the states.
C Comparative Constitutional Analysis of Sub-national

Treaty-making Power
Rules on the external relations of sub-national entities vary quite considerably
between states. In general, federal systems usually recognize a limited treaty-
making power for their constituent entities (for example, in the USA, Germany,
Austria, Belgium, Switzerland, etc.), whereas merely decentralized, so-called ‘re-
gional’, states normally do not constitutionally provide for this power at the re-
gional level.12 The detailed rules on how to engage in external relations differ from
country to country and do not really depend on the self-attributed (‘federal’ or
‘regional’) nature of the states.
However, a deeper analysis of the domestic provisions in several Western fed-
eral and decentralized states (Palermo 1999) clearly shows some fundamental
features that are generally valid regardless of the nature of the state and of the
kind of sub-national entities involved (territorial or even non-territorial, such as
ethnic groups not enjoying a specific territorial self-government). The elements
common to all rules on external relations of non-state entities are twofold. On the
one hand, external relations are not really a competence but rather a procedure:
They are not a power per se but simply the way to exercise powers outside the
realm of the geographic jurisdiction. Thus, the domestic regulations over exter-
nal activities do not determine the powers to be exercised but how to exercise
11 LACE – Linkage Assistance and Cooperation for the European Border Regions. The
operative assistance, e.g., assistance in the preparation of INTERREG-IIA-projects,
is given by the Working Group of European Border Regions (AGEG).
12 There are some exceptions here, also, as in the case of Italy after the 2001 constitu-
tional reform (Constitutional Law 3/2001), which introduced an explicit reference to
a treaty-making power of the regions. Article 117.9 of the Italian Constitution stipu-
lates that “in the areas falling within their responsibilities, regions may enter into
agreements with foreign states and with local authorities of other states in the cases
and according to the forms laid down by state legislation.”
powers that are already vested with the sub-national entities. On the other hand,
this procedural mechanism – regardless of the countless detailed differences
depending on the institutional structure as well as on the legal culture of each
country – is based on cooperation. The sub-national entity (territory or group)
that wants to engage in external relations with other states or sub-national units
(for whatever reason – including the establishment of cross-border relations in
order to foster ethnic ties) needs to cooperate with the central government. The
logical consequence to be drawn from this is that the more efficient the coop-
eration – both in legal terms (efficient procedures) and in socio-political terms
(cooperative attitude) – the more developed and effective the sub-national power
to establish external relations.
More specifically, sub-national entities are allowed to enter into international
treaties or to engage in other forms of external relations only under the following
conditions:
a) there must be the consent of the central government (before or after the
conclusion of the agreement);
b) there is control by the central level (which in case of international treaties
can always prohibit the conclusion or force the sub-national units to de-
nounce them);
c) the external relations of sub-national entities must not affect the interna-
tional liability of the whole state nor its foreign policy.
Especially where ‘foreign policy’ is concerned, the concrete limitations largely de-
pend on discretion, and thereby are limited by means of previous accord between
the sub-national units and the central power.
The central government must be informed about the external activities and can
even terminate them if they violate one of two basic principles: the safeguarding
of the international liability of the state and the unity of the state’s foreign policy.
In a system based on the rule of law, sub-national units may impugn the denial
of consent before the national constitutional or supreme court, which basically
controls only the observation of the cooperation procedures, as the content of ex-
ternal activities is negotiated between the levels of government within the frame-
work of the cooperation dictated by the constitution. These conditions mean, in
practice, that significant autonomy is granted to the sub-national units if they
are able to cooperate with the state, whereas the space for activities without the
consent of the state is very limited, if not completely precluded. Further, there is
normally an independent mechanism to settle disputes about competence, usu-
ally a constitutional court or a functional equivalent.
Thus, the rules on external relations are provided by the set of procedures regu-
lating cooperation within the domestic constitutional order, although influenced
by international (and European) law as well as by practice. The more intense the
cooperation between the central power and the sub-national entities, the more
effective the procedures of external relation. Therefore, due to the supreme posi-
tion of the state in determining what constitutes ‘foreign policy’, the best way to
improve the external capacity of non-state entities and groups is to improve co-
operative mechanisms within the domestic constitutional system, making them

more efficient. The more state control is substituted by participation of all rel-
evant actors in sharing cross-border policy, the more effective the cooperation
and thus the power of non-state actors to engage in external relations.
Having clarified the legal background to any external activity of sub-national
units or minority groups it is now possible to analyze the more recent conflict
settlement agreements in light of this.
III External Relations and CBC in Recent Complex

Power-sharing Agreements
In the context of post-conflict situations, external relations and CBC face partic-
ular difficulties. As has been demonstrated, well-functioning procedures require
a favourable environment, i.e., a cooperative culture between the central state
and the sub-national (cooperating) entity. When mutual trust and efficient pro-
cedural mechanisms are lacking, external relations of the regions and minority
groups can easily be considered dangerous or even as a preparatory step towards
secession. Consequently, only where mutual trust between the majority and mi-
nority population is achieved, and where there is no doubt about the loyalty of
minorities to the state they belong to (Kymlicka 2002: 19), can efficient coopera-
tion mechanisms be established and the external relations of the minority groups
start working effectively. Of course, every situation is unique in this respect, even
those within the same state. For example, German-Danish or German-Dutch
relations are quite different, in terms of the “securitization”13 of the interethnic
relations, from German-Polish or German-Czech relations.14 Similarly, the at-
tempt to create a Euro-region between Tyrol, South Tyrol, and Trentino at the
Italian-Austrian border was initially perceived as a threat to the delicate minor-
ity-protection mechanisms in the area and it could start working only after those
fears were alleviated (Nick and Pallaver 1998).
Against this, especially where CBC is an issue, the case studies can be ana-
lyzed and evaluated – where external relations is one of the instruments for com-
plex power-sharing as an alternative to external self-determination depending
on the degree of cooperation established by the dispute settlement agreements.
Particular regard is given to explicit rules on external relations contained in the
peace agreements but, more generally, the overall framework of cooperation is
also taken into consideration, as well as the specific conditions of each experi-
ence. Instead of a sterile list of the individual countries analyzed, the cases will
13 The term “securitization” is used by Waever (1995).

14 Along the Dutch-German border, despite occupation during the Second World War,
the development of CBC was not only begun very early but has gone so far as to allow
the establishment of institutionalized Euro-regions as public law entities (Scott 1997:
83), whereas on the eastern borders of Germany the confidence-building process
is not yet completed, thus making CBC a very sensitive issue (e.g., along the Czech
border, where the mass-deportation of Sudets took place) (Gerner 1997: 107).
be grouped according to the typology of rules and conditions that can be derived
from the comparative investigation.
With regard to the rules on external relations of minorities and (their) territo-
ries provided by the conflict settlement agreements and their consequent imple-
mentation, the cases studies can be grouped in three main categories:
a) Non-existent external relations. Due to specific conditions, such as the geo-
graphic situation (for example, islands) or the lack of foreign states that can
be considered for cooperation based on the development of ethnic ties, ex-
ternal relations are simply not an issue. This applies to a (small) number of
cases, although some degree of international relations are nevertheless im-
plicitly possible, at least in terms of third party or international involvement
in the conflict resolution.
b) Merely informal relations. This is the case for conflict-resolving agreements
that do not mention the possibility of engaging in external relations but nev-
ertheless some international activities are carried out as a matter of practice
and/or on the basis of the cooperation procedures or the powers laid down
in the constitution.
c) Explicitly mentioned external relations for improving or maintaining the
ethnic characteristics of minority groups. This category of agreements is the
most common and interesting one for our purposes. However, a distinction
should be drawn not only between the intensity of the guarantees in terms
of number and complexity of the specific provisions on this matter but also
between the nature of the guarantee, having regard to whether the powers
are assisted by a merely political or firmly entrenched and substantive legal
guarantee.
A Non-existent External Relations

The first, and easiest, case to deal with is where no power, either explicit or im-
plicit, is allowed to the groups or their territories concerning external relations
of any kind. This is particularly the case of Mindanao, where the only reference
to external relations in the 1996 peace agreement is the reservation of ‘foreign af-
fairs’ as a whole to the central government, thus banning any possible power on
this matter for the regional legislative assembly.15 Also, given the geographical,
geo-political, and ethnic-religious conditions affecting the Mindanao conflict, it
seems quite difficult to identify any possible uses of external relations as an in-
strument to contribute to a more viable solution of the conflict.
15 The same principle (although referred to as ‘foreign policy’ instead of ‘foreign affairs’,
though the two wordings seem to be synonymous) was established also by the 1976
Tripoli agreement.
B Merely Informal External Relations

This second group is of countries that do not explicitly recognize external powers
for groups and/or territories but where cross-border activities are nevertheless
carried out by means of other instruments, legal or (more frequently) factual. In
these cases, the law generally tends to simply tolerate the less visible (and, from
the perspective of the central states, less dangerous) activities instead of explic-
itly regulating them. This is due to the fears arising from a formal recognition of
external relations of potentially secessionist territories and groups, and to the as-
sumption that other, more basic needs are required by the involved minorities.
In Georgia, there are no specific rules on the possible external relations of Ab-
khazia and South Ossetia. Nonetheless, due to the geographic proximity to Rus-
sia – which can be considered (at least informally and somewhat one-sidedly)
the kin state of both entities – informal relations do occur, although merely at a
political or even criminal (smuggling) level. In the case of South Ossetia, there is
the addition of a ‘natural’ partner on the other side of the border, North Ossetia,
although all formal declarations (especially from the North Ossetian side) deny
any particular interest in fostering cooperation across the border (see Ch 11), to
enhance a good neighbourly relationship with Georgia. Moreover, particularly in
the case of Abkhazia, there was international involvement in the conflict resolu-
tion phase.
Much more so than in the Georgian case, the three Balkan experiences ana-
lyzed (Kosovo, Macedonia, Bosnia and Herzegovina) are shaped by the presence
of (more-or-less explicit) kin states for the ‘minorities’. Some groups (possibly
including Albanians in Kosovo and Macedonia, Serbs and Croats in Bosnia) did
have (and probably still have) aspirations of reunification with their ‘homeland’,
although this has been excluded by the peace agreements. It is thus understand-
able that two of the three agreements do not accord, or at least tend to minimize,
powers to the groups or their territorial entities to engage in foreign relations.16
But, at the same time, it is equally understandable that such relations do take
place, in some form or other, given the territorial proximity. Moreover, it is obvi-
ous that the external relations and CBC activities carried out in those areas do
have a quite evident ethno-political connotation.
The peculiar status of Kosovo, and its practical (although not legal) separation
from Belgrade, combined with the need not to foster the aspiration of a unifica-
tion with Albania,17 made it necessary to reduce as much as possible the external
relations of Kosovo, particularly with the ‘kin states’ of the two communities,
Serbia and Albania, at least for the first phase of the conflict settlement. The
Rambouillet agreement did recognize extensive rights of the national communi-
ties, particularly the overall right to “preserve and express their national, cultural,
religious and linguistic identities” (Art VII.1). To do so, the national communities
16 The most outstanding example is provided by UNMIK Regulation 2001/10 of 24 May

2001, prohibiting unauthorized border/boundary crossing from Kosovo.
17 For details, see Ch 8.
may participate “in regional and international non-governmental organizations”

(Art VII.4.x), and the members of national communities have (individually) “the
right to enjoy unhindered contacts with members of their respective national
communities elsewhere in the Federal Republic of Yugoslavia and abroad” (Art
VII.5. lit. a). In addition, “each national community … may exercise additional
rights through federal institutions and institutions of the Republics” (Art VII.6).
This bundle of provisions, where appropriately implemented in light of the
principle of cooperation, may constitute the legal basis for external activities in
order to develop the national identity of each community. Particular attention
should be paid in this regard to the possibility of conducting international rela-
tions through the central authorities, i.e., by means of an effective domestic co-
operation, although, under the present international administration, the external
representation of Kosovo is carried out by the Special Representative of the UN
Secretary-General (UNSRSG).18 However, this authority does not include forms
of cooperation that are not regulated by international law, as are most of the CBC
activities at sub-national (particularly at local) level.
The case of Macedonia is even clearer in excluding formal external relation
powers and procedures from the set of instruments available for complex power-
sharing. One of the fundamental principles of the Ohrid agreement of August
2001 states that “there are no territorial solutions to ethnic issues” (Art 1.1.). This
sharp (and debatable) statement is aimed at eliminating forever the risk of seces-
sion (and also of the establishment of ethnically-based sub-national entities) but
its strict interpretation may indicate the disinclination to any kind of use of the
territory in order to accommodate ethnic claims.19 In fact, the legal framework
is much more complex, and there are some indicators of a possible legal basis
for cross-border activities by local governments and groups as such. In the first
place, the enhancement of decentralization and municipal autonomy with the
constitutional amendments following the Ohrid agreement establishes a frame-
work for territorial cooperation across the border.20 Secondly, the general recog-
nition of the rights of the members of the communities “freely to express, foster
18 Constitutional Framework for Provisional Self-Government Art 48.(i)

19 As the Romans said, ‘excusatio non petita, accusatio manifesta’: am unrequired jus-
tification is proof of the alleged facts. As is well-known, the cross-border contacts
between Albanians in Macedonia, Kosovo, and Albania are intense and they are cer-
tainly not prevented by provisions of this kind. The demonstration of such contacts
is provided by the frequent political statements of the Macedonian government stat-
ing that the crisis was simply imported, brought on by foreign terrorists aiming at
controlling parts of the territory and guaranteeing their revenues from cross-border
smuggling of cigarettes, weapons, and other goods (see Ch 9).
20 See new Articles 114 and 115 of the Macedonian Constitution, regulating local self
government. In particular, Article 114 of the Constitution provides that the law on
local self-government is approved with a qualified majority, thus taking into con-
sideration the aspirations of the two main ethnic groups. Article 115 lists the main
competence fields for the local self-government (e.g., territorial planning, environ-
mental protection, local economic development, local activities and culture, educa-
and develop their identity and community attributes and to use their community
symbols” (Macedonian Constitution Art 48), as well as its future implementation,
may lead to a more flexible management of cross-border cooperation, possibly on
a functional and not merely ethnic, base. However, these quite sharp provisions
did not prevent the establishment of less formalized external activities, such as
the establishment of the Lake-Ohrid Euro-region.21
In both Kosovo and Macedonia, the lack of or the existence of very limited
provisions on external activities for sub-national entities or groups may be con-
nected to the specific causes of the conflict, to the fears of possible secessionist
movements, and, not least, by the absence of a clearly identifiable territory (or-
ganized as a sub-national territorial entity) where the communities are settled.
Given the difficulties for such groups to engage in external activities without the
intermediation of a territorial self-government organization, it is quite clear that
provisions allowing external relations are very poor. Nonetheless, the present le-
gal situation could be sufficient to implement some such relations at some point
in the future when the overall inter-community relationship has normalized and
effective cooperation procedures can start working. To date, some proposals have
been elaborated in this regard and some first steps are being undertaken to in-
clude these countries in cross-border activities.22
C Explicitly Mentioned External Relations

1 Cases
In Bosnia and Herzegovina (BiH), the general power in the field of “foreign
policy” (and “foreign trade policy”) is vested with the central government (BiH
constitution Art III.1 lit. (a-b)). However, the two entities (Federation of Bosnia
and Herzegovina and Republika Srpska) do have “the right to establish special
parallel relationships with neighbouring states” (BiH constitution Art III.2 lit.(a)),
although these relationships shall be “consistent with the sovereignty and territo-
rial integrity of Bosnia and Herzegovina” (BiH constitution Art III.2 lit.(a)), and
these activities are subject to the respect of the general principle of fair coopera-
tion.23 The entities made use of the enabling provision in concluding trans-border
agreements with the ‘motherlands’ of Serbia and Croatia (also going beyond mere
CBC-activities). The decision-making arrangements of the central state institu-
tion, etc.), all showing a high potential to be managed also by means of CBC forms in
the case of municipalities located on the border.
21 This new Euro-region is, however, still only on paper. See http://www.alb-net.com/
pipermail/albsa-info/2002-November/003273.html; and http://www.stabilitypact.
org/newsletter/nl-18.html.
22 For example, US Institute of Peace 2002. For some running examples, see East-West
Institute 2002; East-West Institute 2003.
23 Article III.2 lit. b BiH Constitution: “Each entity shall provide all necessary assistance
to the government of Bosnia and Herzegovina in order to enable it to honour the
international obligations of Bosnia and Herzegovina.”
tions (requiring the consent of the three constituent peoples for each policy deci-
sion) was also used in order to undermine the power of the central government,
as Serbs and Croats preferred to block the decision-making in international rela-
tions at the central level and rather to exercise it at the entity level. The real coun-
terbalance to this ethnically-driven, factual implementation of the entities’ for-
eign policy has been the degree of intervention by the international community
(making Bosnia a quasi-protectorate) (Bieber 2004; Woelk 2004), much more so
than the formally wide central power in the field of foreign relations. Moreover,
in order to avoid possible troubles, especially with the international community,
most of the external relations have been carried out as a matter of fact on an ad
hoc basis rather than on the basis of formally concluded agreements, regardless
of the entities treaty-making power (BiH Constitution Art III.2). For this reason,
it can be affirmed that most of the external activities of sub-national units of BiH
are based on factual cooperation with the ‘motherlands’. Nevertheless, it must be
stated that more recently the central government has signed a considerable num-
ber of international agreements, even in areas that are particularly sensitive for
the ethnic groups.24 This demonstrates an increasing role for the state institutions
in this respect. In a more normalized situation, the treaty-making power of the
entities (and, more generally, their power to engage in different kind of external
relations) could be used as an additional, integrative, cooperative legal instru-
ment to complement the foreign-policy power vested in the state authorities, as
is common practice in federal countries such as Germany, Belgium, Switzerland,
or Austria.
In Gagauzia, a sort of recognition of cross-border powers is also provided, al-
though almost exclusively by means of mere participation of Gagauz representa-
tives in the state’s activities instead of forms of direct action. According to Article
12, al. 3b of the 1994 Law on the Special Legal Status of Gagauzia, its representa-
tive authority – the People’s Assembly – shall have the power, among others, to
participate “in the implementation of internal and external policy of the Republic
of Moldova connected with the interests of Gagauzia.” Such participation prac-
tically consists of the mere inclusion of Gagauz representatives in official del-
egations without granting them a special right to influence the final position of
the Moldovan delegation. In addition, Gagauzia has established formal relations
with Turkey, Transdniestria, Belarus, and the Odessa oblast in Ukraine (Järve
2004), and has special ties with Turkey.25 Arguably, due to the overall situation,
24 Such as the numerous agreements with foreign states in cultural matters (e.g., with
Germany, Sweden, France, Italy, etc.), and particularly with other former Yugoslav
states. See, for example, the agreement with the Republic of Croatia on coopera-
tion in the fields of science, technology, and higher education. The same goes for
other delicate matters such as economic issues. See the Free Trade Agreement be-
tween Bosnia and Herzegovina and the Republic of Croatia (Official Gazette of BiH,
35/00).
25 For example, in 1993 Turkey opened a joint university named ‘1993’ and Turkey also
opened the “Gagauz-Turkish School.” This is due to the fact that the Gagauz language
the implementation of these provisions is still limited, but the existence of a quite
solid legal basis leaves many openings for the future. Moreover, Article 17 of the
same law stipulates that the Executive Committee (the executive body of Gagau-
zia) has, among others, the competence to regulate “in conformity with the law,
on the whole territory of Gagauzia, … the management of economy, social and
cultural development, … environmental protection and the rational use of the
natural resources.” Such a clause could turn into an implicit-power clause for
the management of the community’s own interests, and thus could be used as an
instrument for the implementation of cross-border activities (where necessary to
pursue the indicated goals). This has been the case in many other areas where the
sub-national entities were not vested with any explicit permission to engage in
cross-border activities (but at the same time there was no explicit denial of them),
such as in Spain, Italy, and even France, as it is typical of the very first phase
of developing external relations of sub-national units in previously centralized
states. Additionally, the Law on the Special Legal Status of Gagauzia provides
for an extreme (and merely theoretical) guarantee to the territory in terms of ex-
ternal relations: In case of the union of Moldova with Romania (which has been
excluded by the Romanian authorities), the law confers to Gagauzia the right of
external self-determination and thus the right to separate from Moldova. Being
an organic law, however, it can be changed by the Moldovan parliament under a
special procedure of qualified majority of 80% of the members of parliament.
For our purposes, Bougainville represents a very interesting case where ex-
ternal relations are concerned. Being an island, there cannot be any proper CBC
activity. Nor is there the possibility of autonomous international relations, at least
as long as the island remains under Papua New Guinea’s sovereignty, which im-
plies the exclusive national power for engaging in international treaties.26 Nev-
ertheless, the 2001 peace agreement (BPA) provides for some special rules in
the fields of technical assistance from foreign states and for the establishment of
mechanisms for consultation of Bougainville “in agreed areas of foreign affairs”
(BPA Art 69). Bougainville can be allowed by the National Executive Council to
“send a representative or observer to regional meetings and organizations” (BPA
Art 70), although at its own costs (BPA Art 72), which in fact could practically
impair the effectiveness of the provision considering the economic situation of
the island.
Moreover, where international agreements stricto sensu are concerned, a rep-
resentative of Bougainville can be included in the national delegation “to regional
meetings and organizations of clear special interest to Bougainville” (Article 71
BPA) and the autonomous government can request to participate or even to “en-
gage directly in the negotiation of international agreements of particular rele-
vance to Bougainville” (BPA Art 73). Bougainville may also play a significant role
belongs to the same language family as Turkish. See Bowers et al. (2002).
26 See also Article 74 of the 2001 peace agreement: “the autonomous Bougainville gov-
ernment will respect the national government’s authority regarding international
agreements.”
in international relations with its kin state, the Solomon Islands: the autonomous
Bougainville government has a permanent representative in national delegations
“engaged in the negotiation of new border agreements with Solomon Islands”
(BPA Art 77).
However, in light of what has been argued here (internal cooperation mecha-
nisms as the real legal foundation of external relations), other, apparently more
general, provisions seem even more important in practice than those specifically
mentioned ones. According to the overall provision of Article 68 BPA, “the na-
tional government and the autonomous Bougainville government will establish
agreed procedures to give effect” to the agreement. The safeguard of autonomy
against unilateral changes by means of the treaty-making power of the national
government is provided by Articles 75 and 76 BPA. According to the first of these,
“an international agreement negotiated or signed with a purpose of altering the
agreed autonomy arrangements will take effect only with the agreement of both
the national government and the autonomous Bougainville government” (em-
phasis added). The latter article states that if a disagreement arises between the
national government and Bougainville’s autonomous government “as to whether
an international agreement has been negotiated or signed with such a purpose,
the matter will be resolved through the agreed dispute resolution procedures.”
Furthermore, Article 78 BPA completes the guarantee by affirming that “future
border agreements (other than those concerning defence or national security as-
pects) which affect the jurisdiction of the autonomous Bougainville government
will take effect only by agreement between the national government and the au-
tonomous Bougainville government.”
This rather progressive arrangement, when compared to other conflict settle-
ment agreements, is inspired by the logic of overcoming the majority-minority
positions, stating the principle of equal standing of the involved parties (national
and autonomous government) and thus the obligation to negotiate where the
interests of the minority are concerned, with particular regard to external rela-
tions. It can be said that the BPA extends (at least potentially) power-sharing ar-
rangements to the field of external relations, founding the necessary cooperation
on a paritarian base between the state and the autonomous entity. As long as
such a cooperative framework can be established and maintained, this legal pic-
ture fits perfectly into the principles previously mentioned for an effective man-
agement of external relations. The difficulties in this regard could emerge where
the legal guarantees for the effective implementation of the cooperative rules are
concerned. Many of the clauses safeguarding the paritarian principle in the re-
lationship of the two governments, and particularly regarding external relations,
need practical implementation and, above all, interpretation. The possible weak-
nesses of the entire system may lie not in the provisions but much more in the
mechanisms aimed at resolving possible conflicts of interpretation, particularly
in the lack of a general interpretation power attributed to an impartial judicial
body.27 In this regard, it can be stated that the main guarantees of Bougainville’s
considerable powers in external relations matters are still more of a political than
legal nature, although the system is legally quite sophisticated and founded on
cooperation, which is per se a pr-condition for success.
The most relevant example of an entrenched and legally well-developed system
of external relations as an instrument for peaceful conflict settlement is provided
by Northern Ireland. The Anglo-Irish agreement of 1985 already provided for
cross-border cooperation on security, economic, social, and cultural matters (Art
9). Extensive cross-border cooperation was envisaged as an essential tool in the
compromise found with the Good-Friday-Agreement of 1998 (GFA), also called
the Belfast agreement. The 1998 agreement provides for a bundle of joint bodies
that shall not only allow but even encourage the inter-communal relationship. In
this sense, one can argue that Northern Ireland is the only case where CBC is not
simply tolerated but imposed, on the assumption that it facilitates the establish-
ment of a peaceful culture. In other words, CBC in Northern Ireland is one of the
core mechanisms of the peace agreement. It is an objective per se and not merely
an instrument to pursue other goals.
The first CBC body based on forced cooperation among the parties was the
North-South Ministerial Council, a body that brings together ministers of the
Northern Ireland government and the government of the Republic of Ireland.
They meet in plenary session twice a year, and more often in smaller groups to
discuss technical issues. The intention of the agreement in establishing this body
was precisely to resolve possible disagreements by means of technical discussions
on matters of common interest, which is itself the basic idea underlying CBC
in ethnically tense areas. The decisions should be implemented by cross-border
implementation bodies dealing with an open list of possible subjects, from agri-
culture to education, from environment to tourism. Another CBC body estab-
lished by the agreement was the British-Irish Council, where the governments of
the two states (UK and Ireland) and the devolved governments of UK (including
Northern Ireland) meet.
These institutions did not properly function yet, not even in the (relatively
short) time when the devolved authorities were not suspended. Again, a frequent
misunderstanding (or misinterpretation) of these forms of cooperation has led to
the fear of the unionist community that they might constitute a first step to the
reunification of Ireland.28 In such a context, as stated before, CBC cooperation
27 Articles 265 and followings of the agreement only provide for a dispute settlement
through mediation or arbitration, and eventually through courts, without specifying
more on this crucial aspect. Nevertheless, the safeguard clause laid down in Article
268 BPA, which prescribes that “the national government will not have the power
to withdraw powers from the autonomous Bougainville government or suspend it,”
seems of vital importance.
28 For cross-border cooperation and consociational democracy in Northern Ireland,
see McGarry 2003; O’Leary 2001. See also Alcock 2003 30–31).
loses its very essential feature of being concrete and de-ideologicized, and, in the
absence of this precondition, CBC is bound to fail.
2 Legal vs. Political Guarantees?

The agreements described here differ profoundly from one another in terms of
substantive external relation powers. Perhaps more critical than the assignment
of external affairs power, however, is their entrenchment at the constitutional or
international legal level. Moreover, even if the grant of authority is firmly estab-
lished, a dispute settlement mechanism is crucial. As a subordinate entity will
generally be subject to the overall constitutional order of the state (encapsulated
in the power-sharing settlement), the central authorities may normally be ex-
pected to prevail in cases of conflicts of authority. Exceptionally, where a highly
decentralized system has been adopted (for example, Bosnia and Herzegovina),
the centre may have little power to constrain an excess of foreign affairs powers by
the constituent entities. Hence, it would be important to ensure that conflicts over
the exercise of this authority are resolved according to an objective legal process
– generally a constitutional court. In some of the cases reviewed here, the power
to address clashes of competence may be assigned to an international agency.
One further exceptional example is provided by the Good Friday Agreement
for Northern Ireland. Although this agreement is based on the forced coopera-
tion of the two communities on CBC as one of the main goals of the whole peace
process, if the involved parties fail to cooperate properly the final decision is not
taken by an impartial judicial body. Actually, the British government, as one of
the involved parties, has the final say on the suspension of the local institutions,
the decision being political and not justiciable in a court. As is well known, the
British government has already suspended the Northern Ireland institutions sev-
eral times after claiming that the obligations regarding decommissioning have
failed or not been met.
In turn, as the effective exercise of external powers of any kind by sub-national
entities or groups is dependent on the effectiveness of the domestic coopera-
tion procedures, the lack of legal guarantees might turn out to be positive when
it comes to establishing a web of external relations ‘in the penumbra’ of formal
provisions, i.e., below the level of formalized activities. This might be quite useful
in a first phase of confidence-building between central authorities and autono-
mous entities, allowing the state to realize that sub-national external powers are
not necessarily a threat to national integrity, and the autonomous territories or
communities to become aware of the convenience of cooperation. However, after
such a first, experimental phase, when sub-national external activities tend to
become the rule and not an (at best tolerated) exception, the interest of territories
and groups shifts from being left alone to being safeguarded in their prerogatives.
And since the role of the central power will always prevail in these areas (because
of the very broad range of the realm of concepts like foreign policy, national secu-
rity, etc.), in case of disagreement the sub-national claims do not have any chance
to be affirmed but in a legal litigation before a court based on the principle of the
rule of law.
In other words, the lack of legal guarantees for cooperative instruments can
be a permanent threat to the effectiveness of the arrangements laid down in the
various peace agreements. A cooperation imposed by law but guaranteed merely
by political means runs the risk of being built upon an unstable base, especially in
areas where the conflict is quite recent and the parties have not yet developed the
required trust to freely rely on cooperative mechanisms. Of course, legal guaran-
tees cannot be considered a panacea, particularly in the field of external relations.
As the external relations of sub-national entities are dependent on procedures
of (internal) cooperation, the required cooperative environment cannot be es-
tablished through judicial decisions alone. However, a judicial body in charge of
the settlement of disputes with respect to these cooperation procedures would
appear to offer significant guarantees, reducing the mere political discretion and
acting as a counterweight to dominant political power, power that is usually vest-
ed with the centre.
IV Concluding Remarks
External relations, especially when carried out by minority groups or their rep-
resentative self-governments, are highly symbolic and thus risk neglecting the
functional utility of such relations in favour of the ideological. This is particularly
true in the case of recently settled conflicts, where the parties still lack mutual
trust. However, this chapter has demonstrated that the procedural mechanisms
– as the essence of these relations – must be based upon cooperation between
the central government and the minority group(s). Therefore, the more symbolic
or ideological the relations between the state and the minority (or their repre-
sentative self-governmental unit), the less efficient these mechanisms, as they
would tend to emphasize the differences between the parties and hence not foster
cooperation and acceptance. In simple terms, it can be said that it is the primary
interest of the minorities to promote cooperation with the central authorities,
and external functional activities are one of the most crucial experimental fields
in this regard. It follows then, that only mature autonomies can establish efficient
procedures for entering into external relations, because only stable systems are
able to cooperate – not only with neighbouring territories, but, above all, with
central authorities – which is the most important precondition for effective ex-
ternal relations. This is most evident where the guarantees attached to the whole
machinery are basically of a political instead of legal nature, as in most of the
cases analyzed here.
Moreover, the more stabilization, democratization, and avoidance of the con-
flict (including reference to symbols of the conflict), the more effective the co-
operation, and, consequently, the external activities. In some of the case studies,
this trend becomes more and more visible, as demonstrated by the increasing
number of CBC activities in Southeastern Europe, especially in the Balkans.29
29 For an updated list of the ongoing and planned projects (including Euro-regions),
see http://www.iews.org/programs/projects.cfm?title=Projects&l1=Programs&l2=R
In post-conflict situations, external activities by sub-state units offer a modern,

flexible, and, thus, very attractive tool for general reconstruction or economic
improvement. Nevertheless, horizontal forms of cooperation across borders do
not only require a favourable regional context (especially by cooperation of the
states the entities belong to) but also a cooperative culture within the state as
well as inside the cooperating entity. Where mutual trust and cooperation are
lacking, a cross-border project can easily become (or at least be considered) a
threat to internal stability, and thus be easily and legally blocked by the central
governments. It is particularly important to stress that functional, effective forms
of cooperation across the borders can only work properly in a democratic context
in which the rule of law is established and human rights are effectively protected.
Moreover, this appears to fit into a general tendency towards more functionally
oriented forms of governance. The allocation of powers and functions is moving
to a system of distribution of powers that is determined by functional streams
(economy, efficiency, etc.). This not only explains phenomena such as coopera-
tive federalism (and regionalism) but also supranational forms of cooperation.30
In the context of external relations of minority groups, supranational cooperation
and integration by generating trust and confidence between the states proves to
be a guarantee against fears of ‘revisionist activities’ that might endanger sover-
eignty.
Procedures and institutions of external and cross-border cooperation do not
constitute a new level of governance but provide (additional) horizontal links by
more closely interconnecting existing entities according to functional criteria.
Their acceptance depends primarily on the concrete advantages for the citizens.
Thus, from a long-term perspective, effectiveness becomes the main evaluation
criterion, slowly substituting the ‘mere’ enhancement of ethnic claims. Their inte-
gration into the existing institutional frameworks as complementary elements is
in obvious contrast to the idea of (re)building institutions exclusively along ethnic
lines. Moreover, the functional criterion should not only be seen as paying atten-
tion to concrete rather than ideological or symbolic issues but also as a means
of fostering traditions and identities within and across the borders, thereby re-
ducing the disadvantages of permanent minority positions by opening a broader
context.31
For these reasons, a certain degree of ‘institutional maturity’, a realized climate
of trust, mutual information-sharing, and cooperation between the central state
and the sub-national entities and/or groups seem to be the very pre-conditions
for external activities, given that the issue of borders still remains a sensitive one.
States that see their sovereignty questioned by secessionist tendencies will hardly
egional%20and%20Transfrontier%20Cooperation%20%28RTFC%29&l3=Projects.
30 The European Union, with its functional approach and the oft-quoted principle of
subsidiarity, provides a formidable example of this tendency towards the establish-
ment of multi-level governance.
31 This seems to be the underlying idea of Articles 17 and 18 of the FCNM, as discussed
above.
promote cooperation that create links across their borders. This is especially so
where there is a weak or non-existent legal guarantee of the mutual obligation to
cooperate between the state and the territorial or minority entity.
On the other hand, the more these activities are driven by functional, and not
explicitly ideological aims, the more their adoption can contribute to the estab-
lishment of a cooperative environment. External activities are thus procedural
instruments with a high potential for development and for de-escalating ethnic
conflict. But their effectiveness requires initial goodwill from all the involved par-
ties and the establishment of legal guarantees to prevent non cooperative behav-
iours on the part of the involved parties.
Part Five
Conclusions
Chapter 23
Conclusion
Power-sharing Theory: Lessons from the Complex
Power-sharing Project
John McGarry
Power-sharing theory has been closely linked in recent decades with conso-
ciational theory. The central idea in consociational theory is that if ethnically,
religiously, or linguistically-divided polities are to enjoy political stability, seg-
mental leaders must share power. Consociationalists believe in the potential of
statesmanship. They think that segmental leaders can engage in the politics of
compromise, and persuade their followers to follow suit. Consociational theory
was developed in the 1960s, in response to the dominance of pluralist thinking,
which held that political stability required cross-cutting social cleavages, and
which gave less weight to political will and acumen. Consociationalists assume
that identities in segmented societies are resilient, and not susceptible to short
term transformative (integrationist or assimilationist) enterprises, although they
do not necessarily endorse diversity as a priori desirable, or think that these iden-
tities are primordial and fixed.
Consociational theory is inescapably associated with the Dutch political sci-
entist, Arend Lijphart, whose work developed from a study of his native Neth-
erlands, and from comparative work on Belgium, Switzerland, and Austria (Li-
jphart 1968; 1977). However, many others have espoused consociational ideas,
including Lewis (1965), Nordlinger (1972), McRae (1974), and Lehmbruch (1993),
and consociational theory has been applied to different polities in all parts of the
world. Lijphart’s contribution has been such that he was elected President of the
American Political Science Association in 1996. Consociational theory is now
one of the most influential theories in comparative politics.
Consociational theory is also widely criticized. Most criticisms are from an
integrationist perspective, which sees identities as malleable rather than resilient,
and which is distrustful of power-sharing among segmental elites. Integrationists
believe that agreements among segmental elites are likely to be unattainable or
unworkable. Agreements that are attained, it is argued, are likely to be counter-
productive, as they allocate political resources to elites who have an interest in
maintaining division.
692 John McGarry
Consociations are also seen as lacking central democratic virtues: because all
segments are included in government, there is allegedly no democratic opposi-
tion to hold the government to account or to pose as an alternative government
(Brass 1991: 339).
Integrationists seek to bypass or constrain segmental elites in various ways.
‘Bottom-up’ social transformationists demand that they be challenged by those
in civil society who are committed to integration: particularly trade unions, civic
associations, and non-segmental political parties (Taylor 2001: 47; Bew and Pat-
terson 1990: 217; see also Bew and Patterson 1985: passim). Legal integrationists
seek to constrain ethnocentrism through bills of rights and human rights codes.
The prominent American political scientist Donald Horowitz offers an ‘integra-
tive’ version of power-sharing as an alternative to consociation. Horowitz’s cen-
tral conflict-regulating device is the alternative-vote electoral system in hetero-
geneous constituencies. It, in Horowitz’s view, requires politicians to pool votes
among different groups in order to win seats, which induces them to moderate
their politics. Vote-pooling promotes power-sharing because it facilitates inter-
ethnic coalitions of moderates. According to Horowitz and his supporters, such
integrative power-sharing coalitions are more likely to work than consociations
because their members are more likely to compromise. As rational politicians
normally opt for minimum-winning coalitions, they are also more likely to allow
for democratic oppositions.
In this chapter, I discuss the extent to which the Complex Power-sharing Proj-
ect enhances our understanding of power-sharing theory. Part One discusses
what consociationalists can learn from the project, while Part Two discusses
what integrationist critics of consociation, including those who support integra-
tive power-sharing, can learn.
I Lessons for Consociationalists

Many of the early prominent consociational theorists, including Lijphart and
Nordlinger, developed their understanding of power-sharing from a study of the
small western European democracies of the Netherlands, Belgium, Switzerland,
and Austria. Their work was developed between the 1960s and 1990s. These two
facts help to explain three gaps in traditional consociational theory that are laid
bare by the complex power-sharing project.
A The Neglect of a Positive Role for Outsiders in the Generation,

Implementation, and Operational Maintenance of Power-sharing
Settlements
Conventional power-sharing theory is overly ‘endogenous’ or ‘internalist’; it has
tended to treat states and regions as if they are sealed entities, relatively immune
from exogenous forces. This has produced two related problems. First, there has
been a tendency to downplay the importance of outside factors both when ex-
plaining how power-sharing settlements emerge, and when seeking to engineer
23 Conclusion – Power-sharing Theory: Lessons from the Complex Power-sharing Project 693
their creation. Of the much debated nine factors initially listed by Lijphart as
conducive to a consociational settlement, eight are endogenous (Lijphart 1977:
53-103). According to Lijphart, if a state’s warring factions perceive a common
threat from an external source, this will increase the prospects of internal solidar-
ity. The focus on this sole exogenous factor stemmed, arguably, from Lijphart’s
focus on Belgium, the Netherlands, Switzerland, and Austria, all of which have
been threatened by larger neighbours and have had at least partially consocia-
tional agreements during their recent histories. He may also have been influenced
by Lebanon, another of his favourite case studies, and one where consocation was
also facilitated by an external threat. Lebanon’s ‘pact national’ in 1943 arose from
a united desire on the part of the Lebanese to free themselves from the French. As
a corollary, Lijphart argued that external intervention that was one sided could
antagonize interethnic relations. Nowhere did he consider that outside forces
could facilitate consociation by benign rather than malign intervention, for ex-
ample, through mediation or by using incentives to induce or encourage rival
parties to reach agreement. Lijphart’s view on the role of outsiders was shared by
Nordlinger, who listed a desire to ward off external enemies as one of four crucial
elite motives for power-sharing, but who explicitly omitted the inclusion of ‘reli-
ance upon third parties as neutral arbitrators, mediators, or experts’ from his list
of conflict-regulating practices. Such intervention, according to Nordlinger, was
“rarely applied,” “generally ineffective,” or could be subsumed within his other
categories (Nordlinger 1972: 32).
Lijphart’s (and Nordlinger’s) focus on endogenous factors was, arguably, a
reflection of the Cold War era in which power-sharing theory was developed.
There was little evidence of benign intervention to consider. However, such in-
terventions have been a prominent factor in the cases covered in this project. A
number of different outside parties, including states, international organizations,
and even prominent individuals played positive roles in promoting recent settle-
ments, and in some cases were essential to the achievement of these settlements.
The UN, EU, and OSCE were successfully involved in preventative mediation in
Macedonia during the 1990s and, along with the USA, helped to mediate the
Ohrid Agreement of August 2001. In Regan’s view, Australia and New Zealand
played a ‘vital’ role in mediating a peace settlement in Bougainville in September
2001. In Northern Ireland, the government of the United States, and the former
US Senate Majority leader George Mitchell, mediated successfully between na-
The formal headings of these eight factors on their first elaboration were: (i) no ma-
jority segment; (ii) multiparty systems; (iii) small population size; (iv) appropriately
structured cleavages; (v) overarching loyalties; (vi) representative party systems; (vii)
geographical concentration of segments; and (viii) traditions of elite accommoda-
tion.
Lebanon’s power-sharing arrangements imploded in 1975, largely because of malign
external interventions. There was, then, nothing in Lijphart’s favourite case studies
to suggest that external intervention could be positive.
See Chapter 5.
694 John McGarry
tionalists and unionists, and between nationalists and the British government
(MacGinty 1997; Cox 1999). External mediators can contribute to agreements
through providing good offices or by providing peacekeepers who create space
for negotiations. They can also help instill confidence among weaker parties that
negotiations will be meaningful and commitments delivered on. One important
consequence of American diplomatic involvement in Northern Ireland was an
increase in the confidence of Irish republicans about the merits of negotiations.
Australia and New Zealand’s involvement as mediators in Bougainville appears to
have had the same positive effect on nationalists there.
In addition to mediating, outsiders may have resources that can be used to
influence the calculations of internal parties. The European Union has used the
card of membership to promote minority rights in a number of Eastern Europe-
an countries. There is some evidence that this ‘conditionality’ has had a positive
impact on the position of minorities, even in states like Macedonia, where the
prospects for membership are long-term (Bernier 2001; Kelly 2003). The United
Kingdom government was able to induce Northern Ireland’s unionists to sign
the Good Friday Agreement by making it clear that the alternative was deeper
cooperation with the Irish government in the governance of Northern Ireland
(McGarry and O’Leary 2004).
In extremis, external actors can use military force to bring rival parties to ne-
gotiations. When mediation failed to promote a settlement in Bosnia between
1991 and 1995, armed intervention by NATO prepared the way for negotiations
at Dayton in 1995. Kosovo’s interim settlement in 2001 was negotiated against
the backdrop of western military intervention. Afghanistan’s post-Taliban power-
sharing government was also a product of armed intervention, as was the broad-
based coalition established in Iraq in the wake of the US-led invasion of March
2003.
A second and related lacuna in traditional consociational theory is that it has
neglected the possibility of a positive post-agreement role for outsiders, i.e., one
that involves them both in the implementation and in the active operation of po-
litical settlements. In Macedonia, Mindanao, and Gagauzia, international organi-
zations have monitored, variously, demobilization, decommissioning, elections,
and human rights protections. In Northern Ireland, outsiders monitored the
implementation of paramilitary decommissioning and policing reform. In Bou-
gainville, Australia and New Zealand are to monitor the planned referendum on
self-determination. External agencies have trained new police forces, and funded
A 1994 document on the peace strategy, TUAS (reputedly an acronym for either
Totally Unarmed Strategy or Tactical Use of Armed Struggle), was explicit about the
importance of the American role, noting that “there is potentially a very powerful
Irish-American lobby not in hock to any particular party in Britain or Ireland” and
that “Clinton is perhaps the first US President in decades to be influenced by such a
lobby” (MacGinty 1997: 34).
For more on the role of international organizations in the case studies covered by the
Complex Power-sharing Project, see Chapter 15.
the integration of paramilitaries into normal life. More invasively, external rep-
resentatives are present as experts or arbiters in several of the institutions estab-
lished in Bosnia and Herzegovina and Kosovo, including Bosnia’s Supreme Court
and Central Bank. Both polities are presided over by external High Representa-
tives, and external agents are responsible for breaking political stalemates and for
security. The recent UN plan for Cyprus (the ‘Annan Plan’) envisaged a central
tie-breaking role for outsiders in the island’s supreme court. The fact that several
of the relevant states are signatories to international human rights treaties also
allows outsiders a role in monitoring and enforcing human rights protection.
To highlight the positive role that outsiders have played in promoting and
maintaining settlements is not to argue that such intervention is always right-
headed. In Bosnia, which approximates an international protectorate more than
a sovereign state, outsiders may well have over-intervened. Local leaders have
little responsibility. Arguably, they have reduced incentives to develop a politics
of cooperation, as this carries electoral risks, and as outsiders are available to
take unpopular decisions. The result, as one critic has put it, is a “faked democ-
racy” (Chandler 2000). Since Northern Ireland’s Agreement, the British govern-
ment has over-intervened on a number of occasions. Its decisions to suspend
the Northern Ireland Assembly in 2000, 2001, and 2002, rather than call new
elections, and to suspend the elections that were required by June of 2003, had
a number of negative effects on the political process. The unionist leader, David
Trimble, was encouraged to refuse to compromise with Sinn Féin, knowing that
he could do so without facing voters. The British government’s decision to post-
pone elections was motivated by a concern that these would produce gains for
radicals. However, the delay reduced the pressures on Sinn Féin to moderate, as
the only way it could expand its support was at the expense of the more moderate
nationalist party, the SDLP. Elections would have provided incentives for Sinn
Féin to address unionists’ concerns about decommissioning and police reform,
the two most divisive issues in Northern Ireland politics. Movement on these is-
sues by Republicans would have made it easier for unionists to reciprocate.
We should also be wary about arguments that there is a trend towards benign
international intervention, particularly one that is irreversible. The location of
interventions has been shaped by geopolitical and security considerations at least
as much as normative arguments. Thus, while several of the minorities examined
in this project have derived benefits from the attention of outsiders, others, such
In addition to the High Representative, three members of Bosnia’s Supreme Court

are from abroad. To ensure that Bosnia complies with the human rights standards
in the Dayton Accords, there is an Ombudsperson appointed by the OSCE in con-
sultation with the two entities and a Human Rights Chamber appointed by the two
entities and the Council of Europe. Bosnia’s Commission on Return of Properties has
three members who are appointed by the European Court of Human Rights, while
the job of overseeing the return of refugees is the responsibility of UNHCR.
For more details, see Chapter 21.
For more details, see Chapter 4.
696 John McGarry
as the Palestinians or Chechens, have not. While Ted Gurr has pointed to an
emerging regime of benign intervention in the post-Cold War era (Gurr 2000),
this has run up against countervailing pressures since 11 September 2001. In this
new environment, states that have been successful in depicting minority rebels
as terrorists have received a freer hand to resist their aspirations, as in Chech-
nya and Spain’s Basque country, while geopolitically fortunate groups (the Kurds)
continue to reap benefits.
B Consociational Theory and Self-determination Disputes

Consociational theory developed from a concern with religious and class divi-
sions in a number of European countries – the Netherlands, Belgium, Austria,
and Switzerland (Lijphart 1977). It has neglected the particularity of self-deter-
mination disputes, i.e., those that involve ethno-national communities focused
on contested homelands. Ethno-national minorities are usually described in the
consociational literature as segments, ethnic minorities, ethno-linguistic com-
munities, or territorially-clustered segments. A divided polity is usually equated
with a ‘divided society’ or a ‘plural society,’ with the inference that it is a single
society (or nation), albeit a divided one.
The problems that this can give rise to can be seen from Lijphart’s otherwise
masterly analysis of the Northern Ireland conflict (Lijphart 1975). Lijphart saw
the two groups in conflict as ‘Catholics’ and ‘Protestants,’ and the basis of the
cleavage as ‘religious,’ even though he was fully aware that the groups gave their
support to ‘nationalist’ and ‘unionist’ parties, respectively (Lijphart 1977: 136; Li-
jphart 1975). He argued that the key difficulty was the absence of support for
power-sharing among Protestants because they were capable of exercising hege-
monic power alone, and because they were disposed to Westminster majoritar-
ian practices rather than continental power-sharing norms (Lijphart 1975: 100).
This analysis was accurate, but limited. It overlooked the fundamental fact that
Northern Ireland’s Catholics, as Irish nationalists, were also opposed to inter-
nalist power-sharing within the United Kingdom. Radical Irish nationalists (re-
publicans) wanted national self-determination and a complete withdrawal of the
British state from Ireland, whereas moderate nationalists wanted any consocia-
tion to be internationalized, i.e., to have a linkage to Ireland, and a role for the
Irish government. Even if unionists had proposed a consociation, then, it would
have been insufficient for Irish nationalists. Moreover, a key reason why unionists
opposed a consociation was because they were British nationalists who were pro-
foundly concerned about Irish nationalists’ insistence on links with Ireland. They
also had no incentive to share power for most of the period after 1972, since the
default option was direct rule from Great Britain, their preferred nation-state. An
Lijphart’s masterpiece is entitled Democracy in Plural Societies, while Nordlinger’s

book is entitled Conflict Regulation in Divided Societies. For both of them, a single
polity is seen as comprising a single society, rather than two or more societies, which
is how ethno-national minorities would see it
agreement was reached in 1998 only because, as will be explained, it addressed

this self-determination dispute in a way that satisfied the minimal requirements
of both camps.10
One consequence of the failure to distinguish the needs of ethno-national
communities is that the emphasis in consociational theory is on who should ex-
ercise power at the level of the central government, when the emphasis in self-
determination disputes is on how much power should be exercised by the central
government – and about whether there should be one or more central govern-
ments (i.e., about whether there should be one or more states). While Lijphart
acknowledges that minorities should have autonomy as well as power-sharing,
Nordlinger refused to consider autonomy, because he feared it would lead to se-
cession (Nordlinger 1972: 31).11 He was also concerned that, because autonomous
regions would be invariably heterogeneous, autonomy would allow the region’s
“dominant segment to ignore or negate the demands of the minority segment”
(Nordlinger 1972: 31-32). These views on autonomy for minorities are virtually
identical to those held by integrationist opponents of consociation.12
Many of the settlements in the Complex Power-sharing Project have depended
on the self-determination dimension being squarely addressed. They have diluted
the sovereignty or the unitary nature of the state in ways that were not anticipated
by traditional consociational theory. In several of the recent cases, the main focus
has been on crafting autonomous institutions to allow ‘internal’ self-determina-
tion, and not on power-sharing within the central government. This is true of the
settlements in Bougainville, Gagauzia, and Mindanao. In all three cases, more-
over, the provisions for autonomy are asymmetrical, i.e., they have been con-
ceded only to the ethnic minority and not as part of state-wide arrangements.13
Asymmetrical autonomy is in particular tension with the notion of a nation-state,
as it suggests that the beneficiaries have a special status that distinguishes them
from the rest of the state’s population.
Even in Bosnia, where there are intricate provisions for power-sharing at the
centre, the emphasis in the Dayton settlement is on autonomous institutions for
Serbs, Bosniaks, and Croats. The state of Bosnia is a highly decentralized federa-
tion of two ‘entities,’ Republika Srpska and the Federation of Bosnia and Herze-
govina. The Federation of Bosnia and Herzegovina is, in turn, also decentralized,
10 Also see Chapter 4.

