You are on page 1of 36

Cooperation and Conflict

http://cac.sagepub.com

Just Scraps of Paper?: The Dynamics of Multilateral Treaty-Making


Robert A. Denemark and Matthew J. Hoffmann
Cooperation and Conflict 2008; 43; 185
DOI: 10.1177/0010836708089082

The online version of this article can be found at:


http://cac.sagepub.com/cgi/content/abstract/43/2/185

Published by:

http://www.sagepublications.com

On behalf of:
Nordic International Studies Association

Additional services and information for Cooperation and Conflict can be found at:

Email Alerts: http://cac.sagepub.com/cgi/alerts

Subscriptions: http://cac.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.co.uk/journalsPermissions.nav

Citations http://cac.sagepub.com/cgi/content/refs/43/2/185

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


Just Scraps of Paper?
The Dynamics of Multilateral
Treaty-Making
ROBERT A. DENEMARK AND
MATTHEW J. HOFFMANN

ABSTRACT
Despite its importance in the global system, the literature provides lit-
tle guidance on how treaty-making emerged as a well-accepted prac-
tice. In either assuming the appropriateness of treaty-making (and then
analysing design) or treating treaties as strategic choices in the pursuit
of gains (without analysing how treaties came to be a way to pursue
gains), the current literature discounts the emergence and evolution
of treaty-making. This lacuna contributes to a biased view of treaty-
making as the epiphenomenal result of specific, ahistorical factors,
rather than as a patterned, historical practice. We contend that the evo-
lution of the practice of treaty-making is significant for questions of
design/compliance, the future of multilateral interaction and global
order. In addressing this concern, we pursue two linked goals. The first
is self-consciously descriptive. We introduce a dataset of multilateral
treaties that provides a novel picture of treaty-making across time,
space and issue-areas. The second goal is explanatory. We develop and
test a social constructivist and path-dependent explanation for the
patterns of treaty-making evident in the data, especially 150 years of
exponential growth, the spread of treaty-making across multiple issues
and the diffusion of the practice across the world.
Keywords: increasing returns; multilateral treaties and treaty-making;
path dependence; social constructivism

Introduction

Early in the twentieth century, treaties were famously described as mere


‘scraps of paper’, and in the ensuing years treaties have been criticized as
little more than traps set for naïve liberals by unscrupulous dictators
(Beilenson, 1969). Even if treaties on the allegedly unimportant matters are
sometimes adhered to, it is suggested that important matters cannot be left
to the unreliable confines of pen and ink. Some have gone so far as to sug-
gest that the entire idea of cooperative multilateral interaction is a chimera

Cooperation and Conflict: Journal of the Nordic International Studies Association


Vol. 43(2): 185–219. © NISA 2008 www.nisanet.org
SAGE Publications, Los Angeles, London, New Delhi and Singapore
www.sagepublications.com
0010-8367. DOI: 10.1177/0010836708089082

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


186 COOPERATION AND CONFLICT 43(2)

(Mearsheimer, 1994/5). These criticisms are popular, but unsupportable.


Indeed, it is odd that the ‘scrap of paper’ critique survived, given the con-
text in which it was made. Its author was German Chancellor Theobald von
Bethmann-Hollweg, who believed he had managed relations with Great
Britain sufficiently well to allow Germany’s 1914 invasion of Belgium to
proceed despite the UK pledge to uphold Belgian neutrality. The British
ambassador reported that upon hearing that the UK would live up to its
agreement, the German Chancellor asked if war was to be waged: ‘… just
for a word — “neutrality,” a word which in war time had so often been dis-
regarded — just for a scrap of paper …’ (Horne, 1923: vol. 1, p. 406). The
British honoured their agreement, and went to war against Germany. The
Chancellor’s career entered a steep decline as a result. He was chased from
office and Germany went on to suffer a costly and humiliating defeat.
The British did not support Belgium because of a scrap of paper. They
made an agreement to do so because that is what appeared to best serve
British interests, and they wished this to be well and publicly understood.
We contend that decisions to negotiate individual treaties are neither sep-
arate from other actor strategies in the global system, nor do they begin
with a blank slate. Reus-Smit (1997: 558) asserts that ‘… contractual inter-
national law and multilateralism have become the dominant institutional
practices governing modern international society’ and we argue that, over
time, states have come to accept and internalize treaty-making as the
appropriate foundation for both. Despite its important role in the global
system, the literature provides little guidance on how treaty-making
emerged as a well-accepted practice. From early interest in international
law and formal organizations through the debates between realism and
idealism, then neo-realism and neo-liberalism, onto the current discussions
of institutional design and compliance, extant analyses of multilateral
treaties tend to examine either the specific architecture of agreements to
assess the factors that influence the probability of compliance and effective-
ness, or the strategic use of treaties to attain relative or absolute gains
(Grieco, 1990; Martin, 1993; Goldstein et al., 2000; Leeds et al., 2000; Pahre,
2001; Koremenos et al., 2001; Leeds, 2003; Mitchell, 2003). These foci are use-
ful and have facilitated significant advances in our knowledge. They are
incomplete, however. In either assuming the appropriateness of treaty-mak-
ing (and then getting on with analysing design) or treating treaties as strate-
gic choices in the pursuit of gains (without analysing how treaties came to be
understood as a way to pursue gains), the current literature ignores the
emergence and evolution of this practice. These studies contribute to a
biased view of treaty-making as the epiphenomenal result of specific, ahis-
torical factors rather than as a patterned, historical behavioural practice.
This leaves us unable to appreciate how this practice has evolved, how its
evolution influences questions of design/compliance, how it might change in
the future, and how it connects with larger questions of global order.
To address this concern we pursue two linked goals. The first is self-
consciously descriptive. We introduce a dataset of multilateral treaties with
which to consider the role that treaty-making has played in the global system.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 187