11 Nordlinger, like Lijphart, does not use the term ‘national minorities’ or ‘minority
nations,’ or even ‘ethno-national’ to refer to minority groups that are territorially con-
centrated and that seek self-determination.
12 The bias against minorities is clear enough. It does not seem to have occurred to
Nordlinger that the state as a whole is a territorial unit in which minority demands
could be ignored or negated, or that concerns about minorities within autonomous
regions could be managed in the same way that he suggested for majority-minority
relations in the state, i.e., through power-sharing.
13 This is one of the differences between the 2001 settlement in Bougainville and an
earlier unsuccessful settlement in 1976.
698 John McGarry
with most powers held at the level of its ten, relatively homogeneous, cantons. In
the case of Northern Ireland, Irish nationalists have been given substantial au-
tonomy, albeit within the context of a regional power-sharing government rather
than one that is under the control of their community.
In some of the cases covered in this project, where a minority in one state has
national kin in a neighbouring state, steps have been taken to foster cross-border
linkages. This is a dimension that is missing from traditional consociational theo-
ry, which is focused on prescriptions within discrete sovereign states. In Northern
Ireland, agreement required provisions for ratificatory referenda in both parts of
Ireland, as well as the establishment of a number of inter-state political institu-
tions. The most novel of the inter-state institutions was a North-South Ministe-
rial Council (NSMC), a body nominated by the Irish Republic’s government and
the new Northern Ireland premiers. The Agreement also established the British-
Irish Intergovernmental Conference (B-IIGC), the successor to the Intergovern-
mental Conference established under the Anglo-Irish Agreement. The B-IIGC
guarantees Ireland’s government access to policy formulation on all matters not
– or not yet – devolved to the Northern Ireland Assembly or the NSMC. In the
event of the collapse of the agreement, it will resume the all-encompassing role
its predecessor had prior to 1998.14
Ireland’s cross-border institutions are the most radical to be found in the re-
cent peace settlements. They are not only radical in their scope and formal na-
ture, but in their raison d’etre: they exist primarily to respond to the self-determi-
nation dispute at the heart of its conflict rather than to achieve some functional
benefit.
Cross-border links have been frowned on in other cases, including Macedo-
nia and Kosovo. However, Bosnia and Herzegovina’s two entities, the Federation
of Bosnia and Herzegovina and Republika Srspka, have used powers extended
to them under the Dayton Accords to conclude external confederal agreements
with Croatia and Serbia, respectively.15 Bosnia and Herzegovina’s central govern-
ment has also signed agreements with other Yugoslav states, including Croatia,
on cooperation in matters that are sensitive to its ethnic communities, includ-
ing higher education, science, and technology.16 Bougainville is now entitled to a
14 The final and weakest inter-state institution was a British-Irish Council, included at
the insistence of unionists, in which the Irish and British governments meet with
the governments of the various devolved legislatures in the United Kingdom. It has
been suggested that, in the event of a united Ireland, the British-Irish Council could
develop to perform for unionists the role that the NSMC is intended to perform for
northern nationalists, i.e., it would allow them to maintain links with their national
kin across the new state frontier.
15 Dayton mandated that the use of these powers should be consistent with Bosnia and
Herzegovina’s sovereignty and territorial integrity, and Bosnia’s international overse-
ers have been careful to ensure that they are not abused in a way that ends up dis-
mantling the state.
16 See Chapter 22.
permanent representative in Papua New Guinea’s delegation to its kin state, the
Solomon islands, which deals with new border agreements.
Even where crossborder links have not been included in initial agreements, as
in Macedonia and Kosovo, there have been subsequent proposals and exploratory
steps towards establishing them.17 In the UN Plan for Cyprus, Cyprus is required
under Article 16 to maintain “special ties of friendship” with both Greece and
Turkey. Its component states are permitted to conclude agreements on cultural
and economic matters with other states “provided that such agreements do not
cause prejudice to Cyprus, the authority of the [common state] government, or
the other [component state], and are compatible with the European Union mem-
bership of Cyprus”.
Beyond autonomy and cross-border links, some recent agreements have sup-
plemented autonomy arrangements with other provisions that recognize, at least
implicitly, the bi- or multi-national character of the state. Sometimes there is le-
gal or constitutional recognition that the minority constitutes a distinct people or
nation. The Preamble to the Law on the Special Legal Status of Gagauz states that
the law has the aim of “satisfying the national needs and preserving the identity of
the Gagauzes”. Article 4 of the Bougainville Peace Agreement includes as one of
the objectives of autonomy, the “expression and development of the Bougainville
identity”. After its Framework Agreement, Macedonia altered its constitution to
give its “communities” the “right to establish institutions for culture, art, science
and education, as well as for scholarly and other associations for the expression,
fostering and development of their identity”. While Macedonia has not provided
substantive autonomy to minorities, its Framework Agreement allows local au-
thorities “to place on front of local public buildings emblems marking the identity
of the community”.18 Kosovo’s Rambouillet Agreement, which was superseded by
NATO’s military intervention in 1999, provided for its “national communities” to
be able to “preserve and express their national, cultural, religious, and linguistic
identities”.19
In several cases, including Bosnia and Herzegovina, Kosovo, Macedonia, and
Mindanao, recent accords have taken traditional positions in defence of the
state’s sovereignty and territorial integrity. However, in three other cases, they
have acknowledged the minority’s right of self-determination, including its right
to secede, albeit under carefully specified circumstances. In the Law on the Spe-
cial Status of Gagauzia, the “people of Gagauzia” have been extended “the right of
external self-determination” in the event of a change of status of the Republic of
Moldova, i.e., its union with Romania.20 The Bougainville Peace Agreement pro-
vides for a ‘referendum on Bougainville’s future political status’ to be held within
17 Ibid.
18 See Chapter 21.
19 Rambouillet Agreement, Art VII.
20 Art 1(4).
700 John McGarry
ten to fifteen years, with one of the choices “independence for Bougainville”.21
The Agreement makes the holding of the referendum dependent on “good gover-
nance” on the part of the Bougainvilleans, and gives a veto on the wording of the
question and final decision-making authority (post-referendum) to the national
parliament of Papua New Guinea. In Northern Ireland’s agreement, the signato-
ries, including the UK government, accept the right of “the people of the island
of Ireland by agreement between the two parts respectively and without external
impediment, to exercise their right of self-determination on the basis of consent,
freely and concurrently given, North and South, to bring about a united Ireland”.22
In Jackson-Preece’s view, these various provisions “represent a considerable de-
parture from previous international practice in the area of minority rights and
self-determination”.23
These prescriptive arrangements – asymmetrical autonomy; cross-border
linkages and institutions involving sovereignty-pooling; constitutional recogni-
tion of peoplehood; and, in some cases, a qualified right to secede – are specifi-
cally addressed to self-determination disputes. They blur sovereignty in a number
of ways that are at odds with the classical independent and autarkic state associ-
ated with traditional consociational theory. This gap is worth noting, as virtually
all recent violent conflicts have involved self-determination claims. No recent
conflict, in fact, has been ended by power-sharing alone, although some, such as
Macedonia’s, more closely reflect traditional consociational arrangements than
others.24
C Consociational Theory and Transitions from War to Peace

Consociational theory developed from the study of relatively peaceful western
European democracies. This may explain why consociationalists have given little
attention to the special difficulties faced by territories undergoing transitions
from war to peace. Consociationalists have focused narrowly on the design of
political institutions, including executives, legislatures, and electoral systems. In
21 Bougainville Peace Agreement Art 309-310.

22 Unionist defenders and republican critics of the Agreement correctly point out that
this recognition of the right to self-determination is qualified, i.e., it is at odds with
the traditional republican view that the right be exercised within a single all-Ireland
unit on a majoritarian basis. Nevertheless, the recognition of Ireland’s self-determi-
nation in this qualified manner represented a major concession by the UK govern-
ment. It is the basis for the argument that the Agreement’s institutions rest on a
joint Irish act of self-determination rather than the revisable will of the Westminster
parliament, and the basis on which nationalists reject the legality of the UK’s post-
Agreement unilateral statutory enactment and use of suspension powers.
23 See Chapter 21. In Sudan, an agreement finalized in January 2005 gave the South the
right of secession after six years.
24 Only in Macedonia is there no provision for substantive autonomy, although the Al-
banian minority exercises municipal self-government.
violently divided polities, however, agreement on such institutions is unlikely to

be enough to secure stability. It is difficult to conceive, in fact, of these institutions
being accepted at all, or lasting, unless there is also a viable peace process that
builds confidence and all round security. If there is to be stability in post-conflict
polities, we have to consider prescriptive arrangements that address transitions
from conflict, measures that complement but stretch beyond the consociational
emphasis on political institutions.
The Complex Power-sharing Project, which addresses settlements in violently
divided territories, allows us to identify the sort of issues that are typically central
to peace processes, the difficulties they give rise to, and some ways to deal with
them. The issues include:
• Reform of the police and security forces;
• The handling of paramilitary offenders;
• Decommissioning of paramilitary weapons, and the dissolution of paramili-
tary organizations;
• Demilitarization;
• Transitional justice processes (e.g., truth commissions);
• Return of exiles;
• New human rights protection mechanisms/repeal of emergency laws; and
• Monitoring of ceasefires.
Often, agreement requires more than power-sharing or autonomy. Paramilitaries

that have not been defeated may insist on additional measures, such as leniency
for paramilitary offenders, an overhaul of the security sector, demilitarization,
and measures to protect human rights. In Bougainville, convicted prisoners were
pardoned. In Northern Ireland, it was decided to release prisoners on licence,
which meant that they could be re-incarcerated if they re-offended. The issue
of how to deal with paramilitaries who had offended but not been caught also
had to be faced in both cases. In Bougainville, there was a general amnesty for
all offences committed in relation to the conflict. In Northern Ireland, so-called
‘on the runs’ (OTRs) were allowed to escape incarceration, providing they went
through a formal trial process.
Reforms to security services, particularly the police, are another frequent
priority for rebel groups. In Bougainville, the agreement provided for the with-
drawal of Papua New Guinea’s army and its police riot squads from the island,
and limits were placed on their future deployment. Bougainville was permitted
to establish its own separate police force. In Mindanao, provisions were made to
integrate MNLF paramilitaries into the army. Such integration serves two impor-
tant purposes: it provides a role and employment for former paramilitaries who
might otherwise be inclined to subvert the peace process;25 and it facilitates a
more representative army (or police), thus helping to address the minority’s secu-
25 Rothchild argues that it was an insufficient international commitment to funding

the integration of paramilitaries into a productive life that helped to account for the
breakdown of peace in Liberia in 1995. Rothchild 2003: 11.
702 John McGarry
rity concerns. In Northern Ireland, the inclusion of radical provisions for police
reform was vital to acceptance of the agreement by nationalists. The parties to the
negotiations agreed to hand the issue over to an independent commission, and to
give it terms of reference that included a commitment to a police service that was
representative of the population, nationally impartial, human rights observing,
and routinely unarmed. Northern Ireland’s agreement also contained a general
commitment to demilitarization, i.e., the scaling down of British troop numbers
and fortifications.
Those defending the status quo, on the other hand, will be unlikely to agree to a
settlement unless there are reasonable guarantees that their opponents are com-
mitted to peace. They typically insist on the decommissioning of paramilitary
weapons, and often on the disbanding of paramilitary organizations. Nor is it
simply a matter of denying paramilitaries the use of such weapons: there are also
concerns that these weapons will become available to discharged ex-combatants
or to criminals who will use them to threaten social stability – a serious problem
in many African peace accords. Provisions for arms disposal formed an impor-
tant part of Bougainville’s agreement. Decommissioning of paramilitary weapons
was one, and perhaps the, central demand of Northern Ireland’s unionists, and
was also a vital issue for the British government.
It is often more difficult to reach agreement on these peace-related issues than
on political institutions based on power-sharing or autonomy. Sentiment is likely
to be especially polarized on these questions, as they touch vital issues of group
security. The creation of integrated militaries and police forces is likely to meet
resistance on all sides: from defenders of the status quo who regard the [ex-]
paramilitaries as terrorists and criminals, or as fifth-columnists whose plan is
to demoralize the security forces; and from the rebel leaders themselves, who
baulk at the loss of control and bargaining leverage involved in demobilizing their
fighters and transferring them to an integrated army. As Donald Rothchild has
noted, rebel leaders may see the demobilization of their troops and the creation
of an integrated army as a direct threat to their political survival. He had in mind
Angola, where UNITA leader Jonas Savimbi’s concerns on this matter prompted
him to abandon the peace process.26 Similar fears on the part of Bosnia’s Serbs led
to the negotiators at Dayton agreeing to let Bosnia and Herzegovina’s two entities
retain their own armies, a quite extraordinary state of affairs, unprecedented in
any other ‘federation.’
The Complex Power-sharing Project suggests some ways in which peace is-
sues can be managed. One way forward is to put intractable matters aside un-
til confidence has been built up. This allows progress on issues where there is
consensus. In Northern Ireland, the parties could not agree on policing reform
or decommissioning, but moved forward on prisoner releases and the establish-
ment of political institutions. A second way to proceed is to link concessions
from one side to concessions by the other. The advantage is that neither side loses
26 Savimbi expressed his concerns straightforwardly: “No leader in history that I have
known disarmed and stayed in power”. Cited in Rothchild 2003: 4.
face, and each has an incentive to deliver its side of the bargain. In Bougainville,
the government of Papua New Guinea linked parts of the autonomy agreement
to compliance on arms disposal by the Bougainvilleans. A linkage approach to
peace issues was implicit in Northern Ireland’s agreement, but there was a failure
to spell out details, including the nature of the changes to be made by the differ-
ent parties, and the precise dates by which they would be made.27 The result was
that both republicans and the British government were reluctant to take the first
step. By May of 2003, after a number of political crises, London made a more
explicit and detailed commitment to the linkage principle: demilitarization, the
strengthening of local police boards, and provision for ex-paramilitaries to serve
on such boards were to be linked to, and follow on, “acts of completion”, i.e., the
decommissioning of weapons by the IRA. A third approach to security issues,
already implied above, is to involve outsiders. Outsiders were crucial to breaking
stalemates on decommissioning and police reform in Northern Ireland. They also
assisted the retraining of police forces, the re-integration of paramilitaries, and
the monitoring of compliance with international human rights treaties.
The lessons outlined in the first part of this paper are interconnected. Divided
territories may benefit from external intervention, although they do not always
do so. Those with disputes based on self-determination require these issues to be
squarely addressed. Where there has been violence, the institutional accommo-
dation of self-determination claims will help to pave the way for peace. However,
peace and political stability need more than the creation of power-sharing or
autonomous institutions, they also need a viable peace process.
II Lessons for Critics of Consociational Theory (and Supporters of

Integrative Power-sharing)
It is not just consociationalists who can learn from the lessons outlined in Section
I of this chapter, but also integrationist critics of consociation, including support-
ers of integrative power-sharing. Writing in the post-Cold War era, integration-
ists writers have been more cognizant of the importance of international inter-
vention, and some of them have argued that outsiders should take a proactive
role in facilitating integrative institutions (Horowitz 2000: 277; Wimmer 2003:
137). However, integrationists have been even more likely than consociational-
ists to ignore the particularity of self-determination disputes. Consociationalists
support self-government for ethnic minorities, although they give this less at-
tention than power-sharing. Integrationists, however, strive for integration, not
27 The political parties, including Sinn Féin, were obliged under the terms of the Agree-
ment to use their influence to secure paramilitary decommissioning by May 2000,
in the context of the Agreement’s implementation. The government was obliged to
demilitarize, reform the police, and otherwise implement the Agreement, but there
were no deadlines for its compliance. The Agreement stated that an independent
commission on policing should report by September 1999, but it did not say when, or
indeed if, its recommendations would be implemented.
704 John McGarry
self-determination, and although some support territorial autonomy, they do so

primarily for classical Madisonian reasons, to proliferate points of power and to
establish cross-cutting cleavages across group boundaries rather than to accom-
modate national minorities. They tend to prefer the form of territorial autonomy
that exists in the United States, where there is “little coincidence between ethnic
groups and state boundaries” (Glazer 1983: 276). Integrationists have had little
to offer polarized polities undergoing transitions from war to peace. They are
either as focused on institutions (e.g., on electoral systems and bills of rights) as
consociationalists, or they advocate long-term social transformative agendas that
do not deal with the present. The Complex Power-sharing Project, however, sug-
gests three additional lessons for integrationist critics of consociationalism.
A Consociational Institutions Need Not Privilege Identity Groups or

Entrench Divisions
Integrationists believe that consociation is profoundly counterproductive. By giv-
ing power to ethnocentric elites, it is claimed, consociational reinforces, freezes,
or makes permanent the presumed sources of conflict. Consociation is said to
“institutionalize” divisions or to cast them in “marble” (Rooney 1998: 21; Wilford
1992: 31). Consociational institutions, it is said, allow ethnic elites the resources
to promote ethnocentrism, and to outmanoeuvre integrationist politicians who
preach a more progressive politics based on class, gender, or the environment.
Consociational agreements have been criticized as ethnic or sectarian “pacts”
against those who wish to transcend difference, and as “solidify[ing] intracom-
munal networks,” when the goal should be to promote “inter-communal associa-
tion” (Taylor 2001: 47; McGarry 2001; 134, n.39).
Underlying these views is the belief that consociations ‘privilege’ ethnic over
non-ethnic and trans-ethnic identities. The criticism is not groundless, and there
is evidence from the Complex Power-sharing Project to support it. Under the
Dayton Accords, Bosnia and Herzegovina’s government is presided over by a
rotating presidency, based on one Bosniak and one Croat from the Federation
of Bosnia-Herzegovina, and one Serb from Republika Srpska.28 The indirectly
elected upper chamber of the federal legislature is comprised of five Bosniaks
and five Croats from the Federation of Bosnia-Herzegovina, and five Serbs from
the National Assembly of Republika Srpska. The presidency and vice-presidency
of the Federation of Bosnia and Herzegovina rotates between a Croat and a Bos-
niak. Citizens who do not want to define themselves ethnically are barred from
all of these offices. The institutions also convert Serbs who live in the Federa-
tion of Bosnia and Herzegovina, and Bosniaks and Croats who live in Republika
Srspka, into second-class citizens, and work at cross-purposes with the inter-
national community’s expressed aim of encouraging the ethnically cleansed to
return home.
28 Constitution of Bosnia and Herzegovina 1995 Art V.

Similarly, parts of Northern Ireland’s Agreement privilege unionism and na-

tionalism over other forms of identity. Members of Northern Ireland’s Assembly
are required to designate themselves as ‘unionists,’ ‘nationalists,’ or ‘others’. The
election of Northern Ireland’s First and Deputy First Ministers requires concur-
rent Nationalist and Unionist majorities, as well as an overall majority. The pas-
sage of important laws requires either such a concurrent majority, or a weighted
majority – 60% in the Assembly, including at least 40% of both registered Na-
tionalists and Unionists. While Northern Ireland’s voters have shown no signs
of adopting new (non-unionist and non-nationalist) identities for over a century,
such rules create disincentives for them to change their behaviour. There is an
incentive for voters to choose nationalists or unionists, as members from these
groups will, ceteris paribus, count more than ‘others’ or be more pivotal. The rules
have the effect of pre-determining, in advance of election results, that national-
ists and unionists are to be better protected than ‘others’. The ‘others’, if they were
to become a majority, would be pivotal in the passage of all normal legislation,
but nationalists and unionists would have more pivotality in any key decision
requiring cross-community support (McGarry and O’Leary 2004: 33).
Such privileging of particular identities occurs elsewhere also. During Cyprus’s
brief consociational era in the early 1960s, citizens were required to register on
a Greek Cypriot or Turkish Cypriot electoral roll, regardless of whether they
were from another ethnic community or none. There was to be a Greek-Cypriot
president elected by Greek Cypriots and a Turkish Cypriot president elected by
Turkish Cypriots. The Lebanese National Pact of 1943 gave rise to an informal
convention whereby the presidency, prime-ministership, and speakership were
entrusted to a Maronite Christian, Sunni Muslim, and Shia Muslim, respectively.
Not only were these three groups privileged over other groups, but the Maronites
were the most privileged – as the presidency, at least until 1989, was the most
powerful office. A formal provision gave Christians six out of every eleven seats
in Lebanon’s legislature. The Lebanese experience reveals another disadvantage
of pre-determining the beneficiaries of consociational arrangements: it does not
allow for demographic change in which the beneficiaries share of the popula-
tion increases or decreases. This flaw contributed to the Lebanese civil war that
broke out in 1975: the growing Shia population felt discriminated against by the
country’s political institutions.
The Complex Power-sharing Project also shows, however, that such corporate
mechanisms are not intrinsic to consociational design. Consociational institu-
tions may also be liberal rather than corporate in nature: i.e., they may reward
any party with electoral support, and not just ethnic parties (Lijphart 1995). In
Northern Ireland, all legislative seats are rewarded according to the PR-STV
(proportional representation-single transferable vote) electoral system. It is not
true, as the Dawishas recently claimed in the prestigious journal, Foreign Affairs,
that there are “set aside seats for Catholics and Protestants” (or for nationalists
and unionists for that matter) (Dawisha and Dawisha 2003: 45). Citizens vote on a
common roll; vote for any candidates or parties they prefer; can vote across blocs,
and can express first or lower-order voting preferences outside their blocs. So the
706 John McGarry
election of Assembly members (MLAs) does not privilege particular identities.

With the exception of the First and Deputy First Ministers, ministers win office
through an allocation algorithm, the d’Hondt method, that is ‘difference-blind’: it
operates according to strength of representation won by parties in the Assembly,
not their national identity.29 South Africa’s transitional power-sharing arrange-
ments (1994-1996) allocated deputy presidencies and cabinet seats on the basis of
party strengths and not ascriptive characteristics.
Such liberal consociational mechanisms may be more conducive to the emer-
gence of new parties, including parties that are non-ethnic or trans-ethnic, than
the Westminster system, which is sometimes celebrated for its integrationist out-
comes. The proportional electoral systems associated with consociation allow
parties to win seats with much smaller thresholds than are normally required
under the Westminster electoral system of single-member plurality. As a result,
voters are less likely, ceteris paribus, to consider voting for a new party a waste
of time. By allowing for the ranking of preferences, the Single Transferable Vote
variant of PR also provides an opportunity, though no guarantees, for non-eth-
nic or trans-ethnic voting on lower preferences (O’Leary 2001: 73-74). An execu-
tive constituted by the d’Hondt system allows parties seats in government with
a much smaller share of seats in the legislature than is normally required in the
Westminster system. This means that new parties have a better chance to pro-
mote their visibility, influence public policy, and demonstrate to their supporters
that voting for them is a meaningful exercise. An alternative, the Sainte-Lague
mechanism, is even more advantageous for minorities (McGarry and O’Leary,
1995a: 373-375). In addition, as Jackson-Preece makes clear in her chapter in this
volume, it is perfectly possible to design consociational and autonomy settle-
ments so that they provide liberal protections for individuals, including those
who regard themselves as belonging to no ethnic community, and so that they
outlaw discrimination on ethnic grounds. A constitutional commitment to such
protections is no guarantee, of course, that they will be respected in practice, but
this is also true of settlements that are not based on consociation or autonomy.
Most modern consociationalists eschew corporate devices and prefer liberal
rules that equally protect whatever groups emerge in free elections. They prefer
“self-determination to pre-determination” (Lijphart 1995). Liberal consociation-
alists understand that parties to consociational pacts may make entrenchment
29 This fact has not stopped one critic of the Agreement’s rules from asserting that
d’Hondt privileges certain identities. Peter Emerson, the director of the de Borda
Institute, advocates the replacement of the d’Hondt rule for electing the executive
and its replacement with PR-STV, so that “all assembly members could participate
on an equal basis without using any sectarian labels”. The fact that d’Hondt treats all
members equally and does not require them to use any labels, sectarian or otherwise,
seems to have escaped him. See Emerson 2003. Another integrationist critic of the
Agreement goes further, incorrectly asserting that the Agreement privileges particu-
lar parties. Apparently the “terms of the Agreement require members of Sinn Féin to
be in the executive … [it] provides members of Sinn Féin with the right to be in the
government” (Wilson 2003).
deals, i.e., settlements that institutionally represent (and privilege) certain identi-
ties, and that they may do so both for self-interested reasons and because they
have genuine existential anxieties about the security of the communities they
represent. However, liberal consociationalists think that it is usually desirable
and possible to protect groups without corporatist principles. Thus, Brendan
O’Leary and I have argued that, under the review process provided in Northern
Ireland’s agreement, the parties should make changes that would remove as many
corporate principles as possible. We have recommended the use of the d’Hondt
formula for the nomination of the First and Deputy First Minister, with a built-in
safeguard to ensure that neither Nationalists nor Unionists could assume both
positions (see McGarry and O’Leary 2004: Appendix A). We have also recom-
mended simplifying and changing the current rules used for the passage of ‘key’
legislative measures to a simple weighted majority of at least 60% of MLAs (Mc-
Garry and O’Leary 2004). In the context of a new federal Iraq, we think that
its federal government should be consociational in nature, and should be based
either on Iraq’s regions, or on each party’s share of seats in the legislature, rather
than on pre-determined ethnic categories (McGarry 2003b; O’Leary 2003).
B Consociational Institutions Are Compatible with Democracy, Including

Democratic Oppositions
Another fundamental criticism of consociation is that it is deficient in democratic
virtues. One of its critics describes it as a “macabre” parody of “real democracy”
(McCartney 2000), while another believes that “a fully-developed consociational
system is inherently undemocratic” (Brass 1991: 339). There are two parts to this
complaint. First, consociation with its emphasis on inclusive government, is
said to be bereft of opposition. The price of consociation allegedly involves “aban-
doning a viable opposition politics” (Brass 1991: 334; also see Horowitz 2000:
256-257). If everyone is in office, the criticism goes, who is there to hold the gov-
ernment to account, or to pose as an alternative government? Supporters of inte-
grative power-sharing, which is based on an interethnic but minimum-winning
coalition, argue that one advantage of this is that it allows for opposition. Second,
as consociation focuses on negotiations and accommodation among elites, it is
seen as inconsistent with the development of a modern participatory democracy,
which does not restrict civic responsibility to voting in elections or even mem-
bership in political parties, but encourages active and ongoing political engage-
ment through a wide range of voluntary associations.30
There is some truth in both criticisms. However, the charge that consociational
executives lack democratic opposition is only (partially) accurate when levelled
30 There is a third complaint about the democratic failings of consociationalism, i.e.,

that it infringes the democratic principle of majority rule and that majorities should
be able to govern, even if they are from one bloc in a divided society. Both supporters
of consociational and integrative power-sharing are agreed that majority rule of this
sort would be disastrous.
708 John McGarry
at ‘grand,’ or what O’Leary calls ‘complete,’ consociational executives, i.e., coali-

tions in which all major parties from all major segments are represented (O’Leary
2005). Even in this case, small parties may be in opposition. While Northern Ire-
land’s first post-Agreement government comprised all 4 major parties, with 89
of the Aseembly’s 108 MLAs, this still left 5 parties, and a total of 19 MLAs, in
opposition. Backbenchers from the parties in government may also hold minis-
ters from other parties accountable for their actions. To facilitate this, committee
chairs and ministerial portfolios in distinct functional areas can be allocated to
different parties, as happens in Northern Ireland. Consociational executives may
also be ‘concurrent’ (restricted to parties commanding a majority but not total
support within their respective segments) or ‘weak’ (if at least one of the segmen-
tal parties in office commands only plurality support within its group). They may
also be based on some segments, but not on all. Thus, Israel has had consocia-
tional arrangements between Ashkenazi and Sephardic Jews, while excluding Ar-
abs. In all of these cases, significant parties may remain outside the government
and criticize its policies. Consociational arrangements may even exist within a
single governing party, if that party is internally a multi-segmental coalition, and
if the executive is comprised of representatives of the different segments. Such a
‘consociational party’ can govern within the conventional Westminster model of
government and opposition, as has occurred in Canada and India, and arguably
in South Africa also. It should also be noted that the majoritarian Westminster
model, sometimes positively contrasted with consociation in this respect, does
not necessarily provide strong oppositions. To its defenders, a marked advantage
of the plurality electoral system used in the Westminster model is that it pro-
motes ‘strong’ government by converting electoral minorities into strong legisla-
tive majorities. Indeed, it is so successful in doing this that it can often reduce
the opposition to an ineffective rump. An extreme case occurred in the Canadian
province of New Brunswick in 1987, when the Liberals won every seat in the leg-
islature with 60.4% of the vote (Hyson 1998).
It is true that consociational settlements often rely on the capacity of leaders
to make compromises, and to persuade their followers to accept these, although
such acceptance requires leaders who are authentic. And consociationalists cor-
rectly point out that active participation from civil society may lay bare divisions
and make it difficult to achieve agreement, as happened in Canada between 1987
and 1992 (Russell 2004). However, there is evidence from the Complex Power-
sharing Project that the shaping of consociational settlements, their ratification,
and their aftermaths need not be the exclusive property of political elites. In
Northern Ireland, policing reform resulted from a widespread consultation pro-
cess that involved 2,500 written submissions and public meetings in every district
council area, which were attended by 10,000 people and at which 1,000 spoke
(Independent Commission on Policing in Northern Ireland 1999). The Agree-
ment itself was ratified by simultaneous referenda in both parts of Ireland. As
such, it was a democratic advance on its predecessor, the Sunningdale Agree-
ment of 1973, which was an old-fashioned elite agreement arrived at in a private
British government retreat, and not ratified by referendum. It subsequently be-
came vulnerable after an election to the Westminster parliament in February 1974

returned a significant majority of unionist opponents.31 One of the main assets
of Northern Ireland’s current agreement against its critics is that it is a product
of direct popular endorsement. The current agreement also makes provision for
future referenda to decide the ultimate question of whether Northern Ireland
should remain part of the UK or become part of a united Ireland. In this respect,
it is similar to Bougainville’s settlement, which provides for a referendum on
whether or not it should remain part of Papua New Guinea. Northern Ireland’s
agreement also establishes a publicly funded Civic Forum alongside the elected
Assembly. This institution is made up of representatives of organizations outside
of conventional politics, and presents an opportunity for those who do not feel
represented by conventional political parties to have their voices heard. It has no
counterpart elsewhere in the (non-consociational) UK, including in the newly
devolved regimes in Scotland and Wales.
C Consociationalism, Integrationism, and Realpolitik

Supporters of integrative power-sharing and social transformation argue that the
consociational approach to conflict management is impractical. Horowitz main-
tains that consociationalists are overly optimistic: ethnic leaders in seriously di-
vided polities are unlikely to agree on consociational institutions, and unlikely to
make them work. This is particularly so as consociations are said to entail grand
coalitions, i.e., inclusive executives that contain not just moderates from previ-
ously warring ethnic segments but also radicals. Consociations, Horowitz con-
cludes, are only likely to work where they are unnecessary, that is, where there are
31 The Sunningdale agreement retained the support of a significant majority of the As-
sembly elected in 1973, and the unionists elected to Westminster in February 1974
had no mandate to challenge the earlier Assembly elections. Nonetheless, the West-
minster election began a process of events that resulted in the collapse of the Assem-
bly and Agreement in May of 1974. Had the Sunningdale Agreement been approved
in a referendum, and particularly referenda in both parts of Ireland, it may well have
proved more durable. Its supporters would have argued that a referendum is a more
accurate guide to popular endorsement of the Agreement than an election, particu-
larly one to the UK parliament rather than the Northern Ireland Assembly, and par-
ticularly one held under SMP rather than PR. They would have argued that the result
of a referendum can legitimately only be undone by a subsequent referendum. It is
very doubtful that even the spineless British Secretary of State, Merlyn Rees, would
have permitted such a second referendum (For Rees’s lack of backbone, see O’Leary
in McGarry and O’Leary 2004: Ch. 6). This is partly because of the widely accepted
convention that referenda results should be lasting and that one should not quickly
resort to a second referendum to undo the results of the first, although the Irish re-
cently breached this convention in 2002 in relation to the Treaty of Nice. Even had
Rees permitted such a referendum, it is not at all clear that the anti-Agreement side
would have won. One can make a reasonable case, therefore, that one of the biggest
mistakes of the designers of Sunningdale’s consociations was their failure to seek
public endorsement.
710 John McGarry
only mild divisions. They “are more likely the product of resolved struggles or of
relatively moderate cleavages” (Horowitz 1985: 256). The prospects for integrative
power-sharing are better, in Horowitz’ view, because it is based on cooperation
among moderates.
By contrast, transformationists accuse consociationalists of pessimism. Con-
sociationalists are said to be primordial pessimists, to exaggerate the depth, resil-
ience, and uni-dimensional nature of social divisions, and particularly, to down-
play the capacity of humans to develop new transcendent identities. Ethnicity is
taken by consociationalists as an ‘objective factor’ rather than a choice made by
people (Brass 1991: 338; Taylor 2001: 39). Consociationalists are said to uncriti-
cally accept “the primacy and permanency of ethnicity,” and to convey a “rather
bleak view of humanity” (Taylor 1994: 163; Wilford 1992: 31). In Taylor’s view,
consociationalists ignore the power of human agency, the ability of people to
reject or change their identity: “the point that consociationalism has not grasped,
but that has been central to both liberalism and Marxism, is that human freedom
is a power, a Promethean force” (Taylor 2001: 40). Transformationists prioritize
policies to promote social integration and to remove economic inequalities be-
tween groups lest these augment divisions. They favour a bottom-up approach to
conflict resolution that sees ethnic elites being challenged by groups in civil so-
ciety, including trade unions, peace and conflict resolution groups, and non-eth-
nic political parties that stress cross-cutting social divisions and the construction
of transcendent identities. Some transformers see consociation as incompatible
with transformation, and reject it tout court, but others appear to accept that
transformation can be achieved in spite of consociation. Social transformation is
particularly popular with international NGOs that work in divided polities. They
tend to regard local elites as ethnocentric and backward, and seek to transform
the polities into the more harmonious versions they come from. Despite NATO’s
role in the Dayton Accords, which provided for consociational institutions, most
international (inter-governmental) organizations involved in Bosnia are social
transformationist in perspective. Unhappy with local leaders, they have restrict-
ed Bosnia’s autonomy, while simultaneously funding civic associations that reject
these leaders (Chandler 1999: 138).32
What does the Complex Power-sharing Project suggest about consociation,
integration, and realpolitik? It appears to offer some support to the integration-
ist perspective. Of the three cases of consociation (Northern Ireland, Bosnia and
Herzegovina, and Macedonia), none are thriving. Northern Ireland’s consocia-
tional institutions are currently suspended. There is a widely-held expectation
that Bosnia and Herzegovina would fall apart in the absence of international
32 Local supporters of social transformation welcome this. One, the Marxist Bogdan
Denitch, is so disparaging of Bosnia’s local elites that he appears to believe Bosnia’s
international authorities are more representative of the true interests of the Bos-
nian people than those the Bosnian people vote for (cited in Chandler 1999: 136). For
Denitch, the international authorities seem to have assumed the former role of the
Communist party.
intervention. Macedonia’s power-sharing arrangements are threatened from

within and without. Unsurprisingly, supporters of integrative power-sharing and
social transformation see these cases as vindicating their analysis (see Wilson
and Wilford 2001). However, consociationalists can respond to integrationists in
five ways.
First, consociationalists do not claim that consociation will always work. Con-
sociationalists are not as ‘optimistic’ as Horowitz and his supporters suppose. A
substantial part of the consociational literature is devoted to outlining the con-
ditions that make success more or less likely. Consociations that are largely im-
posed, like Bosnia’s, face more serious problems, ceteris paribus, than those en-
dogenously arrived at. What consociationalists do claim is that in deeply divided
polities, consociation is likely to be the only workable and normatively acceptable
alternative to partition.
Second, a consociationalist might respond that the ‘failure’ of these consocia-
tions has been exaggerated. In each of the three cases, consociation (plus the
explicit treatment of self-determination and security issues) has helped to bring
civil wars to an end. In Bosnia and Herzegovina, the numbers killed in the con-
flict ended by Dayton range between 160,000 and 250,000, while 2.5 to 3 million
people were displaced.33 Peace since 1995 is importantly related to NATO’s pres-
ence, but it has also been facilitated by a political settlement that gives the differ-
ent communities their second, if not their first, preferences. Northern Ireland’s
peace process, with all its attendant difficulties, has also been associated with a
highly significant reduction in political violence. In the seven years that preceded
1994, the year in which republican and loyalist paramilitaries declared ceasefires
and opted for constitutional negotiations, the total loss of life because of the con-
flict was 622 persons. In the seven subsequent years, it has been 140, a fall in the
death toll of nearly four fifths, and this despite a major breakdown in the IRA’s
ceasefire in 1996-1997 and intermittent breakdowns in the loyalists ceasefires
(McGarry and O’Leary 2004: 23). Not one soldier or policeman has been killed in
Northern Ireland since 1998, which makes it, statistically, one of the safest places
in the world for such employment. This decline in violence cannot plausibly be
attributed to the presence of the British army, as it was also present, indeed in
significantly greater numbers, when the violence was much worse. Such facts are
omitted in integrationist accounts, which emphasize stalemates, or which even
distort statistics to create the impression that violence is increasing rather than
decreasing.34
33 The higher figures for killings and refugees come from the US State Department, the
lower from the International Helsinki Federation for Human Rights. See, respec-
tively, http://www.state.gov/www/global/human_rights/1996_hrp_report/bosniahe.
html; and http://www.ihf-hr.org/viewbinary/viewhtml.php?doc_id=4743.
34 Thus Wilson and Wilford, integrationist critics of Northern Ireland’s consociational
agreement, ignore the profound decline in death rates since 1994, and the fact that
virtually nobody has been killed in interethnic clashes, or clashes between either
community and the security forces, since 1998. Instead, they argue that divisions ap-
712 John McGarry
Third, the difficulties that these consociations have experienced have been
related to disagreements over peace-process issues, rather than the consocia-
tions themselves. In Bosnia, one of the chief sources of disagreement concerns
the right of refugees to return to their original place of residence. In Macedonia
and Northern Ireland, there has been serious disagreement around the issue of
decommissioning. Northern Ireland has been polarized on the issue of police re-
form. There is, by contrast, broad support there for consociation itself.. Northern
Ireland’s institutions have been suspended, in 2000, 2001, and 2002, because of
Unionist dissatisfaction over the unwillingness of the IRA to decommission its
weapons, which is in itself a response to the British government’s tardiness in
implementing police reform and demilitarization. If these issues can be resolved,
in the way suggested above, the institutions themselves would be placed on a
more secure footing.
Fourth, it may be possible to engineer or re-engineer consociations to remove
or reduce sources of instability. As integrationists point out, one reason why con-
sociations have been difficult to achieve, and even more difficult to maintain, is
that they have relied largely on trust between leaders, and on voluntary states-
manship, qualities that are often unavailable in sufficient quantities. The criticism
has validity: traditional consociational thinking has been tacitly committed to
power-sharing as a by-product of interparty negotiations over government for-
mation. However, Northern Ireland’s agreement, especially if it endures, prom-
ises to publicize a technique that is not widely known, and that usefully resolves
the disputes that may arise between polarized parties when they must share out
ministerial portfolios (O’Leary, Grofman, and Elklit 2001). It is a technique for
speeding government formation after elections, one that conforms to the pro-
portionality principles of consociational thinking, facilitates power-sharing, and
meets many tests of fairness. This is the d’Hondt allocation process. This takes
advantage of the fact that divisor rules for achieving fair proportions can also
be used to determine the sequence in which parties should be entitled to nomi-
nate ministers. It has the decided advantage of obviating protracted negotiations
over how many ministries each party is entitled to, and over which portfolios
each party is entitled to. It also creates incentives for executive maintenance, as
it provides for the portfolios of parties exiting the executive to be automatically
re-allocated among other parties, including parties from other blocs. In North-
ern Ireland, while the DUP are committed to re-negotiating the agreement, they
have retained their ministerial portfolios throughout. Had the DUP been capable
of preventing the executive, or bringing it down, merely by refusing to take part,
it is very likely that this is what it would have done. Much of Northern Ireland’s
political instability has originated, by contrast, from the office of its co-premiers,
the First and Deputy First Ministers. These are not elected by the d’Hondt pro-
cess, but by a majority of the Northern Ireland Assembly, including a concurrent
pear to be getting worse. A survey is produced that shows “a sense of deterioration in

recent years and a diminishing optimism about future harmony” (Wilson and Wil-
ford 2003: 2-3).
majority of nationalists and unionists. They are, ironically, examples of what sup-
porters of integrative power-sharing recommend: institutions that require inter-
ethnic support, and that are designed to secure the triumph of moderates over
radicals. Resignations and threats to resign have been used by the occupants of
both positions to further their political agendas (see McGarry and O’Leary 2004:
27-29). Elsewhere, O’Leary and I have detailed ways in which the rules for select-
ing the First and Deputy First Ministers could be changed, under the Agreement’s
provisions for review, to make resignations less attractive, and thus to strengthen
the institutions’ stability. We have argued that both offices should be filled by
the d’Hondt procedure, with a proviso that both offices could not be held by
nationalists or unionists. If this happened, the First and Deputy First Ministers
would have the same incentives to keep their positions as the rest of the executive
(McGarry and O’Leary 2004: 48-53). The British government could also enhance
the stability of the institutions if it repealed its Suspension (Northern Ireland) Act
2000, which is, in any case, a breach of the formal British-Irish treaty incorporat-
ing the Agreement. The effect of this would be that an unwillingness to take up
the positions of First and Deputy First Ministers would result in elections, which
might give cause for reflection.35
Fifth, the Complex Power-sharing Project suggests that, when considered rel-
atively, consociation may be more realistic, and more consistent with stability,
than either integrative power-sharing or social transformation. Three consocia-
tional settlements have been negotiated in the cases covered here, in Northern
Ireland, Bosnia-Herzegovina, and Macedonia. This is at direct odds with Horow-
itz’s criticism that consociation is unachievable in deeply divided polities, and
has relevance only for polities with moderate divisions (Horowitz 1985: 572-573).
Consociations have also been negotiated recently in Afghanistan and Congo.
Iraq’s interim governing council is based on consociational principles, and a final
settlement there, and in Cyprus, is likely to be both consociational and federal.
By contrast, integrative power-sharing arrangements are not prominent in any of
our cases. They have not once been agreed to in negotiations between rival par-
ties. They were imposed on one occasion, the 2000 election of Republika Srspka’s
presidency, by Bosnia’s international authorities. However, for reasons that are
discussed below, this experiment was a failure, and was quickly abandoned.
There are a number of reasons why integrative power-sharing is less likely to be
negotiated, and/or less likely to bring stability, than consociational power-shar-
ing. One is that leaders in a divided society are more likely to settle on an electoral
system that allows them to win as they are, i.e., as leaders of ethnic communi-
ties, than on one that requires them to transform themselves into trans-ethnic or
non-ethnic elites. This is particularly true of radical elites, presumably those who
would be most prepared to resist the new arrangements. The adoption of the AV
electoral system that is recommended by supporters of integrative power-sharing
35 The positions of FM and DFM are also interdependent. The resignation or death of
one automatically triggers elections for both positions. The stability of the institu-
tions would be enhanced if the two positions were made independent.
714 John McGarry
is particularly unlikely when the elites involved in negotiations owe their positions
to a prior, proportional electoral system.36 If an integrative power-sharing system
is implemented, it is not as clear as Horowitz thinks that its coalition of moder-
ates will deliver stability. Excluded radicals can destabilize power-sharing institu-
tions. They may accuse included moderates from their bloc of treachery, which
may prevent the latter from making the compromises necessary for successful
power-sharing. Excluded radicals may engage in violence, creating a polarized
atmosphere that pressurizes moderates and makes compromise difficult. This is
what happened during Northern Ireland’s only experiment with a power-sharing
coalition of moderates: the Sunningdale experiment of 1973-1974 (Neuheiser and
Wolff 2003: 1-24; Wolff 2001). The coalition was attacked by radicals on both
sides. It found it difficult to reach substantive internal agreement, amidst mount-
ing violence, and collapsed after less than five months in office.
Inclusion in power-sharing coalitions, on the other hand, can make radicals
less extreme, because it provides them with opportunities to have their con-
cerns addressed constitutionally, and gives them a stake in the system. Inclusion
can strengthen the position of moderates within radical factions, a possibility
Horowitz and others appear to discount. This does not mean that the inclusion
of radicals in government is always a good idea. Radicals, as in Northern Ireland
in 1974, may be virulently opposed to power-sharing and committed to militancy,
in which case they will use their positions in government to destroy it. However,
contrary to Horowitz, it makes political sense to include leaders of radical parties
prepared to participate in power-sharing institutions on the basis of democratic
mandates and methods, particularly when they are waging internal battles with
their hawks on the merits of constitutional politics. This has been the British gov-
ernment’s policy towards Sinn Féin, whom Horowitz would see excluded from
the executive, since 1997. The policy has brought the benefits already referred to,
and it is difficult to imagine Northern Ireland’s peace process being so productive
without Sinn Féin (McGarry and O’Leary, 2004).37 Contrary to our intuitions,
perhaps, an inclusive coalition of rivals may be more consistent with stability
than an exclusive coalition of moderates.
36 As a consequence, the adoption of AV may be more likely in a region that is newly

democratizing.
37 The d’Hondt rule for executive appointment, already referred to, can also mitigate
the effect of having rival parties in government. With d’Hondt, no programme of
government has to be negotiated in advance between the parties entitled to govern-
ment. The design creates strong incentives for parties to take up their entitlements
to ministries, because if they do not then the portfolios go either to their ethnic
rivals or to their rivals in their own bloc. D’Hondt means that no vote of confidence
is required by the legislature either for individual ministers or for the executive as a
whole. These incentives have produced positive results in Northern Ireland (see Mc-
Garry and O’Leary 2004). D’Hondt can be combined with maximum autonomy for
individual ministers acting within their portfolios, a form of ‘power-division’ within
a power-sharing government. Such rules can only go so far, however. There is likely
to be a need for collective agreement on some matters, particularly the budget.
Another difficulty with integrative power-sharing in deeply divided polities

is that AV requires constituencies that are sufficiently heterogeneous. It is also
helpful if no single group enjoys majority status, as otherwise politicians from
this group will have little incentive to vote-pool, and the alternative vote will like-
ly deliver a victory to the ethnic majority. The problem here is that in many deeply
divided polities, especially those emerging from conflict, constituencies lack suf-
ficient heterogeneity. It may be possible to engineer this, through gerrymander-
ing constituency boundaries, or by imposing regional distribution requirements
in state-wide elections. However, as with the AV electoral system itself, extant
political leaders are not likely to endorse such changes.
In addition, the success of vote pooling relies on leaders making trans-ethnic
appeals, and on voters being prepared to vote for moderate politicians from other
ethnic groups. The assumption is that this will happen because it is required to
win elections. However, this appears to make integrative power-sharing depen-
dent on an already existing degree of moderation, the charge, ironically, that in-
tegrationists level at consociationalists. Where divisions are deep (or where con-
stituencies have a dominant group), ethnic leaders may well continue to make
ethnocentric appeals and voters may continue to vote ethnically, or both may
abstain. In fact, the imperative of leaders staying in the count under AV, i.e., get-
ting as big a first or second preference vote as possible, may dictate continuing
ethnocentrism. If this is so, and AV results in the election of leaders who are
not representative of the society, it offends democratic norms, and is likely to
be destabilizing. Proportionality norms, by contrast, better match the parties’
respective bargaining strengths and their conceptions of justice. They ensure that
groups are authentically represented, and represented according to their share of
the population. They are more likely, as a consequence, to ensure stability. One
proportional system (PR-STV) combines the benefits of proportionality with the
possibility of interethnic voting in lower preferences. PR-STV also allows intra-
bloc voting between supporters of moderate and radical parties, which may be a
more likely outcome in deeply divided polities than interbloc voting. Such intra-
bloc voting may encourage moderation, as it allows moderate voters to reward
radical parties for becoming more moderate (McGarry and O’Leary 2004: 31).
Some of the difficulties with AV can be seen from its use for the election of the
presidency of Republika Srspka in 2000. Republika Srspka is almost exclusively
Serb, although around one sixth of those who voted in 2000 were ex-residents,
presumably Bosniaks. This immediately highlights one difficulty with the alterna-
tive vote: a large Serb majority meant there was little prospect of anyone other
than a Serb being elected as president. The rules for the election were drafted by
the OSCE, not the locals. It appears to have calculated that the alternative vote
would facilitate the election of the moderate Serb candidate over a radical rival,
as a sufficient number of Bosniaks would use their lower preferences to support
the moderate Serb.38 As it turned out, lower preferences were rendered nearly
38 The OSCE proclaimed that “candidates for the Republika Srspka ... will have to look
beyond their narrow core of supporters and offer political platforms which appeal
716 John McGarry
superfluous, as the radical Serb candidate won almost a majority (49.8%) in the
first round. He was elected in the second round, on the basis of 35 transfers from
an eliminated fringe Bosniak party (BOSS). Only 3% of BOSS’s transfers went to
candidates from Serb parties, with the rest going to other Bosniak parties (Bose
2002: 233). In Bosnia’s permanent election law, adopted by its politicians in 2001,
the use of AV for the election of Republika Srspka’s president was dropped, and
replaced by a (corporate) consociational alternative.39 Even if it is accepted that
Republika Srspka was not propitious territory for AV because of its dominant
Serb population, this evidence is damaging for AV proponents in three ways: AV
did not promote interethnic voting; it did not appear to help the Serb moderate
candidate against the Serb radical;40 and Bosniak voters had little difficulty wast-
ing their transfers on Bosniak parties that had no prospect of winning.
What about the prospects of social transformation as a realistic alternative
to consociation in divided polities? Social transformation appears to rest on the
assumption that the divisions that are articulated by elites are artificial or su-
perficial. Ethnic political elites are not seen as genuinely representative of their
people. Transformative civic associations are seen, by contrast, as reflecting a
more authentic mass political culture, albeit one that may be located just below
the surface. However, there is little evidence from our cases that social transfor-
mationists better reflect popular aspirations than elected elites. In the elections
that took place in Bosnia and Herzegovina in 1990, in which non-ethnic parties
participated, and which preceded any violence, the three main ethno-national
parties won over three quarters of the vote (Bose 2002: 212). The only two par-
ties that were non-ethnic in nature, Ante Markovic’s Union of Reform Forces
and the Communist League of Bosnia and Herzegovina, won 5.4% and 7.5%, re-
to a wider audience … The preferential voting system will allow ALL citizens of Re-
publika Srspka to have a meaningful voice in the selection of their president’ (cited
in Bose 2002: 222).
39 Article 12.3 of the new election law states as follows: “The candidate from each con-
stituent people receiving the highest number of votes shall be elected. Among these
three candidates, one from each constituent people, the candidate receiving the
highest number of votes shall be elected President, and the two candidates receiving
the second and third highest number of votes shall be elected Vice Presidents”.
40 Bose suggests that AV may in fact have strengthened the radical (SDS) candidate
against the moderate. He bases this on the fact that the SDS candidate achieved a
significantly larger percentage of the vote from residents of Republika Srspka (60%)
than his party achieved in simultaneous elections to Republika Srspka’s parliament
and Bosnia’s parliament (44% and 48%, respectively) (Bose 2002: 230). However,
there are other possible reasons for the voting discrepancy, including the campaign
effectiveness of the main candidates. One could speculate, by contrast, that AV may
have provided an incentive for the radical candidate to moderate, as, in practice, it in-
creased the threshold for victory. However, if this was the result, it was not anticipat-
ed by supporters of AV. In any case, providing incentives to hardliners to moderate
can also be achieved by PR-STV, a preferential but proportional electoral system.
spectively.41 Susan Woodward, an American critic of the nationalist parties and

of consociationalism, remarked that the election results read “like a census of
national identities in the socialist period” (cited in Bose 2002: 212). Social trans-
formationists have responded by pinning responsibility for these results on Tito’s
allegedly perverse institutionalization of ethnic identities, his decision to convert
Yugoslavia into an ethno-federation in which ethnic rights were entrenched at all
levels (Bose 2002: 213; Snyder 2000; Bunce,1999).42 However, Bosnia’s divisions
preceded Tito. The first parties to emerge in the region under the Austro-Hun-
garian empire also reflected national/confessional cleavages.43 They emerged,
moreover, under a regime that was profoundly assimilationist in intent: it sought
to promote a common Bosnian identity to confront the spread of Serbian na-
tionalism among those of the Orthodox religion. This history suggests that Tito
responded to preexisting cleavages, not that he was responsible for creating them
(see McGarry and O’Leary 2003).
The durability of ethno-national cleavages is even more clear-cut in historic
Ulster, and then the Northern Ireland that was carved from it in 1921. Since the
extension of the franchise to the male working class in the 1880s, the region has
divided consistently into two rival ethno-national blocs, and the divisions have
become more pronounced after thirty years of interethnic violence (See McGarry
and O’Leary 2004: 303-313). Non-ethnic parties have performed abysmally, hard-
ly ever winning more than 10% of the vote among them.
Transformers sometimes respond by dismissing this electoral evidence as mis-
focused. A concentration on the share of the vote won by ethno-national parties,
it is argued, exaggerates the degree of polarization. It is said that citizens abstain
from electoral politics, presumably because they are sickened by ethnonational
squabbling;44 or that significant support for social transformation is to be found
among grassroots civic associations. However, there is not much evidence to sup-
port these claims either. Bosnia’s turnout since 1995 has been a little worse than
41 http://www.europeanforum.bot-consult.se/cup/bosnia/develop.htm
42 The blame has also been placed on the fact that the first democratic elections were
held at the republican level rather than at the Yugoslav federal level (see Linz and
Stepan 1992). This apparently gave ethnic republican-level elites an advantage over
their pan-ethnic Yugoslav counterparts. Apparently, had the elections been held first
at the federal level, pan-Yugoslav parties would have had the upper hand. The prob-
lem with this thinking is that minority nationalist parties also did very well around
this time in elections held in unitary states in Eastern Europe , i.e., where there were
no prior regional level elections. For a critique of Linz and Stepan’s reasoning, see
McGarry and O’Leary 2003.
43 The parties represented in the Bosnian parliament organized under Austro-Hungar-
ian rule in 1910 were the Serb National Organization, the Muslim National Organi-
zation, the Croat National Union, and, the smallest of the four, the Croat Catholic
Association (Bose 2002: 214).
44 The problem in Bosnia, according to the OSCE, is that “ordinary people are dissat-
isfied and disillusioned; they thus tend to withdraw from all segments of civic and
political life” (Cited Chandler 2000: 137).
718 John McGarry
the United Kingdom’s, about the same as Canada’s, and much better than the
United States, and none of these other societies have just emerged from civil war
or have fairly impotent institutions.45 Northern Ireland’s turnout is consistently
higher than any of the cases mentioned (McGarry and O’Leary 2004: 88). As for
Bosnia’s transformative associations, the “central problem” in the words of the
OSCE officer responsible for developing them was “how to encourage participa-
tion” in them(Cited, Chandler 2000: 149). A study of Bosnia’s civic associations
concludes: “There seems to be a large gap between the civil society associations
funded and supported by the OSCE, and other international institutions, and
Bosnian people. For the OSCE and other institutions, this gap demonstrates the
lack of a democratic culture in Bosnia. While few people are actively involved in
civil society associations, leaving them predominantly middle-class based, the
main Nationalist parties still easily attain the majority of the votes in elections”
(Chandler 2000: 150). The same can be said for transformative associations in
Northern Ireland and in other violently-divided territories (see Cochrane 2001).
All this suggests that social transformation may not be a realistic option in the
short to medium term. Where it is not, a more immediate strategy is required to
ensure that rival communities can live together in peace, and this would seem to
entail some form of accommodation among representative leaders. Transforma-
tionists are correct to point out that political institutions should be designed so
that they do not privilege existing identities and so that they allow for the emer-
gence of new interethnic or non-ethnic identities. This is a point that academic
supporters of consociation accept. Consociationalists are also happy to encour-
age the development of voluntary associations based on different interests and
identities. In fact, a period of successful consociational governance should lay the
basis for the organic development of normal politics and the related development
of such associations around bread and butter issues.
III Conclusion
The Complex Power-sharing Project provides three lessons for supporters of con-
sociational theory, although others, including supporters of integrative power-
sharing should also take note. First, outsiders can play a benign role in facilitating
and maintaining consociations. This does not mean that all such interventions
are benign, or that all well-intentioned interventions are right-headed. Second, if
ethno-national conflicts are to be effectively managed, self-determination issues
must be squarely grasped and addressed in the new institutions. Third, effective
conflict resolution requires not simply a focus on political (legislative, executive,
electoral) institutions but also attention to a range of issues connected with peace
45 Bosnia’s electoral turnout rate in the 1990s, when measured as a proportion of the
voting age population, i.e., not registered voters, varied between 52% and 58%. Dur-
ing this same period, the United Kingdom’s turnout rate varied between 57% and
69%, Canada’s between 55% and 57%, and the United States’ between 34% and 49%.
Data from http://www.idea.int/index.htm.
and security. These include police reform, demilitarization/ decommissioning,

and the protection of human rights. A failure to satisfactorily resolve any such
issue can destabilize and ultimately destroy political institutions, even if there is
widespread support for them.
The project also provides three lessons for critics of consociation, including
supporters of social transformation and integrative power-sharing. First, it is
possible to design liberal consociations, i.e., consociations that do not privilege
certain ethnic identities over other ethnic, non-ethnic or trans-ethnic identities,
and that are respectful of both individual and group rights. Second, consocia-
tion is consistent with strong democratic values. Third, consociation is neither
utopian nor primordially pessimistic, as its detractors suggest. Consociationalists
do not think that consociations are universally applicable panaceas, nor do they
believe that divisions are etched in stone. Rather, they argue that consociations
are more likely to be successful under certain conditions than others, that conso-
ciation may be the only normatively defensible alternative to partition, and that a
successful period of consociational governance can pave the way for the erosion
of divisions.
Biographies
Florian Bieber
Florian Bieber is a Senior Non-resident Research Associate of the European Cen-
tre for Minority Issues, based in Belgrade, and a recurrent Visiting Professor at
the Central European University, Budapest. He also teaches at the Regional Mas-
ters Programme for Democracy and Human Rights, Sarajevo. He received his
Ph.D. in Political Science from the University of Vienna. He published articles on
nationalism and politics in Southeastern Europe and edited Montenegro in Tran-
sition. Problems of Identity and Statehood (Baden-Baden: Nomos, 2003); (with
Džemal Sokolović) Reconstructing Multiethnic Societies: The Case of Bosnia-Her-
zegovina (Aldershot: Ashgate, 2001); and (with Židas Daskalovski) Understand-
ing the War in Kosovo (London: Frank Cass, 2003).
John Bradley
John Bradley is a Professor at the Economic and Social Research Institute (ESRI)
in Dublin. His research explores the impact of the Single European Market, re-
gional aid programmes (or Structural Funds), and Economic and Monetary
Union on the smaller, less developed peripheral states and regions of the Eu-
ropean Union (Greece, Ireland, Portugal, East Germany, Northern Ireland), as
well as the transition of former Communist countries of Eastern Europe to EU
membership. He has carried out a wide range of international consultancy as-
signments for the Irish government, the European Commission, and many other
international organizations.
Paul Cornish
Paul Cornish is head of the International Security Programme at Chatham House,
and has previously worked at Kings College London and the Centre for Inter-
national Studies in Cambridge. His areas of specialization include international
security and defence, post-conflict reconstruction, and the European security in-
stitutions. Professor Cornish’s recent publications include “NATO: the Practice
and Politics of Transformation”, International Affairs 80(1) (2004), and, as editor,
The Conflict in Iraq 2003 (London: Palgrave Macmillan, 2004).
722 Biographies
Farimah Daftary
Farimah Daftary is a Senior Non-resident Research Associate of the European
Centre for Minority Issues (ECMI) in Flensburg, Germany. She received her Mas-
ters of International Affairs (M.I.A.) with a specialization in East Central Europe
from the School of International and Public Affairs (SIPA), Columbia University,
New York in 1991. As a (Senior) Research Associate at ECMI from 1997 to 2002,
her research concerned various aspects of interethnic relations and minority is-
sues in Europe as well as the OSCE’s ‘Human Dimension’. From April 2002 to June
2003 she was affiliated with the EU Accession Monitoring Program (EUMAP) of
the Open Society Institute, Budapest, dealing specifically with the reports on mi-
nority protection in the EU candidate states of Central and Eastern Europe. She
has published a number of articles on Corsica’s autonomy arrangement, power-
sharing in Macedonia, and language issues in Slovakia. She has also co-edited two
books: Nation-building, Ethnicity and Language Politics in Transition Countries
(with François Grin) (Budapest: Open Society Institute/LGI, 2003) and Radical
Ethnic Movements in Contemporary Europe (with Stefan Troebst) (Oxford/New
York: Berghahn Books, 2003).
Eben Friedman
Employed as Research Associate at the European Centre for Minority Issues
(ECMI) in Flensburg, Germany, Eben Friedman completed the Ph.D. programme
in political science at the University of California, San Diego in September 2002.
His doctoral thesis, entitled “Explaining the Political Integration of Minorities:
Roms as a Hard Case”, drew on two years of field research in Slovakia and Mace-
donia. He also holds an M.A. in Political Science from the Johns Hopkins Univer-
sity, where he specialized in international relations. At ECMI, his activities focus
mainly on Macedonia and Roms.
Angela Hegarty
Angela Hegarty is Senior Lecturer in Law at the University of Ulster at Magee.
Born and raised in Derry, she graduated in law from QUB in 1985 and then quali-
fied as a solicitor. She obtained an LLM in Human Rights, Emergency Law and
Discrimination, also from QUB, in 1991. She is the Director of the LLB Pro-
gramme at Magee College, where she teaches Public Law and runs a Clinical Law
programme. Her research interests are in human rights, transitional justice, and
policing. She is currently engaged in research and writing about truth processes
and the Bloody Sunday Inquiry and to that end is a Research Fellow at the Tran-
sitional Justice Institute at the University of Ulster.
She is a past Chair of the Committee on the Administration of Justice, which
won the Council of Europe Human Rights Prize in 1998. She is a board member
of a number of human rights NGOs and is also a trustee of the Bloody Sunday
Centre. She was previously appointed to the Equal Opportunities Commission
for Northern Ireland, the Standing Advisory Commission on Human Rights, and
to the Northern Ireland Human Rights Commission.
Biographies 723
Jennifer Jackson-Preece is currently Lecturer on Nationalism in Europe at the
London School of Economics and Political Science. She has published widely in
the area of human and minority rights, including National Minorities and the Eu-
ropean Nation-States System (Clarendon Press, 1998); and Minorities and Their
Rights (Polity Press, forthcoming).
Priit Järve
Priit Järve Ph.D. is Senior Research Associate at the European Centre for Minority
Issues (ECMI), Flensburg, Germany. Priit Järve graduated from Tartu University,
Estonia. From 1989 to 1997, he was the Director of the Institute of Internation-
al and Social Studies in Tallinn, Estonia, specializing in political science. From
1995 to 1997, he also acted as the Plenipotentiary of the Estonian President to
the Estonian Roundtable on Minorities (1995-1997) and edited Estonian Human
Development Reports of 1995, 1996, and 1997 under the auspices of the UNDP.
Priit Järve joined ECMI in 1997. He has published on democratization of transi-
tional societies, security of small states, multiculturalism, language policies, and
interethnic relations in the post-Soviet area in American Studies in Scandinavia,
Nationalities Papers, Politiikka, Journal of Baltic Studies and in various ECMI
publications. As an ECMI staff member, he has done project work in Moldova,
Ukraine, Estonia, Latvia, and in the Kaliningrad oblast of Russia.
John McGarry
John McGarry is Professor and Canada Research Chair in Nationalism and De-
mocracy in the Department of Political Studies at Queen’s University (Kingston,
Ontario). He is the editor, co-editor, and co-author of several books, including
Essays on the Northern Ireland Conflict: Consociational Engagements (forthcom-
ing, late 2003); Northern Ireland and the Divided World (2001); Minority Nation-
alism and the Changing International Order (2001); Explaining Northern Ireland
(1995); and The Politics of Ethnic Conflict Regulation (1993). He has also published
in journals such as Ethnic and Racial Studies, Government and Opposition, Na-
tionalism and Ethnic Politics, Nations and Nationalism, Political Studies, Parlia-
mentary Affairs, Journal of Conflict Studies, and the Journal of Commonwealth
and Comparative Politics.
James Mayall
James Mayall is the Sir Patrick Sheehy Professor of International Relations and
Director of the Centre of International Studies at the University of Cambridge.
He is a fellow of Sidney Sussex College and of the British Academy. Amongst his
recent publications are (editor and contributor): The New Interventionism, 1991-
1994: United Nations Experience in Cambodia, Former Yugoslavia and Somalia
(Cambridge University Press, 1996), World Politics: Progress and Its Limits (Polity
Press, 2000); and (editor with Gene Lyons) International Human Rights in the
Twenty-first Century: Protecting the Rights of Groups (Rowman and Littlefield,
2003).
724 Biographies
Barbara Metzger
Barbara Metzger is currently an Associate at the Centre of International Studies
in Cambridge and received her Ph.D. in history from the University of Cambridge
in 2001. From 2001 to 2005, she was a Senior Researcher with the Cambridge
Carnegie Project on Complex Power-sharing and Self-determination Disputes
and has lectured at the University of Cambridge on human rights history, law
and institutions in the international arena. More recently, she has been prepar-
ing her thesis on the League of Nations and Human Rights for publication. Her
research and teaching interests are the history of human rights and international
governance.
Brendan O’Leary
Brendan O’Leary, B.A. (Oxon.), Ph.D. (LSE), is Lauder Professor of Political Sci-
ence and Director of the Penn Program in Ethnic Conflict. Recent co-authored
and co-edited books include Terror, Insugency and the State (2007); The Future of
Kurdistan in Iraq (2005); The Northern Ireland Conflict: Consociational Engage-
ments (2004); and Right-Sizing the State: The Politics of Moving Borders (2001).
O’Leary has acted as a constitutional and political advisor to governments, par-
ties and international organizations on and in Northern Ireland, Somalia, Kwa-
Zulu Natal, Nepal, and Kurdistan. He is presently running a Sawyer-Mellon semi-
nar series on Power-sharing in Deeply Divided Places.
Francesco Palermo
Francesco Palermo is Associate Professor of Comparative Constitutional Law in
the Faculty of Law of the University of Verona. He is also a Lecturer at the Uni-
versity of Trento, Senior Researcher at the European Academy of Bolzano/Bozen
and Visiting Professor at Vermont Law School. He obtained his Ph.D. in com-
parative constitutional law at the University of Innsbruck. Main research fields
are comparative, European, and sub-national constitutional law, minority rights,
and European integration.
Anthony Regan
Anthony Regan is a Fellow in the State, Society and Governance in Melanesia
Project in the Research School of Pacific and Asian Studies at the Australian Na-
tional University, Canberra, Australia. His main field of research is the law and
politics of constitutions, conflict and reconciliation, and the design of the state
as part of post-conflict political settlements. He lived and worked in Papua New
Guinea and Uganda from 1981 to 1996, and has been living in Bougainville since
August 2002. He has been a constitutional adviser to the governments of Papua
New Guinea and Uganda. He has worked in Bougainville since 1981, and advised
Bougainvillean parties to the negotiations with the PNG government on the po-
litical future of Bougainville (1999–2004). He has had some involvement in the
Solomon Islands and Sri Lankan peace processes, and the East Timor constitu-
tion-making process.
Biographies 725
Andrew Reynolds
Andrew Reynolds, Associate Professor of Political Science, received his M.A.
from the University of Cape Town and his Ph.D. from the University of Califor-
nia, San Diego.
Reynolds is interested in democratization, constitutional engineering, and
electoral politics and has worked for the United Nations, the International Insti-
tute for Democracy and Electoral Assistance (IDEA), the National Democratic
Institute (NDI), the International Republican Institute (IRI), the Organization for
Security and Co-operation in Europe (OSCE), and has served as a consultant on
issues of electoral and constitutional design for Afghanistan, Angola, Burma, Fiji,
Guyana, Iraq, Indonesia, Jordan, Kenya, Liberia, Northern Ireland, Sierra Leone,
South Africa, the Sudan, and Zimbabwe. He has received research awards from
the U.S. Institute of Peace, National Science Foundation, and Institute on Global
Conflict. Among his books are Electoral Systems and Democratization in South-
ern Africa (Oxford, 1999); Election ’99 South Africa: From Mandela to Mbeki (St.
Martin’s, 1999); Elections and Conflict Management in Africa (USIP, 1998), co-
edited with T. Sisk; and the edited book The Architecture of Democracy: Constitu-
tional Design, Conflict Management, and Democracy (Oxford, 2001).
Ulrich Schneckener
Senior Fellow at the German Institute for International and Security Affairs
(Stiftung Wissenschaft und Politik), Berlin, Germany. Research interests: inter-
national conflict and crisis management, civil wars, international terrorism, and
counter-terrorism. Recent Publications: (ed. with Stefan Wolff ) Managing and
Settling Ethnic Conflicts (London: Hurst, 2003). Auswege aus dem Bürgerkrieg
(Frankfurt a.M.: Suhrkamp, 2002).
Philip Towle
Philip Towle is Reader in International Relations in the Centre of International
Studies at Cambridge University. He has worked for the Foreign and Common-
wealth Office and the Australian National University, and has taught in Cam-
bridge since 1980. His recent books include Enforced Disarmament (Clarendon
1997); and Democracy and Peacemaking (Routledge 2000).
Ketevan Tsikhelashvili
Ketevan Tsikhelashvili holds a B.A. degree in International Relations from Tbilisi
State University (Georgia) and an M.A. degree in International Relations and Eu-
ropean Studies from the Central European University, Budapest, Hungary. She
has an extensive experience of working with international and local organiza-
tions and as a researcher, journalist, and an expert on conflict and ethnic minor-
ity issues (NATO Parliamentary Assembly, Brussels, Belgium; European Center
on Minority Issues, Flensburg, Germany; and Transitions Online, Prague, Chech
Republic). She has also worked as a senior political analyst at the Foreign Policy
Research and Analysis Center, Ministry of Foreign Affairs of Georgia.
726 Biographies
Currently, she works with the Center for OSCE Research (CORE) at the Uni-
versity of Hamburg as an external researcher/staff member, as a researcher at
the International Center on Conflict and Negotiation (ICCN) in Tbilisi, and as a
consultant of Friedrich Naumann Foundation in Georgia.
Mark Turner
Mark Turner is Professor of Development Policy and Management at the Uni-
versity of Canberra. He has extensive experience of research and consultancy
in politics and public sector management in many countries of the Asia-Pacific
region, including the Philippines, Vietnam, Laos, Cambodia, Thailand, Vietnam,
Indonesia, Kazakhstan, Papua New Guinea, and Vanuatu. His recent books in-
clude Decentralisation in Indonesia: Redesigning the State (Canberra: Asia Pacific
Press, 2003) and Central-Local Relations in Asia Pacific: Convergence or Diver-
gence? (Houndmills and New York: Macmillan and St Martin’s, 1999).
Marc Weller
Marc Weller is the Director of the European Centre for Minority Issues, in Flens-
burg, Germany. His writings mainly focus on conflict management, issues of in-
ternational law, and minority rights. He has acted as legal advisor to several gov-
ernments and organizations, and a member of international peace processes. Dr
Weller is also a Reader in International Law at the University of Cambridge and
a Fellow of the Lauterpacht Research Centre for International Law and Hughes
Hall. He is also the Director of the Carnegie Project on Resolving Self-determina-
tion Disputes through Complex Power-sharing and of the Cambridge Rockefeller
Project on Restoring an International Consensus of the Rules Governing the Use
of Force.
Stefan Wolff
A political scientist by background, Stefan Wolff is Professor of Politics at the
University of Nottingham. In his research and consulting work, he specializes in
the prevention, management, and settlement of ethnic and religious self-deter-
mination conflicts and in post-conflict reconstruction of deeply divided and war-
torn societies, with a particular focus on institutional design. He has extensive
expertise in Northern Ireland, the Balkans, and the Middle East, and has also
worked on a wide range of other conflicts elsewhere, including Central and East-
ern Europe, Africa, and Southeast Asia. Wolff ’s publications to date include six
books and over twenty journal articles and book chapters. Among his books are
Disputed Territories: The Transnational Dynamics of Ethnic Conflict Settlement
(2002); The German Question since 1919 (2003); Managing and Settling Ethnic
Conflicts (co-edited with Ulrich Schneckener, 2003); and Peace at Last? The Im-
pact of the Good Friday Agreement on Northern Ireland (with a foreword by Lord
Alderdice; co-edited with Jörg Neuheiser, 2002). In addition, he is the founding
editor of The Global Review of Ethnopolitics (www.ethnopolitics.org), a quarterly,
peer-reviewed online journal dedicated to the study of ethnic conflicts and their
management around the globe.
Biographies 727
Nathalie Ubilava
Nathalie Ubilava is a freelance consultant focusing on financial services industry
and multilateral international donor organizations. Her corporate consulting ex-
perience includes strategic marketing and market research both in developed and
emerging economies. Nathalie has completed projects ranging from economet-
ric modelling of retail network performance to offer construction for web-based
financial portal. While at Corporate Value Associates (Boston, MA, USA), she
formulated customer segmentation to support cross-sales and customer acquisi-
tions strategy for a global provider of financial services.
Nathalie’s work in emerging markets involves multiple projects in her native
Georgia, mainly under the US Agency for International Development. In this re-
spect, while with Booz-Allen & Hamilton Inc., Nathalie designed and executed a
survey to qualify and quantify the economic impact of land reform in the Repub-
lic of Georgia. The results ultimately informed the drafting of national legislation
on land reform. In Georgia, Nathalie has also devised and implemented several
surveys and econometric analyses of local real estate markets and transport-gen-
erating industries.
Niall Johnson
After undergraduate studies at Macquarie University and the University of New
South Wales (both in Sydney, Australia) and an M.A. from Wilfrid Laurier Uni-
versity (Canada), Niall Johnson completed his Ph.D. at the University of Cam-
bridge.
Bibliography
Abubakar, C.A. (2000). “SPCPD and Economic Development in SZOPAD: High Expecta-
tions, Low Output.” Kasarinlan 15(2): 125-164.
Adams, R. (ed.). (2001). Gudpela Nius Bilong Pis. Peace on Bougainville: Truce Monitoring
Group. Wellington: Victoria University Press.
Aguja, M. J. (2000). “The Aftermath of Ethnic Violence – Post War Reconstruction in
the Southern Philippines: A Preliminary Assessment of the Role of the International
Community.” INCORE Ethnic Studies Network. http://www.incore.ulst.ac.uk/home/
esn/joyo.html (21 June).
Akhavan, P. (1997). “Justice And Reconciliation in The Great Lakes Region of Africa: The
Contribution Of The International Criminal Tribunal For Rwanda.” Duke Journal of
Comparative and International Law 7: 325–349.
Alcock, A. (2003). “The British-Irish Agreement on Northern Ireland 10 April (Good Fri-
day) 1998. Triumph or Tragedy?” A. Alcock et al. (eds.). Peace Process Through Con-
stitutional Arrangements in Northern Ireland. Bolzano/Bozen: European Academy.
Allan, T.R.S. (2001). Constitutional Justice: A Liberal Theory of the Rule of Law. Oxford:
Oxford University Press.
Allen, J., and C. Hurd. (1963). Languages of Bougainville District. Ukarumpa: Summer In-
stitute of Linguistics.
Amnesty International. (2001). “Macedonia (FYROM): A Durable Peace Depends on Re-
spect for Human Rights.” http://web.amnesty.org/ai.nsf/Index/EUR650052001?Ope
nDocument&of=COUNTRIES\MACEDONIA (12 October 2002).
Amorim Neto, O. and G.W. Cox. (1997). “Electoral Institutions, Cleavage Structures, and
the Number of Parties.” American Journal of Political Science 41: 149-174.
Anderson, B. (1983). Imagined Communities, Reflections on the Origins and Spread of Naa-
tionalism. London: Verso.
Anderson, J. (1995). “The Exaggerated Death of the Nation-State.” J. Anderson, C. Brook,
and A. Cochrane (eds.). A Global World? Re-ordering Political Space. Milton Keynes/
Oxford: The Open University/Oxford University Press.
Anderson, J., L. O’Dowd, and T.M. Wilson. (2002). “Introduction: Why Study Borders
Now?” Regional and Federal Studies 12(4): 1–12.
Anderson, N. (1999). “OSCE Preventive Diplomacy in the Former Yugoslav Republic of
Macedonia.” Helsinki Monitor 10(2): 49-64.
Andjelić, N. (2003). Bosnia-Herzegovina: The End of a Legacy. London: Frank Cass.
730 Bibliography
Anstee, M.J. (2003). Never Learn to Type: A Woman at the United Nations. Chichester:
John Wiley and Sons.
Arnautović, S. (1996). Izbori u Bosni i Hercegovini ‘90. Sarajevo: Promocult.
Asia Foundation (2002). Indonesian Raid Decentralization Appraisal: First Report. Jakar-
ta: Asia Foundation.
Asian Development Bank (2000). Filipino Report Card on Pro-poor Services. Manila: Asian
Development Bank.
Babić, D. (2001). “Journalism in Post-Dayton Bosnia: How to Make the Media More Re-
sponsible?” D. Sokolović and F. Bieber (eds.). Reconstructing Multiethnic Societies:
The Case of Bosnia-Herzegovina. Aldershot: Ashgate.
Bagayaua, G. (2002). “Math, Science and the Koran,” Newsbreak Magazine. http://www.
inq7.net/nwsbrk/2002/ (22 March 2003).
Bajtay, P. (ed.). (1996). Regional Cooperation and the European Integration Process: Nordic
and Central European Experiences. Budapest: Hungarian Institute of International
Affairs.
Ball, N. (2001). “Transforming Security Sectors: the IMF and World Bank Approaches.”
Journal of Conflict, Security and Development 1(1).
Ballard, J. (1981). “Policy-making as Trauma: the Provincial Government Issue.” J. Ballard
(ed.). Policy-making in a New State: Papua New Guinea 1972-77. St Lucia: University
of Queensland Press.
BASA-PRESS (2001). “Comrat Authorities Celebrate 11th Anniversary of Separatist Move-
ment”. http://news.ournet.md//cgi-bin/ournet/mnews/page.cgi?ID=13370&From_
Cat=41&d=1 (20 August).
Bassouini, M.C. (1996). “Searching For Peace And Achieving Justice: The Need For Ac-
countability.” Law and Contemporary Problems 59: 9–29.
Bedford, R. and A. Mamak. (1977). Compensation for Development: The Bougainville Case.
Bougainville Special Publication No. 2. Christchurch: University of Canterbury, De-
partment of Geography.
Begić, K. (1997). Bosna i Hercegovina: Od Vanceove Misjie do Daytonskog Sporazuma.
Sarajevo: Bosanska Knjiga.
Bell, C. (2000). Peace Agreements and Human Rights. Oxford: Oxford University Press.
Belloni, R. (2001). “Civil Society and Peacebuilding in Bosnia and Herzegovina.” Journal of
Peace Research 38(2): 163-180.
Benn, S.I. and R.S. Peters. (1964). The Principles of Political Thought. London: Collier-
Macmillan.
Bennett, C. (1998). “Bosnia-Herzegovina.” P. Harris and B. Reilly (eds.). Democracy and
Deep Rooted Conflicts. Stockholm: International IDEA.
Beran, H. (1987). The Consent Theory of Political Obligation. London: Croom Helm.
Bercovitch, J. (2002). “Introduction: Putting Mediation in Context.” J. Bercovitch (ed.).
Studies in International Mediation. Basingstoke: Palgrave.
Bernier, J. (2001). “Nationalism in Transition: Nationalizing Impulses and International
Counterweights in Latvia and Estonia”. M. Keating and J. McGarry (eds.). Minor-
ity Nationalism and the Changing International Order. Oxford: Oxford University
Press.
Bew, P. (2000). “The Belfast Agreement of 1998: From Ethnic Democracy to a Multicultural
Consociational Settlement?” M. Cox, A. Guelke, and F. Stephen (eds.).” A Farewell to
Bibliography 731
Arms? From ‘Long War’ to Long Peace in Northern Ireland. Manchester: Manchester
University Press.
Bew. P. and H. Patterson. (1985). The British State and the Ulster Crisis: From Wilson to
Thatcher. London: Verso.
Bew, P and H. Patterson. (1990). Scenarios for Progress in Northern Ireland. London:
Clarendon.
Bew, P., P. Gibbon, and H. Patterson. (1996). The State in Northern Ireland 1921-1972: Po-
litical Forces and Social Classes. Revised and updated ed. Manchester: Manchester
University Press.
Beyani, C. (2000). Human Rights Standards and the Movement of People Within States.
Oxford: Oxford University Press.
Beyerlin, U. (1998) “Neue rechtliche Entwicklungen der regionalen und lokalen grenzüber-
schreitenden Zusammenarbeit”. Brunn , G. and P. Scmitt-Egner (eds.). Grenzübersch-
reitenden Zusammenarbeit in Europa. Nomos Verlagsgesellschaft: Baden-Baden.
Bieber, F. (1999a). Bosnien-Herzegovina und der Libanon im Vergleich. Historische Entwick-
lung und Politisches System vor dem Bürgerkrieg. Sinzheim: Pro Universitate Verlag.
Bieber, F. (1999b). “Consociationalism – Prerequisite or Hurdle for Democratization in
Bosnia? The Case of Belgium as a Possible Example.” Southeast Europe Review 2(3):
79-94.
Bieber, F. (2001). “Croat Self-Government in Bosnia – A Challenge for Dayton?” ECMI
Brief, 5 May.
Bieber, F. (2002). “Bosnia-Herzegovina: Developments Towards a More Integrated State?”
Journal of Muslim Minority Affairs 22(1): 205-218.
Bieber, F. (2005). Post-war Bosnia: Ethnicity, Inequality and Public Sector Governance.
London: Palgrave.
Bildt, C. (1998). Peace Journey: The Struggle for Peace in Bosnia. London: Weidenfeld and
Nicolson.
Birch, J. (1996). “The Georgian/South Ossetian Territorial and Boundary Dispute.” J.
Wright, S. Goldenberg, and R. Schofield (eds.). Transcaucasian Boundaries. London:
UCL Press Limited.
Blais, A., L. Mascotte, and A. Dobrzynska. (1997). “Direct Presidential Elections: A World
Summary.” Electoral Studies 16: 441-455.
Blanchard, O. (1997). The Economics of Post-Communist Transition. Oxford: Clarendon
Press.
Bloomfield, D. and B. Reilly. (1998). “Characteristics of Deep-rooted Conflict.” Harris, P.
and B. Reilly. (eds.). Democracy and Deep-rooted Conflicts. Stockholm: International
IDEA.
Bonazzi, T. and M. Dunne (eds.). (1995). Citizenship and Rights in Multicultural Societies.
Keele: Keele University Press.
Bose, S. (2002). Bosnia after Dayton: Nationalist Partition and International Intervention.
London: Hurst.
Bosnia and Herzegovina Ministry of Education. (2002). Reforming Education to Give BiH
a Better Future. Sarajevo: Ministry of Education.
Bougainville Administration Officers’ Working Group. (2003). Law and Justice for Bou-
gainville - Transition to New Bougainville Policing, Courts, CUS. Community Based
Dispute Settlement.
732 Bibliography
Bougarel, X. (1996). Bosnie: Anatomie d’un conflit. Paris: La Découverte.

Bowers, S.R., M.T. Doss, and V. Ciobanu. (2002). Gagauzia and Transdniestria: The
Moldovan Confederation Conudrum. The William R. Nelson Institute for Public Af-
fairs. http://www.jmu.edu/orgs/wrni/gagauzia18.htm (8 February 2004).
Bradley, J. (1996). An Island Economy: Exploring Long-term Economic and Social Conse-
quences of Peace and Reconciliation in the Island of Ireland. Dublin: Forum for Peace
and Reconciliation.
Bradley, J. (2001). “The Island Economy: Past, Present and Future.” A. Pollak. (ed.). Can the
Celtic Tiger Cross the Irish Border? Cork: Cork University Press.
Brass, P.R. (1991). Ethnicity and Nationalism. New Delhi: Sage.
Bruce, S. (2001). “Terrorists and Politics: The Case of Northern Ireland’s Loyalist Paramili-
taries.” Terrorism and Political Violence 13(2): 27–48.
Bucheit, B. (1981). Secession: The Legitimacy of Self-determination. New Haven: Yale Uni-
versity Press.
Bugajski, J. (1994). Ethnic Politics in Eastern Europe: A Guide to Nationality Policies, Or-
ganizations, and Parties. Armonk: M.E. Sharpe.
Buha, A. (1999). Argument en faveur de la République Serbe. Lausanne: L’Age d’Homme.
Bulatao, R.A. (1975). “Ethnic Attitudes in Five Philippine Cities.” M.R. Hollnsteiner (ed.).
Society, Culture and the Filipino. Quezon City: Institute of Philippine Culture, At-
eneo de Manila University.
Bunce, V. (1999). Subversive Institutions: The Design and the Destruction of Socialism and
the State. Cambridge: Cambridge University Press.
Bunreacht na hÉireann. (1937). Constitution of Ireland. Dublin: Stationery Office.
Burg, S.L. and P.S. Shoup. (1999). The War in Bosnia-Herzegovina. Ethnic Conflict and
International Intervention. Armonk: M.E. Sharpe.
Burns, A. (1957). In Defence of Colonies. London: Allen and Unwin.
Cagoco-Guiam, R. (1999). “A Critical Partnership: Civil Society and the Peace Process.”
Accord 6.
Campbell, C., F. Ní Aoláin, and C. Harvey. (2003) “The Frontiers of Legal Analysis: Re-
framing the Transition in Northern Ireland”. Modern Law Review 66(3): 317–345.
Carl, A. and L. Garasu. (eds.). (2002). Weaving Consensus: The Papua New Guinea – Bou-
gainville Peace Process. Accord 12. London: Conciliation Resources.
Carothers, T. (1998). “The Rule of Law Revival.” Foreign Affairs 77(2): 95–106.
Carothers, T. (1999). Aiding Democracy Abroad: The Learning Curve. Washington D.C.:
Carnegie Endowment for International Peace.
Carty, A. (1996). Was Ireland Conquered? International Law and the Irish Question. Lon-
don: Pluto Press.
Casesse, A. (1995). Self-determination of Peoples: A Legal Reappraisal. Cambridge: Cam-
bridge University Press.
Central and Eastern Europe Initiative (CEELI). (2002a). Judicial Reform Index for Macedo-
nia. American Bar Association, March.
CEELI. (2002b). Structure of the Courts in Judicial Reform Index for Kosovo. American Bar
Association, April.
Centre for Russian Studies. (2001). “South Ossetia Not to Seek Early Unification with
North Ossetia.” Events 2001. http://www.nupi.no/cgi-win/Russland/krono.exe?278.
Bibliography 733
Centre for Strategic and International Studies. (2002). Post-conflict Reconstruction Task
Framework. Washington: CSIS.
Chandler, D. (2000). Bosnia: Faking Democracy After Dayton. London: Pluto Press.
Chandler, D. (2001). “Bosnia: The Democracy Paradox”. Current History 100(644): 114–
119.
Che Man, W.K. (1990). Muslim Separatism: The Moros of Southern Philippines and the
Malays of Southern Thailand. Quezon City: Ateneo de Manila Press.
Chinn, J. and S.D. Roper. (1998). “Territorial Autonomy in Gagauzia.” Nationalities Papers
26(1): 87–101.
Chuter, D. (2000). Defence Transformation: A Short Guide to the Issues. Monograph No.
49. Pretoria: Institute for Security Studies.
Cigar, N. (1996). Genocide in Bosnia: The Policy of “Ethnic Cleansing.” College Station: A
and M University Press.
Clifford, R.C.L. and T.J. Wilton. (2002). “Media Operations and the ARRC”. S. Badsey (ed.).
The Media and International Security. London: Frank Cass.
Cochrane, F. (2001). “Unsung Heroes? The Role of Peace and Conflict Resolution Organi-
zations in the Northern Ireland Conflict.” J. McGarry (ed.). Northern Ireland and
the Divided World: Post-Agreement Northern Ireland in Comparative Perspective.
Cohen, L. (1993). Broken Bonds: The Disintegration of Yugoslavia. Boulder: Westview.
Cohen, S. (2001). States of Denial: Knowing About Atrocities and Suffering. London: Polity
Press.
Colley, L. (1992). Britons: Forging the Nation, 1707-1837. New Haven: Yale University
Press.
Collins, E. and M. McGovern. (1997). Killing Rage. London: Granta Books.
Commission on Higher Education. (2002). “Statistics: Distribution of Higher Education
Institutions.” Manila. http://www.ched.gov.ph/statistics/ (9 March 2003).
Conference on Security and Cooperation in Europe. (1993). “Report No. 13 by the CSCE
Mission to Moldova.” 13 November.
Connell J. (ed.). (1977). Local Government Councils in Bougainville. Bougainville Special
Publication No.3. Christchurch: Department of Geography, University of Canter-
bury.
Connell, J. (1991). “Compensation and Conflict: The Bougainville Copper Mine, Papua
New Guinea.” J. Connell and R. Howitt (eds.). Mining and Indigenous Peoples in Aus-
tralasia. Sydney: Sydney University Press.
Connell, J. (1997). Papua New Guinea: The Struggle for Development. London: Routledge.
Conyers, D. (1976). The Provincial Government Debate: Central Control Versus Local Par-
ticipation in Papua New Guinea. Port Moresby: Papua New Guinea Institute of Ap-
plied Social and Economic Research.
Cooper, N. and M. Pugh. (2002). Security-sector Transformation in Post-conflict Societies.
Working Paper No. 5. London: Centre for Defence Studies.
Coppieters, B. (1999). “The Roots of the Conflict.” Accord 7: 14-19.
Costa, C. and M. Medeiros. (2002). “Police Demilitarization: Cops, Soldiers and Democ-
racy.” Journal of Conflict, Security and Development 2(2): 25-45.
Costello, M.A. (1992). “The Demography of Mindanao.” M. Turner, R.J. May, and L.R. Turn-
er. (eds.). Mindanao: Land of Unfulfilled Promise. Quezon City: New Day Publishers.
734 Bibliography
Council of Europe. (1995). Examples of Good Practice of Trans-Frontier Cooperation Con-

cerning Members of Ethnic Groups on the Territory of Several States. Strasbourg:
Council of Europe Press.
Council of Europe. (1997). Documents and Agreements Concerning Trans-Frontier-Coop-
eration in Europe. Strasbourg: Council of Europe Press.
Council of Europe. (1998). “Opinion on the Legal Code of Gagauzia.” Working Group of
the Venice Commission and the Congress of Local and Regional Authorities of Eu-
rope, Strasbourg, 7 May. http://www.venice.coe.int/site/interface/english.htm.
Council of Europe. (1999a). “Georgia’s Application for Membership of the Council of Eu-
rope.” Opinion 12 January 1999. Doc. 8296.
Council of Europe. (1999b). “Education in Bosnia and Herzegovina: Governance, Finance
and Administration.” Strasbourg: Council of Europe.
Council of Europe. (2001). “Proposal for A Law On The Merger Of The Human Rights
Chamber and the Constitutional Court of Bosnia and Herzegovina.” Adopted by the
Venice Commission At its 48th Plenary Meeting.
Council of Europe. (2002). “Functioning of Democratic Institutions in Moldova.” Doc.
9418. 23 April.
Council of Europe. (2002a). “Consolidated Opinion on the Law on Modification and Ad-
dition in the Constitution of the Republic Of Moldova.” European Commission for
Democracy Through Law (Venice Commission). Strasbourg, 12 March.
Council of Europe. (2002b). “Opinion on the Law on Modification and Addition in the
Constitution of the Republic of Moldova in Particular Concerning the Status of
Gagauzia”. Adopted by the Venice Commission at its 50th Plenary Session. Venice,
8-9 March 2002. http://www.venice.coe.int/site/interface/english.htm.
Cousens, E. and C. Carter. (2001). Toward Peace in Bosnia: Implementing the Dayton Ac-
cords. Boulder: Lynne Rienner.
Cox, G. (1997). Making Votes Count. Cambridge: Cambridge University Press.
Cox, M. (1999). “The War that Came in from the Cold: Clinton and the Irish Question.”
World Policy Journal 16(1): 59-67.
Cox, M. (2001). “State-building and Post-war Reconstruction: Lessons from Bosnia.” Ge-
neva: CASIN.
Criminal Justice System Review Group. (2000). Review Of The Criminal Justice System In
Northern Ireland: The Report Of The Criminal Justice System Review. HMSO: Bel-
fast.
Crow, S. (1993). “Russia Seeks Leadership in Regional Peacekeeping.” RFE/RL Research
Report 15. 9 April.
Crowther, W. (1998). “Ethnic Politics and the Post-Communist Transition in Moldova.”
Nationalities Papers 26(1): 147–164.
http://www.osce.org/documents/mm/1993/11/454_en.pdf.
Cunningham, M.J. (2000). British Government Policy in Northern Ireland, 1969-2000.
Manchester: Manchester University Press.
Cvetkovski, N. (1996). “Georgian-South Ossetian Conflict.” Dissertation, Aalborg Univer-
sity, Denmark. Danish Association for Research in Caucasus. http://www.caucasus.
dk/publication5.htm.
Daftary, F. (2001). “Conflict Resolution in FYR Macedonia: Power-Sharing or the ‘Civic
Approach.’” Helsinki Monitor 4: 291-312.
Bibliography 735
Dahrendorf, N. (ed.). (2003). A Review of Peace Operations: A Case for Change. London:
International Policy Institute.
Davis, D.E. and A.W. Pereira. (2003). Irregular Armed Forces and Their Role in Politics and
State Formation. Cambridge: Cambridge University Press.
Dawisha, K. and A. Dawisha. (2003). “How to Build a Democratic Iraq.” Foreign Affairs
82(3): 36-50
Denoon, D. (2000). Getting Under the Skin: The Bougainville Copper Agreement and the
Creation of the Panguna Mine. Melbourne: Melbourne University Press.
Department of Education – Northern Ireland. (2003). “About the Department of Educa-
tion.” http://www.deni.gov.uk/about/d_ed_system.htm (13 March).
Department of Education and Science. (2001). “The New Kosovo Framework Curriculum:
Preschool, Primary and Secondary Education.” September 2001.
Department for International Development. (2002). Understanding and Supporting Secu-
rity Sector Reform. London: Department for International Development.
Department of Justice. (1997). “Backgrounder.” 27 February.
Department of State. (2000), “1999 Country Reports on Human Rights Practices: Bosnia-
Herzegovina.” 25 February.
Dicey, A.V. (1885). The Law Of The Constitution. 10th ed. London: Macmillan.
“Die Plattform des Präsidiumsvorsitzenden der SR Bosnien Hercegovina und des Prä-
sidium der Republik Mazedonien über die zukünftige jugoslawische Gemeinschaft.”
Internationale Politik 42(989): 22-24.
Dinnen, S. (2002). “Winners and Losers: Politics and Disorder in the Solomon Islands
2000-2002.” The Journal of Pacific History 37(3): 285-298.
Dinnen, S., R. May, and A. Regan (eds.). (1997). Challenging the State: The Sandline Affair
in Papua New Guinea. Canberra: Asia Pacific Press, Australian National University.
Dixon, P. (1997). “Paths to Peace in Northern Ireland (II): The Peace Processes 1973-74 and
1994-96.” Democratization 4(3): 1-25.
Domi, T. (2001). “‘Cowboy’ Toughs it out in Bosnia.” Balkan Crisis Report 219.
Donais, T. (2002). “The Politics of Privatization in Post-Dayton Bosnia,” South East Euro-
pean Politics 3(1): 1-19.
Donia, R. (2000). “The Search for Tolerance in Sarajevo’s Textbooks.” Human Rights Re-
view 1(2): 38-55.
Donnelly, J. (2003). “In Defence of the Universal Declaration Model.” G. Lyons and J. May-
all (eds.). International Human Rights in the Twenty-First Century: Protecting the
Rights of Groups. Oxford: Rowman and Littlefield.
Dorney, S. (1998). The Sandline Affair: Politics and Mercenaries and the Bougainville Cri-
sis. Sydney: ABC Books.
Downs, A. (1957). An Economic Theory of Democracy. New York: Harper and Row.
Downs, G. and S.J. Stedman. (2002). “Evaluation Issues in Peace Implementation.” S.J.
Stedman, D. Rothchild, and E. Cousens (eds.). Ending Civil Wars. Boulder: Lynne
Rienner.
Doyle, M. and N. Sambanis. (2000). “International Peacebuilding: A Theoretical and
Quantitative Analysis.” American Political Science Review 94(4): 779-801.
Duffield, M. (2001). Global Governance and the New Wars. London: Zed Books.
Dunford, M. and R. Hudson. (1996). Successful European Regions: Northern Ireland Learn-
ing From Others. Belfast: Northern Ireland Economic Council.
736 Bibliography
Durverger, M. (1954). Political Parties: Their Organization and Activity in the Modern
World. London: Methuen. (Translation. Originally published in French, 1951).
Dworkin, R. (1986). Law’s Empire. Cambridge: Harvard University Press.
Eagles, J. (2002). “Aid as an Instrument for Peace: A Civil Society Perspective.” Weaving
Consensus: The Papua New Guinea-Bougainville Peace Process. Accord 12(2).
East-West Institute. (2002). Southern Adriatic Transfrontier Cooperation Forum. http://
www.ewi.info/publications.cfm?title=Publications&l1=Publications&l2=&l3=&l4=
&view=detail&pubid=145 (8 February 2004).
East-West Institute. (2003). Cross Border Trafficking in South Eastern Europe: Assessing
Trafficking Activities in the Southern Adriatic Region. http://www.ewi.info/publica-
tions.cfm?title=Publications&l1=Publications&l2=&l3=&l4=&view=detail&pubid=
228 (8 February 2004).
Economic Commission for Europe. (2003). Economic Survey of Europe 1, Statistical Ap-
pendix. New York and Geneva: United Nations.
Eide, A. (1983). “Dynamics of Human Rights and the Role of the Educator.” A. Eide and M.
Thee (eds.). Frontiers of Human Rights Education. Oslo: Universitets Forlaget.
von Einsiedel, S. (2002). You, the People: Transitional Administration, State-building and
the United Nations. New York: International Peace Academy.
Elazar, D. (1987). Exploring Federalism. Tuscaloosa: University of Alabama Press.
Elazar, D. (1999). Federal Systems of the World: A Handbook of Federal, Confederal and
Autonomy Arrangements. http://www.jcpa.org/dje/books/fedsysworld-intro.htm (15
October 2003).
Elklit, J. and A. Reynolds. (2002). “The Impact of Election Administration on the Legiti-
macy of Emerging Democracies.” Journal of Commonwealth and Comparative Poli-
tics 57: 86-119.
Elklit, J. and P. Svensson. (1997). “What Makes Elections Free and Fair?” Journal of Democ-
racy 8(3): 32-46.
Engström, J. (2002). “Multi-ethnicity or Bi-nationalism? The Framework Agreement and
the Future of the Macedonian State.” Journal on Ethnopolitics and Minority Issues in
Europe 2(1). http://www.ecmi.de/jemie/special_1_2002.html.
European Commission for Democracy Through Law (Venice Commission). (2002). “Con-
solidated Opinion on the Law on Modification and Addition in the Constitution of
the Republic Of Moldova”, Strasbourg, 12 March. http://www.venice.coe.int/site/in-
terface/english.htm.
European Commission. (1992). A Practical Guide to Cross-border Cooperation. Luxem-
bourg: European Commission.
European Parliament. (1996). Cross-border and Interregional Cooperation in the European
Union. Luxembourg: European Parliament Regional Policy Series.
European Stability Initiative (ESI). (2000). “Turning Point: The Brussels PIC Declaration
and a State-building Agenda for Bosnia and Herzegovina.” Brussels: European Stabil-
ity Initiative.
ESI. (2001). “Reshaping International Priorities in Bosnia and Herzegovina (Part 3): The
End of the Nationalist Regimes and the Future of the Bosnian State.” Brussels: Euro-
pean Stability Initiative.
ESI. (2002). “Western Balkans 2004: Assistance, Cohesion and the New Boundaries of
Europe.” Brussels: European Stability Initiative.
Bibliography 737
ESI. (2003). “The Road to Thessaloniki: Cohesion and the Western Balkans.” Brussels: Eu-
ropean Stability Initiative.
Evans, G. and B. O’Leary. (1997). “Intransigence and Flexibility on the Way to Two Forums:
The Northern Ireland Elections of 30 May 1996 and Public Opinion.” Representation
34(3-4): 208–218.
Evans, G. and B. O’Leary. (2000). “Northern Irish Voters and the British-Irish Agreement:
Foundations of a Stable Consociational Settlement?” Political Quarterly 71(1): 78–101.
Fair, J.D. (1972). “The Anglo-Irish Treaty of 1921: Unionist Aspects of the Peace.” Journal of
British Studies XII.
Farrell, D.M. (1997). Comparing Electoral Systems. London: Prentice Hall.
Faundez, J. (2000). “Legal Reform in Developing and Transition Countries - Making Haste
Slowly.” Law, Social Justice and Global Development 1. http://elj.warwick.ac.uk/glo-
bal/issue/2000-1/faundez.html (1 December 2003).
Fay, M.-T., M. Morrisey, and M. Smyth. (1997). The Cost of the Troubles Study: Mapping
Troubles-related Deaths in Northern Ireland, 1969-1994. Derry/Londonderry: IN-
CORE.
Fay, M.-T., M. Morissey, and M. Smyth. (1999). Northern Ireland’s Troubles: The Human
Costs. London: Pluto Press.
Ferrer, M.C. (2000). “Recycled Autonomy? Enacting the New Organic Act for a Regional
Autonomous Government in Southern Philippines.” Kasarinlan 15(2): 165-190.
Financial Times. (2003). “National Army Takes Shape in Afghanistan.” 7 August.
Finer, S.E. (1988). The Man on Horseback: The Role of the Military in Politics. London:
Pinter.
Frankel, J. (2001). “The Alternative Vote System in Fiji: Electoral Engineering or Ballot-rig-
ging?” Journal of Commonwealth and Comparative Politics 39(2): 1-31.
French, H. (2000). Vanishing Borders: Protecting the Planet in the Age of Globalization.
New York: W.W. Norton.
Friedman, E. (1996). The Bosnian Muslims: Denial of a Nation. Boulder: Westview.
Fujita, M., P. Krugman, and A. Venables. (1999). The Spatial Economy: Cities, Regions, and
International Trade. Cambridge, M.A.: The MIT Press.
Fukuyama, F. (1992). The End of History and the Last Man. London: Penguin Books.
Fuller, E. (1991). “South Ossetia: Analysis of the Permanent Crisis.” Report on the USSR.
Radio Free Europe/Radio Liberty, 15 February.
Gabbe, J. (1995). European Models of Inter-regional and Cross-border Cooperation in the
European Union. Gronau: LACE.
Gallagher, M. (1987). “Does Ireland Need a New Electoral System?” Irish Political Studies
2: 27-48.
Gallagher, M. (1991). “Proportionality, Disproportionality and Electoral Systems.” Electoral
Studies 10: 33-51.
Gellner, E. (1983). Nations and Nationalism. Oxford: Blackwell.
Gellner, E. (1997). Nationalism. London: Weidenfeld and Nicholson.
George, T.J.S. (1980). Revolt in Mindanao: The Rise of Islam in Philippine Politics. Kuala
Lumpur: Oxford University Press.
Gerner, K. (1997). “The Evolving Political Role of Borders and Border Regions in Central
Europe.” P. Ganster, A. Sweedler, and J. Scott (eds.). Borders and Border Regions in
Europe and North America. San Diego: State University Press.
738 Bibliography
Ghai, Y. (2000). “Ethnicity and Autonomy: A Framework for Analysis.” Y. Ghai (ed.). Au-
tonomy and Ethnicity: Negotiating Competing Claims in Multi-ethnic States. Cam-
bridge: Cambridge University Press.
Ghai Y. and A. Regan. (1992). The Law, Politics and Administration of Decentralization
in Papua New Guinea. NRI Monograph 30. Port Moresby: The National Research
Institute.
Ghai Y. and A. Regan. (2000). “Bougainville and the Dialectics of Ethnicity, Autonomy and
Separation.” Y. Ghai (ed.). Autonomy and Ethnicity: Negotiating Competing Claims in
Multi-ethnic States. Cambridge: Cambridge University Press.
Glang, A.C. (1972). “Why the Shooting Won’t Stop.” Solidarity 7(4): 6-8.
Glas Srpski. (1998). 22-23 August.
Glazer, N. (1983). “Federalism and Ethnicity: The American Solution.” N. Glazer (ed.). Eth-
nic Dilemmas, 1964-82. Cambridge: Harvard University Press.
Global IDP. (2001). “Discrimination and Segregation in the Education System Hampers
the Return of Minority Families (2000-2001).” http://ww.idpproject.org (6 March
2003).
Global IDP. (2002). “Georgia: Issues of Language and Education in Abkhazia: a Decisive Fac-
tor on Return.” http://www.db.idpproject.org/Sites/idpSurvey.nsf/wViewCountries/
9376043CFA (20 March 2003).
Goltz, T. (1993). “Letter from Eurasia: The Hidden Russian Hand.” Foreign Policy 92: 92-
113.
Goodwin-Gill, G. (1998). The Refugee in International Law. Oxford: Oxford University
Press.
Gow, J. (1997). The Triumph of the Lack of Will: International Diplomacy and the Yugoslav
War. London: Hurst.
Gowing, P.G. (1979). Muslim Filipinos–Heritage and Horizon. Quezon City: New Day Pub-
lishers.
Government of the Republic of the Philippines. (2000). “Status of Implementation of the
1996 GRP-MNLF Peace Agreement.” Kalinaw Mindanao. 2000. http://mindanao.
com/kalinaw/peaceproc/pa-status-grp-032k.htm (7 July 2001).
Griffin, J. (1977). “Local Government Councils as Instruments of Political Mobilization in
Bougainville.” J. Connell (ed.). Local Government Councils in Bougainville. Bougain-
ville Special Publication No.3. Christchurch: Department of Geography, University
of Canterbury.
Griffin, J. (1982). “Napidakoe Navitu.” R.J. May (ed.). Micronationalist Movements in Papua
New Guinea: Political and Social Change. Monograph No.1. Canberra: Australian
National University.
Griffin, J., and M. Togolo. (1997). “North Solomons Province, 1974-2000.” R.J. May and A.J.
Regan (eds). Political Decentralization in a New State: The Experience of Provincial
Government in Papua New Guinea. Bathurst: Crawford House Press.
Grin, F. (1991). “The Estonian Language Law: Presentation with Comments.” Language
Problems and Language Planning 15(2): 191–201.
Groom, A.J.R. and P. Taylor. (2000). “The United Nations System and the Kosovo Crisis.”
A. Schnabel and R. Thakur (eds.). Kosovo and the Challenge of Humanitarian Inter-
vention. New York: UN University Press.
Guelke, A. (1996). “The United States, Irish Americans and the Northern Ireland Peace
Process”. International Affairs 72(3): 521-536.
Bibliography 739
Gurr, T.R. (1993). Minorities At Risk: A Global View. Washington D.C.: US Institute of
Peace Research.
Gurr, T.R. (2000). People Versus States: Minorities at Risk in the New Century. Washington
D.C.: US Institute of Peace Press.
Gutierrez, E. “The Politics of Transition.” Accord 6.
de Guzman, R.P. (1993). “Is there a Philippine Public Administration?” V.A. Bautista,
M.C.P. Afiler, D.R. Reyes, and P.D. Tapales (eds.). Introduction to Public Administra-
tion in the Philippines: A Reader. Quezon City: University of the Philippines Press
and College of Public Administration.
Gwynn, D. (1950). The Partition of Ireland, 1912-25. Dublin: Browne and Nolan.
Hadden, T. and K. Boyle. (1989). The Anglo-Irish Agreement: Commentary, Text and Of-
ficial Review. London: Sweet and Maxwell Ltd.
Hagman, L. and Z. Nielsen. (2002). A Framework for Lasting Disarmament, Demobiliza-
tion, and Reintegration of Former Combatants in Crisis Situations. New York: Inter-
national Peace Academy.
Hampson, F.O. (1996). Nurturing Peace: Why Peace Settlements Succeed or Fail. Washing-
ton D.C.: US Institute for Peace Press.
Hanf, T. (1991a). “Konfliktminderung durch Kulturautonomie. Karl Renners Beitrag zur
Frage der Konfliktregelung in multi-ethnischen Staaten.” E. Fröschl, M. Mesner, and
U. Ra’anan (eds). Staat und Nation in multi-ethnischen Gesellschaften. Wien: Pas-
sagen Verlag.
Hanf, T. (1991b). “Reducing Conflict Through Cultural Autonomy: Karl Renner‘s Contri-
bution.” Raa‘anan, U., M. Mesner, K. Armes, and K. Martin (eds.). State and Nation
in Multi-ethnic Societies: The Breakup of Multinational States. Manchester: Man-
chester University Press.
Hannum, H. (1996). Autonomy, Sovereignty, and Self-determination: The Accommodation
of Conflicting Rights. Philadelphia: University of Pennsylvania Press.
Hansen, A.S. (2002). From Congo to Kosovo: Civilian Police in Peace Operations. Adelphi
Paper 343. Oxford: Oxford University Press.
Hansen, G. (1998). “Humanitarian Action in the Caucasus: A Guide for Practitioners.” Hu-
manitarianism and War Project and Local Capacities for Peace Project. Occasional
Papers of the Watson Institute 32. Brown University.
Hansen, G. (2000). “Final Evaluation.” Independent Report GEO/96/019 for Rehabilitation
of Tskhinvali Region. United Nations Development Programme. June 26.
Harvey, C. (2001). “Building a Human Rights Culture in a Political Democracy: The Role
of the Northern Ireland Human Rights Commission.” C. Harvey (ed.). Human Rights,
Equality and Democratic Renewal in Northern Ireland. Oxford: Hart Publishing.
Harvey, C. (2003). “On Law, Politics and Contemporary Constitutionalism.” Fordham In-
ternational Law Journal 26: 996–1014.
Hastings, M. (2002). Going to the Wars. London: Macmillan.
Hayner, P.B. (1994). “15 Truth Commissions - 1974 to 1994: A Comparative Study.” Human
Rights Quarterly 16(4): 597–655.
Hazell, R. and B. O’Leary. (1999). “A Rolling Programme of Devolution: Slippery Slope
or Safeguard of the Union?” R. Hazell (ed.). Constitutional Futures: A History of the
Next Ten Years. Oxford: Oxford University Press.
Hedl, D. (2000). “Bosnian Trade Off.” Balkan Crisis Report 169.
740 Bibliography
Hegarty, A. (2003). “The Government of Memory: Public Inquiries and the Limits of Jus-
tice in Northern Ireland”. Fordham International Law Journal 26(4): 1148–1192.
Heinemann-Grüder, A. and T. Pietz. (2003). Turning Soldiers into a Work Force: Demobi-
lization and Reintegration in Post-Dayton Bosnia and Herzegovina. Brief 27, Bonn
International Center for Conversion.
Helsinki Committee for Human Rights in BH. (1999a). “Position of Media in BH within the
Context of Human Rights 1999.”
Helsinki Committee for Human Rights in BH. (1999b). “Report about Human Rights Situ-
ation in Bosnia and Herzegovina January - December 1999.”
Hendrickson, D. and A. Karkoszka. (2002). “The Security Sector and the Challenges of its
Transformation.” SIPRI Yearbook 2002: Armaments, Disarmament and International
Security. Oxford: Oxford University Press.
Hendrickson, D. and N. Ball. (2002). Off-budget Military Expenditure and Revenue: Issues and
Policy Perspectives for Donors. CSDG Occasional Papers 1. London: King’s College.
Hewitt, B.G. (1996). “Abkhazia: a Problem of Identity and Ownership.” J. Wright, S. Gold-
enberg, and R. Schofield (eds.). Transcaucasian Boundaries. London: UCL Press
Limited.
Hewko, J. (2002/3). “Foreign Direct Investment in Transitional Economies: Does the Rule
of Law Matter?” East European Constitutional Review 11/12: 71–80.
Hillyard, P. (2000). “Police Bill is Not Faithful Reflection of Patten.” Irish Times, 2 August.
Hirst, P. and G. Thompson. (1995). Globalization in Question: The International Economy
and the Possibilities of Governance. Oxford: Polity Press.
Holbrooke, R. (1998). To End a War. New York: Random House.
Holmes, S. (2002). “Reforming Russia’s Courts: Introduction.” East European Constitu-
tional Review 11(1/2): 90–92.
Holsti, K.J. (1996). The State, War, and the State of War. Cambridge: Cambridge
University Press.
Horowitz, D.L. (1985). Ethnic Groups in Conflict. Berkeley: University of California Press.
Horowitz, D.L. (1989a). “Ethnic Conflict Management for Policymakers.” J. P. Montville
(ed.). Conflict and Peacemaking in Multiethnic Societies. Lexington: Heath.
Horowitz, D.L. (1989b). “Making Moderation Pay: The Comparative Politics of Ethnic
Conflict Management.” J. P. Montville (ed.). Conflict and Peacemaking in Multiethnic
Societies. Lexington: Heath.
Horowitz, D.L. (1991). A Democratic South Africa? Constitutional Engineering in a Di-
vided Society. Berkeley: University of California Press.
Horowitz, D.L. (2000). “Constitutional Design: An Oxymoron?” I. Shapiro and S. Macedo
(eds.). Designing Democratic Institutions. New York: New York University Press.
Horowitz, D.L. (2001). “The Agreement: Clear, Consociational and Risky.” J. McGarry
(ed.). Northern Ireland in a Divided World. Oxford: Oxford University Press.
Horowitz, D.L. (2002). “Constitutional Design: Proposals Versus Processes.” Andrew Rey-
nolds (ed.). The Architecture of Democracy. Oxford: Oxford University Press.
Horwitz, M. J. (1977). “The Rule of Law: An Unqualified Human Good?” Yale Law Journal
86: 561-566.
House of Commons. (1998). Hansard: Bound Volume 319. 18 November.
House of Commons. (2003). “The Northern Ireland Assembly (Election and Periods of
Suspension) Bill.” Bill 104 of 2002-2003. 9 May.
Bibliography 741
Hristova, L. (2000). “La démocratie post-communiste: Identité politique.” Manuscript ob-

tained from author. Skopje.
Huitfeldt, I. (2000). “Independent Judiciary in Bosnia and Herzegovina.” Ljudska Prava
1(3/4): 30-45.
Human Rights Chamber. (2000). Annual Report 1999.
Human Rights Watch/Helsinki. (1992). Bloodshed in the Caucasus: Violations of Humani-
tarian Law and Human Rights in the Georgia-South Ossetia Conflict. Helsinki and
New York: Human Rights Watch, April.
Human Rights Watch/Helsinki. (1993). War or Peace? Human Rights and Russian Military
Involvement in the “Near Abroad. Helsinki and New York: Human Rights Watch.
December.
Human Rights Watch. (1995). Georgia/Abkhazia: Violations of the Laws of War and Rus-
sia’s Role in the Conflict. Arms Project, Human Rights Watch/Helsinki 7(7).
Human Rights Watch. (2001). “Torture, Kidnappings by Albanians in Macedonia.” http://
www.hrw.org/press/2001/08/macedonia-0811.htm (12 October 2002).
Huntington, S.P. (1957). The Soldier and the State: The Theory and Politics of Civil-Military
Relations. Cambridge: Harvard University Press.
Hutchinson, J. (1987). The Dynamics of Cultural Nationalism: The Gaelic Revival and the
Creation of the Irish Nation State. London: Allen and Unwin.
Hutchman, L. (2000). Beyond Borders. Fredericton, New Brunswick: Broken Jaw Press.
Hyson, S. (1998a). “The Horrible Example”. Policy Options 9(8): 25-27
Hyson, S. (1998b). “A One Party Legislature: Where’s ‘Her Majesty’s Loyal Opposition’ in
the Loyalist Province?” Canadian Parliamentary Review 11(2): 22-25.
Ibrahimagić, O. (1999). Supremacija Bosne i Hercegovine nad entitetama. 2nd ed. Sarajevo:
Kult B andVijeće Kongresa bošnjačkih intelektualaca.
Independent Commission on Policing in Northern Ireland. (1999). “A New Beginning: Po-
licing in Northern Ireland - The Report of the Independent Commission on Policing in
Northern Ireland” (Patten Report). Belfast: Independent Commission on Policing in
Northern Ireland.
Information Telegraph Agency of Russia/ Telegraph Agency of the Soviet Union (ITAR/
TASS). (2000). 22 November.
Inman, R. and D. Rubinfeld. (1997). “Rethinking Federalism.” Journal of Economic Perspec-
tives 11(4): 43-64.
Institute for War and Peace Reporting (IWPR). (2001a). “NLA Harms Albanians’ Aspira-
tions.” IWPR Balkan Crisis Report 228, 21 March. http://www.iwpr.net (12 October
2002).
IWPR. (2001b). “Futile Dialogue Exposed.”
IPWR Balkan Crisis Report 228, 21 March. http://www.iwpr.net (12 October 2002).
IWPR. (2001c). IWPR Balkan Crisis Report 251, 31 May. http://www.iwpr.net (12 October
2002).
IWPR. (2001d). IWPR Balkan Crisis Report 264 (Part I), 20 July. http://www.iwpr.net (12
October 2002).
IWPR. (2001e). “Albanians Back Macedonian Unity,” IWPR Balkan Crisis Report 250, 25
May. http://www.iwpr.net (12 October 2002).
IWPR. (2003). “Militants Threaten Renewed Conflict.” IWPR Balkan Crisis Report 403, 3
February. http://www.iwpr.net (3 February).
742 Bibliography
International Bureau of Education. (2003). “Curriculum Change and Social Cohesion in

Conflict-affected Societies: Proposal of Northern Ireland.” http://www.ibe.unesco.
org/Regional/social_cohesion/sc_nire.htm (6 May).
International Criminal Investigative Training Assistance Programme (ICITAP). (2003).
“Macedonia”. http://www.usdoj.gov/criminal/icitap/macedonia.html (11 July).
International Crisis Group (ICG). (1999a). “Republika Srpska - Poplasen, Brcko and Kos-
ovo: Three Crises and Out.” 6 April.
ICG. (1999b). “Preventing Minority Return in Bosnia and Herzegovina: An Anatomy of
Fear and Hate.” 2 August.
ICG. (1999c). “Is Dayton Failing?” 28 October.
ICG. (2000). “Bosnia’s Refugee Logjam Breaks: Is the International Community Ready?”
30 May.
ICG. (2001a). “No Early Exit: NATO’s Continuing Challenge in Bosnia.” 22 May.
ICG. (2001b). “Macedonia: The Last Chance for Peace.” Skopje and Brussels: ICG.
ICG. (2001c). “Macedonia: War on Hold.” Skopje and Brussels: ICG.
ICG. (2001d). “The Macedonian Question: Reform or Rebellion.” Skopje and Brussels:
ICG.
ICG. (2002a). “Macedonia’s Public Secret: How Corruption Drags the Country Down.”
http://www.crisisweb.org/ (12 October).
ICG. (2002b). “Courting Disaster: The Misrule of Law in Bosnia and Herzegovina.” Sara-
jevo/Brussels: ICG.
ICG. (2003). “Moldova: No Quick Fix.” 12 August.
International Institute for Democracy and Electoral Assistance. (2002). The United Na-
tions and Democracy: Towards Sustainable Peace-Building. International IDEA Dis-
cussion Paper. New York: International IDEA.
International Monetary Fund. (2000). Republic of Moldova: Interim Poverty Reduction
Strategy Paper. Washington: International Monetary Fund.
Irish Business and Employers Confederation/Confederation of British Industry. (1998).
Position Paper on Strand Two of the Belfast Agreement (North-South Ministerial
Council). Dublin and Belfast: IBEC-CBI Joint Business Council.
Irwin, C. (2003). “Devolution and the State of the Northern Ireland Peace Process.” The
Global Review of Ethnopolitics 2(3-4): 71–91.
Jackson, R. (2000). The Global Covenant. Oxford: Oxford University Press.
Jackson-Preece, J. (1998). National Minorities and the European Nation-states System. Ox-
ford: Oxford University Press.
Jackson-Preece, J. (2003). “Human Rights and Cultural Pluralism.” G. Lyons and J. Mayall
(eds.). International Human Rights in the Twenty-first Century: Protecting the Rights
of Groups. Oxford: Rowman and Littlefield.
Järve, P. (1998). From Ethnopolitical Conflict to Inter-ethnic Accord in Moldova. ECMI Re-
port 1. Flensburg: European Centre for Minority Issues.
Jennings, W.I. (1956). The Approach to Self Government. Cambridge: Cambridge Univer-
sity Press.
Joint Control Commission on the Georgia-South Ossetia Conflict Regulation (JCCG-
SOCR). (1996). Memorandum on Measures to Provide Security and Strengthen Mu-
tual Trust Between Sides in the Georgia-South Ossetia Conflict. http:ecmi.de/cps/
documents_gum_memo.html.
Bibliography 743
JCCGSOCR. (1997). Protocol on Interaction of Law-enforcement Bodies in the Georgian-

South Ossetian Conflict Zone. JCC Working Group on Security and Military Issues.
26 September.
JCCGSOCRR. (2000). Regulations on the Special Coordinating Center under the JCC on
Interaction Between the Law Enforcement Bodies in the Zone of the Georgian-Osse-
tian Conflict Zone. [Translated from Russian]. 29 February.
JCCGSOCRR. (2001a). Protocol of the 11th Plenary Meeting of the Joint Control Commis-
sions. Vladikavkaz, Russia. 21 April.
JCCGSOCRR. (2001b). Protocol on Elaboration of Russian-Georgian Intergovernmental
Program on Cooperation for Economic Rehabilitation of the Georgian-South Ossetian
Conflict Zone. Georgia. 19 July.
Judah, T. (2001). “A Greater Albania?” The New York Review of Books 48(8): 35-37.
Juzhnaja Osetija. (2000). 9 June.
Kabui, J. (2002). “Peacebuilding and Consolidation.” A. Carl and L. Garasu (eds). Weaving
Consensus: The Papua New Guinea – Bougainville Peace Process. London: Concilia-
tion Resources.
Kaldor, M. (1999). Old and New Wars. London: Polity Press.
Kamkia, F. (1996). “The Difficulties of Education in Post-war Abkhazia.” Paper presented
at the Conference on the Role of International Organizations in the Process of Post-
Conflict Peace-Building in Abkhazia, Pitsunda, October. http://www.apsny.org/pit-
sunda8.html (6 March 2003).
Kassof, A.H. (2001). “President’s Report.” Project of Ethnic Relations 16.
Katz, R. (1980). A Theory of Parties and Electoral Systems. Baltimore: Johns Hopkins Uni-
versity Press.
Katzenstein, P. (1985). Small States in World Markets. Ithaca: CUP.
Kelly, P. (2003). “Neighbourhood Saviour.” The Weekend Australian. June 14-15.
Kelman, H.C. (2002). “Interactive Problem-solving: Informal Mediation by the Scholar-
Practitioner.” J. Bercovitch (ed.). Studies in International Mediation. Basingstoke:
Palgrave.
Khachikian, A. (2000). “Multilateral Mediation in Intrastate Conflicts: Russia, the United
Nations, and the War in Abkhazia.” M.C. Greenburg, J.H. Barton, and M.E. McGuin-
ness (eds.). Mediation and Arbitration to Prevent Deadly Conflict: Carnegie Commis-
sion on Preventing Deadly Conflict. New York: Carnegie Corporation of New York.
Kicker, R., J. Marko, and M. Steiner (eds.). (1998). Changing Borders: Legal and Economic
Aspects of European Enlargement. Frankfurt: Lang.
Kiki, A. (1968). Ten Thousand Years in a Lifetime. Melbourne: F.W. Cheshire.
King, A. (2001). Does the United Kingdom Still Have a Constitution? The Hamlyn Lectures:
Fifty-Second Series. London: Sweet and Maxwell.
King, C. (1995). “Gagauz Yeri and the Dilemmas of Self-determination.” Transition 1(19):
21-25.
King, C. (1998). “Ethnicity and Institutional Reform: The Dynamics of ‘Indiginization’ in
the Moldovan ASSR.” Nationalities Papers 26(1): 57-72.
Kirk-Greene, A.H.M. (ed.). (1971). Crisis and Conflict in Nigeria: A Documentary Source-
book. Oxford: Oxford University Press.
Knaus, G. and F. Martin (2003). “Lessons from Bosnia and Herzegovina: Travails of the
European Raj.” Journal of Democracy 14(3): 60-74.
744 Bibliography
Kontakt. (1999). Alternativa: Public Information Center on Regulation of Georgian-South

Ossetian Conflict.
Kostovicova, D. (2002). The Politics of Identity and Space: Albanian-language Education in
Kosovo, 1992-1998. Ph.D. dissertation. University of Cambridge.
Koštunica, V. (1999). “Postface.” A. Buha (ed.). Argument en faveur de la République Serbe.
Lausanne: L’Age d’Homme.
Krause, S. (2000). “Surviving a Year of Crises.” Transitions Online. Prague. http://www.tol.
cz (12 October 2002).
Kriesberg, L. (2003). Constructive Conflicts: From Escalation to Resolution. 2nd ed. Lan-
ham: Rowman and Littlefield.
Kristioglo, I. (2000). “Dinamika regional’nyhh etnopoliticeskihh protsessov na primere
Gagauzii (Gagauz-Eri).” (Gagauzia as an Example of Dynamics of Regional Ethno-
political Processes).
Kritz, N. (1996). “The Rule of Law in the Post-conflict Phase.” C.A. Crocker, F.O. Hampson,
and P. Aall (eds.). Managing Global Chaos: Sources and Responses to International
Conflict. Washington D.C.: US Institute of Peace Press.
Krugman, P. (1997). “Good News from Ireland: A Geographical Perspective.” A. Gray (ed.).
International Perspectives on the Irish Economy. Dublin: Indecon Economic Consult-
ants.
Kviris Palitra. (2001). June.
Kymlicka, W. (1996). Multicultural Citizenship: A Liberal Theory of Minority Rights. Ox-
ford: Oxford University Press.
Kymlicka, W. (2002). Multiculturalism and Minority Rights: West and East. JEMIE paper,
Issue 4/2002. Flensburg: European Centre for Minority Issues.
Lantschner, E. and R. Medda-Windischer. (2003). “Protection of National Minorities
through Bilateral Agreements in South Eastern Europe.” European Yearbook of Mi-
nority Issues 1: 535–561. The Hague: Kluwer.
Laakso, M. and R. Taagepera. (1979). “Effective Number of Parties: A Measure with Ap-
plication to West Europe.” Comparative Political Studies 12: 3-27.
Laarhoven, R. (1989). Triumph of Moro Diplomacy: The Maguindanao Sultanate in the
17th Century. Quezon City: New Day Publishers.
Laffan, M. (1983). The Partition of Ireland, 1911-25. Dundalk: Dundalgan Press.
Laffan, M. (1999). The Resurrection of Ireland: The Sinn Féin Party, 1916-1923. Cambridge:
Cambridge University Press.
Lal, B. (2002). “Constitutional Engineering in Post Coup Fiji.” Andrew Reynolds (ed.). The
Architecture of Democracy. Oxford: Oxford University Press.
Lapidoth, R. (1997). Autonomy: Flexible Solutions to Ethnic Conflicts: United States Insti-
tute of Peace Press.
Laponce, J. (1960). The Protection of Minorities. Berkeley: University of California Press.
Lawlor, S.M. (1980-81). “Ireland from Truce to Treaty: War or Peace? July to October 1921.”
Irish Historical Studies 22: 49–64.
Leanen, R. (1998). “Russian Peacekeeping in the ‘Near Abroad’: Human Rights Protection
or Neo-Imperialism? South Ossetia – A Case Study”. K. Malfliet and R. Leanen (eds.).
Minority Policy in Central and Eastern Europe: The Link Between Domestic Policy,
Foreign Policy and European Integration. Leuven: Katholike Universiteit Leuven, In-
stituut voor Europees Beleid.
Bibliography 745
Lehmbruch, G. (1993). “Consociational Democracy and Corporatism in Switzerland”. Pub-

lius 23: 43-60.
Lewis, W.A. (1965). Politics in West Africa. London: Allen and Unwin.
Lijphart, A. (1968). The Politics of Accommodation. Yale University Press.
Lijphart, A. (1975). “The Northern Ireland Problem: Cases, Theories and Solutions”. British
Journal of Political Science 5: 83–106.
Lijphart, A. (1977). Democracy in Plural Societies: A Comparative Exploration. New Ha-
ven: Yale University Press.
Lijphart, A. (1984). Democracies: Patterns of Majoritarian and Consensus Government in
Twenty-one Countries. New Haven, London: Yale University Press.
Lijphart, A. (1985). Power-sharing in South Africa. Berkeley: University of California
Press.
Lijphart, A. (1990a). “Electoral Systems, Party Systems and Conflict Management in Di-
vided Societies.” R. Schrire (ed.). Critical Choices for South Africa. Cape Town: Ox-
ford University Press.
Lijphart, A. (1990b). “Foreword: One Basic Problem, Many Theoretical Options - And a
Practical Solution?” J. McGarry and B. O’Leary (eds.). The Future of Northern Ire-
land. Oxford: Clarendon Press.
Lijphart, A. (1994). Electoral Systems and Party Systems. Oxford: Oxford University
Press.
Lijphart, A. (1995). “Self-determination versus Pre-determination of Ethnic Minorities in
Power-sharing Systems.” W. Kymlicka (ed.). The Rights of Minority Cultures. Oxford:
Lijphart, A. (1996). “The Puzzle of Indian Democracy: A Consociational Interpretation.”
American Political Science Review 90(2): 258–268.
Linz, J. and A. Stepan. (1992). “Political Identities and Electoral Sequences: Spain, the So-
viet Union, and Yugoslavia.” Daedalus 121(2): 123-139.
Livingstone, S. (2001). “The Judiciary and Legal Profession in Transition.” C. Harvey (ed.).
Human Rights, Equality and Democratic Renewal in Northern Ireland. Oxford: Hart
Publishing.
Lloyd, A. (1999). My War Gone By, I Miss it So. London: Transworld Publishers.
Low-Beer, A. (2001). “Politics, School Textbooks and Cultural Identity: The Struggle in
Bosnia and Hercegovina.” International Textbook Research 23(2): 215-223.
Lund, M. (2000). “Preventive Diplomacy for Macedonia, 1992-1999: From Containment to
Nation-Building.” B. Jentleson (ed.). Opportunities Missed, Opportunities Seized. B.
Lanham: Rowman and Littlefield.
Lustick, I.S. (1979). “Stability in Deeply Divided Societies: Consociationalism Versus Con-
trol.” World Politics 31(3): 325–344.
Lustick, I.S. (1985). State-building Failure in British Ireland and French Algeria. Berkeley:
University of California Press.
Lustick, I.S. (1993). Unsettled States, Disputed Lands: Britain and Ireland, France and Al-
geria, Israel and the West Bank-Gaza. Ithaca: Cornell University Press.
Luverà, B. (1996). Oltre il Confine. Euregio e conflitto etnico: tra regionalismo europeo e
nuovi nazionalismi in Trentino-Alto Adige. Bologna: Il Mulino.
Lyons, F.S.L. (1973). “From War to Civil War in Ireland: Three Essays on the Treaty Debate.”
B. Farrell (ed.). The Irish Parliamentary Tradition. Dublin: Gill and MacMillan.
746 Bibliography
Lyons, G. and J. Mayall (eds.) (2003). International Human Rights in the Twenty-First Cen-
tury: Protecting the Rights of Groups. Oxford: Rowman and Littlefield.
Mable, J. A. (2001). “Toward Peace in the South.” Philippines Free Press 92(29): 18-20.
Macedonia, Ministry of Education and Science (MMoES). (2001). National Report: Devel-
opment of Education in the Republic of Macedonia. Skopje: UNESCO.
Macedonia.org (MO). (2001). “Education in Macedonia.” http://faq.Macedonia.org.infor-
mation/education.html (6 March 2003).
MacFarlane, S.N., L. Minear, and S. Shenfield. (1996). Armed Conflict in Georgia: A Case
Study in Humanitarian Action and Peacekeeping. Occasional Paper 21, The Thomas
J. Watson Jr. Institute for International Studies, Brown University.
MacGinty, R. (1997a) “American Influences on the Northern Ireland Peace Process”. Jour-
nal of Conflict Studies 17: 31-50
MacGinty, R. (1997b). “Bill Clinton and the Northern Ireland Peace Process.” Aussenpolitik
48(3): 237–244.
Macrory, R. and Turner, S. (2002). “Cross-border Environmental Governance.” Regional
and Federal Studies 12(4): 59–87.
Mageean, P. and M. O’ Brien. (1999) “From the Margins to the Mainstream: Human Rights
and the Good Friday Agreement.” Fordham International Law Journal 22: 1499–
1538.
Malfliet, K. and R. Leanen (eds.). Minority Policy in Central and Eastern Europe: The Link
Between Domestic Policy, Foreign Policy and European Integration. Leuven: Katho-
like Universiteit Leuven, Instituut voor Europees Beleid.
Mallie, E. and D. McKittrick. (1996). The Fight for Peace: The Secret Story Behind the Irish
Peace Process. London: Heinemann.
Mamak, A. and R. Bedford. (1974). Bougainvillean Nationalism: Aspects of Unity and Dis-
cord. Bougainville Special Publication No. 1. Cristchurch: University of Canterbury,
Department of Geography.
Mansergh, N. (1997). “The Prelude to Partition: Concepts and Aims in India and Ireland.”
D. Mansergh (ed.). Nationalism and Independence: Selected Irish Papers. Cork: Cork
University Press.
Mantovani, G. (2000). Exploring Borders: Understanding Culture and Psychology. Phila-
delphia: Routledge.
March, J. and H. Simon. (1993). Organizations. Oxford: Blackwell.
Marsh, N.E. (1959). The Rule Of Law in A Free Society: A Report Of The International Con-
gress Of Jurists, New Delhi, India, January 5–10, 1959. New Delhi, India: International
Congress of Jurists.
Massicotte, L. and A. Blais. (1999). “Mixed Electoral Systems: A Conceptual and Empirical
Survey.” Electoral Studies 18: 341-366.
Matoshi, H. (2001). “The Albanians’ New Model Army.” IWPR Balkan Crisis Report 274,
24 August. http://www.iwpr.net (12 October 2002).
May, R.J. (1992). “The Wild West in the South: a Recent Political History of Mindanao.”
M. Turner, R.J. May, and L.R. Turner (eds.). Mindanao: Land of Unfulfilled Promise.
Quezon City: New Day Publishers.
May, R.J. (1999). “Decentralization in Papua New Guinea: Two Steps Forward, One Step
Back.” M. Turner (ed.). Central-Local Relations in Asia-Pacific: Convergence or Diver-
gence? London and New York: Macmillan Press Ltd, and St. Martin’s Press, Inc.
Bibliography 747
May, R.J. (2001). “Muslim Mindanao: Four Years After the Peace Agreement.” Southeast
Asian Affairs 2001: 263-275. Singapore: Institute of Southeast Asian Studies.
May, R.J. and A.J. Regan (eds.). (1997). Political Decentralization in a New State: the Expe-
rience of Provincial Government in Papua New Guinea. Bathurst: Crawford House
Press.
May, R.J. and M. Spriggs (eds.). (1990). The Bougainville Crisis. Bathurst NSW: Crawford
House Publishing.
Mayall, J. (1990). Nationalism and International Society. Cambridge: Cambridge Univer-
sity Press.
Mayall, J. (2000) “National Self-determination and International Order: The Asian Experi-
ence.” M. Leifer (ed.). Asian Nationalism. London: Routledge.
McCartney, R. (2000). “Devolution is a Sham.” Observer, 20 February.
McCrudden, C. (1999a). “Equality and the Good Friday Agreement.” J. Ruane and J. Todd
(eds.). After the Good Friday Agreement: Analysing Political Change in Northern Ire-
land. Dublin: University College Dublin Press.
McCrudden, C. (1999b). “Mainstreaming Equality in the Governance of Northern Ireland.”
Fordham International Law Journal 22: 1696–1775.
McCrudden, C. (2001). “Equality.” C. Harvey (ed.). Human Rights, Equality and Demo-
cratic Renewal in Northern Ireland. Oxford: Hart Publishing.
McEvoy, K. (2001). Paramilitary Imprisonment in Northern Ireland: Resistance, Manage-
ment and Release. New York: Oxford University Press.
McGarry, J. (1990). “Northern Ireland and the Option of Consociationalism.” Plural Socie-
ties XX: 1–20.
McGarry, J. (ed.). (2001a). Northern Ireland and the Divided World: Post-Agreement
Northern Ireland in Comparative Perspective. Oxford: Oxford University Press
McGarry, J. (2001b). “Globalization, European Integration and the Northern Ireland Con-
flict.” M. Keating and J. McGarry (eds.). Minority Nationalism and the Changing In-
ternational Order. Oxford: Oxford University Press.
McGarry, J. (2003a). “Consociational Theory and the Good Friday Agreement.” European
Yearbook of Minority Issues 1: 288–312. The Hague: Kluwer Law International.
McGarry, J. (2003b). “Constructing a Federation in the New Iraq: The Lessons from Can-
ada and Other Multi-ethnic Federations.” Paper delivered to the Conference “Multi-
Nationalism, Power-sharing and the Kurds in a New Iraq.” Cafritz Foundation Con-
ference Center, George Washington University. September 12.
McGarry, J. and B. O’Leary (eds.). (1990a). The Future of Northern Ireland. Oxford: Oxford
University Press.
McGarry, J. and B. O’Leary. (1990b) “Northern Ireland’s Future: What is to be Done?”
Conflict Quarterly 10(2): 42-62.
McGarry, J. and B. O’Leary (eds.). (1993). The Politics of Ethnic Conflict Regulation. Lon-
don: Routledge.
McGarry, J. and B. O’Leary. (1995a). Explaining Northern Ireland: Broken Images. Oxford,
Cambridge M.A.: Basil Blackwell.
McGarry, J. and B. O’Leary. (1995b). “Five Fallacies: Northern Ireland and the Liabilities of
Liberalism.” Ethnic and Racial Studies 18(4): 837–61.
McGarry, J. and B. O’Leary. (1999). Policing Northern Ireland: Proposals for a New Start.
Belfast: Blackstaff.
748 Bibliography
McGarry, J. and B. O’Leary. (2003). “Federalism, Conflict-regulation and National and

Ethnic Power-sharing.” Paper presented at the Annual Meeting of the American Po-
litical Science Association, Philadelphia, August 28-31.
McGarry, J. and B. O’Leary (eds.) (2004). The Northern Ireland Conflict: Consociational
Engagements. Oxford: Oxford University Press.
McGrath, M. (2000). The Catholic Church and Catholic Schools in Northern Ireland: The
Price of Faith. Dublin: Irish Academic Press.
McKenna, T.M. (1998). Muslim Rulers and Rebels: Everyday Politics and Armed Separa-
tism in the Southern Philippines. Pasig City: Anvil Publishing.
McKittrick, D. (2001). “The Running Man of Drumcree.” Independent on Sunday, 1 July.
McKittrick, D., S. Kelters, B. Feeney, and C. Thornton. (2001). Lost Lives: The Stories of the
Men, Women and Children Who Died as a Result of the Northern Ireland Troubles.
Edinburgh: Mainstream Publishing.
McRae, K. (ed.). (1974). Consociational Democracy: Political Accommodation in Segment-
ed Societies. Toronto: McClelland and Stewart.
McRae, K. (1975). “The Principle of Territoriality and the Principle of Personality in Multi-
lingual States.” International Journal of the Sociology of Language 4: 33-54.
Meeting of the Conference of the Ministers of Education of Bosnia and Herzegovina
(2000a). “Declaration.” 10 May.
Meeting of the Conference of the Ministers of Education of Bosnia and Herzegovina
(2000b). “Agreement.” 10 May.
de Mello, L. (1999). Fiscal Federalism and Government Size in Transition Economies: the
Case of Moldova. IMF Working Paper, WP/99/176. Washington: International Mon-
etary Fund.
Méndez, J.E. (1997). “In Defense of Transitional Justice.” A. J. McAdams (ed.). Transitional Jus-
tice and the Rule of Law in New Democracies. Indiana: University of Notre Dame Press.
Mertus, J. (1999). From Legal Transplants to Transformative Justice: Human Rights and the
Promise of Transnational Civil Society. Atlanta: Halle Institute Occasional Paper.
Mercado, E.R. (1992). “The Moro People’s Struggle for Self-determination.” M. Turner, R.
May, and L.R. Turner (eds.). Mindanao: Land of Unfulfilled Promise. Quezon City:
New Day Publishers.
Mercado, E.R. (1998). “Peace and Development: The MNLF and the SPCPD Experience.”
Kalinaw Mindanao. http://mindanao.com/kalinaw/peaceproc/pa-imp-mercado.htm
(7 July 2001).
Mill, J.S. (1859). Utilitarianism, On Liberty, and Considerations on Representative Govern-
ment. London: Everyman.
Mill, J.S. (1865). Considerations on Representative Government. London: Longman.
Mitchell, G.C. (2000). Making Peace. Berkeley: University of California Press.
Mitchell, G.C., J. de Chastelain, and H. Holkeri. (1996). Report of the International Body on
Arms Decommissioning (The Mitchell Report). Dublin and London.
Mitchell, P. (1991). “Conflict Regulation and Party Competition in Northern Ireland.” Eu-
ropean Journal of Political Research 20(1): 67–92.
Mitchell, P. (1995). “Party Competition in an Ethnic Dual Party System.” Ethnic and Racial
Studies 18(4): 773–96.
Mitchell, P. (1999). “The Party-system and Party Competition.” P. Mitchell and R. Wilford
(eds.). Politics in Northern Ireland. Boulder: Westview Press.
Bibliography 749
Mitchell, P. (2001). “Transcending an Ethnic Party System?” R. Wilford (ed.). The Impact
of Consociational Governance on Electoral Dynamics and the Party System. Oxford:
Mitchell, P., B. O’Leary, and G. Evans. (2001). “Northern Ireland: Flanking Extremists Bite
the Moderates and Emerge in Their Clothes.” Parliamentary Affairs 54(4): 725–742.
Mitchell, P., B. O’Leary, and G. Evans. (2002). “The 2001 Elections in Northern Ireland: Mod-
erating ‘Extremists’ and the Squeezing of the Moderates.” Representation 39(1): 23–36.
Mitchell, P. and R. Wilford (eds.). (1999). Politics in Northern Ireland. Boulder: Westview
Press.
Monroe, B. and A. Rose. (2002). “Electoral Systems and Unimagined Consequences: Parti-
san Effects of Districted Proportional Representation.” American Journal of Political
Science 46(1): 67-89.
Morris, V. and M. Scharf. (1998). The International Criminal Tribunal for Rwanda. New
York: Transnational Publishers.
Moser, R. (1999). “Electoral Systems and the Number of Parties in Post-Communist States.”
World Politics 51(3): 359-384.
Musgrave, T. (1997). Self-determination and National Minorities. Oxford: Oxford Univer-
sity Press.
Muslim, M.A. and R. Cagoco-Guiam. “Mindanao Land of Promise.” Accord 6.
Naegele, J. (2001). “UCK Guerrillas Flex Their Muscles.” RFE/RL Balkan Report 5 (22-23).
http://www.rferl.org/balkan-report/ (12 October 2002).
Najčevska, M. and V. Petroska Beška. (1999). “Inter-Ethnic Relations in the Republic of
Macedonia.” Helsinki Monitor 10(3): 8-14.
Nanevska, B. (2002). “Ball Extravaganza, Soup Kitchens and Garbage Containers.” AIM
Skopje. http://www.aimpress.org (12 October 2002).
Nash, J. and E. Ogan. (1990). “The Red and the Black: Bougainvillean Perceptions of Other
Papua New Guineans.” Pacific Studies 13(2): 1-17.
National Statistics Office (NSO). (2001). “Analysis of the Results of the Functional Literacy,
Education and Mass Media Survey.” Manila. http://www.census.gov.ph/data/sector-
data/fl94-funclit.html (9 March 2003).
NSO. (2003). “Population and Literacy Situation in the Philippines, in Maguindanao and
the Autonomous Region of Muslim Mindanao (ARMM).” Manila. http://www.accu.
or.jp/litdbase/literacy/wed/statis1.htm (6 March 2003).
Neuheiser, J. and S. Wolff (eds.). (2002). Peace at Last? New York and Oxford: Berghahn.
Neukirch, C. (2000) “Autonomy and Conflict Transformation: The Case of the Gagauz
Territorial Autonomy in the Republic of Moldova”. Paper at the seminar “Concepts of
Minority Governance in Europe on the Threshold of the 21st Century.” 6–7 October
2000, Bozen/Bolzano, South Tyrol.
Neukirch, C. (n.d). National Minorities in the Republic of Moldova: Some Lessons Learned,
Some Not? Manuscript.
Ní Aoláin, F. (2000). The Politics of Force: Conflict Management and State Violence in
Northern Ireland. Belfast: The Blackstaff Press.
Nick, R. and G. Pallaver. (1998). Jenseits von Grenzen. Die Europaregion aus der Sicht der
Bevölkerung. Innsbruck: Studia Universitätsverlag.
Nodia, G. (1998). Causes and Visions of Conflict in Abkhazia. Working paper for the Berkley
Program in Soviet and Post-Soviet Studies. Berkeley: University of California.
750 Bibliography
Nordlinger, E. (1972). Conflict Regulation in Divided Societies. Cambridge, M.A.: Harvard

Studies in International Affairs.
North Atlantic Treaty Organization (NATO). (2001). “NATO’s Second Mission in
the Former Yugoslav Republic of Macedonia”. http://www.nato.int/docu/up-
date/2001/0924/e0927a.htm (12 October 2002).
North, D. (1990). Institutions, Institutional Change and Economic Performance. Cam-
bridge: Cambridge University Press.
Northern Ireland Assembly (1999). Hansard: Bound Volume II. 15 July.
Northern Ireland Office. (2000). “Review of the Criminal Justice System in Northern
Ireland”. http://www.nio.gov.uk/review_of_the_criminal_justice_system_in_north-
ern_ireland.pdf.
Nowak, M. (1999). “Is Bosnia and Herzegovina Ready for Membership in the Council of
Europe?” Human Rights Law Journal 20(7-11): 285-289.
O’Callaghan, M. (1999). Enemies Within: Papua New Guinea, Australia, and the Sandline
Crisis: The Inside Story. Sydney: Doubleday.
O’Callaghan, M. (2003). “Pacific Policemen.” The Weekend Australian. June 14-15.
O’Clery, C. (1997 (1996)). The Greening of the White House: The Inside Story of How Amer-
ica Tried to Bring Peace to Ireland. Dublin: Gill and Macmillan.
O’Duffy, B., and B. O’Leary. (1990). “Appendix 3. Political Violence in Northern Ireland,
1969-June 1989.” J. McGarry and B. O’Leary (eds.). The Future of Northern Ireland.
O’Duffy, B. and B. O’Leary. (1995). “Tales from Elsewhere and an Hibernian Sermon.” H.
Margetts and G. Smyth (eds.). Turning Japanese? Britain with a Permanent Party of
Government. London: Lawrence and Wishart.
Office of the High Representative of the Secretary-General (OHR). (2000a). Newsletter,
Economic Reform and Reconstruction in Bosnia and Herzegovina (BiH) 4(2), April.
OHR. (2000b). “High Representative’s Decision: Decision Establishing the BiH State
Court.” Sarajevo, 12 November.
OHR. (2001a). “Education Policy in Bosnia.” OHR Human Rights/ Rule of Law Depart-
ment. 10 May. http://www.ohr.int (6 March 2003).
OHR. (2001b). “High Representative’s Decision: Decision Providing the Independent Judi-
cial Commission (IJC) with a Comprehensive Mandate.” Sarajevo, 14 March.
OHR. (2001c). “High Representative Decides SPRA Between RS and FRY Does Not Vio-
late Vital Interests.” Sarajevo, 6 June.
OHR. (2001d). “Independent Judicial Commission Signs Memorandum of Understanding
on Appointment Procedures in the Judiciary.” Sarajevo, 4 July.
OHR. (2002a). Press Release: High Representative Issues Decision on Judicial and Prosecu-
torial Appointments. OHR Sarajevo, Thursday, 4 April 2002.
OHR. (2002b). Press Release: High Representative Imposes Package to Strengthen Democ-
racy and the Rule of Law. OHR Sarajevo, Friday, 24 May 2002.
OHR. (2002c). “Decision on Constitutional Amendments in Republika Srpska.” Sarajevo,
19 April.
OHR. (2002d). “Decision on Constitutional Amendments in the Federation of Bosnia and
Herzegovina.” Sarajevo, 19 April.
OHR. (2002e). “Decision Amending the BiH Election Law, Placing it in Accord with the
New Entity Constitutions.” Sarajevo, 19 April.
Bibliography 751
Ogan, E. (1972). “Business and Cargo: Socio-economic Change among the Nasioi of Bou-
gainville.” New Guinea Research Bulletin No. 44. Canberra: The Australian National
University.
Ogan, E. (1991). “The Cultural Background to the Bougainville Crisis.” Journal de la Société
des Océanistes 92/93(1/2): 61-67.
Ohmae, K. (1996). The End of the Nation State. London: Harper Collins.
O’Leary, B. (1987). “The Anglo-Irish Agreement: Statecraft or Folly ?” West European Poli-
tics 10(1): 5–32.
O’Leary, B. (1989). “The Limits to Coercive Consociationalism in Northern Ireland.” Politi-
cal Studies 37(4): 452–68.
O’Leary, B. (1990). “Appendix 4. Party Support in Northern Ireland, 1969-89. J. McGarry
and B. O’Leary (eds.). The Future of Northern Ireland. Oxford: Oxford University
Press.
O’Leary, B. (1992). “Public Opinion and Northern Irish Futures.” Political Quarterly 63(2):
143–170.
O’Leary, B. and J. McGarry. (1993). The Politics of Antagonism: Understanding Northern
Ireland. London and Atlantic Heights: Athlone.
O’Leary, B. (1995a). “Afterword: What is Framed in the Framework Documents?” Ethnic
and Racial Studies 18(4): 862–872.
O’Leary, B. (1995b). “Brave New Worlds: The UK and Irish Governments’ Proposals for
Northern Ireland.” The Parliamentary Policy Forum 1: 29–31.
O’Leary, B. (1995c). “Introduction: Reflections on a Cold Peace.” Ethnic and Racial Studies
18(4): 695–714.
O’Leary, B. (1997a). “The Conservative Stewardship of Northern Ireland 1979-97: Sound
Bottomed Contradictions or Slow Learning?” Political Studies 45(4): 663–676.
O’Leary, B. (1997b). “Unionists Will Lose Electoral Dominance.” Irish Times, 2 July.
O’Leary, B. (1999a). “The 1998 British-Irish Agreement: Consociation Plus!” Scottish Af-
fairs 26: 1–22.
O’Leary, B. (1999b). “A Bright Future and Less Orange.” Review of ‘A New Beginning’ by
the Independent Commission on Policing for Northern Ireland.” Times Higher Edu-
cation Supplement. 19 November.
O’Leary, B. (1999c). “The Implications for Political Accommodation in Northern Ireland
of Reforming the Electoral System for the Westminster Parliament.” Representation
35(2-3): 106–113.
O’Leary, B. (1999d). “The Nature of the Agreement.” Fordham Journal of International Law
22(4): 1628–1667.
O’Leary, B. (1999e). “The Nature of the British-Irish Agreement.” New Left Review 233:
66–96.
O’Leary, B. (2000a). “3,636 so Far, and Counting.” Review of D. McKittrick et al. Lost Lives:
The Stories of the Men, Women and Children who Died as a Result of the Northern
Ireland Troubles. Times Higher Education Supplement, 3 March.
O’Leary, B. (2000b). “Letter on Patten Report and Police Bill.” Irish Times, 9 August.
O’Leary, B. (2000c). “The Patten, the Whole Patten and Nothing but the Patten.” Irish
Times, 28 July.
O’Leary, B. (2000d). “Suspension Would be a Potentially Fatal Blow.” Irish Times, 5 Febru-
ary.
752 Bibliography
O’Leary, B. (2000e). “What a Travesty: Police Bill is Just a Parody of Patten.” Sunday Busi-
ness Post, 30 April.
O’Leary, B. (2001a). “Agreement is Not Devolution in Unitary State.” Irish Times. 11 Oc-
tober.
O’Leary, B. (2001b). “The Belfast Agreement and the Labour Government: How to Handle
and Mishandle History’s Hand.” A. Seldon (ed.). The Blair Effect: The Blair Govern-
ment 1997-2001. London: Little, Brown & Company.
O’Leary, B. (2001c). “The Character of the 1998 Agreement: Results and Prospects.” R.
Wilford (ed.). Aspects of the Belfast Agreement. Oxford: Oxford University Press.
O’Leary, B. (2001d). “Comparative Political Science and the British-Irish Agreement.” J. Mc-
Garry (ed.). Northern Ireland in a Divided World. Oxford: Oxford University Press.
O’Leary, B. (2001e). “Elections, Not Suspensions.” The Guardian, 13 July.
O’Leary, B. (2001f ). “Personal View: Ignore the Prophets of Doom.” Financial Times, 11
June.
O’Leary, B. (2001g). “The Protection of Human Rights under the Belfast Agreement.” Po-
litical Quarterly 72(3): 353–365.
O’Leary, B. (2001h). “Reid Should Not Allow a New Unionist Boycott.” Irish Times, 4 Oc-
tober.
O’Leary, B. (2001i). “The Character of the 1998 Agreement: Results and Prospects.” R. Wil-
ford (ed.). Aspects of the Belfast Agreement. Oxford: Oxford University Press.
O’Leary, B. (2001j). “An Iron Law of Nationalism and Federation? A (neo-Diceyian) the-
ory of the Necessity of a Federal Staatsvolk, and of Consociational Rescue.” Nations
and Nationalism 7(3): 273-296.
O’Leary, B. (2002a). “The Belfast Agreement and the British-Irish Agreement: Consocia-
tion, Confederal Institutions, a Federacy and a Peace Process.” A. Reynolds (ed.). The
Architecture of Democracy. Oxford: Oxford University Press.
O’Leary, B. (2002b). “The Belfast Agreement: The Making, the Management and the Mis-
management of a Complex Association.” C. Hay (ed.). British Politics Today. Oxford:
Polity.
O’Leary, B. (2003). “Multi-national Federalism, Federacy, Power-Sharing and the Kurds of
Iraq.” Paper delivered to the Conference “Multi-Nationalism, Power-Sharing and the
Kurds in a New Iraq”. Cafritz Foundation Conference Center, George Washington
University. September 12.
O’Leary, B. (2005). “Debating Consociation: Normative and Explanatory Arguments.”
S.J.R. Noel (ed.). From Power-sharing to Democracy: Post-conflict Institutions in Eth-
nically Divided Societies. Montreal: McGill-Queens University Press.
O’Leary, B., B. Grofman, and J. Elklit. (2001). The Use of Divisor Mechanisms to Allocate
and Sequence Ministerial Portfolio Allocations: Theory and Evidence from Northern
Ireland. Working paper presented at the European Public Choice conference in Par-
is, 19 April.
O’Leary, B. and G. Evans. (1997c). “Northern Ireland: La Fin de Siècle, The Twilight of the
Second Protestant Ascendancy and Sinn Féin’s Second Coming.” Parliamentary Af-
fairs 50(4): 672–680.
O’Leary, B. and J. McGarry. (1996). The Politics of Antagonism: Understanding Northern
Ireland. 2nd expanded ed. London and Atlantic Heights: Athlone.
O’Leary, B., J. McGarry, and K. Salih. (eds.). (2005). The Future of Kurdistan in Iraq. Phila-
delphia: University of Pennsylvania Press.
Bibliography 753
Oliver, M. (1991). Black Islanders: A Personal Perspective of Bougainville 1937-1991. Mel-

bourne: Hyland House.
Olson, M. (2000). Power and Prosperity: Outgrowing Communist and Capitalist Dictator-
ships. New York: Basic Books.
O’Rawe, M. (2003). “Transitional Policing Arrangements in Northern Ireland: The Can’t
and Won’t of the Change Dialetic.” Fordham International Law Journal 26(4): 1015–
1073.
Ordeshook, P.C. and O. Shvetsova. (1994). “Ethnic Heterogeneity, District Magnitude, and
the Number of Parties.” American Journal of Political Science 38(1): 100-123
Organization for Economic Cooperation and Development (OECD). (2001). Thematic Re-
view of National Policies for Education – Kosovo. Paris: OECD.
Organization of Islamic Conference (OIC). (1972) Third Islamic Foreign Ministers Con-
ference (ICFM). Resolution No. 12, “Resolution on the Situation of Moslems in the
Philippines.” Jeddah, Saudi Arabia. 29 February–4 March.
OIC. (1998). “The Highlights of Implementation of the Final Peace Agreement.” Kalinaw
Mindanao. 26 August 1998. http://mindanao.com/kalinaw/peaceproc/pa-imp-oic.
htm (7 July 2001).
Organization for Security and Cooperation in Europe (OSCE). (1989). “Concluding Docu-
ment of the Vienna Meeting 1986.”
OSCE. (1990). “Charter of Paris for a New Europe.” Meeting of Heads of State and Govern-
ment.
OSCE. (1991). “Document of the Moscow Meeting of the Conference on the Human Di-
mension of the CSCE.”
OSCE. (1994a). Transdniestrian Conflict: Origins and Issues. Background information pa-
per. http://www.osce.org/moldova/documents/files/background.pdf (10 June).
OSCE. (1994b). OSCE Mission to Georgia. Constitutional Status of South Ossetia. Draft
Proposal. Tbilisi. 15 August.
OSCE. (1996). Declaration of the OSCE Summit in Lisbon. December 1.
OSCE. (1998). OSCE Mission to Georgia. Position of OSCE Mission on the Issues Related
to Georgia-South Ossetia Conflict Resolution, Statement of the Head of the OSCE
Mission to Georgia. June 20.
OSCE. (1999a). OSCE Mission to Georgia. Activity Report. Number 19. 1–17 November.
OSCE. (1999b). Declaration of the OSCE Summit in Istanbul. November 19.
OSCE. (2000a). OSCE Mission to Georgia. Activity Report. Number 4. 15–29 February.
OSCE. (2000b). OSCE Mission to Georgia. Activity Report. Number 13. 1–15 July.
OSCE. (2000c). OSCE Mission to Georgia. Activity Report. Number18. 15–30 September.
OSCE. (2000d). “Chisinau-Comrat Centre-Region Relations in the Fields of Fiscal-Budg-
etary Policy, Public Property and Adaptation of Legislation.” Materials of the work-
shop organized by the OSCE Mission to Moldova, Chisinau. 11–12 December 2000.
OSCE. (2001). OSCE Mission to Georgia. Brief Introduction. June.
Ortino, S. (2002). The Nomos of the Earth. Baden-Baden: Nomos.
Owen, D. (1995). Balkan Odyssey. London: Indigo.
Pajić, Z. (2001). “The Role of Institutions in Peacebuilding.” Z. Papić (ed.). International
Support Policies to SEE Countries—Lessons (Not) Learnt in Bosnia-Herzegovina.
Sarajevo: Open Society Fund Bosnia-Herzegovina.
754 Bibliography
Pakenham, F. (1935). Peace by Ordeal: The Making of the Anglo-Irish Treaty. London.
Palermo, F. (1999). Die Außenbeziehungen der italienischen Regionen in rechtsver-
gleichender Sicht. Frankfurt: Peter Lang.
Papua New Guinea Government Department of Education (PNGDoE). (1991). Education
Sector Review. National Capital District: DoE.
Patten, C. (1999). “A New Beginning: The Report of the Independent Commission on
Policing for Northern Ireland.” Belfast: Independent Commission on Policing for
Northern Ireland.
Paye, O. and E. Remacle. (1996). “UN and CSCE Policies in Transcaucasia.” B. Coppieters
(ed.). Contested Borders in the Caucasus. Brussels: VUB Press.
Peace Implementation Council. (2000). “Declaration of the Peace Implementation Coun-
cil.” Brussels, 23/24 May. www.ohr.int/docu/p20000524a.htm.
Pearson, B. (2002). “Putting Peace into Practice: Can Macedonia’s New Government Meet
the Challenge?” Washington D.C.: US Institute of Peace.
Pellet, A. (1992). “The Opinions of the Badinter Arbitration Committee: A Second Breath
for the Self-determination of Peoples.” European Journal for International Law 3(1):
178-181.
Pernthaler, P. and S. Ortino (eds.). (1997). Europaregion Tirol/Euregio Tirolo. Rechtliche
Voraussetzungen und Schranken der Institutionalisierung/Le basi giuridiche ed i lim-
iti della sua istituzionalizzazione. Trento: Autonome Region Trentino-Südtirol/Re-
gione Autonoma Trentino-Alto Adige.
Perry, V. (2002). “Constitutional Reform and the ‘Spirit’ of Bosnia and Herzegovina.” ECMI
Brief 7, February. Flensburg: European Centre for Minority Issues.
Perry, V. (2003). “Reading, Writing and Reconciliation: Educational Reform in Bosnia and
Herzegovina.” ECMI Working Paper 18. Flensburg: European Centre for Minority Is-
sues.
Petritsch, W. (2000). “Bosnien und Herzegowina fünf Jahre nach Dayton.” Südosteuropa
Mitteilungen 40(4): 297-312.
Pobrić, N. (2000). Ustavno pravo. Mostar: Slovo.
Popov, N. (1994) “Le populisme Serbe (suite).” Les Temps Modernes 49(574): 22-84.
Popovski, V. and M. Panov (1998). Opštini vo Republika Makedonija spored novata podelba
vo 1996 godina. Skopje.
Porter, M. (1998). “Clusters and Competition: New Agendas for Companies, Governments and
Institutions.” M. Porter (ed.). On Competition. Boston: Harvard Business Review Book.
Potier, T. (2001). Conflict in Nagorno-Karabakh, Abkhazia and South Ossetia: A Legal Ap-
praisal. The Hague: Kluwer Law International.
Poulton, H. (1989). Minorities in the Balkans. London: Minority Rights.
Poulton, H. (1995 ). Who are the Macedonians? Indiana: Indiana University Press.
Powell, G.B. (1982). Contemporary Democracies. Cambridge, M.A: Harvard University
Press.
Prihodko, N. (2001). “Moldavany ili rumyny? Za desyat’ let nezavisimosti process
natsional’noi samoidentifikacii ne zavershilsya” (Moldovans or Romanians? Ten
years of Independence Have Not Completed the Process of National Self-identifica-
tion). Nezavisimaya gazeta, 24 August 2001.
Przeworski, A. (1991). Democracy and the Market. Cambridge: Cambridge University
Press.
Bibliography 755
Radio Free Europe/Radio Liberty (RFE/RL) Daily Report. (1994a). “Moldovan Concessions
to Gagauz Deemed to be Excessive.” No. 127, 7 July.
RFE/RL Daily Report. (1994b). “Demirel in Moldova.” No.105, 6 June.
RFE/RL. (1998). “Turkish President in Moldova.” RFE/RL Newsline, 26 June. http://www.
rferl.org/newsline/1998/06/260698.asp.
RFE/RL. (2001). Newsline, 24 August.
Rae, D. (1971). The Political Consequences of Electoral Laws. New Haven: Yale University
Press.Raich, S. (1995). Grenzüberschreitende und interregionale Zusammenarbeit in
einem “Europa der Regionen”. Dargestellt anhand der Fallbeispiele Grossregion Saar-
Lor-Lux, EUREGIO und “Vier Motoren für Europa”; ein Beitrag zum europäischen
Integrationsprozeß. Baden Baden: Nomos.
Ratner, S.R. (1996) “Drawing a Better Line: Uti Possidetis and the Borders of New States.”
American Journal of International Law 90(4): 590–624.
Ratti, R. (1991). Regioni di frontiera. Teoria dello sviluppo e saggi politico economici. Luga-
no: CCM.
Raz, J. (1977). “The Rule of Law and its Virtue.” Law Quarterly Review 93: 195–202.
Rees, E. (2002). “Security-sector Reform and Transitional Administrations.” Journal of
Conflict, Security and Development 2(1): 151-156.
Regan, A. (1997). “ Preparation for War and Progress Towards Peace - Bougainville Dimen-
sions of the Sandline Affair.” S. Dinnen, R. May, and A.J. Regan (eds.). Challenging
the State: The Sandline Affair in Papua New Guinea. Canberra: Research School of
Pacific and Asian Studies, The Australian National University.
Regan, A. (1998). “Current Developments in the Pacific: Causes and Course of the Bou-
gainville Conflict.” The Journal of Pacific History 33(3): 269-285.
Regan, A. (2001a). “Why a Neutral Peace Monitoring Force? The Bougainville Conflict and
the Peace Process.” M. Wehner and D. Denoon (eds.). Without a Gun: Australia’s Ex-
perience of Monitoring Peace in Bougainville, 1997-2001. Canberra: Pandanus Books.
Regan, A. (2001b). “Establishing the Truce Monitoring Group and the Peace Monitoring
Group.” M. Wehner and D. Denoon. (eds.). Without a Gun: Australia’s Experience of
Monitoring Peace in Bougainville, 1997-2001. Canberra: Pandanus Books.
Regan, A. (2002a). “The Bougainville Political Settlement and the Prospects for Sustain-
able Peace.” Pacific Economic Bulletin 17(1): 114-129.
Regan, A. (2002b). “Resolving Two Dimensions of Conflict: The Dynamics of Consent,
Consensus and Compromise.” A. Carl. and L. Garasu (eds.). Weaving Consensus: The
Papua New Guinea – Bougainville Peace Process. London: Conciliation Resources.
Regan, A. (2002c). “Bougainville: Beyond Survival.” Cultural Survival Quarterly 26(3): 20-24.
Regan, A. (2002d). “Phases of the Negotiation Process.” Weaving Consensus: The Papua
New Guinea-Bougainville Peace Process. Accord 12(2).
Regan, A. (2003). “The Bougainville Conflict: Political and Economic Agendas.” K. Ballen-
tine and J. Sherman (eds.). The Political Economy of Armed Conflict: Beyond Greed &
Grievance. Boulder: Lynne Rienner Publishers.
Regan, A. and R. May. (2001). “Reassessing Australia’s Role in Papua New Guinea and the
Island Pacific.” J. Cotton and J. Ravenhill (eds.). The National Interest in a Global Era
– Australia in World Affairs 1996-2000. Melbourne: Oxford University Press.
Reilly, B. (1997). “Preferential Voting and Political Engineering: A Comparative Study.”
Journal of Commonwealth and Comparative Politics 35: 1-17.
756 Bibliography
Reilly, B. (2001). Democracy in Divided Societies: Electoral Engineering for Conflict Man-
agement. Cambridge: Cambridge University Press.
Reilly, B. (2002). “Papua New Guinea.” D. Nohlen, F. Grotz, and C. Hartmann (eds.). Elec-
tions in Asia II. Oxford: Oxford University Press.
Reilly, B. and A. Reynolds. (1999). Electoral Systems and Conflict in a Divided Society.
Washington: National Academy Press.
Renan, E. (1882) Qu’est-ce qu’une nation? M. Snyder (trans.). Paris: Calmann-Levy.
Report of the Secretary-General (2000). “On the United Nations Mission in Bosnia and
Herzegovina (UNMIBH).” 7 June.
Republic of the Philippines. (1986). The Constitution of the Republic of the Philippines.
Manila: Government of the Republic of the Philippines.
Reuter, J. (1992). “Die politische Entwicklung in Bosnien-Herzegowina.” Südosteuropa
41(11-12): 672-673.
Review Body on Post-Primary Education (2001). “Right to Education: Report of the Re-
view Body on Post-Primary Education (The Burns Report).”
Reynolds, A. (1999). Electoral Systems and Democratization in Southern Africa. Oxford:
Reynolds, A. (1999). “Women in the Legislatures and Executives of the World.” World Poli-
tics 51(4): 547–572.
Reynolds, A.(ed.). (2002). The Architecture of Democracy. Oxford: Oxford University
Press.
Ricq, C. (1996). Handbook on Transfrontier Cooperation of Local and Regional Authorities
in Europe. Strasbourg: Council of Europe Press.
Rimoldi, M. and E. Rimoldi. (1992). Hahalis and the Labour of Love: A Social Movement on
Buka Island. Oxford and Providence: Berg Publishers.
Roberts A. and B. Kingsbury (eds.). (1983). United Nations, Divided World: The UN’s Roles
in International Relations 2nd ed. Oxford: Clarendon Press.
Robinson, M. (2003). “From Rhetoric to Reality: Making Human Rights Work.” European
Human Rights Law Review 2003(1): 1–9.
Rodil, B.R. (2001). Kalinaw Mindanao: The Story of the GRP-MNLF Peace Process, 1975-
1996. Davao City: Alternative Forum for Research in Mindanao.
Rodrik, D. (1999). The New Global Economy and Developing Countries: Making Openness
Work. Washington: Overseas Development Council.
Rooney, K. (1998). “Institutionalizing Division”. Fortnight. June 1998
Rothchild, D. (2003). “Liberalism, Democracy and Conflict Management: The African Ex-
perience.” A. Wimmer et al. (eds.). Facing Ethnic Conflict: Toward a New Realism.
Boulder: Rowman and Littlefield.
Rousseau, J.-J. (1968). The Social Contract. London: Penguin.
Ruane, J. and J. Todd.(1996). The Dynamics of Conflict in Northern Ireland. Cambridge:
Rüb, M. (2001). “Gibt es eine neue, eine großalbanische Rebellengruppe auf dem Balkan?”
Frankfurter Allgemeine Zeitung, 13 August.
Rummel, R.J. (1994). Death by Government. New Jersey: Transaction Press
Rusi, I. (2001). “Comment: Last Chance for Peace”. IWPR Balkan Crisis Report, 271 (Part
I), 14 August. http://www.iwpr.net (12 October 2002).
Bibliography 757
Russell. P.H. (2004). Constitutional Odyssey: Can Canadians Become a Sovereign People?
Toronto: University of Toronto Press.
Sammut, D. and N. Cvetkovski. (1996). Confidence-Building Matters: The Georgia-South
Ossetia Conflict. Verification Technology Information Center (VERTIC).
Santos, S.M. (2001). The Moro Islamic Challenge: Constitutional Rethinking of the Mind-
anao Peace Process. Quezon City: University of the Philippines Press.
Sartori, G. (1968). “The Sociology of Parties: A Critical Review”. O. Stammer (ed.). Party
Systems, Party Organizations and the Politics of the New Masses. Berlin: Free Uni-
versity of Berlin.
Sartori, G. (1994, 2nd ed. 1996). Comparative Constitutional Engineering: An Inquiry into
Structures, Incentives and Outcomes. New York: New York University Press.
Sasajkovski, S. (1999). “Ideološkata profilacija na izbornite programi na relevantnite
politički partii”. Parlamentarni izbori ‘98 vo Makedonija, 104-123. Skopje.
Saskova, E. and J.M. Janeva (2003). “Ohrid Breakthrough.” IWPR Balkan Crisis Report 396,
9 January. http://www.iwpr.net (9 January 2003).
Sbornik zakonodatel’nyh aktov Gagauzii (Collection of Legal Acts of Gagauzia). (2001).
No.3, April 1999—January 2001, Comrat, 2001.
Schneckener, U. (2002a). Auswege aus dem Bürgerkrieg. Modelle zur Regulierung ethno-
nationalistischer Konflikte. Frankfurt: Suhrkamp.
Schneckener, U. (2002b). “Making Power-Sharing Work: Lessons from Successes and Fail-
ures in Ethnic Conflict Regulation.” Journal of Peace Research 39(2): 203-228.
Schneckener, U. (2003). “Theory and Practice of European Crisis Management: Test Case
Macedonia.” European Yearbook of Minority Issues 1. The Hague: Kluwer Law Inter-
national.
Schneider, H. (1996). “Friede für Bosnien-Herzegowina? Das Vertragswerk von Dayton als
Herausforderung für Europa.” Integration 19(1): 1-13.
Schöpflin, G. (2001). “Minorities and Democracy.” A.M. Biro and P. Kovacs (eds.). Diver-
sity in Action. LGI/OSI.
Schwarm, P. (1994). “Leaving the Womb.” Balkan War Report. 3 September.
Scott, J. (1997). “Dutch-German Euroregions: A Model for Transboundary Cooperation?”
P. Ganster, A. Sweedler, and J. Scott (eds.). Borders and Border Regions in Europe and
North America. San Diego: State University Press.
Sell, L. (2002). “The Serb Flight from Sarajevo: Dayton’s First Failure.” East European Poli-
tics and Societies 14(1): 179-202.
Seton Watson, H. (1960). Neither War nor Peace: The Struggle for Power in the Post-war
World. London: Methuen.
Shkolnikov, V. (1999). “Conflicts in the Caucasus, Human Rights, and the Role of OSCE.” T.
Trier and L.F. Hansen (eds.). Conflict and the Forced Displacement in the Caucasus-
Perspectives, Challenges and Responses. International Conference on the Caucasus
Region, Danish Refugee Council, January.
Short, C. (1999). “Security Sector Reform and the Elimination of Poverty.” London, De-
partment for International Development, Speeches. 9 March 1999.
Shoup, P. (1997). “The Elections in Bosnia and Herzegovina: The End of an Illusion.” Prob-
lems of Post-Communism 44(1): 3-15.
Shugart, M.S. and J.M. Carey. (1992). Presidents and Assemblies: Constitutional Design
and Electoral Dynamics. New York: Cambridge University Press.
758 Bibliography
Shugart, M.S. and M.P. Wattenberg (eds.). (2001). Mixed-member Electoral Systems: The
Best of Both Worlds? Oxford: Oxford University Press.
Shvetsova, O. (1999). “A Survey of Post-Communist Electoral Institutions, 1990-1998.”
Electoral Studies 18: 397-409.
Sieghart, P. (1986). The Lawful Rights of Mankind. Oxford: Oxford University Press.
Simić, D. (2001). “The Algerian Case: Between Politics and Principles.” AIM. 25 January.
Sisk, T.D. (1996). Power Sharing and International Mediation in Ethnic Conflicts. Wash-
ington D.C.: US Institute of Peace.
Smith, A.D. (1985). “Ethnie and Nation in the Modern World.” Millenium 14: 127-142.
Smith, C. (2001). “Security-sector Reform: Development Breakthrough or Institutional
Engineering?” Journal of Conflict, Security and Development 1(1): 5-19.
Smith, S. (2002). “The Role of the United Nations Observer Missions. Weaving Consen-
sus: The Papua New Guinea-Bougainville Peace Process. Accord 12(2). http://www.
c-r.org/accord/accord12/index.shtml.
Snyder, J. (2000). From Voting to Violence: Democratization and Nationalist Conflict. New
York: Norton.
Sokolović, D. (2001). “Social Reconstruction and Moral Restoration.” D. Sokolović and F.
Bieber (eds.). Reconstructing Multiethnic Societies: The Case of Bosnia Herzegovina.
Aldershot: Ashgate.
Spriggs, M. and D. Denoon. (eds.). (1992). The Bougainville Crisis: 1991 Update. Canberra:
Research School of Pacific Studies, Australian National University in association
with Crawford House Press.
Starmer, K. (2003). “Two Years of the Human Rights Act.” European Human Rights Law
Review 1: 14–23.
Stedman, S.J. and D. Rothchild. (1996). “Peace Operations: From Short-term to Long-term
Commitment.” International Peacekeeping 3(2): 17-35.
Stedman, S.J. (1997). “Spoiler Problems in Peace Processes.” International Security 22(2):
5-23.
Stedman, S.J. (2001). “Implementing Peace Agreements in Civil Wars: Lessons and Rec-
ommendations for Policymakers.” IPA Policy Paper Series on Peace Implementation.
May 2001. New York.
Stedman, S.J., D. Rothchild, and E. Cousens (eds.). (2002). Ending Civil Wars: The Imple-
mentation of Peace Agreements. Boulder: Lynne Rienner.
Stevens, J. (2003). Stevens Enquiry 3: Overview and Recommendations. Metropolitan Po-
lice: London.
Strassoldo, R. and G. Delli Zotti. (1982). Cooperation and Conflict in Border Areas. Milano:
Angeli.
Suberu, R. and L. Diamond. (2002). Institutional Design, Ethnic Conflict Management
and Democracy in Nigeria. Andrew Reynolds (ed.). The Architecture of Democracy.
Sutton, M. (1994). An Index of Deaths from the Conflict in Northern Ireland, 1969-93. Bel-
fast: Beyond the Pale Publications.
Taagepera, R. (1999). “The Number of Parties as a Function of Heterogeneity and Electoral
System.” Comparative Political Studies 32: 531-548.
Taagepera, R. and M. Shugart. (1989). Seats and Votes: The Effects and Determinants of
Electoral Systems. New Haven: Yale University Press.
Bibliography 759
Tan, S.K. (1993). Internationalization of the Bangsamoro Struggle. Quezon City: Center for
Integrative and Development Studies, University of the Philippines.
Tan, S.K. (2000). “Three Wars and the President.” Kasarinlan 15(2): 221-226.
Tannam, E. (1999). Cross-border Cooperation in the Republic of Ireland and Northern Ire-
land. Basingstoke: Macmillan.
Tapales, P.D. (1996). “The Philippines.” McCarney, P.L. (ed.). The Changing Nature of Local
Government in Developing Countries. Toronto: Centre for Urban and Community
Studies, University of Toronto.
Taylor, R. (1994). “A Consociational Path to Peace in Northern Ireland and South Africa?”
A. Guelke (ed.). New Perspectives on the Northern Ireland Conflict. Aldershot: Ave-
bury.
Taylor, R. (2001), “Northern Ireland: Consociation or Social Transformation”. J. McGarry
(ed.). Northern Ireland and the Divided World. Oxford: Oxford University Press.
Teitel, R.G. (2000). Transitional Justice. Oxford: Oxford University Press.
Terrell, J. (1986). Pre-history in the Pacific Islands. Cambridge: Cambridge University
Press.
Thompson, H. (1991). “The Economic Causes and Consequences of the Bougainville Cri-
sis.” Resource Policy 17(1): 69-85.
Thompson, M. (1994). Forging War. The Media in Serbia, Croatia and Bosnia-Hercegovina.
Avon: Article 19.
Thompson, P. (1997). “The Gagauz of Moldova: The Road to Autonomy.” M. Opalski (ed.).
Managing Diversity in Plural Societies: Ethnic Relations and Ethnopolitics in Post-
Communist Eastern Europe. Ottawa: Forum Eastern Europe..
Thornberry, P. (1991). International Law and the Rights of Minorities. Oxford: Oxford Uni-
versity Press.
Tkachuk, T. (2001). “Tak est’ li u kommunistov plan?” (Do Communists Have a Plan?)
Moldavskie vedomosti. 26 September.
Toniatti, R. (1997). “La bozza di statuto della Regione europea.” P. Pernthaler and S. Ortino
(eds.). Europaregion Tirol/Euregio Tirolo. Trento: Autonome Region Trentino-Südti-
rol/Regione Autonoma Trentino-Alto Adige.
Trenin, D. (1996). “Russia’s Security Interests and Policies in the Caucasus Region.”
B.Coppieters (ed.). Contested Borders in the Caucasus. Brussels: VUB Press.
Trier, T. and L.F. Hansen (eds.). (1999). Conflict and Forced Displacement in the Caucasus.
International Conference on the Caucasus Region, Danish Refugee Council, Janu-
ary.
Trimble, D. (1998). Speech by the First Minister to the Chamber of Commerce of Ireland.
Dublin. 19 November.
Trimble, D. and S. Mallon. (1998). Statement Issued by the First Minister and Deputy First
Minister Towards Implementing the Belfast Agreement. Irish Times, 19 December.
Troebst, S. (1994). “Visit to Moldova of the CSCE High Commissioner on National Mi-
norities, 7 to 10 December 1994.” Manuscript.
Troebst, S. (1997). “An Ethnic War That Did Not Take Place: Macedonia, Its Minorities and
Its Neighbours in the 1990s.” D. Turton (ed.). War and Ethnicity: Global Connections
and Local Violence. Rochester: University of Rochester Press.
Troebst, S. (2000).“‘Großalbanien,’ ein Trugbild.” Frankfurter Allgemeine Zeitung, 25 Oc-
tober 2000.
760 Bibliography
Turkish Press Review. 26 June 1998. http://www.hri.org/news/turkey/trkpr/1998/98-06-

26.trkpr.html#06.
Turner, M. (1990). Papua New Guinea: The Challenge of Independence. Ringwood: Pen-
guin.
Turner, M. (1995a). “Terrorism and Secession in the Southern Philippines: The Rise of the
Abu Sayaff.” Contemporary Southeast Asia 17(1): 1-19.
Turner, M. (1995b). “Subregional Economic Zones, Politics and Development: The Philip-
pine Involvement in the East ASEAN Growth Area (EAGA).” The Pacific Review 8(4):
637–648.
Turner, M. (1999). “Central-Local Relations: Themes and Issues.” M. Turner (ed.). Central-
Local Relations in Asia-Pacific: Convergence or Divergence. Houndmills and New
York: Macmillan and St Martin’s.
Turner, M. (2001). Choosing Items from the Menu: New Public Management in Southeast
Asia. Paper presented at the Fifth International Research Symposium in Public Man-
agement. Barcelona. 9–11 April.
Turner, M., R.J. May, and L.R. Turner (eds.). (1992). Mindanao: Land of Unfulfilled Prom-
ise. Quezon City: New Day Publishers.
Tushnet, M. (1983). “Following the Rules Laid Down: A Critique of Interpretivism and
Neutral Principles.” Harvard Law Review 96: 781–827.
Tushnet, M. (1988). Red, White And Blue: A Critical Account Of American Constitutional
Law. Boston: Harvard University Press.
Twining, W. (2000). Generalizing About Law: The Case Of Legal Transplants. The Tilburg-
Warwick Lectures. http://www.ucl.ac.uk/laws/jurisprudence/docs/twi_til_1.pdf (24
November 2003). http://www.ucl.ac.uk/laws/jurisprudence/docs/twi_til_4.pdf.
United Nations (UN). (1994). “United Nations Framework for Strengthening of the Rule
Of Law.” Annex 11-1 Report of the United Nations Secretary General to the Forty-ninth
Session of the United Nations General Assembly. (United Nations General Assembly
Document A/49/512).
UN. (1996). Report of the Expert on the Impact of Armed Conflict on Children (The Machel
Report). New York: United Nations.
UN Commission on Crime Prevention and Criminal Justice (CCPCJ). (2001). “10th Ses-
sion of the Commission on Crime Prevention”. 8-17 May 2001.
UNCCPCJ. (2002). “Reform of the Criminal Justice System: Achieving Effectiveness and
Equity”. 26 February.
UN Development Programme (UNDP). (2001a). “National Human Development Report.
Republic of Moldova 2000.”
UNDP. (2001b). Regional Development Programme “Gagauz-Yeri.” Chisinau: UNDP.
UNDP. (2001c). “Early Warning Report, Quarterly Report October-December”.
UNDP. (2002a). Human Development Report. New York: UNDP.
UNDP. (2002b). Human Development Report – Kosovo. Prishtinë: UNDP.
UNDP. (2002c). Human Development Report – Bosnia and Herzegovina. Sarajevo: UNDP.
UNDP. (2003). “Early Warning Report, Quarterly Report April-June.”
UN Economic and Social Council (ECOSOC). (2001). Specific Groups and Individuals.
Commission on Human Rights, 57th session. 25 January.
UN Educational, Scientific and Cultural Organization (UNESCO). (2001). Education for All
Thematic Study: Education in Situations of Emergency and Crisis. Paris: UNESCO.
Bibliography 761
UN High Commissioner for Refugees (UNHCR)). (2000a). “Minority Returns to the Re-
publika Srpska-Bosnia-Herzegovina”. Sarajevo, June.
UNHCR. (2000b). “UNHCR Mid-Year Report 2000.”
UNHCR. (2005a). “Returns Summary to Bosnia and Herzegovina From 01/01/96 to
31/7/2005.”
UNHCR. (2005b). “Total Minority Returns in/to BiH from 1996 to 31 July 2005.”
UN International Children’s Emergency Fund (UNICEF). (1999). Education in Emergen-
cies and for Reconstruction. New York: UNICEF.
UN Mission in Bosnia and Herzegovina (UNMIBH). (2002a). “About us.” http://www.un-
mibh.org/unmibh/abutus.asp (11 July 2003).
UNMIBH. (2002b). “Building a Better Future Together, December 1995 to June 2002.”
UNMIBH. (2002c). “UN Radio Headlines (The last programme) for Friday, 13 December
2002.” http://www.unmibh.org/radio/view.asp?Radioid=235 (11 July 2003).
UNMIBH. (2002d). “International Police Task Force (IPTF).” http://www.unmibh.org/un-
mibh/iptf/ (11 July 2003).
UN Mission in Kosovo (UNMIK). (2001). “Constitutional Framework for Provisional Self-
Government.” Regulation No. 2001/9. Kosovo: UNMIK.
UNMIK. (2002). “Police in Kosovo 2001.”
UNMIK. (2003). “Despite Progress, Kosovo Still a Considerable Way from UN Targets -
Annan”. http://www.unmikonline.org/archives/news02_03.htm (11 July 2003).
UNOCHA (United Nations Office for Coordination of Humanitarian Affairs). (2001).
South Ossetia. Briefing Note, March.
United States (US) Institute of Peace. (2002). Kosovo Final Status - Options and Cross-bor-
der Requirements. Special Report No. 91. http://www.usip.org/pubs/specialreports/
sr91.html (8 February 2004).
US State Department. (1993). Country Reports on Human rights Practices for 1993. Wash-
ington: Bureau of Democracy, Human Rights, and Labor.
US State Department. (2000). 1999 Country Reports on Human Rights Practices, Former
Yugoslav Republic of Macedonia. http://www.state.gov/www/global/human_rights/
1999_hrp_report/macedoni.html) (12 October 2002).
Vankovska, B. (2002). Part I: The Path from ‘Oasis of Peace’ to ‘Powder Keg’ of the Balkans.
Berlin: Henrich Böll Foundation.
Varela, A.P. (1996). Administrative Culture and Political Change. Quezon City: College of
Public Administration, University of the Philippines.
Venice Commission (European Commission for Democracy through Law). (2005). “Opin-
ion on the Constitutional Situation in Bosnia and Herzegovina and the Powers of the
High Representative.” CDL-AD (2005) 004, 11 March.
Vincent, J. (1986). Human Rights and International Relations. Cambridge: Cambridge Uni-
versity Press.
Vitug, M.D. and G.M. Gloria. (2000). Under the Crescent Moon: Rebellion in Mindanao.
Quezon City: Ateneo Center for Social Policy and Public Affairs and Institute for
Popular Democracy.
Wadi, J.P. (1993). Islamic Diplomacy: A Case Study of the OIC and the Pacific Settlement
of the Bangsamoro Question (1972-1992). Thesis presented for the degree of M.A. Is-
lamic Studies, Institute of Islamic Studies, University of the Philippines.
762 Bibliography
Waever, O. (1995). “Securitization and Desecuritization.” R. Lipschutz (ed.). On Security.

New York: Columbia University Press.
Wainwright, E. (2003). Our Failing Neighbour: Australia and the Future of the Solomon
Islands. Canberra: Australian Strategic Policy Institute.
Walker, T. (2001). “NATO Troops Caught in a Balkan Ulster.” The Sunday Times, 18
March.
Walter, B. (1997). “The Critical Barrier to Civil War Settlement.” International Organiza-
tion 51(3): 340-363.
Warren, J.F. (1985). The Sulu Zone 1768-1898. Quezon City: New Day Publishers.
Weber, M. (1978). Economy and Society: An Outline of Interpretive Sociology. G. Roth and
C. Wittich (eds.). Berkeley: University of California Press.
Wehner, M. and D. Denoon (eds.). (2001). Without a Gun: Australia’s Experience of Moni-
toring Peace in Bougainville, 1997-2001. Canberra: Pandanus Books.
Weller, M., F. Bieber, and E.M. Christiansen. (2001). Power-sharing in Bosnia and Herze-
govina. Flensburg: European Centre for Minority Issues.
Wesley-Smith, T. (1992). “Development and Crisis in Bougainville: A Bibliographic Essay.”
The Contemporary Pacific 4(2): 407-432.
Wetlaufer, G.B. (1999). “Systems Of Belief In Modern American Law: A View From Centu-
ry’s End.” American University Law Review 49(1): 1–80.
Whyte, J. (1993). “Dynamics of Social and Political Change in Northern Ireland”. D. Keogh
and M. Haltzel (eds.). Northern Ireland and the Politics of Reconciliation. Cambridge:
Wieland, C. (2000). Nationalstaat wider Willen. Politisierung von Ethnien und Ethnis-
ierung der Politik: Bosnien, Indien, Pakistan. Frankfurt: Campus.
Wilford, R. (1992). “Inverting Consociationalism? Policy, Pluralism and the Post-modern”.
Hadfield, B. (ed.). Northern Ireland: Politics and the Constitution. Buckhingham:
Open University Press.
Wilford, R. (2001). “The Assembly and the Executive.” R. Wilford (ed.). Aspects of the Bel-
fast Agreement. Oxford: Oxford University Press.
Wilford, R. and R. Wilson. (2001). “A ‘Bare Knuckle Ride’: Northern Ireland.” R. Hazell
(ed.). The State and the Nations: The First Year of Devolution in the United Kingdom.
Thorverton: Imprint.
Wilford, R. and R. Wilson. (2003). A Route to Stability: The Review of the Belfast Agree-
ment. Belfast: Democratic Dialogue.
Wilson, C. (2003). “Rejection of the Belfast Agreement is Entirely Compatible with the
Unionist Commitment to ‘Equal Citizenship’”. Belfast Telegraph, 28 October.
Wimmer, A. (2003). “Democracy and Ethno-religious Conflict in Iraq”. Survival 45(4): 111-
134.
Woelk, J. (2004). “Federalism and Consociationalism as Tools for State Reconstruction:
Experiences From Bosnia and Herzegovina.” J. Marko, A. Tarr, and R. Williams (eds.).
Minority Rights and Sub-national Constitutional Law. Westport: Praeger.
Wolfers, E. (2002). “‘Joint Creation’: The Bougainville Peace Agreement – and Beyond.” A.
Carl and L. Garasu (eds.). Weaving Consensus: The Papua New Guinea – Bougain-
ville Peace Process. London: Conciliation Resources.
Wolff, S. (2000). “Managing Disputed Territories, External Minorities, and the Stability
of Conflict Settlements.” Ph.D. dissertation. London: London School of Economics,
Government Department.
Bibliography 763
Wolff, S. (2001). “Context and Content: Sunningdale and Belfast Compared. ” R. Wilford
(ed.). Aspects of the Belfast Agreement. Oxford: Oxford University Press.
Wolff, S. (2003). Disputed Territories: The Transnational Dynamics of Ethnic Conflict Set-
tlement. London: Berghahn Books.
Wood, N. (2001). “Macedonian Rebels Retreat to Kosovo.” The Observer, 2 September.
Woodward, S. (2001) “On War and Peace-building: Unfinished Legacy of the 1990s.” http://
ww.ssrc.org/sept11/essays (28 Dec. 2003).
Woodward, W. (2002). “Eleven-minus for Northern Ireland”. The Guardian, 25 June.
World Bank (1992). World Development Report 1992. New York: Oxford University Press.
World Bank (1995). Priorities and Strategies for Education. Washington: World Bank.
World Bank (2003). World Development Report 2003. Washington and New York: World
Bank and Oxford University Press.
World Education Forum (WEF). (2000a). “The EFA 2000 Assessment: Country Reports:
Moldova.” http://www2.unesco.org/wef/countryreports/moldova/rapport_1.html (9
March 2003).
WEF. (2000b). “The EFA 2000 Assessment: Country Reports: Georgia.” http://www2.
unesco.org/wef/countryreports/georgia/rapport_1.html (9 March 2003).
Wright, J. (1996). “The Geopolitics of Georgia.” J.Wright, S. Goldenberg, and R. Schofield
(eds.). Transcaucasian Boundaries. London: UCL Press Limited.
Zahir, S. A. (1998). “The Moro Jihad: Continuous Struggle for Islamic Independence in
Southern Philippines.” Nida’ul Islam 23, April-May. http://www.fas.org/irp/world/
para/docs/ph1.htm (29 August 2001).
Zartman, W.I. (1985). Ripe for Resolution: Conflict and Intervention in Africa. Oxford: Ox-
ford University Press.
Zverev, A. (1996). “Ethnic Conflicts in the Caucasus 1988-1994.” B. Coppieters (ed.). Con-
tested Borders in the Caucasus. Brussels: VUB Press.
Index
A America/n, see also United States of

America 6, 14, 15, 47, 50, 63, 68, 70, 93,
Aaland Islands xiii, 316
114, 161, 162-163, 166, 189, 191, 219, 468,
Abkhaz 19, 349 n. 7, 350, 351, 352, 389 495, 619, 621, 636, 691, 692, 694, 717
Abkhazia 41, 347, 348, 349, 350-353, 354, Latin 57
355, 356-357, 358, 359-361, 363, 366, 374,
South 9
377-378, 379, 389-391, 392, 428, 447, 458-
459, 505, 506, 515-516, 525, 526, 547-548, Spanish 27
612, 628, 638 n. 1, 650 n. 2, 653, 654, 658, Argentina 9, 568, 611
660, 661, 334, 678 American Convention on Human
accountability 12, 14, 48, 84, 86, 92, 93, Rights 631
95-97, 105, 116, 189, 190, 419, 442, 443, ASEAN (Association of Southeast
449, 452, 491-492, 577, 578, 610, 611, 613, Asian Nations) 166, 168, 181, 470, 534,
615-616, 617, 620, 622, 624, 625, 643, 707, 550
708 Asia(n) 8, 10, 168
Aceh xiii, 388, 470 Asian Development Bank 174, 473
affirmative action, see also quotas 53 n. assimilate 5, 57, 67, 604
3, 54, 92, 94, 537, 549-550
assimilation(ist) 10, 15, 47, 48, 58, 67, 97,
Africa(n) xi, xiii, 9, 10, 11, 12, 27-28, 577, 109, 126, 156, 351, 643, 652, 691, 717
614, 616, 619, 702
association (with a/other state/s) xvi,
American 638 21, 28, 29, 33, 49, 55, 136, 237, 391, 394, 445
African National Congress (ANC) 472 partial 27, 49
African Union (formerly Australia/n 34, 127, 129, 133-135, 144-145,
Organization of African States) 10, 27, 149-150, 413, 472-473, 476, 477, 479, 481,
470 482, 496-497, 534, 564, 569, 620, 693-694
Albania(n), see also Albanian popula- Austria/n 50, 72 n. 7, 335 n. 22, 373, 568,
tion in Bosnia and Herzegovina, Alba- 667, 670, 674, 676, 681, 691, 692, 693, 696
nian population in Kosovo, Albanian authoritarian 9, 202
population in Macedonia 266 n. 4,
autonomy ix, x, xiii, xiv-xv, xvi, xvii, 5,
270, 393, 423, 585, 676, 678, 679 n. 19
34, 43, 45, 49, 51, 52, 54-55, 56, 64, 72, 106,
Greater 266, 268, 301, 518, 588, 589 116, 117, 125, 128, 136, 141, 142, 144, 145,
diaspora 467-468 146, 147, 155, 156, 157, 159, 161, 164, 167,
apartheid 6, 24, 468 168, 180, 203, 205, 209, 214, 221, 239, 244,
247, 256, 257, 260, 263, 307, 308, 310, 311,
766 Index
312, 313, 314, 315-316, 317, 318-319, 320, 391, 393, 408, 423, 440 n. 3, 441, 448,
323, 327, 329, 331, 334, 335-336, 341, 343, 633, 704
349, 350, 391, 392, 396, 399, 403, 412, 413, autonomous, see also autonomous
418-419, 423, 427, 428, 432, 434, 440 n. government xiii, xiv, xvi, 8, 23, 32, 33
3, 441, 444, 446, 448, 449, 464, 481, 489, n.7, 36, 43, 49, 55, 56, 101, 136, 137, 138,
520, 538, 557, 563, 568, 581, 583, 587, 602, 139, 142-143,144, 146, 148, 149, 150, 151,
629, 630, 633, 634-635, 638, 642, 651, 664, 152, 154, 165, 168, 169, 170, 173, 174, 175,
665, 668, 670, 675, 683, 697, 698, 699, 700 177, 178, 182, 183, 195, 198, 208, 214, 244,
n. 24, 701, 702, 706, 710, 714 n. 37 256, 263, 273 n. 18, 308, 311, 313, 316, 317,
agreement/arrangements/settlements x, 318, 319, 320, 322, 323, 325, 326, 328, 329,
xiii, xiv, xvi-xvii, 33 n. 7, 41, 125, 136, 331, 332, 333, 334, 335, 336, 338, 339, 341,
137-138, 140, 143, 144, 145, 147, 150, 151, 342, 343, 349, 351, 360, 391, 392, 402, 413,
152, 155, 157, 158, 159, 245, 313, 316, 317, 414, 415, 416, 417, 418, 421, 422, 423, 426,
319-320, 326, 341, 388, 391, 405, 417, 423, 434, 436, 437, 440, 441, 442, 443, 444,
440 n. 3, 633, 642, 664, 665, 683, 699, 446, 447, 448, 473, 484, 505, 506, 508, 513,
703, 706, 722 521, 526, 542, 548, 551, 563, 568, 581, 582,
asymmetrical xiii, 125, 388, 697, 700 583, 584, 602, 608, 610, 633, 635, 640, 642,
communal 97-98 644, 645, 652, 665, 682, 683, 684 n. 27,
685, 697, 703
constitutional(ly entrenched) 56, 101-102
corporate legal 55
cultural 21, 203, 209, 407 B
de facto 128 Badinter Arbitration Commission 28,
dominion 64 29, 35, 36, 38, 39-40, 197, 198-199, 278, 388,
economic 328, 538 389, 395 n. 20
educational 505 Balkans 19, 355, 455, 475, 479, 480, 495,
496, 540, 541, 544-546, 584, 595, 610, 619,
ethnic 196, 232, 238, 239
686, 726
federal 239
Baltic 8, 546
fiscal 171, 173, 509, 511
Republics 33, 38, 395, 396
group 56, 287
Bangladesh 9, 12, 30, 40, 166, 568, 642
institutional 673
Banthustans 24
law-making 101
Basque(s) 29, 112, 316, 696
local/municipal 188, 319, 434, 679
Belgium 50, 53, 54, 55, 57, 203, 468, 674,
non-territorial 55, 209 681, 691, 692, 693, 696
operational 144 Biafra 12, 18, 24, 30
personal 55, 423, 440, 441, 633 blocking mechanism/powers/rights, see
policy 539, 549, 552, 553 also veto xiv, 50, 51, 54, 55-56, 262, 263,
political 21, 551, 583 400, 401, 487
regional 539, 551 border(s), see also frontier 6, 10, 43, 61,
religious 209 n. 14 64, 110, 134, 143, 199, 200, 206, 240, 266,
right of 645 268, 295, 304, 319, 321, 332, 339, 365, 369,
370, 371, 414, 415, 423, 429, 467, 484, 547,
segmental 203, 205, 285, 287-288, 412
566, 579, 586, 587, 541, 654, 667, 668-672,
symmetrical 125 673-674, 676, 678, 687, 688
territorial 21, 29, 55, 125, 126, 137, 138, colonial 12
155, 156, 157, 158, 195, 196, 208, 283, 285,
287, 309, 311, 313, 316-317, 318, 335-336,
Index 767
cross-/trans- 82, 93, 101, 105, 106, 107, 660, 694, 695 n. 96, 697, 698, 702, 704,
111, 121, 142-143, 152, 225, 236-237, 268, 710, 711
274, 326, 332, 352, 362, 369, 377, 408, Muslim population in BiH 195, 196, 198,
467, 468, 479, 531, 537, 554, 555, 556, 667, 199, 200, 209 n. 14, 409, 486, 566, 717
668-669, 670-672, 673, 674, 675, 676, n. 43
678, 679, 680, 681, 682, 683, 684, 687, Serb population in 193, 194-202, 205,
698-699, 700 207, 208, 209, 210, 212-214, 215, 220,
external 218 222, 226, 227, 228, 230, 232, 237, 239,
control/guard/patrol 207, 218, 229, 234, 240, 245, 409-410, 455, 458, 459, 483,
254, 256, 266, 268, 274, 369, 390, 542, 488, 566, 580, 652, 678, 681, 697, 702,
565, 567, 571, 572, 574, 580, 586, 588, 592 704, 715-716, 717
internal 10 Vance-Owen Plan 1993 199-200, 220,
international 8, 10, 175, 467, 468, 565 473, 483, 579
internationally recognized 10, 365, 396 Bosniak 200, 201, 202, 205, 206, 208,
Bosnia and Herzegovina (BiH) xiii, xiv, 209, 210, 212, 213, 214, 222, 226, 228, 233,
14, 19, 23, 31, 41, 51, 58, 167, 193-241, 244, 237, 261, 282 n. 31, 295, 300, 455, 458, 459,
245, 249, 254, 255, 263, 282, 292, 389, 393, 519, 579, 697, 704, 715-716
394, 395, 396, 408, 409-413, 424, 430, 431, Bougainville xiii, xvi, 29, 43, 125-159, 399,
432, 433, 434, 435, 437, 438, 439, 440, 443, 408, 413-416, 420, 423, 426, 430, 431, 432,
444, 445, 446, 449, 452, 454, 455, 457, 458, 433, 435-436, 437, 438, 439, 440, 443, 445,
459, 460, 461, 462, 463, 464, 465, 466, 446-447, 448, 449, 452, 458, 460, 462,
468, 470, 473-474, 475, 476, 477, 478, 479, 463, 464, 466, 468, 472-473, 473 n. 9, 476,
480, 483, 484, 486, 487, 488, 489, 491, 477, 478, 479, 481, 482, 485, 493, 494, 495,
492, 493, 494, 495, 496, 497, 499, 505, 506, 496, 497, 505, 506, 507-509, 523, 525, 526,
518-520, 522, 525, 526, 528, 529, 532, 534, 532, 534, 540, 550, 551, 563-564, 569, 572,
540, 541, 542, 543 n. 9, 544, 560, 566-568, 581-582, 596, 597, 601, 602, 607, 608, 610,
572, 575 n. 1, 577, 579-581, 586, 596, 601, 612, 628, 633, 638, 640, 642, 643, 645, 650,
603, 604, 607, 608, 609, 611, 612, 613, 621, 658, 659, 661, 665, 682-684, 693-694, 697,
623, 633, 641, 642, 643, 645, 646, 647, 648, 698-700, 701, 702, 703, 709
650, 651, 652, 653, 658, 660, 663, 678, 680- Bougainville Peace Agreement 2001 43,
681, 685, 694, 695, 697-698, 699, 702, 704, 125, 132, 135-150, 399, 414, 416, 423, 440,
710, 711, 712, 713, 716, 717, 718 485, 508, 563, 582, 602, 608, 610, 614 n.
Albanian population in xiv 49, 633, 638, 640, 642, 644, 645, 647,
Croat population in 194-202, 205, 207, 658, 659, 661, 665, 682, 699-700, 702,
208, 209, 210, 212, 213, 214, 216, 220, 221, 709
223, 226, 227, 228, 232, 233, 235, 237, 239, British 14, 62-67, 69-111, 114-115, 119,
240, 394, 409, 410, 411, 412, 434, 447, 121-124, 127, 375, 401, 408, 419-420, 438,
448, 449, 455, 458, 459, 473, 483, 486, 439, 446, 471-472, 477, 482, 552, 554, 561,
519, 566, 579, 580, 678, 681, 697, 704, 717 568, 591-593, 603, 611, 615, 623, 639, 645,
n. 43 651, 684-685, 694, 695, 696, 698, 702, 703,
“Dayton Agreement/Peace Accords” 708, 709 n. 31, 711, 712, 713, 714
(General Framework Agreement for Bulgaria/n 270, 309 n. 4, 310, 316, 327,
Peace in Bosnia and Herzegovina) 427, 540, 547, 568
1995 xiv, 41, 194, 198, 199, 201-238, 240, Burma/Myanmar xi, 9, 19, 29
245, 250, 254, 394, 369, 410, 412, 431, 455,
474, 486, 488, 518, 519, 542, 566-567, 579,
586, 601, 604 n. 11, 605, 607 n. 23, 613 n. C
47, 633, 640, 641, 644, 646, 647, 648, 658, Cameroon 10
768 Index
Canada xv, 9, 13, 39, 50, 53, 54, 57, 636, anti- 9, 30
650, 652, 708, 718 entity xii, 24-26
Catholic/s 52, 63, 65, 67, 68, 70, 74, 93, post- xvi, 9, 10, 50, 405, 620
95, 96, 97, 100, 284, 295, 296, 456, 463, self-determination xii, 23, 24-26, 28, 29,
465, 523-524, 543 n. 10, 560-561, 568, 591, 31, 33, 40, 42
618, 640, 645, 661, 662, 696, 705, 717 n.
colonialism xii, 12, 13, 25, 26, 29, 30, 44,
43
61, 65, 128, 163-164, 620
Caucasus 19, 347, 353 n. 14, 355, 360, 366,
communism/t 167, 194-195, 197 n. 5, 234-
369 n. 39, 429, 547, 594
235, 244, 270 n. 11, 273, 307, 312, 318, 329,
ceasefire 37, 43, 67, 70, 71, 89, 100, 111, 333, 335-336, 338, 339, 342, 532-534, 540,
112, 113, 115, 116, 118, 119, 130, 133, 182, 201, 541, 542, 546, 548, 710 n. 32, 716
215, 219, 222, 253, 256, 274, 277, 278, 279,
community/ies xiv, xvi, 13, 15, 32, 34, 43,
311, 345, 348, 350, 353, 356, 358, 362, 363,
47, 49, 50, 51, 52, 53, 54, 55-56, 58, 61, 72,
367, 409, 429, 440, 477, 484, 485, 490,
84, 92, 93, 94, 97-98, 110, 114, 119, 120, 122
580, 585, 586, 587, 588, 592, 593, 594, 701,
n. 31, 131, 149, 161, 163, 178, 180, 184, 185,
711
190, 197, 198, 202, 203, 204, 205, 208, 210,
Chechens 29, 42, 389, 397, 696 212, 222, 228-230, 233, 237, 238, 240, 251,
Chechnya xii, 8, 24, 25, 31, 36-37, 38-39, 259-260, 261, 263, 280, 284, 285, 286, 287,
42, 388 n. 1, 389, 397, 429, 653, 696 293, 295, 296, 304 n. 68, 323, 325, 349, 350,
China 8, 41, 256 352, 354, 355, 357, 377, 393, 407, 419, 421,
Christianity 72, 162 424, 434, 443, 444, 445, 454, 455, 456,
Christian/s 53, 54, 55, 97, 162, 163-164, 460, 463, 464, 467, 468, 470, 476, 480,
165, 167, 169, 171, 182, 184, 186, 191, 309, 490, 510, 511, 517, 521, 525, 527, 530, 531,
398, 416, 512, 527, 550, 590, 661, 705 539, 544, 555, 556, 559, 560, 561, 562, 564,
566, 569, 573, 582, 588, 591, 592, 595, 608,
citizens 7, 12, 14, 17, 18, 48, 51, 53, 57, 67,
615, 627, 633, 634, 635, 636, 638, 639, 640,
92, 97, 107, 113, 175, 176, 201, 208, 218, 229,
641, 642, 648, 649, 670, 674, 678, 679-680,
232, 236, 239, 250, 261, 270 n. 12, 271, 276,
682, 684, 685, 698, 699, 707, 711, 718
280, 283, 285, 289, 295, 296, 298, 299, 301,
304, 305, 310, 320, 321, 327, 335, 410, 428, autonomous 635
430, 434, 440, 503, 504, 526, 561, 562, 577, civic 634, 637, 644, 653
579, 621, 629, 633, 634, 636, 644, 645, 646, cross- 50, 51, 55-56, 72, 73, 74, 75, 79, 80,
648, 653, 660, 663, 687, 704, 705, 716 n. 83, 85, 86, 90 n. 21, 103, 106, 107, 119,
38, 717 463, 479, 545, 705
citizenship 170, 270 n. 12, 299, 320 n. 13, cultural 49, 169, 171, 175, 176, 177, 459,
417, 503, 524, 527, 528, 634, 637, 643, 646 512, 631, 641
coalition 15, 48, 51, 64, 72 n. 6, 73 n. 8, 74, ethnic xiv, 48, 49, 61, 73, 87, 208, 243,
78, 83, 84, 85, 87, 90, 91, 93, 109, 117, 130, 263, 272, 276, 283, 287, 288, 289, 290,
131, 132, 139, 171, 194, 195, 204 n. 9, 208, 300, 375, 434, 441, 455, 484, 522, 528,
220, 223, 228, 230, 262, 273, 274, 275, 294, 569, 588, 634, 635, 637, 641, 643-644,
300, 311, 655 n. 20, 433, 434, 448, 456, 459, 653, 697, 698, 705, 706, 713
470, 487, 492, 694, 707, 708, 709, 714 imagined 7
Cold War xiii, 8, 9, 12, 13, 16, 19, 41, 45, indigenous 171, 176, 177, 178, 459, 512
533, 573, 577, 630, 636, 669, 670, 693, 696, international xv, xvi, 35, 37, 39, 40, 125,
703 132, 135, 143, 144, 149, 150, 157, 159, 193,
colonial xii, xvi, 8, 10, 12, 25, 26, 27, 28, 198, 199, 201, 203, 204, 207, 214-220,
29, 30, 40, 44, 49, 62, 63, 127, 128, 161, 162, 221, 222, 223, 224, 228, 229, 233, 235, 238,
163, 164, 166, 405, 588, 615 246, 252, 266, 267, 268, 270, 276, 277,
Index 769
289, 291, 292, 293, 300-301, 303, 304, compliance xiv, 148, 172, 173, 191, 252 n.
305, 313, 317, 355 n. 19, 378 n. 55, 402, 6, 253, 254, 412, 443, 449, 484, 485, 513,
424, 429, 436, 437, 438, 439, 443, 444, 519, 563, 585, 586, 632, 646, 648, 649-652,
467, 468, 473, 474, 479, 480, 483, 484, 659, 662, 663, 703
487, 496, 508, 541, 543, 569, 579, 588, confederal 16, 37, 56, 61-124, 139, 392,
590, 662, 681, 704 431, 446, 698
linguistic 49, 639, 696 confederation/s 39, 56, 106, 108, 109, 315,
local 127, 170, 191, 362, 408, 418, 423, 389, 399, 445, 532
427, 441, 448, 562, 565, 569, 671 conflict/s, see also dispute/s, ethnic con-
minority 261, 262, 265, 267, 269, 274, flict, self-determination conflict x, xi,
280, 281-283, 285, 286, 294, 303, 448, xii, xiii, xv, xvi, 5, 8, 12, 14, 15, 16, 18-19, 27,
449, 454, 560, 562, 568, 589, 629, 630, 29, 37, 42, 45, 47, 48, 49, 57, 58, 61, 64-67,
631, 637, 638 72, 88, 89 n. 20, 92, 96, 99, 100, 104, 111,
majority 11, 53, 265, 272, 285, 560 115, 123, 124, 125, 126, 127, 129, 130, 131,
national 48, 49, 97, 98, 100, 205, 208, 132, 133, 134, 135, 136, 137, 151, 152, 153, 154,
210, 212, 248, 249, 250-251, 255, 272, 403, 155, 156, 157, 159, 161-166, 167, 173, 174, 176,
539, 565, 640, 643, 678-679, 696, 697, 186, 188, 191, 193, 194, 196, 198, 199, 200,
699 201, 202, 208, 215, 216, 234, 239, 244, 245,
253, 258, 265, 266, 267, 268-279, 280, 289,
of states 13
290, 297, 298, 301-302, 303, 304, 308, 309,
political 41, 539 310, 311, 312, 313, 315, 316, 336, 341, 345-
religious 48, 73, 209, 233, 284, 295-296, 381, 387, 388, 390, 396, 398, 403, 404, 407,
441, 661 408, 409, 410, 413, 414, 416, 419, 421, 424,
rural 128 427, 428, 429, 433, 436, 438, 439, 440,
self-government 52, 72 444, 447, 451, 465, 467, 468, 469, 470, 471,
472, 473, 474, 475, 477, 478, 480, 481, 483
territorial 671
n. 22, 484, 488, 493, 494-495, 496, 497,
tribal 171, 178, 441, 512, 608 498, 499, 503, 504, 505, 508, 509, 513, 515,
Comores and Mayotte 28 518, 523, 525, 526, 527, 530, 531, 532, 533,
complex power-sharing, see also power- 534, 535-539, 540-551, 559, 560, 562, 563,
sharing and consociational power- 564, 565, 566, 575, 578, 579, 580, 581, 582,
sharing xiii, ix, xiv, xvii, 5, 13, 61-124, 583, 584-585, 586, 587, 588, 589, 590, 591,
125-159, 161-192, 239, 285-288, 346, 347, 592, 593, 594, 595, 596, 599, 600, 602, 604,
361, 363, 364, 376, 377, 378, 379, 387, 388, 609, 610, 611, 613, 616, 621, 622, 623, 624,
405, 408, 409-429, 430, 431, 434, 436, 437, 627, 628, 630, 632-635, 639, 641, 645, 647,
442-445, 449, 451-466, 477, 531-557, 569, 653, 659, 662, 663, 664, 665, 669, 677,
571, 572, 596, 597, 627-665, 676, 692, 701, 680, 683, 686, 696, 698, 700, 701, 704, 711,
705, 713 715, 718
agreements 387, 504, 569, 599-626, 627, armed xi, 29, 30, 92, 111, 163, 174, 186,
628, 630, 635, 637, 638, 639, 641, 643, 191, 194, 201, 224, 239, 243, 269 n. 8, 279,
645, 646, 652, 653, 676-686 289, 308 n. 2, 309, 310, 311, 313, 314, 315,
arrangements xvi, 156, 157, 203, 346, 388, 342, 468, 503, 471, 592, 596
396, 401-402, 429-447, 453, 464, 568, internal ix, xi, 30, 129, 156, 343, 413, 416,
569, 571-597, 653 532, 533, 535, 547, 559, 610
confederal 16 international 29, 30, 469 n. 1
federal 16 management 67-69, 378, 397, 445, 454,
settlements ix, 19, 447 474 n. 12, 475, 709
770 Index
post- xiv, 144, 153, 154, 193-241, 317, 347, 105, 108, 109, 117, 121, 136, 138, 139, 140-
351, 356, 363, 379, 406, 414, 438, 444, 146, 147, 150, 151, 152, 154, 165, 177, 179,
453, 474, 479, 486, 497, 503-504, 507, 189, 195, 205, 206, 207, 208, 209, 210, 211,
522, 523, 526, 527, 529, 530, 533, 535, 544, 212, 213, 214, 219, 224, 228, 229, 230, 232,
556, 571-572, 577, 578, 579, 589, 596, 597, 240, 244, 246, 247, 250, 258, 259, 261, 263,
676, 687, 701 265, 270, 271, 280, 283, 284, 287 n. 37, 290
prevention 238, 276, 588, 647 n. 40, 294, 295-296, 310, 311, 318, 319-320,
regulation 351, 357, 360, 365, 448, 628, 322, 323-325, 333, 334, 335, 336, 337, 341,
635, 692, 693 343, 349 n. 4, 388, 389, 390, 391, 394, 396,
398, 399, 413, 414, 416, 423, 424, 426, 430,
resolution 11, 168, 291, 304, 307, 346,
433, 438, 440, 482, 488, 491, 516, 518, 548,
347, 358, 379, 445, 449, 468, 469, 470,
560, 583, 599, 600, 601, 602-604, 605,
534, 535, 537, 539, 541, 551, 571, 578, 585,
607, 614 n. 49, 617, 621, 624, 627, 632, 633,
647, 677, 678, 710, 718
638, 639, 644, 648, 658, 664, 668 n. 1, 675,
secessionist ix, xvii, 37, 151, 413, 468 677, 679 n. 20, 680, 681, 699, 704 n. 28
separatist 467 constitutional, see also constitutional
settlement ix, 346, 363-366, 377, 378, self-determination 39, 48, 51, 55, 56,
419, 442, 444, 449, 477, 579, 670, 676, 57, 64, 67, 71, 73, 79, 85, 89, 97, 100-102,
677, 678, 683, 684 104, 105, 107, 111, 118, 121, 125, 126, 128,
violent 18, 50, 125, 127, 132, 408, 532, 545, 129, 136, 137, 139, 147, 151, 152, 157, 158, 171,
595, 700 197 n. 4, 198, 203, 207, 208 n. 13, 210, 211,
Congo 12, 19, 468, 713 213, 214, 229, 237, 240, 243-263, 265 n. 1,
271 n. 14, 279, 280, 282, 284, 288, 291, 292,
consociation 47-58, 61-124, 408, 445,
293, 294, 301, 303, 305, 323, 333, 335, 341,
455, 572, 592, 633, 692, 696, 697, 703, 704,
343, 364, 365, 389, 394, 395, 399, 404, 409,
706, 707, 709, 710, 711, 712, 713, 716, 718,
412, 414, 419, 420, 423, 424, 427, 432, 439,
719
440, 445, 446, 447, 448, 449, 451, 452, 479
consociational 15, 16, 49, 50, 51, 52, 54, n. 16, 490, 516, 521, 522, 539, 552, 566, 583,
55, 56, 57, 68, 69, 70, 72, 73, 79 n. 12, 83, 600, 601, 602, 603, 605, 621, 627, 633, 638,
87, 88, 90, 97, 100, 109, 110, 114, 378, 394, 639, 640-641, 646, 647, 648, 658, 659, 660,
407, 453, 628, 635, 684 n. 28, 691, 692, 665, 669, 671, 674-676, 685, 699, 700, 706,
693, 694, 696-718 711, 714
agreement 87, 693, 704, 711 n. 34 amendment/s 51, 213, 280, 282, 284, 288,
arrangements 52, 58, 86, 120, 124, 203, 291, 292, 293, 294, 295, 296, 481 n. 20,
267, 700, 705, 708 614 n. 50, 638, 639, 641, 648, 679
decision-making 51, 57, 123 arrangement/s xv, 15, 31, 53, 55, 146-147,
democracy 50, 205, 285, 287, 684 n. 28 205, 568, 580, 588, 593, 604
executive 51, 707-708 court 211-212, 213, 214, 218-219, 223,
institution 51, 705 229, 231, 235, 236, 248, 249, 250, 271 n.
power-sharing, see also power-sharing 13, 283, 322, 338, 341, 391, 392, 393, 411,
and complex power-sharing 16, 51, 427, 487, 488, 489, 600, 606, 607, 612,
204, 408, 713 675, 685
settlement 56, 68, 693, 708 law 32, 37, 139, 141, 147, 148, 150, 151, 152,
153, 414, 668
consociationalism/t ix, xiv, xvii, 203,
204, 263, 454, 456, 459, 469, 691, 692-703, reform 12
704, 706, 707, 708, 709-718, 719 settlement(s) 18, 44, 88,
constitution xv, xvi, 5, 14, 16, 22, 23, 31, un- 103, 213, 229, 376, 595
32, 33, 34, 36, 38, 39, 42, 53, 54, 56, 65, 100, Corsica/n 29, 388
Index 771
Council of Europe 232, 240 n. 27, 313, democratic xii, xv, 7, 8, 9, 10, 11, 13, 15,
315, 316, 318, 325, 331, 335, 396, 474, 475, 20, 22, 28, 39, 50, 51, 56, 58, 66, 70, 71, 79,
477, 481, 488, 489, 520, 583, 617, 620, 631, 83, 87, 91, 92, 93, 95, 116, 118, 119, 147, 157,
639, 641, 643, 670-671, 672, 695 n. 6, 186, 194, 195, 202, 216, 217, 220, 221, 234,
Committee of Ministers 218, 231, 649, 238, 239, 246, 247, 249, 256, 257, 259, 266,
671 n. 2 273, 275, 283, 285, 308, 318, 338 n. 28, 339,
Parliamentary Assembly (PACE) 336, 340, 357, 376, 377, 388, 389, 404, 422, 443,
338 n. 28, 343, 642 449, 451, 452, 453, 455, 459, 465, 492, 539,
573, 577, 592, 595, 599, 602, 610, 620, 621,
Croatia 14, 23, 34-36, 45, 194-198, 200,
622, 640, 642, 643, 644, 649, 652, 662,
201, 202, 209, 218, 220, 221, 223, 224, 225,
687, 692, 707-709, 714, 715, 717 n. 42, 718,
228, 232, 233, 237, 244, 245, 395, 396, 409,
719
486, 495, 567, 579, 642, 652, 680, 681 n.
24, 698 democratize/ation 7, 13, 58, 63, 123, 217,
235, 237, 238, 454, 455, 489, 616, 618, 619,
Croatian Declaration of Independ-
620, 622, 669, 686, 714 n. 36
ence 14
devolution 68, 69 n. 3, 73, 77 n. 11, 80, 92
Serb population in 14, 29, 198
n. 22, 101, 117, 120, 121, 122, 179, 316, 400,
Cypriot 53, 705 408, 410, 412, 416, 433, 434, 435, 448, 518,
Cyprus 24, 38 n. 12, 53, 54, 55, 58, 388- 529, 534 n. 3, 538, 550
389, 469, 470, 495, 699, 705, 713 diplomacy 65, 114, 161, 168, 265, 276, 348,
Annan Plan 38, 695, 699 356, 357, 363, 375, 378 n. 55, 482, 486-487,
Constitution 1960 53-54, 705 491, 493, 496, 579
Greek Cypriots 54, 705 disenfranchise 10, 18, 20, 23, 25, 26, 27,
Turkish Cypriots 53, 54 29, 30, 244, 405
Czech/ Republic 9, 22, 547, 642, 668, disintegrate/ion xi, 11, 194, 198, 307, 312,
670, 676 409, 444, 578
Czechoslovak/ia 22, 654 dispute/s, see also conflict, self-determi-
nation dispute 5, 11, 28, 49, 70, 102 n.
24, 126-128, 130, 145, 146, 156, 168, 196,
D 198, 199, 212, 239, 240, 249, 250, 286, 298,
302, 305, 311, 317, 343, 347, 367, 414, 416,
decision-making xiv, 51, 54, 57, 79, 81,
426, 427, 428, 436, 438, 468, 491, 507, 508,
123, 128, 137, 173, 194, 204 n. 7, 210, 217,
516, 518, 531, 532, 542, 551, 560, 564, 571,
228, 283, 343, 365, 371, 372, 374, 394, 407,
581, 582, 583, 584, 585, 587, 588, 589, 592,
424, 429, 431, 443, 449, 464, 476, 537, 628,
594, 595, 601, 618, 623, 645, 649, 653, 669-
638, 680, 681, 700
670, 675, 703, 712
decolonization, see also colonial 9, 10,
armed 350
11, 13, 14, 19, 25, 26, 29, 44, 193
internal 413, 532,
demilitarisation 94, 104, 111, 113, 117,
122, 125, 136, 147-149, 255, 257, 368, 404, resolution 394, 683
586, 587, 592, 593, 594, 701, 702, 703, 712, secessionist 126
719 settlement ix, 136, 145-146, 149, 172, 249,
democracy/ies 9, 13, 14-15, 43, 50, 58, 69 346, 359, 379, 391, 393, 399, 400, 527,
n. 3, 204, 205, 222, 265, 279, 285, 287, 342, 584, 593, 676, 684 n. 27, 685, 686
343, 419, 443, 449, 451, 454, 464, 481 n. dissolution xvi, 10, 22, 28, 31, 32, 34, 35,
20, 543, 599, 600, 603, 617, 619, 620, 621, 36, 38-39, 40, 41, 117, 119, 193, 194, 195,
622, 630, 643, 663, 684 n. 28, 692, 695, 291, 309, 395, 421, 428, 446, 468, 517, 550,
700, 707-709 642, 701
772 Index
549, 550, 551, 552, 553, 555, 556, 557, 578,

E
621, 667, 682, 687
East Bengal 12 education, see also right to educa-
East Timor 8, 10, 13, 25, 26, 135, 150, 193, tion 15, 16, 52, 54, 55, 68, 73, 81-83, 97,
559, 572 98, 105 n. 26, 106, 138, 153, 162, 163, 169,
economic, see also economic autonomy, 170, 172-174, 178, 190, 191, 203, 209 n. 14,
economic rights xvii, 11, 13, 15, 20, 44, 225, 232-234, 250, 251, 260, 271-272, 282,
65, 66 n. 1, 126, 127, 130, 131, 140, 143, 144, 283, 285, 286, 287, 288, 296, 310, 321, 325,
153, 154, 155, 165, 166, 168, 169, 172, 173, 326-327, 332, 334, 338, 375, 377, 393, 410,
174, 181, 198, 219, 226, 228, 234, 235, 272 412, 417, 419, 420, 422, 424, 427, 440 n. 3,
n. 17, 297, 305, 307, 312, 314, 319, 320, 328, 476, 489, 490, 503-530, 531, 538, 539, 543
329, 331, 332, 334, 337, 342, 343, 349, 350, n. 10, 544, 545, 546, 548, 550, 555, 563, 569,
351, 352 n. 13, 357-358, 360, 361, 362, 364 590, 596, 625, 631, 632, 635, 639, 660, 661,
n. 27, 365, 367, 368-369, 371, 373, 374, 375 662, 663, 681 n. 24, 684, 698, 699
n. 51, 376, 378 n. 55, 390, 393, 397, 413, election(s) xvi, 14, 43, 47-48, 71, 74, 75,
428, 429, 467, 468, 470, 471, 475, 479, 486, 76, 77, 78 n. 12, 81, 83, 84, 85, 86, 87, 89,
495, 496, 497, 498, 506-507, 510, 514, 515, 90, 91, 103, 106, 114, 116, 119, 122, 123, 137,
528, 529, 531-557, 644-645, 664, 667, 669, 139, 148, 150, 164, 165, 170, 179, 180, 181,
672, 681 n. 24, 682, 684, 687, 699 182, 186, 188, 189, 194, 195, 196, 204, 208
development 73, 107, 154, 166, 174, 183, n. 13, 210 n. 16, 213, 217-218, 219, 220, 221,
184, 185, 194, 283, 288, 296, 320, 324, 222, 223, 224, 228, 234, 239, 248, 249, 253,
326, 417, 420, 427, 439, 489, 507, 258, 255, 256, 258, 259, 260, 262, 273, 274, 275,
534, 541, 542, 545, 546, 553, 555, 591, 596, 278, 282, 283, 284 n. 33, 285, 286, 287,
670, 673, 674, 679 n.20 290, 291, 298, 299-300, 302, 303, 304, 311,
314, 317, 318, 321, 325-326, 331, 332, 333,
growth 174, 181, 504, 514, 534, 536, 537,
334, 336, 337, 338, 349, 354, 368, 376 n. 53,
542, 556
399, 422, 425, 426, 452-456, 459, 461-465,
in/equality 114, 129, 130, 347, 644, 710 466, 484, 485, 488, 489, 490, 491, 499,
exploitation xii, 25, 645 548, 554, 578, 620, 621, 630, 663, 694, 695,
institutions 54 705, 706, 707, 709, 712, 713, 715-718
opportunities 128, 163, 228, 530, 543, 644 electoral 52, 65, 89, 99, 109, 110, 114, 115,
policy/ies 78, 82, 153-154, 174-176, 234- 118, 119, 122, 136, 169, 188, 195, 218, 222,
235, 328-330, 374, 422, 424, 491, 531-557, 224, 274, 284, 298, 321, 325-326, 340, 433,
669 444, 451, 452, 456, 464-465, 695, 705, 708,
reconstruction 193, 218, 488, 489, 490, 717, 718
491, 542, 670 arrangements 451-466
reform 219, 220, 234, 480, 572 commission 87 n. 15, 318, 339, 465, 466
resources xiv, 129 competition 52
socio- 163, 165, 169, 183, 185, 271-272, defeat 14
289, 334, 351, 504-507, 525, 532, 534, 535, support 68, 89
543, 545, 551, 556, 563, 590 systems 52-54, 86-91, 109, 204, 287, 300,
Economic Community of West African 317, 326, 333, 407, 434, 451-454, 457-464,
States (ECOWAS) 470 627, 628, 692, 700, 704, 705-706, 708,
economy/ies 11-12, 107, 127, 174, 235, 236, 713-714, 715, 716 n. 40
272, 287, 321, 324, 326, 328, 331, 334, 335, emigration 65
342, 362, 364, 370, 377, 453, 490, 504, 507, enfranchise, see also disenfranchise 18,
508, 523, 525, 528, 532-533, 534, 536, 537, 23, 29
538, 539, 540, 541, 542, 544, 545, 546, 547,
Index 773
equality 37, 49, 66, 71, 72, 78, 83, 95, 97- cross-/trans- 2, 88, 204, 286, 463, 464
98, 106, 114, 117, 209 n. 14, 244, 280, 320, 704, 706, 713, 715, 719
397, 423, 599, 600, 604 n. 15, 620, 644, expulsion 48
647 identity 125, 127, 172, 193, 260, 413, 455,
in- xi, xii, 129, 130, 347, 513, 645, 646 456, 504, 510, 528, 551, 581, 638, 672, 704,
Eritrea 13, 22-23, 29, 42, 396 717, 718, 719
Ethiopia xv, 13, 22-23, 29, 32, 42 inter- 88, 213, 216, 217, 219, 222, 234, 235,
ethnic(ally) xvi, xvii, 10, 13, 14, 15, 26, 27, 238, 240, 241, 266, 267, 269, 270-272,
28, 29, 32, 42, 51, 52, 53, 54, 55, 56, 57, 58, 274, 276, 300, 303-304, 309, 310, 317,
63 , 69 n. 3, 98, 130, 134, 138, 165, 178, 185, 342, 350, 351, 355, 397, 412, 423, 437, 444,
193, 196, 197, 199, 200, 204, 205, 207, 208, 448, 463, 474, 475, 476, 484, 530, 532,
210, 212, 213, 214, 218, 221, 224, 226, 232, 662, 676, 692, 693, 707, 711 n. 34, 713,
233, 238, 239, 240, 244, 245, 246, 248, 249, 715, 716, 717, 718
250, 258, 259, 262, 263, 266, 267, 268, 269, multi- 37, 48, 88 n. 18, 193, 195, 238, 265,
270, 271, 272, 273, 274, 275, 276, 277, 278, 280, 304, 307, 313, 455, 456, 457, 459,
279, 280, 281, 282, 283, 286, 287, 288, 289, 464, 567, 587
290, 291, 292, 293, 294, 295, 296, 297, 298, peoples xi
299, 300, 301, 302, 303, 304, 307, 309, 312, parties 52, 88, 89 n. 20, 243, 258, 259,
320 n. 13, 342, 346, 347, 348, 349, 350, 351, 262, 272, 275, 277, 278, 285, 288, 293 n.
352, 353, 354, 360, 368, 393, 394, 403, 405, 44, 295, 302, 303, 426, 433, 434, 463, 589,
407, 412, 413, 416, 419, 421, 426, 427, 428, 705, 710, 716, 717
433, 434, 440 n. 3, 442, 448, 453, 454, 456,
politics xvii, 240, 241, 249, 522
460, 462, 464, 504, 516, 517, 518, 519, 526,
527, 531, 532, 533, 540, 543, 545, 548, 550, rights 98, 717
559, 560, 564, 566, 568, 569, 572, 582, 584, self-determination 19, 348
588, 589, 590, 603, 624, 628, 634, 643, 644, violence xv, 134, 274, 424, 541, 551, 559-
653, 664, 668, 669, 670, 675, 677, 679, 569, 663
680, 681, 684, 687, 691, 704, 706, 707, ethnicity 7, 57, 194, 199-200, 204, 213,
709, 710, 713, 714 n. 37, 715, 716, 717 307, 348, 353, 376, 464, 515, 516, 517, 596,
cleansing 14, 24, 40, 130, 138, 197, 200, 608, 615, 630, 710
221, 224, 225, 240, 278, 352, 579, 584, ethno-
628, 641, 704 centric/ism 668, 692, 704, 710, 715
communities/groups, see also ethnic cultural 307, 345 n. 1, 348, 360, 377, 631,
minorities iv, 7, 20, 27, 48, 49, 61, 73, 641
87, 138, 162, 208, 243, 245, 248, 249, 259,
national 66, 68, 72 n. 4, 78, 87, 88, 89,
263, 270, 272, 274, 276, 283, 286, 287,
96, 99, 100, 111, 115, 217, 307, 492 n. 30,
288, 289, 290, 293, 300, 307, 311, 312, 313,
515, 548, 583, 696, 697, 716, 718
33, 342, 351, 375, 434, 440 n. 3, 441, 453,
455, 456, 459, 464, 484, 505, 522, 528, political 19, 307, 308, 309, 312, 313, 348,
530, 543, 559, 562, 567, 569, 572, 588, 612, 678
618, 627, 634, 635, 637, 643, 644, 647, territorial xiv, 19
653, 667, 669, 674, 679 n. 20, 681, 698, Europe/an 6, 8, 9, 36, 41, 50, 66, 73 n. 8,
704, 705, 706, 713, 715 89 n. 19, 104, 108, 121, 127, 161, 214, 215,
conflict/dispute 7, 58, 89 n. 20, 377, 413, 229, 240-241, 307, 316, 317, 496, 516, 517,
424, 530, 532, 541, 542, 560, 568, 610, 525, 537, 540, 544, 550, 553, 556, 567, 591,
627-628, 632, 635, 641, 645, 646, 647, 639, 669, 670-674, 675, 696, 700
653, 668, 669-670, 688 Central 7, 19, 311, 426, 545, 546-547, 548,
577, 609, 672
774 Index
East/ern 7, 10, 19, 199, 270 n. 11, 271, 311, 574, 583, 628, 662, 682, 700, 706, 707-708,
320 n. 13, 426, 459, 465, 479, 480 n. 18, 709, 712, 713, 714, 718
518, 522, 545, 546-547, 548, 577, 603, 609,
613, 616, 672, 686, 694, 717 n. 42
F
West/ern 467, 523, 566, 667, 692
European Charter for Regional or Mi- Falklands 9
nority Languages 232, 271 n. 14, 631, federal/ative xvi, 16, 22, 32, 33, 34, 35, 36,
633, 649, 659 37, 38, 39, 43, 45, 49, 53, 55, 56, 57, 73, 101,
European Convention on Human Rights 102, 105, 108, 109, 110, 124, 136, 146, 193,
(ECHR) 73, 94, 98, 250, 251, 599, 603- 194, 207, 209, 212, 239, 244, 316, 341, 347,
604, 606, 609, 612, 617, 631, 642, 644, 348, 349, 359, 366, 371, 388, 389, 392, 394,
649, 651, 658, 659, 662 399, 402, 407, 423, 431, 434, 446, 538, 542,
European Court of Human Rights 13 543, 580, 581, 584, 628, 642, 668, 674, 679,
n. 3, 212, 218, 250-251, 474, 477, 487-489, 681, 704, 707, 713, 717 n. 42
617, 698 n. 6 federalism/t 47, 108, 204, 205, 283, 445,
European Framework Convention on 446, 549, 668, 687
the Protection of National Minorities federacy/tion 8, 10, 19, 22, 33, 34, 35, 36,
(FCNM) 98-99, 232, 271 n. 14, 331, 631, 38, 39, 40, 47-58, 61-124, 196, 198, 200-
633, 641, 644, 649, 658, 659, 662, 672, 687 201, 203, 205, 206, 208, 209, 210, 211,
n. 31 212, 213, 214, 219, 221, 223, 225, 228, 231,
European Union (EU), formerly Euro- 232 233, 235, 236, 237, 238, 245, 247 n. 2,
pean Community 5, 28, 35, 37, 51, 73, 255, 263, 273 n. 18, 341, 407, 412, 423, 433,
78, 105, 106, 107, 108, 110, 121, 199, 214, 445-447, 534, 540, 541 n. 7, 636, 668, 697,
218, 229, 255, 256, 263, 267, 276, 278, 280, 702, 717
291, 294, 295, 299, 300-301, 302, 303, 305 integrative 56-57
n. 69, 355, 357-358, 365, 372-374, 379, 394, pluralist 49, 56, 57-58
395 n. 20, 369, 399, 400, 427, 440, 470, Finland xiii, 650, 652
473, 475, 474, 477, 478, 479-480, 482, 483, Flemish 53, 55
485 n. 25, 486, 488, 489, 490-491, 496,
497, 514, 520, 528, 534, 535, 538, 540, 541 fragmentation 13, 28, 50, 87 n. 15, 89, 193,
n. 7, 542, 543, 544-545, 547-548, 550, 553, 204, 460, 464, 518, 540 n. 5
554-555, 556, 562, 577, 589, 595, 618, 620, France 10, 12, 199, 388, 470 n. 4, 483, 533,
668, 673-674, 687, 693, 694, 699 681 n. 24, 682
European Commission 301, 370 freedom xiii, 13, 65
European Court of Justice 439 of assembly 628
European Parliament 73 n. 8, 80, 89 n. of association 12, 628, 636, 637, 697
19, 122 of speech 12
executive, see also executive power-shar- French 6, 11, 53, 55, 197 n. 4, 198, 276, 278,
ing 47, 51, 53, 54, 55, 56, 57, 73, 76-80, 456, 483, 693
83, 84-86, 87 n. 16, 92 n. 22, 93, 94, 102, frontier 27, 35, 638, 646, 672, 698 n. 14
104, 105, 106, 107, 111, 112, 116-117, 120,
123, 165, 167, 171, 179, 181, 183, 208, 217,
258, 285, 316, 322, 324-325, 326, 331, 334, G
340-341, 343, 368 n. 36, 372, 374, 392, 394, Gabon 12
409, 410, 411, 415, 416, 417, 418, 420, 421,
Gagauzia/n xvi, 32, 43-44, 307-343, 391,
422, 425-426, 427, 437, 446, 452, 456, 459-
400, 408, 412, 418, 419, 420, 421-423, 426,
460, 464, 488, 489, 490, 513, 549, 554, 559,
430, 431, 432, 434, 435, 437, 438, 439, 440,
Index 775
441, 442, 444, 445, 446, 447, 449, 452, co- xiv, 45, 114, 243, 244, 246,
459, 464, 475, 476, 477, 478, 479, 481, 482, democratic 10, 20, 28, 247, 644
483, 484, 493, 494, 495, 496, 497, 505, 506, devolved 69, 70, 108, 449, 684
513-514, 523, 526, 528, 548, 549, 550, 568,
domestic 67
572, 582-584, 596, 602, 603, 628, 641, 650,
651, 658, 660, 664, 681, 682, 694, 697, 699 federal 57, 101, 347, 349, 359, 407, 580,
707
Law on the Special Legal Status of
Gagauzia 1994 xv-xvi, 32, 43, 311, 314- good xiii, xvi, 12, 43, 137, 139, 262, 399,
33, 335, 337, 339, 343, 391, 400, 568, 583, 424, 615, 620, 624, 627, 628, 646, 659,
633, 638, 639, 641, 644, 658, 660, 681, 682, 700
699 effective xv, 40, 41
Gagauz/es xvi, 32, 308, 309-310, 311, 312, interim 44, 243-263, 392, 397, 581, 713
313, 314, 315, 316, 317, 318, 319, 320, 321, international x, 36
322, 323, 325, 326, 327, 328, 329, 331, 332, local 47, 82, 89, 93, 94, 114, 122, 127, 131,
333, 334, 335, 336, 337, 338, 339, 341, 391, 132, 140, 145, 169, 170-171, 172, 176, 179-
414, 421, 422 , 423, 433, 448 180, 181-182, 188, 191, 271 n. 13, 282, 283,
genocide 12, 14, 24, 40, 48, 58, 164, 209 287, 288, 299, 338, 407, 415, 416, 418, 419,
n. 15, 358 423, 430, 431, 441, 543, 679
Georgia/n x, xiii, 19, 41, 345 n. 1, 346-379, minority 24
388, 389, 390-391, 426, 429, 438, 447, 452, multi-ethnic 2, 8
454, 457, 458, 459, 460, 462, 464, 465,
multi-national 205
466, 468, 475, 478, 479, 481, 483 n. 22,
484, 490-491, 495, 505, 506, 515, 526, 532, multi-party 51
534, 540, 546 n. 11, 547, 548, 594-595, 604, municipal 281, 297, 413
609, 612, 629, 658, 659, 664, 665, 678 national 129, 130, 131, 133, 135, 136, 137,
Germany 22, 199, 302, 404, 454, 470 n. 4, 138, 139, 140, 141, 142, 143, 144, 145, 146,
553, 652, 670, 674, 676 n. 14, 681 147, 148, 149, 150, 151, 152, 153, 155, 156,
Gibraltar 9, 21 157, 158, 159, 170, 171, 173, 178, 191, 308,
389, 393, 398, 399, 414, 415, 416, 417, 418-
Goa 8
419, 427, 428, 430, 433, 437, 440, 447,
governance/government 479, 508, 509, 511, 512, 514, 518, 539, 549,
autonomous xvi, 43, 56, 101, 143, 144, 550, 551, 555, 563, 582, 590, 596, 608, 682
146, 148, 149, 151, 154, 183, 392, 415, 416, n. 26, 683
442, 508, 610, 635, 665, 682, 683 parliamentary 51, 67
central, see also central state xi, xii, xiii, provincial 128-129, 132 n. 7, 138, 144,
xiv, xvi, 18, 19, 22, 23, 24, 30, 32, 35, 37, 146, 151, 152, 153, 154, 170, 171, 174, 414,
39, 40, 43, 44, 45, 47, 56, 102, 126, 128, 418, 508, 551
130, 138, 155, 165, 170, 171, 173, 176, 178,
regional 170-172, 173, 174, 175-176, 177,
180, 184, 185, 190, 195, 200, 202, 205,
178, 190-191, 430, 433, 440, 441, 443,
263, 280, 281, 287, 308, 309, 311, 317, 328,
509, 510, 511, 512, 513, 553, 555, 590
331, 332, 333, 337, 341, 342, 347, 393, 395
n. 18, 396, 397, 400, 405, 407, 408, 412, representative 47-48, 205, 653
415, 419, 420, 422, 428, 430, 432, 435, self- xiii, xiii, xiv, xvi, 16, 25, 42, 43, 45,
437, 443, 445, 446, 447, 467, 468, 496, 49, 50, 52, 54-55, 56, 57, 68, 72, 73, 76,
498, 509, 510, 513, 522, 538, 539, 549, 115, 168, 214, 221, 223, 240, 243, 247, 248,
590, 672, 675, 677, 680, 681, 686, 687, 256, 257, 258, 259, 260, 261, 262, 271,
697, 698 280, 281, 283, 284, 286, 287, 296-297,
coalition 78, 83, 110, 230, 262, 273, 274, 299, 349, 356, 364, 388, 389, 392, 393,
294, 311, 433, 434 402, 403-404, 405, 419, 423, 424, 428,
776 Index
434, 441, 464, 489, 490, 517, 521, 543, 551, 528, 530, 555, 584, 631, 636, 638-639, 643,
630, 633, 635, 638, 639, 640, 641, 642, 660, 680, 699, 704-707, 710, 717
646, 652, 653, 674, 679 n. 20, 680, 686, community 283
700 n. 24, 703 cultural 21
territorial 49, 56, 672 ethnic 125, 172, 193, 413, 455, 504, 510,
governmental ix, xiv, 52, 65, 72, 108, 346, 528, 551, 581
410, 439, 650 group 21
inter- 61-124, 136, 145, 158, 171, 190, 346, legal 39, 41, 43, 243, 392, 400, 403
359, 374-375, 429, 471, 478, 698, 710
national 6, 10, 193, 320, 428, 679
non- 167, 238, 251, 266, 272, 373, 375,
political 7, 13, 100
410, 439, 628, 650, 672, 679
regional 555
Great Britain/Britain, see also United
Kingdom 10, 43, 63-67, 100-102, 108, religious 672
121, 124, 199, 400, 414, 419, 430, 464, 470, territorial 28
552, 591-592, 593, 594 n. 4, 696 imperial 8, 10, 15
Greece 270, 305, 427, 469, 541 n. 7, 544, imperialism/t xii, 25, 64, 162, 551
545, 699 India/n xiii, 8, 9, 12, 14, 19, 29, 30, 40, 55,
Greek 54, 388, 705 72 n. 7, 146, 708
Muslim population in 14, 55
H indigenous 363, 495, 508, 546, 559, 578,
602, 609, 624
human rights, see also freedom, minority
indigenous people(s)/population xii,
rights, and rights ix, xii, xiii, xvii, 11, 12,
21, 161, 171, 172, 176, 177, 178, 179, 416, 459,
13, 20, 30, 31, 37, 44, 68, 71, 92, 93, 94, 99,
512, 622, 631, 640, 642
117, 120, 138, 141, 143, 144, 145-146, 150,
151, 152, 157, 158, 177, 178-179, 198, 206, Indonesia(n) 8, 10, 13, 26, 164, 166, 167-
207, 218, 219, 230-231, 232, 235, 245, 250- 168, 175, 187, 388, 468, 473, 476, 477, 481,
251, 257, 272, 276, 288, 291, 317, 326, 357, 550
362, 390, 394, 396, 397, 398, 400, 425, 471, insecurity 150, 167, 225, 231, 276, 305, 351
474, 476, 484, 488, 489, 503, 504, 521, 525, insurgency/ies 16, 19, 270-272, 591
531, 560, 561, 564, 567, 574, 578, 579, 582, insurgent(s) 67, 269 n. 7, 360, 370, 416,
599-600, 601, 603, 604, 606, 609, 610, 525, 560, 562, 563, 565, 568, 588
611, 612, 615-618, 619-620, 621, 622, 623, integration 28, 29, 47, 48, 49, 54, 58, 63,
624, 625, 627-665, 687, 692, 694, 695, 701, 67, 68, 97, 123, 126, 127, 132, 136, 137, 138,
702, 703, 719 139, 142, 155, 156, 157, 158, 169, 185, 215,
humanitarian 19, 29, 30, 31, 36, 37, 40, 240, 241, 263, 277, 285, 305, 348, 354, 373,
44, 45, 154, 245, 253, 255, 257, 299, 340, 391, 392, 400, 413, 455, 479, 480, 504, 522,
352 n. 12, 358, 362, 379, 397, 470, 471, 473, 526, 530, 537, 540, 545, 581, 607, 611, 612,
474, 479, 480, 485, 489, 490, 491, 542, 578, 640, 661, 667, 670, 687, 692, 695, 701, 710
585, 606, 610, 612, 646, 647 dis- xi, 194, 198, 312, 409, 444
re- 12, 221, 232, 234, 352, 363, 373, 426,
I 480 n. 19, 485, 525, 575, 578, 703
integrationism/t xvii, 48, 89, 155, 157,
Iceland 34, 395 n. 15
378, 628, 635, 691, 692, 697, 703, 704,
identity 6, 7, 10, 18, 20, 26, 32, 45, 56, 65, 706, 709-718
97, 98, 100, 123, 127, 128, 155, 156, 162, 244,
integrative x, xiii, xvii, 57, 204, 238, 239,
261, 283, 392, 393, 397, 399, 405, 455, 456,
240, 285, 398, 528, 681, 692, 703-718, 719
Index 777
international, see also international in- 470, 471, 474, 475, 476, 485, 496, 497,
volvement 498, 560, 584, 586, 588, 639, 693, 694
actor(s)/agents xi, xiii, xvi, 19, 35, 42, recognition 8, 9, 12, 14, 24, 131, 197, 198,
134, 183, 204, 222, 240, 293, 299-300, 200
315-317, 334, 341, 353, 356, 358, 378, 379, relations 6, 7, 34, 140, 206, 516, 542, 668,
405, 438, 439, 468, 469, 484, 488, 489, 673, 677, 679, 681, 682, 683
490, 491, 492, 493, 494, 495, 497, 498, society 5-6, 7, 12, 14, 16
499, 543, 636
support 30, 129, 135, 150, 183, 197, 299,
agency/ies 38, 41, 204 n. 9, 216, 217, 219, 304-305, 363, 530
228, 520, 535, 609, 681, 682, 683
system xii, xiii, 17-18, 22, 29, 35, 45, 646
agreement/s 143, 146, 260, 425, 614, 628,
International Labour Organization
681
(ILO) 631
assistance xiii, 304, 336, 343, 520
Convention No. 169 Concerning
body 164, 180, 481, 488
Indigenous and Tribal Peoples in
borders 8, 10, 175, 467, 468, 565
Independent Countries 21, 631, 640
community xv, xvi, 35, 37, 39, 40, 125,
intervention 30, 31, 40, 58, 67, 134, 145,
132, 135, 143, 144, 149, 150, 157, 159, 193,
201, 203, 217, 219, 220 n. 20, 240, 292,
198, 199, 201, 203, 204, 207, 214-220,
427, 429, 436, 438, 442, 443, 453, 468,
221, 222, 223, 224, 228, 229, 233, 235, 238,
469, 482, 487, 491, 492 n. 30, 494, 496,
246, 252, 266, 267, 268, 270, 276, 277,
498, 533, 536, 557, 575, 578, 584, 591, 635,
289, 291, 292, 293, 300-301, 303, 304,
681, 693-695, 696, 703, 718
305, 313, 317, 355 n. 19, 378 n. 55, 402,
424, 429, 436, 437, 438, 439, 443, 444, armed/military 31, 36, 41, 485, 585, 694,
467, 468, 473, 474, 479, 480, 483, 484, 699
487, 496, 508, 541, 543, 569, 579, 588, international 193, 204, 217, 218, 222, 238,
590, 662, 681, 704 695, 703, 710-711
humanitarian law 29, 30, 253, 352, 578, non- xii, 18, 30, 31, 44
585, 606, 610, 612 interventionism/ist 58, 193, 217, 358
institutions 203, 207, 214, 218, 443, 542, involvement xiii, 30, 74, 129, 133, 134, 135,
569, 718 146, 168, 176, 179, 219, 232, 255, 259, 290,
law xii, 17, 21, 23, 24, 27, 28, 31, 40, 42, 292, 298, 299 n. 57, 319, 347, 352, 355, 359,
65, 104, 313, 317, 356, 359, 390, 392, 396, 360, 362, 363, 378, 390, 391, 397, 419, 438,
397, 405, 483, 609, 610, 617, 621, 630, 449, 498, 542, 568, 578, 581, 583, 584, 589,
638, 642, 645, 653, 654, 668, 669, 671, 590, 591, 595, 596, 597, 632, 673, 694
673, 679 international x, xiv, xvii, 42, 45, 125, 126,
legal personality 22, 30, 38, 389, 397, 134, 159, 180, 198, 214, 234, 237, 245, 315,
mediation/or 199, 200, 278, 379, 398, 318, 356, 358, 360, 388, 393-394, 396, 439,
429, 474, 498 444, 447, 495, 496, 497, 498, 533, 590,
677, 678
norms 24, 58, 94, 392, 467, 599, 603
third-party 467-499
obligation/s 143, 206, 207, 317, 414, 484,
680 n. 23 Iraq 5, 31, 40, 58, 571, 578, 579, 694, 707,
713
observers 149, 253, 325, 326, 354 n. 17,
464, 472, 585 Ireland, see also Northern Ireland xv,
43, 61, 62, 63, 64, 65, 66, 67, 68, 70, 72, 74,
organizations 58, 203, 214, 216, 217, 218,
90 n. 21, 100, 101, 104, 105-106, 107, 108,
219, 230, 233, 238, 240 n. 27, 256, 265,
109, 110, 120, 121, 124, 400, 401, 408, 414,
302 n. 63, 355, 340, 348, 355 n. 20, 356,
419, 420, 438, 446, 449, 470, 479, 538, 544,
363, 365, 370, 378, 379, 437, 439, 468,
778 Index
552-553, 554, 555, 592, 603, 614, 621, 642, 525, 526, 528, 532, 534, 540, 541, 542-543,
649, 684, 694 n. 4, 696, 698, 700, 708-709 544, 559, 560, 565-566, 568, 569, 572, 584-
irredenta/ism/t 8, 9, 11, 58, 65, 420 588, 589, 596, 602, 603, 604, 607, 608,
Israel 50, 225, 468, 470, 610, 708, 609, 612, 613, 614, 621, 633, 635, 636, 638,
639, 640-641, 643, 645, 646, 647, 648, 649,
Italy xiii, 199, 334, 470 n. 4, 553, 568, 674
650, 651, 652, 653, 658, 659, 660, 661, 662,
n. 12, 681 n. 24, 682
678, 679, 680, 694, 695, 698, 699
Ivory Coast 12
Albanian population in 11, 42, 243, 244,
245-246, 248, 249, 258-259, 261, 263,
J 266, 268, 269 n. 7, 299, 393, 422-423,
463-464, 483, 491, 495, 520, 522-523,
Jews 50, 213, 296, 708 526, 565-566, 584, 587, 636, 646, 655,
judicial/ry 17, 47, 54, 56, 57, 92, 99, 107, 660, 678, 679 n. 19
131 n. 6, 138, 140, 141, 142, 151, 152, 155, Constitutional Framework for Provi-
177, 220, 235, 236, 245, 246, 248, 249, 250, sional Self-government in Kosovo
250, 279, 330, 393, 409, 416, 417, 419, 421, 2001 229, 258, 260, 262, 263, 424, 490,
422, 424, 426, 427, 436, 437, 441, 452, 484, 521, 522, 566, 605, 633, 635, 638, 639,
489, 572, 573, 574, 599, 600, 604, 606, 640, 641, 646, 647, 648, 658, 660
608, 609, 612, 613, 614, 617, 618, 619, 620, Muslim population in 248, 566
624, 627, 632, 640, 648, 652, 662, 663,
669, 683, 685, 686, 699 Rambouillet Interim Agreement for
Peace and Self-Government in Ko-
jurisdiction 5, 22, 33 n. 8, 55, 73, 84, 86, sovo 1999 xv, 42, 243-263, 402, 403,
97, 100, 101, 102 n. 24, 103, 104, 105, 106, 404, 486, 585, 586, 678, 699
109, 114, 120, 175, 177, 212, 245, 247, 288, Serb population in 244, 248, 249, 250,
297, 319 n. 10, 371, 390, 399, 414, 422, 428, 258-259, 261, 262, 263, 403, 423-424,
539, 549, 550, 554, 555, 567, 601, 606, 608, 434, 448, 455, 456, 458, 463-464, 489,
634, 637, 642, 653, 659, 668, 669, 674, 683 491, 520, 522, 526, 565-566, 584, 586, 587,
636, 660, 662
K Kurdish/s 5, 31, 40, 696
Kurd/s 5, 696
Kashmir 29, 470
Katanga 12, 19, 24, 387
Kenya 398, 456-457, 568 L
kin state 27, 204, 244, 310, 313, 319, 419, Laos 550
420, 467, 470, 496, 610, 668, 678, 683, 699 Latin America 57
Kosovars/ns 29, 247, 250, 251, 455, 489- Latvia 34, 38, 395, 547 n. 12, 642
490, 491, 521, 522, 646
Lebanon 50, 55, 115, 468, 693, 705
Kosovo x, xiv, xv, xvi, 11, 19, 25, 31, 34, 35,
36, 38, 39, 40, 42, 193, 196, 229, 243-263, Muslim population in 53, 54, 705
266, 268-270, 277, 292, 297, 299, 304, 305 League of Nations 8, 584
n. 69, 393, 397, 402-405, 408, 410, 412, legislative 47, 51, 52, 56, 72, 73, 74, 81, 83,
413, 423-426, 427, 430, 431, 432, 433, 434, 87 n. 15, 92, 99, 102, 103, 104, 108, 129, 152,
435, 437, 438, 439, 440, 443, 444, 445, 171, 179, 186, 217, 248, 249, 258, 261, 263,
446, 447, 448, 449, 452, 454, 455, 456, 457, 284, 293, 299, 311, 315, 316, 322, 326, 333,
458, 460, 461, 462, 463, 464, 465, 466, 334, 336, 341, 374, 391, 392, 394, 409, 410,
468, 470, 474, 475, 476, 477, 478, 479, 483, 412, 421, 453, 454, 455, 457, 458, 488, 489,
484-485, 486-487, 488, 489-490, 491, 492, 535 n. 4, 574, 601, 628, 640, 677, 705, 707,
493, 494, 495, 496, 497, 505, 506, 520-523, 708, 718
Index 779
legislature xiv, 53, 54, 56, 57, 105, 136, 640-641, 644, 647, 648, 658, 660, 661,
144, 147, 157, 167, 170, 249, 322, 407, 411, 679-680, 693, 699
416, 417, 421, 422, 434, 452, 453, 454, 457, Serb population in 270, 295, 300
459, 460, 461, 464, 574, 604, 615, 698, 700, majoritarian 51, 56, 57, 65, 88, 89, 279,
704, 705, 706, 707, 708, 714 n. 37 419, 449, 457, 458, 461, 543, 696, 697 n. 12,
legitimate xii, 5, 7, 10, 13, 54, 71, 89, 180, 700 n. 22, 708
183, 268, 289, 329, 412, 413, 440, 445, 447, majorities 47, 48, 54, 55, 56, 67, 70, 73,
465, 591, 633, 638, 641, 709 n. 13 74, 77, 88, 110, 281-282, 286, 287, 296, 307,
il- 30 442, 451, 453, 458, 705, 707 n. 30, 708
legitimacy xii 12, 15, 17, 18, 39, 43, 134, majority 11, 14, 15, 21, 33, 43, 53, 54, 56,
168, 194, 197, 268-269, 290-291, 302, 307, 64, 65, 67, 68, 70, 71, 72, 74, 75, 76, 77, 80,
352, 360, 376, 377, 400, 443, 444, 449, 459, 91, 94, 95, 97, 99, 100, 101, 103, 104, 109,
464, 465, 496, 498, 503, 555, 615, 630, 646 114, 116, 119, 120, 124, 135, 147, 155, 163,
Liechtenstein 32 167, 169, 171, 186, 191, 197, 206, 207, 208
Lithuania 33, 395, 547 n. 12, 642 n. 13, 210, 211, 212, 213, 225, 227, 239, 246,
loyalist/s 66, 67, 68, 70, 71, 86, 89, 96, 111, 249, 258, 261, 262, 263, 265, 266, 268, 269,
113, 114, 118, 119, 524, 591, 592, 662, 711, 271, 273, 277, 281, 282, 283, 285, 286, 287,
288, 293, 294, 295, 296, 298, 300, 303, 320,
321, 323, 324, 336, 339, 341, 343, 351, 388,
M 400, 401, 407, 409, 412, 414, 416, 419, 420,
423, 454, 456, 457, 459, 460, 464, 465,
Macedonia, Former Yugoslav Republic 466, 488, 512, 516, 517, 519, 560, 561, 562,
of ix, xiv, 42, 58, 193, 196, 244, 265-305, 591, 599, 622, 630, 635, 636, 639, 641, 644,
389, 393-394, 395, 396, 407, 408, 412, 426- 654, 667, 676, 679 n. 20, 682, 683, 693,
428, 430, 431, 432, 433, 434, 435, 437, 438, 705, 707, 708, 709, 712, 713, 715, 716, 718
439, 440, 444, 445, 446, 447, 448, 449,
452, 454, 457, 458, 459, 460, 461, 462, 464, Malaysia/n 9, 22, 40, 72 n. 7, 146, 164,
465, 466, 468, 474-475, 476, 477, 478, 479, 166, 175, 183, 187-188, 189, 191, 468, 473,
480, 482, 483, 484, 485, 487 n. 26, 492, 477, 481, 550
493, 494, 495, 496, 497, 505, 506, 516-518, Mali 40, 188, 388
525, 526, 528, 532, 534, 540, 541, 543-544, Maori 53
561-563, 566, 568, 572, 585, 588-590, 596, Mauritania 9
603, 607, 608, 609, 612, 613, 614 n. 50, military xiv, 11, 19, 29, 30, 31, 36, 47, 64,
633, 638, 639-640, 641, 642, 645, 646, 648, 65, 66, 70, 92, 111, 114, 130, 131, 134, 137,
649, 650, 651, 658, 660, 661, 663, 678, 161, 162, 164, 167, 176, 182, 183, 187, 189,
679-680, 693, 694, 698, 699, 700, 710, 711, 191, 193, 197, 198, 201, 202, 203, 214-216,
712, 713 219, 222, 224, 240, 245, 253, 254, 255, 256,
Albanian population in 265, 266, 267, 266, 269 n. 7, 274, 275, 276, 279, 302, 303,
268-281,282-284, 285-291, 292-299, 300, 349, 352, 353, 357, 358, 362, 364, 367, 376,
301-302, 303-304, 305, 393, 426-427, 433, 378, 397, 400, 428, 429, 467, 471, 474, 476,
434, 444, 448, 475, 485, 516-518, 526, 543, 477, 483, 484, 485, 486, 487, 488, 489,
561-562, 568, 587, 588-590, 678, 679 n. 19, 490, 491, 492, 494, 496, 531, 559, 566, 567,
700 n. 24 571-597, 694, 699
Muslim population in 274, 284, 295 Mindanao 161-192, 408, 412, 413, 414,
Ohrid Framework Agreement 2001 42, 416-419, 422, 430, 431, 432, 433, 435, 436,
267, 273 n. 19, 277-305, 393-394, 426- 437, 438, 439, 440, 441, 442, 444, 445,
427, 434, 475, 479 n. 16, 483, 516-517, 446, 447, 448, 448, 449, 452, 458, 459,
543-544, 562, 589-590, 633, 638-639, 460, 462, 464, 466, 473, 476, 477, 478,
479, 480 n. 19, 481, 485, 492, 493, 494, 495,
780 Index
496, 497, 505, 506, 507, 509-513, 523, 525, returns 225, 231
526, 527, 528, 529, 530, 532, 534, 540, 550- rights ix, xiv, 8, 13, 20-21, 35, 67, 72, 98,
551, 563, 572, 590-591, 596, 602, 604, 608, 99, 125, 143-144, 152, 158, 231, 232, 238,
612, 613, 628, 633, 638 239, 245, 259, 266, 268, 270, 271, 272,
Mindanao Peace Agreement 1996 166- 274, 277, 280, 285, 288, 291, 300, 317,
189, 191, 447, 509-510, 512, 527, 528, 326, 332, 377, 390, 396, 476, 484, 486,
590-591, 606, 633, 677 490, 491, 525, 526, 531, 543, 588, 618, 627-
Muslim population in/Moro 33 n. 7, 161- 665, 672, 678, 694, 700
165, 167, 169, 171, 176, 177, 181-182, 183, Moldova x, xiii, xvi, 19, 32-33, 41, 43, 307-
188, 189, 191, 416-417, 418, 422, 431, 433, 343, 388, 389, 391-392, 400, 418, 421-423,
441, 447, 448, 449, 459, 473, 481, 483, 433, 434, 435, 437, 438, 440, 441, 444, 447
494, 496, 505, 506, 507, 509-512, 525, n. 5, 448, 449, 452, 457, 458, 459, 460, 461,
527, 528, 550-551, 536, 590, 608, 639, 651, 462, 464, 465, 466, 468, 475, 478, 479, 481
652, 653, 661, 663, 648, 650, 651, 652, n. 20, 484, 495, 505, 506, 513, 514, 526, 532,
653, 658, 665, 677, 694, 697, 699, 701 534, 540, 547, 548-550, 551, 568, 582-584,
minority/ies xii, xiv, 5, 10, 11, 14, 15, 20- 607, 633, 641, 643, 651, 658, 659, 660, 664,
21, 24, 47, 53, 54, 55, 56, 58, 64, 65, 66, 67, 665, 681, 682, 699
70, 72, 88, 97, 99, 109, 110, 120, 125, 138, Morocco 9, 26
155, 157, 158, 163, 165, 178, 200, 208, 225, Montenegro, see also Union of Serbia
227, 229, 232, 233, 236, 239, 244, 245-246, and Montenegro 22, 23, 196, 197, 244,
261, 262, 263, 265, 266, 267, 268, 270, 271, 399, 400, 447, 560
272, 280, 281, 282, 283, 286, 287, 288, 289, municipal 170, 172, 177, 179, 258, 260,
290, 294, 299, 300, 303, 313, 315, 328, 331, 271, 278, 281, 284, 296-297, 300, 330, 408,
351-352, 396, 403, 416, 419, 426, 434, 438, 410, 411, 412, 413, 417, 418, 419, 425, 426,
441, 442, 448, 449, 451, 452, 453, 454, 455, 427, 428, 434, 435, 448, 454, 464, 518, 522,
456, 458, 459, 463, 464, 465, 466, 489, 491, 562, 568, 635, 679, 700 n. 24
494, 512, 514, 516, 517, 519, 520, 526, 527,
municipality/ies xv, 32, 174, 179, 182, 190,
560, 561, 562, 568, 589, 606, 608, 612, 618,
225 n. 23, 232, 259, 260, 263, 281, 282, 283,
624, 629, 630, 631, 634, 635-637, 638, 639,
287, 288, 290, 296-297, 300, 321, 323, 393,
641, 642, 643, 644, 645, 647, 648, 652, 653,
410, 411, 425, 426, 427, 428, 522, 529, 672
654, 667-688, 694, 695, 696, 697, 698,
699, 700, 701, 706, 708, 717 n. 42 Muslim/s, see also Muslim population
in Bosnia and Herzegovina, Muslim
cultural 65, 100
population in India, Muslim popula-
disenfranchised 8 tion in Lebanon, Muslim population
ethnic 20, 47, 138, 272, 282, 291, 309, 316, in Macedonia, Muslim population in
342, 516, 560, 562, 567, 618, 630, 631, 637, Mindanao, Muslim population in Su-
651, 651, 652, 662, 663, 696, 697, 703 dan 14, 53, 54, 55, 163, 392, 394
language 271, 280-281, 283, 296, 310, 332, Muslim League 14
526, 638, 662
linguistic 20, 631, 637
N
national 20, 21, 47, 58, 67, 98, 99, 100,
231, 313, 317, 318, 332, 642, 672, 696, 697 nation/s xv, 6, 7, 8, 15, 28, 32, 33, 34, 39,
n. 11, 704 42, 48, 49, 57, 61, 63, 98, 109, 156, 162, 172,
protection 54, 55, 99, 143, 198, 232, 396, 187, 193, 195, 199, 203, 205, 208, 209, 210,
639, 641, 673 211, 212, 213, 216, 232, 233, 234, 237, 239,
religion 639, 640 240, 244, 270, 312, 348, 349, 390, 395 n.
20, 451, 488, 514, 546, 619, 696, 697 n. 11,
religious 20, 100, 639, 664
699
Index 781
-building 63, 65, 456 nationalist xi, 6, 7, 11, 14, 16, 61, 63-72,
-state 6, 15, 48, 62, 105, 212, 214, 280, 74-79, 81-83, 87, 88, 90-92, 95-100, 104-
295, 456, 534, 538, 539, 553, 667, 668, 106, 108-111, 113-121, 123, 163, 194, 195-196,
696, 697 203, 204, 205, 208 n. 13, 215, 217, 220, 221,
national, see also ethno-national, na- 222, 223, 224, 227, 228, 230, 231, 234, 235,
tional government, national minori- 237, 240, 307, 314, 347, 348, 349, 350, 353,
ties, national self-determination 6, 354, 360, 419, 455, 456, 472 n 6, 474, 482,
7, 8, 47, 48, 49, 51, 52, 54, 56, 57, 58, 65, 76, 518, 519, 528, 555, 580, 609, 694, 695, 696,
98, 100, 108, 126, 129, 132 n. 7, 135, 136, 698, 700 n. 22, 702, 705, 707, 713, 717, 718
137, 138, 139, 140, 142, 144, 145-146, 147 n. nationality/ies xv, 14, 32, 34, 42, 49, 57,
17, 148, 150, 151, 155, 156, 158, 159, 164, 165, 244, 249, 271, 280, 287, 294, 295, 309, 320,
167, 169, 170, 171, 172, 173, 174, 176, 177, 327, 333, 336 n. 24, 347, 348-349, 397, 514,
178, 180, 183, 184, 191, 194, 195, 199, 201, 516, 519, 527, 628, 643, 646, 654
206, 208, 209, 210, 212, 213, 214, 228, 231, negotiation xiv, xv, 31, 34, 37, 39, 43, 67,
233, 236, 249, 251, 259, 273 n. 18, 275, 276, 69-71, 72 n. 5, 73 n. 8, 76, 79 n. 12, 82, 83,
278, 280, 281 n. 30, 290, 295, 300, 301, 89, 103, 104, 111, 112, 114, 117, 119, 126, 129,
303, 308, 310, 311 n. 6, 312, 319, 320, 322, 132-133, 134, 135, 139, 140, 141, 143, 145,
324, 325, 326, 327, 331, 333, 336, 337, 339, 146, 150, 154, 155, 157, 166-168, 182, 186,
343, 349 n. 5, 354 n. 18, 391, 393, 395, 398, 187, 189, 198, 199, 200, 202, 207 n. 12, 213,
399, 411, 413, 414, 415, 417, 418, 419, 426, 220, 237, 240, 246, 255, 257, 262, 267, 277,
433, 434, 435, 437, 440, 441, 444, 447, 448, 279, 289, 290, 302, 303, 314, 315, 317-318,
449, 452, 455, 456, 457, 459, 464, 468, 489, 335, 338, 341, 345, 348, 352, 354, 356, 357,
503, 506, 507, 508, 509, 510, 512, 513, 514, 360, 361, 362, 363, 364, 366, 371, 372, 373,
517, 519, 538, 539, 549, 550, 551, 553, 556, 375, 378, 388, 396, 398, 404, 405, 414, 416,
561, 562, 563, 564, 567, 569, 571, 574, 582, 419, 421, 427, 436, 439, 445, 469, 470, 472,
589, 590, 600, 602, 608, 611 n. 40, 617, 474, 476, 478, 480, 481, 482, 483, 486, 488,
630, 633, 635, 638, 640, 641, 645, 648, 658, 494, 495, 496, 498, 515, 526, 544, 555, 580,
660, 672, 675, 682, 683, 685, 693, 698, 699, 583, 585, 586, 589, 590, 591, 592, 593, 594,
700, 704, 705, 717 n. 43 595, 596, 602, 603, 614, 621, 623, 624, 627,
community/ies 48, 49, 73, 97, 98, 100, 633, 624, 665, 682, 683, 694, 702, 707, 711,
110, 205, 208, 210, 212, 248, 249, 250, 712, 713, 714
251, 255, 403, 539, 565, 640, 643, 660, Netherlands 49, 50, 55, 468, 691, 692,
678, 679, 697, 699 693, 696
culture 6, 15, 50, 251, 320, 638 New Zealand 53, 133-134, 145 n. 16, 149,
identity 6, 10, 98, 193, 251, 320, 428, 531, 413, 454, 470, 472, 473, 477, 479, 481, 482,
678, 679, 699, 706, 717 496, 497, 534, 564, 569, 693, 694
liberation xi, xii, 29-30, 33 n. 7, 34, 65, Nigeria 9, 19, 456, 457, 576
164-166, 266, 416, 427, 473, 533, 563, 588, Norway 9, 470
590, 628, 663
North Atlantic Treaty Organization
security 11, 418, 574, 583, 590, 622, 645, (NATO) 36, 201, 214, 218, 243, 244, 245,
683, 685 254-255, 257, 268, 276, 277, 278, 290, 291,
sovereignty 9, 135, 581, 582, 667 292, 293, 294, 295, 301, 302, 304, 605 n.
sub- 136, 447, 458-459, 509, 517, 519, 549- 69, 335, 403, 409, 470, 474, 475, 477, 478,
550, 635, 667-688 482, 485, 486, 487, 488, 489, 490, 497, 577,
nationalism xi, 5-16, 63, 83, 123, 172, 197 584, 585, 586, 589, 590, 594
n. 5, 269 n. 8, 308, 513, 518, 539, 548, 705, Northern Ireland, see also Ireland xiii,
717 xiv, xv, 31, 42, 43, 54, 55, 56, 58, 61-124,
ultra- 9 203, 400-402, 407, 408, 410, 413, 414,
782 Index
419-421, 426, 430, 431, 432, 433, 453, 436,

P
437, 438, 439, 440, 443, 445, 446-447,
448, 449, 452, 456, 458, 459, 460, 461, Palestine/ian 25, 225, 468, 470, 653, 696
462, 464-465, 466, 468, 470, 471-472, 476, Pakistan 12, 14, 19, 29, 30
477, 478, 479, 481, 482, 483, 485, 494, 495,
Papua New Guinea xvi, 125-159, 399,
496, 497, 498, 505, 506, 523-524, 525, 526,
413-415, 430, 435, 438, 449, 452, 457, 459,
527, 528, 529, 531, 532, 535, 537, 538, 541,
461, 462, 463, 465, 472, 505, 506, 507, 534,
543, 550, 551-556, 560-561, 562, 568, 572,
541, 546 n. 11, 551, 563, 581, 596, 602, 608,
591-593, 596, 597, 601, 602, 603, 064, 607,
682, 699-700, 701, 703, 709
608-610, 611, 612, 613, 614, 615, 618, 621-
622, 623, 628, 633, 641, 643, 649, 651, 658, Paraguay 9
660, 662, 684, 685, 693-694, 695, 696, paramilitary/ies 65, 66, 67, 68, 70, 77,
698, 701, 702-703, 705, 708-709, 710, 711, 80, 96, 100, 111, 112, 113, 118, 119, 121, 197,
712-713, 714, 717, 718 224, 245, 253, 268, 269 n. 7, 302, 312, 314,
Anglo-Irish Agreement 1985 69, 72, 110- 353, 485, 487, 495, 531, 535 n. 4, 550, 552,
111, 113-114, 121, 471, 479, 684, 698 572, 573, 574, 476, 582, 584, 586, 587, 589,
591, 592-593, 597, 610, 613, 651, 662, 694,
Good Friday (Belfast) Peace
695, 701-703, 711
Agreement 1998 xiv, xv, 43, 69-121, 400-
partition 12, 14-15, 48, 58, 64, 65, 114, 121,
402, 419, 471, 482, 495, 535 n. 4, 543,
292, 389, 419, 434, 538, 711, 719
551-556, 592, 603, 609, 614, 618, 633, 639,
640, 641, 644, 647, 648, 649, 660, 684, party/ies (political) 5, 15, 48, 51, 52, 65,
685, 694, 695, 700, 705, 707 66, 68, 70, 71, 74, 75, 78, 79, 80, 81, 82, 84,
86, 87, 88, 89, 90, 91, 92, 95, 102, 104, 114,
119, 120, 121, 171, 181, 194-196, 200, 210 n.
O 16, 218, 220-221, 223, 234, 246, 261, 262,
269 n. 9, 273-274, 275, 282, 285, 288, 294,
opposition 14, 15, 18, 22, 24, 32, 50, 51,
300, 302, 312, 318, 329, 333, 335, 336, 349,
65, 84, 115, 131, 135, 156, 162, 165, 167, 175,
357, 397, 433, 454, 455, 456, 460, 461, 462,
181-183, 186, 217, 221, 239, 240, 245, 253,
463, 464, 548, 575, 592, 593, 594 n. 4, 695,
262, 273, 293, 295, 312, 318, 336, 339, 341,
705, 706, 707, 708, 710 n. 32, 712, 716
354, 396, 487, 492, 524, 550, 593, 663, 692,
707, 708 list 52, 87-88, 90, 298, 454, 457, 461
Organization for Security and Coopera- multi- 51, 67, 72 n. 5, 114, 222, 342, 454,
tion in Europe (OSCE) 19, 214, 217- 462, 693 n. 1
218, 233, 246, 254, 255, 265, 266, 267, 271, system 52, 87, 88, 89, 407, 433, 434, 448,
276-277, 278, 295, 299, 302, 310, 313 n. 8, 451, 454, 461, 463, 464, 693 n. 1
316, 317, 320 n. 12, 325, 328, 333, 334, 335, peace xiii, 5, 13, 16, 18, 24, 37, 45, 111, 119,
336, 339, 340, 341, 352 n. 12, 355, 356-358, 136, 146, 162, 165, 166, 168, 169, 180, 181,
364, 365-366, 367-368, 372, 373, 375, 376 182, 183, 184, 186, 187, 188, 191, 192, 194-
n. 53, 379, 390, 391, 392, 397, 428, 249, 201, 202, 203, 215, 216, 235, 236, 257, 275,
440, 465, 466, 470, 474-475, 477, 478, 481, 297, 301-304, 345, 347, 348, 352, 354, 355,
483, 484-485, 488, 489, 490, 491, 496, 497, 356, 362, 363, 367, 369, 376, 378, 392, 410,
520, 548, 562, 565, 567, 569, 583, 584-585, 413, 416, 439, 453, 455, 467, 468, 472, 473,
586, 589, 595, 617, 618, 620, 631, 638, 641, 474, 476, 478, 479, 481, 482, 483, 485, 486,
643, 648, 649, 658, 661, 662, 663, 664, 493, 494, 497, 504, 510, 515, 518, 521, 526,
665, 693, 695 n. 6, 715, 717 n. 44, 718 527, 530, 551, 559, 560, 568, 569, 571, 579,
High Commissioner on National Minori- 584, 587, 603, 623, 624, 627, 694 n. 4, 700-
ties (HCNM) 271, 276, 277, 313 n. 8, 703, 704, 710, 711, 718
316-318, 333, 340, 475, 477, 483, 649, 663 agreement, see also Bougainville Peace
Agreement, Dayton Peace Agreement,
Index 783
Good Friday (Belfast) Peace Agree- rights of 20, 21, 26, 27, 32, 65, 100, 397,
ment, Mindanao Peace Agreement, 629-630
Sudan Peace Agreement 439, 525, self-determination of 20, 21-22, 23, 26,
526, 661 27-28, 32, 37, 65, 100, 387, 397, 629-630
process 61-124, 125, 132-135, 137, 139, Philippines iii, 9, 33 n. 7, 161-192, 392,
146, 149-150, 151, 156, 157, 159, 167, 183, 416, 417, 418, 435, 437, 438, 444, 447, 452,
185, 186, 187, 211, 216, 219, 221, 237, 239, 454, 457, 458, 459, 460, 462, 464, 465,
240, 348, 353, 356, 357, 358, 360, 364, 466, 468, 473, 505, 506, 507, 509, 511, 529,
365, 366, 367, 379, 443, 475, 478, 482, 534, 541, 546 n. 11, 550-551, 563, 590, 602,
484, 486, 488, 493, 494, 495, 497, 548, 607, 608, 612, 640, 641, 645, 651, 653, 658,
552, 568, 664, 685, 701, 702, 711, 712, 714 665
settlement 24, 111, 187, 193, 216, 453, 560, plebisicite, see also referendum/a 7, 21,
565, 568, 587, 693, 698 33 n. 7, 165, 170, 171, 186, 188, 376, 392,
peaceful 35, 36, 37, 49, 65, 70, 71, 83, 109, 642
120, 164, 172, 177, 188, 189, 245, 257, 259, plurality 51, 53, 459, 460, 706, 708
301, 304, 307, 311, 335, 336, 342, 347, 350, polls/ters 15, 115, 182, 196, 268
354, 355, 376, 377, 416, 426, 479, 480, 496,
population/s xi, xv, xvi, 5, 6, 8, 9, 10, 11,
532, 543, 551, 559, 563, 584, 592, 594, 667,
12, 17, 20, 21, 23, 24, 25, 26, 27, 29, 30, 31,
672, 684, 700
32, 33, 37, 40, 41, 42, 43, 45, 53, 54, 64, 65,
peacekeeping 12, 19, 26, 193, 214, 215, 67, 93, 97, 125, 131, 132, 137, 138, 155, 156,
255, 308 n. 2, 311, 356, 357, 358-359, 361, 157, 162, 163, 167, 172, 174, 178, 182, 186,
363, 365, 367, 409, 429, 439, 472, 473, 475, 189, 197, 201, 225, 228, 232, 238, 239, 240,
478, 486, 487, 491, 559, 567-568, 569, 594, 244, 245, 248, 262, 263, 266, 268, 269,
694 270, 271, 272, 273 n. 19, 274, 275-276, 277,
people/s xii, xv, xvi, 6, 7-8, 11, 12, 14, 15, 278, 280, 281, 282, 284, 286, 287, 288, 289,
17, 18, 21-22, 25, 26, 27, 28, 32, 33, 34, 39, 290-291, 292, 293, 295, 297, 298, 300, 301,
42, 43, 44, 49, 53, 57, 58, 61-64, 66, 83, 84, 302, 303, 304, 307, 309, 310, 311, 319, 320,
97, 98, 100, 101, 109, 121, 123, 125, 128, 129, 323, 326, 327, 334, 343, 349, 351, 352, 353,
130, 131, 137, 138, 151, 152, 154, 161, 164, 369, 370, 388, 389, 392, 400, 401, 405,
165, 170, 173, 174, 176, 177, 178, 179, 182, 409, 412, 413, 416, 420, 421, 426, 440-442,
183, 184, 185, 186, 187, 188, 189, 195, 197, 455, 492, 504, 505, 507, 508, 509, 510, 513,
206, 207, 209, 210, 212, 213, 225, 228, 232, 514, 516, 520, 522, 523, 525, 527, 528, 529,
239, 252, 257, 259, 270, 272, 275, 278, 280, 539, 541 n. 6, 543 n. 10, 544, 550, 551, 560,
290, 295, 303, 305, 309, 311, 312, 313, 314, 565, 569, 581, 584, 587, 588, 591, 594, 622,
317, 318, 319, 320, 321, 322-332, 333, 337, 623, 629, 640, 643, 645, 646, 651, 660,
338-341, 343, 348, 349, 352 n. 13, 397, 391, 670, 676, 693 n. 1, 697, 702, 705, 715, 716,
392, 395, 398, 400, 401, 402, 403, 404, 411, 718 n. 45
413, 414, 416, 417, 420, 421, 422, 426, 428, Portugal 9, 25, 26, 544, 545
440, 443, 445, 450, 452, 458, 509, 513, 524,
power, see also power-sharing xii, xiii, 6,
527, 529, 548, 549, 550, 551, 555, 556, 559,
14, 15, 30, 47-49, 56, 57, 65, 66, 67, 71, 76,
560, 561, 564, 565, 568, 581, 583, 587, 594,
85, 89, 93, 95, 97, 98, 99, 102, 103, 104, 105,
613, 628, 631, 640, 641, 643, 653, 654, 660,
115, 116, 117, 129, 136, 137, 138, 139, 140,
661, 664, 669, 681, 699, 700, 708, 710, 711,
141, 142, 143, 144, 145, 146, 147, 151, 152,
716, 718
163, 165, 174, 179, 187, 199, 203, 204, 205,
colonial 29 207, 210, 211, 213, 216, 219, 222, 223, 224,
ethnic xi, 37 228, 229, 230, 235, 240, 248, 249, 251, 258,
indigenous xii, 21, 631, 640, 642 260, 263, 278, 284, 286, 288, 297, 307, 315,
316, 321, 322, 323, 326, 329, 333, 342, 348
784 Index
n. 2, 349, 350, 352, 353, 354, 356, 360, 362, executive ix, 50, 51, 55, 72, 76-86, 285,
377, 389, 392, 396, 398, 407, 410, 412, 415, 287, 288, 407, 418, 421, 454, 482
416, 419, 425, 433, 434, 435-436, 440, 444, federative 49
446, 447, 448, 449, 452, 453, 487, 488, 491, formal 288, 365, 390, 407, 434, 447, 448
492, 515, 520, 529, 542, 547, 550, 554, 561,
governmental xiv
568, 572, 580, 584, 588, 597, 599, 611, 615,
619, 621, 622, 628, 640, 643, 645, 663, 674- horizontal 348, 363, 376, 377, 379, 407,
676, 677, 680, 681, 682, 683, 684 n. 27, 408, 409, 412, 432-434, 438, 439, 441,
685, 696, 697, 702 n. 26, 710 447, 448, 449
balance of 113-114, 215, 412, 432 informal 274, 285, 287, 288, 345, 346, 355,
376, 390, 407
colonial xii, 127, 615
institutions 240, 408, 410, 419, 420, 431,
competition for 47-48, 50, 57
435, 436, 437, 439, 440, 442, 443, 446,
division of 47-48, 320-323, 412, 418 447, 448, 448, 449, 535 n. 4, 541, 543, 714
economic 342 integrative 204, 692, 703-718, 719
executive 93, 208 internal 71, 543, 696
imperial 10 local 68, 491
military 47, 358, 582, 590, 596 mandatory 433
political 47, 50, 54, 244, 259, 342, 440, national 433, 434, 448
686
regional 332, 433, 434, 435, 447, 448, 698
public xiv, 157, 244, 206, 263, 391
settlements ix, 370, 376, 387, 390, 451,
power-sharing, see also complex power- 456, 469, 685, 692-696
sharing, consociational power-shar-
territorial 56, 58
ing xiv, 45, 47-58, 66, 67, 68, 69, 83, 85,
86, 122, 123, 137, 157, 158, 168-179, 193-241, theory 211, 238-239, 346, 691-719
244, 263, 265-305, 308, 311, 335, 336-341, vertical 408, 409, 432-434, 438, 439, 441
342, 343, 346, 347, 361, 363, 364, 371, 372, voluntary 68
373, 376, 377, 379, 397, 407-450, 452-453, presence
454, 458, 468-469, 474, 475, 476, 487, 488,
civil/ian 193, 240, 256, 257, 584, 586, 587,
490, 491, 492, 498, 499, 515, 517, 530, 533,
588
548, 551-556, 572, 588, 594, 597, 599, 652,
654, 691, 692, 693, 696, 697, 700, 701, 703, EU 299
712, 714 international xiv, 19, 193, 214, 215, 218,
agreements 138, 195, 262, 289, 314-318, 219, 232, 240, 256, 302, 439, 444, 485,
319, 342, 343, 476, 483, 530, 584, 599, 631, 581, 586, 587, 596
632, 633, 637, 638, 644, 647, 648, 649, military/security xiv, 19, 67, 131, 214, 215,
659, 660, 661, 668 216, 240, 256, 303, 358, 439, 587, 588, 711
arrangement(s) 126, 137, 158, 159, 193, NATO 301, 302, 304, 487 n. 26, 580, 581,
203-220, 238, 239, 240, 259, 263, 285, 584, 711
286, 287, 288, 291, 293, 304, 305, 308, OSCE 391, 397
311, 312, 319-334, 342, 346, 352, 355, 361- UN 214, 215, 218, 243, 389, 584
375, 378, 390, 398, 433, 452, 454, 469, presidency/ies 35, 36, 37, 51, 57, 78, 167,
492, 493, 499, 530, 535, 542, 543, 548, 181, 187, 194, 195, 199, 207, 208-209, 211,
569, 572, 582, 583, 589, 594, 595, 596, 567, 212, 213, 216, 221, 228, 229, 244, 249, 261,
631, 633, 637, 650, 683, 693 n. 2, 705, 711, 262, 411, 473, 704, 705, 706, 713, 715
713
president/s 12, 34, 50, 54, 56, 93, 133, 164,
devices/mechanisms xv, 41, 63 n. 8, 204, 166-168, 169, 170, 171, 176, 177, 178, 179,
211, 212, 287, 378, 419, 441, 668 180, 181, 182, 186, 187, 188, 190, 191, 195,
Index 785
197 n. 4, 198, 202, 208, 212, 213, 216, 221,

Q
248, 249, 250, 261, 262, 273, 275, 276, 277,
278, 279, 286, 287 n. 37, 289, 292, 294, Quebec 13, 39
297, 302-303, 309, 310, 311, 312, 317, 319, Quebecois 53, 636, 652
322, 323, 329, 330, 335, 337, 338, 339, 340,
quota/s, see also affirmative action 53,
341, 343, 347 n. 2, 353, 354, 355 n. 20, 359
54, 86, 88, 89, 95, 115, 185, 261, 262, 272,
n. 23, 370 n. 43, 372, 391, 392, 395, 397,
333, 461, 516, 517, 608, 618
398, 400, 405, 411, 418, 422, 425, 426, 427,
456, 457, 459, 460, 471, 472, 479 n. 17, 486,
585, 691, 694 n. 4, 705, 715, 716 R
presidential 51, 76, 179, 181, 188, 208,
racist 25, 49
220, 258, 274, 354, 376 n. 53, 42, 460, 574
rebel/s xii, 18, 23, 30, 42, 130, 167, 191,
principality xv, 32
416, 481, 562, 696, 701, 702
Protestant/s 52, 63, 65, 66, 67, 74, 94, 97,
rebellion 12, 18, 30, 75, 178, 189, 590
100, 113, 115, 456, 463, 524, 560, 561, 591,
618, 645, 696, 705 recognition, see also international recog-
nition 5, 7, 11, 21, 24, 28, 30, 34, 38, 41,
public 13, 54, 55, 84, 97, 98, 104, 106, 113,
48, 57, 64, 100, 101, 113, 115, 131, 167, 198-
119, 121, 128, 167, 171, 173, 174, 179, 182, 185,
199, 206 n. 10, 212, 232, 233, 239, 270, 277,
204 n. 9, 229, 251, 275, 282, 283, 284, 286,
393, 401, 454, 456, 498, 518, 544, 573, 613
287, 288, 294, 303, 310, 321, 323, 324, 325,
n. 47, 627, 628, 630, 636, 637, 638, 640,
327, 328, 336, 352, 354, 356, 357, 361, 363,
642, 643, 647, 652, 661, 670, 678, 679, 681,
370, 375, 376, 377, 378, 424, 425, 427, 441,
699, 700
464, 510, 512, 513, 539, 542, 543, 546, 549,
550, 565, 566, 574, 604, 607, 610, 611, 612, referendum/a, see also plebiscite xv,
613, 618, 621, 623, 630, 638, 639, 668 n. xvi, 10, 17, 23, 25, 26, 33, 35, 39, 42, 43,
1, 669, 670, 672, 674, 676 n. 14, 699, 708, 44, 70, 72, 73, 79, 100, 101, 102, 103, 104,
709 n. 31 108, 112, 115, 117, 119, 125, 126, 135-137, 145,
146-147, 149, 150, 157, 158-159, 197, 198,
authority/ies, see also public power xiv,
202, 220, 252, 293-294, 311, 318, 321, 323,
98, 158, 243, 259-261, 393, 398, 409, 411,
325, 332, 333, 338, 339-340, 343, 389, 396,
412, 414, 415, 416, 418, 419-420, 421, 422,
397, 398, 399, 400, 401, 403, 404, 413-414,
423, 424, 425, 427, 428, 435, 437, 476,
416, 417, 420, 422, 423, 424, 426, 434, 436,
484, 635, 648, 651
441, 445, 446, 448, 449, 538, 603, 614, 642,
expenditure 54, 538, 544, 550, 552 694, 698, 699-700, 708, 709
funds 54, 343 refugee/s 131, 164, 206, 215, 219, 220, 221,
institutions 229, 251, 648, 650-651 224, 225-228, 230, 231, 233, 240, 257, 266,
life 20, 54, 283, 296, 419, 639 269 n. 7, 276, 299, 301, 350, 351, 352 n. 13,
officials 47, 121, 262, 647 355, 362, 363, 364, 374, 375, 416, 452, 470,
opinion 15, 115, 222, 420 474, 475, 488, 489, 490, 491, 567, 587, 606,
607, 612, 613, 637, 645-647, 651, 652, 660,
order 325, 334, 468, 491, 583, 645 662, 663, 664, 665, 695 n. 6, 711 n. 33, 712
policy 48, 69, 74, 108, 260, 422, 424, 503, representation 52-54, 57, 76, 79 n. 12, 87,
538, 583, 628, 650, 706 88, 139-140, 155, 178, 183, 188, 200, 202,
safety/security 82, 83, 177, 253, 256, 587 208, 209-210, 212, 236, 238, 244, 245-246,
sector 54, 72, 86, 92-97, 138, 153, 503, 541 248-249, 251, 261-262, 263, 272, 282-283,
n. 6, 543, 545, 546, 552 284, 287, 294, 303, 332, 336, 357, 358, 362,
services xiv, 55, 92, 138, 140, 141, 142, 151, 377, 392, 393, 394, 400, 407, 434, 437, 448,
155, 283, 288, 296, 427, 543, 552, 589, 609 452, 453, 454, 457, 458, 460, 461, 463, 464,
786 Index
485, 511, 562, 589, 608, 648, 649, 661, 664, group 20-21, 109, 232, 313, 629-630, 635,
679, 706 719
effective 40 indigenous 21, 172, 177, 178
federal 36 individual 109, 110, 629-630, 635, 642
over- 53, 54, 454, 463, 464 land 21
parliamentary xiv, 139 language 5, 271 n. 13, 280-281, 327, 332,
political 199, 213, 226, 245 334, 393, 526, 637-638, 639
proportional 51, 53, 54, 56, 80, 86, 203, media 271
214, 286, 288, 296, 298, 595, 628, 705 of consultation 69
regional xiv of due process 618, 623
under- 54, 88, 517, 662 of free association 14
repress/ion xii, 18, 29, 30, 36, 63, 66, 244, of free speech 14, 623
424, 587, 599, 600, 622, 623-624 of return (refugees) 211, 225, 232, 254,
repressive 245, 376, 609 613, 637, 646-647, 712
Republic Srpska 24, 194, 195, 197, 201, of women 172, 178
220, 223, 232 n. 26, 233, 394, 410, 412, 447, political 97, 231, 503, 617, 621, 629, 630-
455, 488, 518-520, 542, 580, 608 n. 32, 647, 631, 632, 636, 637, 641, 643-644
651, 653, 663, 680, 697, 698, 704 pre-state 35, 40
republican 43, 64, 65, 66, 68, 70, 71, 92, property 225, 324, 351, 537, 538, 629, 630,
104, 111, 112, 113, 114, 115, 117, 118, 119, 123, 632, 647
194, 423, 472, 552, 592, 694, 695, 696, 700
religious 231, 631, 639-640
n. 22, 703, 711, 717 n. 42
social 97, 98, 362, 621, 624, 629, 631-631,
right(s), see also human rights and mi-
632, 636, 637, 639, 644-645
nority rights
autonomy 313, 316, 350, 440 n. 3, 645 to a fair trial 618, 623, 630
citizens’ 250, 320, 327, 634, 635 to education 271, 327, 332, 393, 516, 526,
625, 631, 632
civil 65, 68, 97, 325, 397, 611, 617, 621,
625, 628, 629, 630-631, 632, 636, 637, to health care 625, 631
641, 643-644 to housing 625, 631, 632
collective 15, 98, 99, 110, 272, 642 to legislative initiative 332, 325
consultation 69, 248 to legislative power 332
cultural 5, 97, 98, 271, 334, 503, 621, 624, to liberty 630
629, 631-631, 632, 637-643 to life 630, 632
community 1, 55, 178, 203, 251, 260, 261, to participation xii, 20, 140, 630, 672
300, 393, 490, 521, 525, 631, 635, 640, to privacy 630
648, 649, 680
to secede/ssion 9, 13, 20, 23, 31, 32, 33,
democratic governance 20 34, 36, 57, 100, 323, 394-395, 399, 423,
economic 98, 503, 621, 624, 629, 630- 699-700
631, 632, 636, 637, 639, 644-645 to security of the person 630
electoral 99, 630 to self-determination xii, xv, xvi, 5, 6, 8,
equal 98, 213, 248, 285, 419, 636-637, 645 9, 13, 17, 20-31, 32, 33, 34, 36, 37, 38, 40,
ethnic 98, 300, 313, 634, 717 42, 43, 44, 65, 100-101, 109, 161, 195, 196,
fishing 415 244, 246, 315, 318, 320, 335, 349 n. 4, 387,
390, 391, 394-395, 397, 398, 400-401,
fundamental 14, 178, 249 n. 5, 599, 623,
405, 467, 629, 630, 641-643, 653, 654,
653
682, 699-700
Index 787
to social security 631

S
to territorial integrity 35, 246
to use force 45 Sabah 9, 164
to work 631 Scotland xv, 39, 108, 419, 538, 552, 709
veto xiv, 56, 72, 86, 99, 203, 210-211, 212, secede, see also right to secede 19, 23,
214, 228, 285, 286, 287, 606, 612 24, 35, 39, 42, 193, 205, 265 n. 1, 337, 449
voting 49 secession, see also right to secession xi,
xii, xv, xvi, 9, 10, 13, 15, 16, 18, 19, 22, 23,
regimes ix
24, 28, 31, 32, 33, 34, 35, 38, 39, 40, 41, 45,
Roma 213, 244, 261, 266, 270, 272, 280, 57, 58, 125, 126
282, 286, 289, 294, 299-300, 303, 309 n. 4,
secessionism/t
426, 455, 458, 566, 584, 636, 662
ix, xi, xii, xiii, xiv, xvii, 5, 9, 12, 13, 18, 19, 30,
Romania/n xvi, 32, 43, 277, 309, 310, 311,
31, 37, 38, 41, 125, 126
315, 318, 320, 335, 338, 400, 421, 449, 513,
540, 547, 548, 682, 699 security xiv, 10, 11, 13, 56, 66, 73, 95, 113,
122, 148, 149, 170, 172, 178, 192, 203, 210,
rule (government) xiii, 10, 12, 14, 16, 63,
218, 222, 224, 227, 228, 253, 255, 256, 263,
64, 66, 68, 96, 127, 129, 197, 273, 339, 347,
276, 279, 286, 287, 294, 301, 302-303, 304,
348, 349, 354, 421, 423, 518, 543 n. 9, 696,
322, 324, 330, 332, 339, 340, 351, 355, 357,
717 n. 43
362, 364, 367-368, 389, 390, 393, 417, 429,
authoritarian 9 441, 467, 471, 474, 475, 486, 491, 497, 531,
civilian 11 452, 561, 564, 572-577, 579-581, 585, 587-
colonial 44 588, 590-592, 594-596, 621, 630, 662, 664,
democratic 9, 11, 599 684, 695, 701, 702, 703, 707, 711, 719
home 63-64 force/s 66, 177, 266, 268, 292, 413, 424,
majority 15, 56, 68, 74, 599, 707 n. 30 475, 573, 574, 577, 585, 588, 590, 651, 662,
701, 702, 711 n. 34
minority 55
institutions 54
of law
national 11, 574, 583, 590, 622, 645, 683,
43, 219, 235-236, 340, 342, 437, 484, 490,
685
491, 565, 599-626, 652, 662, 675, 687
policies 68, 578, 592
right of governments to 12
sector xvii, 54, 571-597, 701
self- 315-316, 422
self-determination, see also right to self-
Russia/n Federation, see also USSR xii,
determination xi-xvii, 5-16, 17- 45, 64,
8, 22, 25, 36-37, 38-39, 42, 199, 254, 262,
74, 100, 102, 104, 109, 125, 135, 137, 147,
308 n. 2, 310-311, 312, 316, 320 n. 12, 332,
155, 158, 193, 196, 197, 201, 239-241, 245,
334, 335, 336, 347, 348, 350, 355-357, 358-
246, 311, 320, 346, 349, 387, 390, 391, 392,
360, 362, 636-364, 365, 366-367, 369, 374
394, 396, 402, 403, 405, 407-409, 468, 531,
n. 50, 389-390, 391, 392, 397, 405, 421,
551, 558, 623, 638, 641-643, 653, 669, 670,
428-429, 470 n. 4, 474, 475, 476, 477, 483,
694, 697, 700, 703, 704, 706, 711, 718
486, 490, 534, 540, 547, 548, 568, 594-595,
678 agreement 61-124
Russian 254, 309, 310, 312, 318, 326, 327, classical xii, 19, 20-31, 33, 42
335, 336, 338, 350, 353, 354, 355 n. 20, 358, colonial xii, 10, 12, 13, 24, 25-26, 28, 29,
359, 360, 364, 365, 366, 369, 370, 374-375, 33, 42
389, 396, 429, 438, 475, 478, 487, 491, 494, conditional xv-xvi, 32-33, 43
513, 514, 540, 586, 594-595, 660 conflict/struggle, see also self-determi-
Rwanda 11, 14, 577, 610, 616 nation disputes ix, xi, xiii, xv, 18, 19,
29, 30, 31, 42, 44, 45, 156, 159, 244, 346,
788 Index
405, 407, 408, 417, 419, 423, 431, 434, 584, 585, 586, 587, 622, 704, 715-716, 717
436, 437, 438, 442, 443, 444, 447, 449, n. 43
467-499, 531, 532, 533, 534, 541, 545 n. 11, settlement/s, see also conflict settlement,
547, 556, 557, 571, 584, 602, 629, 647 consociational settlement, dispute
constitutional xv, 19, 23, 25, 26, 28 n. 4, settlement, power-sharing settlement,
29, 31-40, 42, 45, 388, 394, 396, 401, 402 self-determination settlement ix, x,
disputes 5, 13, 17- 45, 68, 125-159, 161- xiii, xiv, xvi, xvii, 14, 19, 24, 31, 37, 38, 42,
192, 193, 194, 196, 197, 238, 308-311, 347, 43, 44, 45, 49, 63, 65, 67, 68, 69, 70, 71, 80,
348, 387-406, 413, 414, 416, 438, 531-557, 89, 90 n. 21, 106, 111, 112, 114, 115, 116, 120,
572, 581, 582, 592, 616, 628, 630, 633, 635, 123, 125, 126, 130, 132, 133, 138, 178, 182,
637, 643, 644, 646, 647, 649, 650, 652, 187, 193, 200, 216, 225, 236, 238, 243, 257,
654, 667-688, 696-700, 703 305, 308, 313, 317, 319, 331, 332, 345, 346,
entity/unit xi, xv, 24, 26, 28 n. 4, 29, 42, 351, 354, 356, 361, 366, 368, 376, 378, 379,
43, 64, 259, 387, 400, 401, 405 387, 388, 389, 391, 392, 393, 394, 396, 397,
398, 399, 400, 402, 403, 404, 406, 409,
external xvi, 20, 21-22, 28, 32, 33 n., 7,
427, 429, 437, 441, 442, 444, 447, 448,
42, 43, 268, 318, 320, 335, 388, 389, 391,
450, 453, 468, 471, 472, 473, 474, 476, 477,
393, 641, 654, 660, 676, 682, 699
478, 480, 481, 482, 484, 486, 494, 495-496,
internal 20, 28, 642, 652, 697 497, 498, 499, 527, 548, 559, 560, 562, 563,
movements xi, xii, 19, 421 565, 568, 571, 581, 584, 585, 586, 587-588,
national 6, 9, 48, 65, 68, 73, 101, 121, 334, 590, 592, 593, 594-595, 596, 597, 618, 623,
696 653, 664, 669, 686, 693, 694, 695, 697,
settlements ix, xiii-xiv, 19, 42, 73, 100- 698, 701, 702, 706-707, 709
104, 388 autonomy x, xiv, xvi, 33 n. 7, 41, 391, 405,
separation (of a state) 9, 13, 22-23, 45, 706
126, 128, 156, 163, 221, 543, 587, 635, 678 constitutional 18, 44, 88
separatism/t 10, 18, 309, 310, 347, 467, final xv, 243, 252, 256, 262, 367, 402-403,
468, 495, 498, 548, 652, 653, 654 428, 444, 472, 491, 713
Serbia/n, see also Serb population in interim xvii, 246, 247, 404, 694
BiH, Serb population in Croatia, Serb internationalised 41-44, 396
population in Kosovo, Serb population political 67, 69, 111, 114, 116, 125, 132, 133,
in Macedonia, Union of Serbia and 135 n. 12, 151, 245, 256, 317, 346, 362, 363,
Montenegro 11, 19, 22, 25, 34, 36, 194, 364, 366, 367, 377, 389, 390, 483 n. 22,
195, 196, 197, 198, 200, 202, 220, 224, 237, 492, 498, 602, 613, 694, 711
243, 244, 246, 247, 249, 251, 254, 255, 258,
status 31, 431, 445
262, 263, 266 n. 4, 268, 269, 292, 396, 399,
402, 403, 404, 405, 409, 423, 424, 447, Singapore 9, 22, 40, 550
467, 486, 495, 496, 520, 534, 541, 542, 543, Slovak Republic 9, 22, 281 n. 30, 547, 642
565, 566, 579, 585, 586, 636, 678, 680 Slovenia/n 23, 34-36, 38, 45, 194, 198,
Greater 197 244, 395, 541 n. 7, 547, 642, 667, 670
Serbia and Montenegro, Union of 19, Somalia 11, 29, 166
22, 42, 243, 396, 399-400, 426, 540 n. 5, Somaliland 24, 29, 40
567, 641, 642, 658, 659 sovereignty, see also sovereign states xvi,
Serb/s 29, 195-198, 200-202, 208, 209 n. 3, 4, 5, 6, 7, 11, 13, 33, 36, 37, 38, 41, 49, 55,
15, 213-214, 215, 220, 222, 226, 227, 228, 63, 65, 67, 70, 73, 98, 101, 104, 108, 110,
230, 232, 237, 240 n. 27, 245-250, 254-255, 121, 123, 129, 135, 143, 157, 162, 197, 203,
258-259, 261, 262, 263, 403, 455, 458, 459, 206, 237, 246-247, 256, 257, 279, 310, 340,
463, 464, 489, 491, 496, 520, 532, 565, 579, 388, 389, 392, 393, 394, 395 n. 16, 401, 402,
Index 789
403, 404, 413, 422, 423, 467, 547, 580, 581, national 144, 173, 327
582, 586, 589, 602, 603, 619, 628, 639, 641, state/s, see also nation/s ix, xii, 17-18, 19,
651, 665, 660, 667, 678, 680, 685, 695, 696 20, 21, 22, 24- 27, 29-39, 40, 44, 45, 48, 49,
n. 14, 697, 698 51, 52, 55, 57, 58, 61, 63, 65-68, 73, 89 n. 20,
South Africa/n 8, 25, 44, 71, 72 n. 7, 453, 96, 98, 100, 101, 108, 110-111, 117, 121, 124,
468-469, 472, 578, 611, 706, 708 126, 127, 128, 130, 134, 140, 146, 150, 155,
South Ossetia 19, 345-381, 389-390, 408, 156, 161, 163, 164, 173, 182, 186, 187, 193,
413, 426, 428-429, 430, 432, 433, 435, 437, 194, 195, 196, 197, 199, 200, 203, 205, 206,
438, 439, 440, 444, 445, 446, 447, 452, 207, 208, 209, 210-211, 212, 213, 215, 218,
464, 475, 476, 477, 478, 479, 481, 482, 483, 219, 221-222, 223, 224, 227, 228, 229-230,
484, 487, 490-491, 495, 505, 506, 515-516, 234-235, 236-237, 238, 239, 240, 251, 254,
528, 547-548, 572, 594-595, 596, 605, 628, 256, 260, 265, 266, 268, 270, 271, 272, 275,
638 n. 1, 650 n. 2, 658, 660, 661, 664, 678 278, 280-281, 283, 284, 285, 286, 287, 290,
South Tyrol xiii, 10, 203, 316, 468, 676 291, 298, 300, 304, 307, 309, 310, 315, 316,
320, 321, 326, 327, 328, 329, 332, 336, 337,
Southern Rhodesia 24
338, 341, 342, 359 n. 22, 365, 366, 367, 374,
Soviet 8, 12, 310, 312, 314-315, 321, 327, 377, 387, 388, 391, 392, 393, 394, 396, 397,
329, 332 n. 19, 347, 348-349, 351, 353, 358 398, 399-400, 403, 405, 407, 409, 410,
n. 21 412, 413, 414, 421, 423, 426, 427, 428, 429,
post- 307, 320 n. 13, 335, 347, 360, 377, 431, 432, 438, 440, 441, 442, 443, 444,
378, 421, 428, 429, 515, 532 n. 1, 548, 582, 445, 446, 452, 453, 454, 457, 458, 463, 465,
583-584, 631 467, 468, 470-471, 473, 474, 476, 487, 488,
Soviet Union see also USSR xiii, 8, 10, 495, 503, 506, 511, 516, 517, 518, 529, 534
22, 33, 199, 308, 311, 312, 314, 347, 348 n. 3, n. 3, 535, 538, 539, 540, 541, 543, 544, 545,
349, 355, 421, 428, 513, 515, 550, 582, 594, 546, 547, 548, 551, 553, 556, 557, 560, 565,
616, 673 567, 568, 569, 571, 572, 573, 577, 578, 580,
Spain/Spanish 9, 21, 25, 26, 27, 112, 161- 584, 588, 592, 593, 596, 610, 611, 615, 616,
162, 316, 544, 682, 696 618, 619, 621, 622, 623, 624, 627, 628, 629,
630, 631, 632, 633, 634, 637-638, 639, 640,
Sri Lanka x, xi, 9, 28, 29, 31, 388, 456
641, 642, 643, 644, 645, 646, 649, 650, 651,
stability ii, xvi, 9, 18, 27, 44, 45, 49, 58, 652, 653, 667, 668, 669, 670, 671, 672, 673,
68, 69, 120, 134, 237, 265, 266, 304, 305, 674, 675, 676, 677, 680, 681, 682, 683, 684,
343, 355, 358, 359, 366 n. 30, 368, 405, 441, 685, 686, 687-688, 692, 693, 694, 695, 696,
442-445, 464, 467, 470, 479, 489, 492, 697, 698, 699, 700, 704, 715, 717 n. 42
496, 534, 550, 551, 556, 569, 571, 573, 576,
central, see also central government xii,
580, 585, 588, 591, 594, 595, 596, 597, 600,
xv, 18, 19 23, 30, 42, 207, 436, 513, 676,
602, 613, 615, 619, 624, 627, 646, 669, 687,
678, 680, 687
691, 701, 702, 703, 713, 714, 715
civic 285, 295, 303
in- 31, 116, 134, 427, 429, 585, 712
collapse xi
standards 12, 43, 94, 137, 139, 142, 171,
173, 191, 198, 251, 262, 326, 438, 492, 508, consent 22-23
512-513, 515, 517, 519, 532, 550, 562, 563, construction x, 403, 407, 408, 423, 431,
564, 566, 567, 569, 576, 577, 593, 596, 599, 436, 438, 443, 444, 445
600, 604, 610, 617, 618, 619, 620, 622, 647- independent/ce 8, 28, 32, 36, 43, 44, 197,
648, 650, 651, 659, 695 n. 6 239, 263, 313, 320, 389, 391, 396, 405, 532,
European 516, 522 540, 541 n. 6, 631, 642, 659
international 232, 251, 253, 600, 604, institutions 110-111, 196, 205, 207, 210-
609, 610, 611, 612, 616, 617, 619, 622, 628, 211, 213, 223, 229, 230, 238, 240, 270,
633, 644, 646
790 Index
286, 288, 294, 307, 342, 409, 410, 415, Comprehensive Peace Agreement
422, 434, 435, 511, 548, 579, 680, 681 2005 398
kin- 27, 204, 221, 244, 310, 313, 319, 419, Machakos Protocol 2002 44, 397, 399
420, 467, 470, 496, 497, 498, 610, 621, Muslim population in 397-398
668, 678, 683, 699 Surinam 50
practice 35, 40
sovereign, see also sovereignty 6, 8, 17,
28, 37, 39, 61, 69, 108, 389, 398, 540, 695, T
698 Taiwan 9, 41
successor 10, 41, 399, 673 Tanzania 12
weak 150, 155, 208, 307-308, 342 terrorism/t xi, xii, 5, 30, 42, 71, 122, 163,
statehood xii, 18, 19, 24, 37, 38, 40, 41, 164, 183, 189, 191, 219, 245, 266, 268, 274,
216, 239, 258, 394, 394, 413, 642 289, 397, 468, 492, 495, 512, 560, 567, 584,
status (geopolitical) xv, xvi, 10, 20, 21, 586, 587, 588, 591, 593, 609, 610 n. 35, 662,
22, 23, 24, 26, 29, 30, 31, 39, 40, 41, 42, 43, 663, 679 n. 19, 696, 702
44, 45, 55, 56, 73, 100, 104, 111, 114, 118-121, territorial xiv, 7, 8, 9, 11, 12, 17, 18, 20,
125, 137, 149, 168, 189-192, 195, 197, 238- 22, 26, 27, 30, 33, 43, 49, 53, 55, 56, 58, 93,
241, 244, 245, 246-248, 259-261, 262, 263, 143, 170, 184, 191, 196, 197, 199, 200-201,
308, 313, 315, 318, 320, 323, 325, 329, 334- 208-209, 211, 212, 259, 268, 271, 285, 287,
342, 343, 348, 349, 361, 364, 376, 388, 389, 288, 292, 308, 321, 329, 338, 341, 348, 365,
390, 391, 392, 399, 400, 401, 402, 403-404, 368, 390-391, 393, 401, 407, 416, 417, 422,
405, 419, 423, 426, 428, 443, 445, 446-447, 423, 426, 427, 432, 434, 440-442, 444,
483 n. 22, 498, 513, 543, 548, 551, 583, 602, 448, 459, 548, 588, 633, 636, 642, 668, 669,
605, 641-642, 643, 652, 678, 699 670-671, 672, 674, 678, 679, 680, 682, 696,
dual 36 697 n. 11
express self-determination 32-38 autonomy 29, 55, 125, 126, 137, 138, 155,
federal 36, 38, 39, 45 156, 157, 158, 195, 196, 199, 205, 209, 283,
285, 287, 307, 309, 311, 313, 316, 318, 320,
final 251-253, 262, 404-405, 408, 413,
322, 323, 391, 393, 396, 408, 423, 426,
426, 430, 445, 491
440, 441, 448, 467, 633, 685, 704
future xv, 5, 43, 256, 259, 365, 398, 404,
integrity ix, xii, 19, 23, 28, 31, 35, 37, 40,
421, 430, 434, 447
41, 42, 44, 182, 198, 206, 228, 237, 243,
implied self-determination 32, 39 245, 246-247, 256, 257, 274, 277, 279, 313,
self-determination 20, 28 n. 4, 39-40, 42, 315, 354, 364, 366, 371, 379, 388, 389, 390,
391, 396, 400 392, 393, 394, 402, 403-404, 428, 589,
sovereign 37 601, 627, 641, 643, 659, 698 n. 15, 699
special xiii, xvi, 43, 156, 311, 313, 315-318, unity 5-16, 18, 19, 20, 23, 25, 28, 34, 35,
319-320, 332-333, 337, 341, 391, 568, 583, 36, 39, 40, 41, 44, 246, 392, 403, 405
592, 633, 636, 638, 639, 641, 642, 644, territory/ies xi, xvi, 5, 8, 9, 10, 11, 17, 21-
652, 658, 660, 681-682, 697, 699 23, 24, 25, 26, 27, 28, 30, 32, 33, 36, 37, 39,
Sweden 9, 470, 650, 681 n. 24 40, 41, 43, 47, 48, 56, 61-64, 65, 101, 105,
Switzerland 7, 50, 51, 57, 468, 470, 568, 108, 109, 123, 143, 145, 162, 165, 168, 170,
674, 681, 691, 692, 693, 696 179, 188, 189, 190, 191, 195, 196, 197, 198,
199, 200-201, 202, 212, 213, 215, 221, 243,
subsidiarity 54, 204, 412, 448, 687 n. 30
244, 245, 260, 263, 266, 274, 292, 297, 309,
Sudan x, xi, xiii, xvi, 9, 10, 29, 31, 44, 388 310, 311, 314, 316, 319-320, 321, 324, 325,
n. 1, 397-399, 700 n. 23 328, 329, 331, 337, 339, 345 n. 1, 350, 368,
388, 390, 397, 399, 403, 409, 421, 423, 429,
Index 791
434, 439, 440-442, 448, 455, 503, 506, 508, 389, 392, 395, 399, 403-405, 409, 413, 424,
515, 517, 523, 525, 526, 528, 529, 546 n. 11, 426, 454, 465, 466, 469, 470, 472-474, 477,
549, 559-569, 579, 583, 584, 634, 635, 638, 478, 482-483, 487, 488, 489, 490, 496-497,
642, 645, 659, 669, 670, 672, 673, 675, 677, 515, 543, 567, 569, 571, 572, 582, 584, 587,
678, 679, 680, 682, 685, 686, 688, 700, 588, 605, 606, 609, 610, 617, 620, 622, 638,
701, 703, 716, 718 642, 643, 693, 695, 699
Tibet 8 Development Programme (UNDP) 180,
Togo 10 183, 185, 334, 355, 358, 371-373, 375 n. 52,
Transdniestria/n 19, 43, 44, 308-311, 313- 440, 470, 473, 475, 477, 479, 488, 489,
316, 321, 332, 335-341, 343, 360, 361 n. 25, 490, 491, 528, 595
391-392, 400, 423, 447 n. 5, 582-584, 628, General Assembly 28, 40, 149, 600
638 n. 1, 644, 650, 658, 659, 661, 665, 681 High Commissioner for Human Rights
transparent/cy xv, 217, 342, 361, 372, (UNHCHR) 652, 662
376-377, 443, 553, 577, 578 High Commissioner for Refugees (UN-
Turkey 5, 317, 319, 325, 327, 332, 334, 340, HCR) 225, 355, 358, 364, 375, 440, 470,
469, 479, 568, 681, 699 474, 475, 477, 488, 489, 490, 491, 548,
595, 645, 651, 662, 663
International Court of Justice 26-27, 209
U
n. 15
Ulster 9, 61-67, 80, 89, 91, 93, 95, 97, 102, Mission in Bosnia and Herzegovina (UN-
106, 266, 543, 560-561, 568, 592, 660, 717 MIBH) 567
unification 57, 65, 66, 67, 87 n. 16, 104, Mission in Kosovo (UNMIK) 243, 258,
105, 109-110, 114, 124, 310, 320, 335, 360, 259, 262, 263, 404, 474, 478, 487, 489-
367 n. 32, 409, 420, 449, 678, 684 491, 521-522, 542, 565-566, 652, 653, 659,
unify/ied 26, 88, 100, 110, 132, 156, 162, 662, 678 n. 16
167, 181, 188, 196, 233, 275, 276-277, 469 n. Secretary General 12-13, 14, 256, 258,
1, 510, 588 n. 4 303, 410, 424, 425, 474, 566
Union of Soviet Socialist Republics Observer Mission on Bougainville (UN-
(USSR), see also Soviet Union 33-34, OMB) 133, 413, 439, 472, 497
36, 38, 41, 58, 309, 347, 394, 395, 532, 534, Security Council (UNSC) xv, 12, 36, 42,
540, 546-550 171, 243, 246, 253-254, 256-257, 263, 303,
unionists 61, 62, 64-72, 74-77, 78 n. 12, 352 n. 12, 390, 392, 394, 402, 403, 424,
79-83, 87, 88, 89 n. 19, 91-92, 94-96, 99- 430, 439, 469, 474, 486-487, 489, 521,
106, 108-111, 113-121, 123-124, 311, 401, 419 585-586
United Arab Republic 40 UN Charter 7-8, 28, 179, 256, 390, 467,
United Kingdom (UK), see also Great 604, 642, 679
Britain xv, 9, 21, 39, 43, 61-68, 70-74, UN Declaration on the Rights of
76, 78, 79-80, 83, 85, 91, 94-104, 108, 110,
Indigenous Peoples 179
112, 113, 116-118, 120-124, 400-402, 414,
419, 420, 430, 439, 440, 446, 466, 469, UN Declaration on the Rights of Per-
505, 506, 523, 532, 535, 538, 541, 549, 552- sons Belonging to National or Ethnic,
553, 555, 573, 574, 577, 591, 599, 603, 609, Linguistic or Religious Minori-
611, 618, 620-622 633, 649, 658, 684, 694, ties 631, 640, 641
696, 698 n. 14, 700, 709, 718 UN Framework for Strengthening the
United Nations (UN) 8, 12-13, 19, 20, 25 Rule of Law 617
n. 2, 28, 36, 133, 135, 150, 168, 184, 198-199, UN International Covenant on Civil and
201, 214, 218, 243-263, 266, 291, 303, 313, Political Rights (ICCPR) 28, 327, 467,
352 n. 12, 355 n. 20, 356-357, 371, 375, 379, 604, 617, 629, 658
792 Index
Human Rights Committee (HRC) 617 violent, see also violent conflict 18, 29,
UN International Covenant on Social, 39, 100, 122, 151, 176, 178, 182-183, 194,
Economic and Cultural Rights (ICE- 267, 276, 312, 413, 417, 427, 428, 429, 450,
SCR) 604, 629 455, 468, 471, 494, 509, 527, 542, 562, 572,
United States of America (US/A), see 582, 589, 591, 652, 662, 701, 718
also American/n 5, 6, 8, 12, 15, 31, 57, vote/s 6, 7, 39, 48, 52, 68, 75, 76, 78-91, 99,
65, 70-71, 80, 93, 163, 189, 199, 200, 202, 104, 107, 109, 114, 115, 120, 123, 136-137,
213, 215, 219, 245, 252, 258, 267, 268, 275, 147, 150, 159, 165, 179, 181, 186, 188, 194,
276, 277, 278, 280, 291, 299, 319, 375, 398, 196-197, 204, 210, 212, 217, 223, 249, 256,
403, 427, 468, 470, 471-472, 473-477, 479, 261, 274, 281, 282-283, 286-287, 290, 293,
482, 483, 486, 496-497, 562, 566, 576, 577, 301, 311, 319, 321, 323, 324, 336, 339, 377,
585, 584, 585, 589, 620, 621, 624, 636, 674, 398, 441, 443, 452, 456-457, 460, 461, 463-
693, 694, 704, 711 n. 33, 718 465, 488, 692, 705, 706, 708, 710, 714 n.
unity, see also territorial unity xvi, 28, 37, 715-718
38, 44, 100, 133, 139, 156, 188, 244, 322, voter/s 51, 52, 74, 87 n. 16, 88, 89, 90-91,
398, 399, 402, 675 113 n. 28, 115, 139, 188, 197, 290, 294, 295,
national 273, 275, 276, 278 296, 298, 321, 325, 340, 449, 454-456, 462-
465, 695, 705, 706, 715-716, 718 n. 45
state xiv, 18, 19, 37
voting, see also voting rights ix, 72, 74,
Universal Declaration of Human
77, 85, 88, 99, 113 n. 28, 158, 210, 261, 293,
Rights 7-8, 179, 629
398, 401, 407, 412, 451, 455, 456, 459, 462,
use of force 9, 15, 24, 31, 34, 35, 41, 243, 463-464, 706, 707, 716, 718 n. 45
253, 364, 469, 486, 487, 492, 585
uti possidetis 9,10, 13, 27-28, 29, 44, 193,
405 W
Wales xv, 97, 108, 419, 538, 552, 709
V war/s, see also Cold War, conflict/s, dis-
pute/s 5, 6, 9, 11, 14, 16, 34, 41, 48, 53,
veto/es, see also blocking powers and veto 64, 67, 112, 114, 123, 132, 156, 162, 164, 168,
rights ix, xiv, 56, 78 n. 12, 79, 80, 95, 96, 182, 184, 193, 194-201, 202, 203, 204, 205,
202, 205, 210-211, 228-229, 237, 246, 248- 207, 211, 214, 215, 216, 217, 219, 221, 222,
249, 261-262, 277, 278, 285, 286, 303 n. 66, 224, 225, 226, 227, 228, 232, 233, 234, 237,
393, 394, 453, 455, 499, 700 238, 239, 240, 266, 269 n. 9, 274, 276, 350,
violence xi, xiii, xv, 16, 31, 35, 36, 43, 66, 351, 353, 354, 355 n. 19, 358, 359, 362, 368,
67, 113, 114, 116, 117, 122, 125, 130, 177, 178, 409-410, 473, 474, 483, 484, 504, 507, 509,
194, 220, 231, 240, 245, 276, 279, 295, 297, 520, 538, 542, 556, 559-569, 565, 566, 567,
302, 314, 334, 349, 352, 419, 421, 424, 427, 578, 579, 581, 591, 595, 601, 610, 663, 700-
428, 445, 453, 463, 467, 472 n. 6, 474, 477, 703, 704
486, 487, 493, 494, 495, 498, 523, 533, 535 civil 15, 29, 48 n.1, 53, 65, 131, 267, 275,
n. 4, 536, 541 n. 7, 543, 550-551, 562-563, 289, 416, 455, 468, 496, 610, 705, 711, 718
566, 568-569, 576, 583, 584, 587, 591, 594,
crime(s)/criminal(s) 202, 215, 224, 231,
595, 613, 647, 651, 654, 660, 662, 663, 664,
294, 368, 471
665, 703, 711, 714, 716
First World 6, 7, 9, 64, 532, 546, 547
ethnic xv, 134, 266-267, 274, 424, 451,
551, 559-569, 663, 717 guerrilla 65, 130, 131
political 66, 189, 711 prisoners of 29, 30
renunciation of 43, 255, 586 Second World 9, 45, 148, 421, 423, 516,
532, 546, 547, 616, 645, 667, 670, 676 n.
14
Index 793
warfare xi, 30, 65, 352-353, 467, 486, 492, Yugoslavia (the former Socialist Federal
567 Republic of ) xiii, xvi, 10, 14, 28, 29, 31,
West Irian 8 34, 36, 41, 45, 58, 194-199, 202, 208-209,
Western Sahara 17, 25, 26 218, 220-221, 223, 225, 228, 232-234, 237,
239, 243-247, 248 n. 2, 253, 255-257, 265-
Westphalia 11
266, 268, 271-272, 292, 355 n. 19, 395-396,
403-404, 409, 423, 426-427, 455, 486,
Y 518, 520, 532 n. 1, 534, 540, 541-546, 584,
586-587, 612, 613, 616, 628, 641, 652, 679,
Yemen 40 717
Yugoslav xvi, 10, 19, 34-35, 37, 195-197,
201, 208, 220, 235, 237, 244-245, 268,
304, 402-403, 423-424, 465, 485, 487, Z
532, 540-542, 565, 579, 580, 586, 681 n. 24, Zambia 12
698, 717 n. 42
Zimbabwe xi

(Publications On Ocean Development) Marc Weller, Barbara Metzger, Niall Johnson - Settling Self-Determination Disputes - Complex Power-Sharing in Theory and Practice-Hotei Publishing (2008)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Publications On Ocean Development) Marc Weller, Barbara Metzger, Niall Johnson - Settling Self-Determination Disputes - Complex Power-Sharing in Theory and Practice-Hotei Publishing (2008)

Uploaded by

Copyright:

Available Formats

Settling Self-determination Disputes

Complex Power-sharing in Theory and Practice

Marc Weller and Barbara Metzger

Published under the Auspices of the

isbn: 978 9004 16482 6

© 2008 by Koninklijke Brill nv, Leiden, The Netherlands.

Printed and bound in The Netherlands.

Settling Self-determination Conflicts: An Introduction xi

Section One Framework and Case Studies 1

Part One Historical and Theoretical Framework 3

Chapter 2 Why the Legal Rules on Self-determination Do Not Resolve

Chapter 3 The Logics of Power-sharing, Consociation and Pluralist

Part Two Case Studies 59

Chapter 5 Resolving the Bougainville Self-determination Dispute:

Chapter 6 Resolving Self-determination Disputes Through Complex

Chapter 7 Power-sharing and International Intervention: Overcoming

Chapter 8 Interim-governance for Kosovo: The Rambouillet

Chapter 9 Power-sharing in Macedonia? 265

Chapter 10 Gagauzia and Moldova: Experiences in Power-sharing 307

Chapter 11 Case Study of the Conflict in South Ossetia 345

Section Two Cross-cutting Analysis 383

Part Three Vertical Power-sharing 385

Chapter 13 Power-sharing and the Vertical Layering of Authority: A

Chapter 14 Electoral Arrangements in Systems of Complex Power-

Chapter 15 Third-party Involvement in Self-determination Conflicts 467

Part Four Functional Power-sharing 501

Chapter 17 Resolving Self-determination Disputes Using Complex

Chapter 18 Policing Territories Previously Subject to Civil War and

Chapter 19 The Military Dimension of Security Sector Governance in

Chapter 20 Transforming Justice, Reclaiming the Rule of Law: Legal

Chapter 21 The Role of Human and Minority Rights in Complex Power-

Chapter 22 ‘Bridges’ in Self-determination Disputes? External Relations

Part Five Conclusions 689

This study is the result of an international collaborative project supported and

ment mechanisms, and the involvement of a layer of international involvement

The claim to self-determination encapsulates the hopes of ethnic peoples and

appeared to be unsolvable. The doctrine of self-determination has traditionally

ciation of the self-determination claim is meant to be permanent. However, in a

agreement on the status of the autonomous territory of Gagauzia within Mol-

ety is composed of solipsists, whose cooperation is confined to recognizing each

prevented secession where it could be agreed. There were precedents – Norway’s

answer was that, in the absence of a negotiated boundary adjustment, successor

building? These symptoms generally presage a slide towards a form of patho-

stitutions’ emerge together, the absence of an overarching political culture will

* An earlier version of this chapter appeared in 4(2) Ethnopolitics (2005).

humanitarian catastrophe. The United Nations refused to intervene during the

II The Classical Right to Self-determination

A A Concept with Multiple Meanings

by members of minorities individually, it is clear that they can exercise these

Manifestly, the doctrine of self-determination has different legal consequences

B The Issue of State Consent

The manifestation of an act of will of the population is necessary even where a

C Self-determination as an Exceptional Right

D The Definition of the Entity

In the UN framework, there exists a technical identification of ‘non-self-governing

pendence are therefore an example of colonial self-determination in the classical

Burkina Faso/Mali Frontier Dispute, Merits, 1986 ICJ 564.

and to take account of it in the interpretation of the principle of self-determination of

Subsequently, this doctrine was endorsed by the Badinter Commission appointed

An exception to this rule would relate to a self-determination entity that decides to

G Modalities of Reaching the Point of Self-determination

The treatment by the international system of classical colonial cases of self-deter-

tion fighters in neighbouring territories, and training, equipping and supplying