Multilateral treaties, broadly defined as those that have been signed by


three or more actors, tend to represent areas in which there exists substan-
tial agreement, and are more often available to the public. To discover what
patterns of treaty-making might exist across time, space and issue-area, we
created the Multilateral Agreement and Treaty Record Set (MATRS). The
dataset includes 6976 multilateral treaties signed between 1595 and 1995.
The MATRS dataset was constructed by consolidating information from
several compendia of multilateral agreements and specialty lists, and con-
tains data on the title, date of signature and two substantive categorizations
(one broader and one more specific). For about 6000 treaties we also have
information on signators and place of signing (see Appendix 1 for more
information on the construction of the database).
Our treatment of the multilateral treaties in the MATRS dataset differs
from other orientations. The MATRS dataset is designed to facilitate
research into the patterns of multilateral treaty-making. Treaties are con-
sidered here as discrete and observable instances of a given form of co-
operation, as opposed to indicators of the status of agreements in specific
issue-areas, or examples of specific elements of design. We recognize, of
course, that not all treaties are of equal importance. The founding instru-
ments of the United Nations play a very different role in global politics than
the 1934 agreement unifying methods of analysing cheeses. But we are
interested in the evolution of the practice of treaty-making and it is there-
fore crucial to understand both system-defining agreements and the day-to-
day business of global affairs that are manifested through multilateral
treaties. The global distribution of treaty-making, rather than the role or
efficacy of individual treaties, is our analytical target.1 This systemic exam-
ination of treaty-making makes visible the entrenched-ness of the practice
in a way that examining treaties individually or sectorally cannot.
In brief, the descriptive enterprise finds that multilateral treaty-making
has increased over time, and has been doing so systematically, and expo-
nentially, from the 1850s until quite recently. The empirical pattern of
treaty-making across time, space and issue-area confirms the sense of the
literature that multilateral interaction has been increasing, but confounds
much of what has been suggested about the reasons for this increase.
Interdependence (as traditionally measured by trade volume) does not cap-
ture the treaty-making dynamic, nor is there a simple relation to the growth
of the number of actors or the presence of a hegemon. Most surprisingly,
multilateral treaty-making activity does not emerge as a response to crisis.
Far from being crisis-driven, the consistent increase in treaty-making
appears to be only temporarily crisis-interrupted.
The second goal of this work is explanatory. The patterns that emerge in
the first part of this work, especially the 150 years of exponential growth,
appear to be most plausibly understood through the lens of social con-
structivism. Several additional patterns (transference across time and issue
area, geographic spread, persistence of locations of treaty-signing and per-
haps even the recent downturn in multilateral treaties) are also
consistent with constructivist explanations.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


188 COOPERATION AND CONFLICT 43(2)

Patterns of Treaty-Making and Conventional Wisdom

We begin with the broadest picture of multilateral treaty activity over the
past 400 years. Figure 1 presents the number of multilateral treaties signed
each year from 1595 to 1995, and immediately challenges traditional wis-
dom. Those who assume that multilateralism has increased over time are
correct, yet the explosion of treaty-making in the past 150 years is quite dra-
matic. Exponential growth is notable for nearly the entire period, and
not just following the Second World War, as is conventionally argued.
Furthermore, while multilateralism as a mode of interaction might well
have suffered political attacks from the 1980s, multilateral treaty signings
peaked in the 1960s and, following a decline, levelled off in the 1980s and
enjoyed a mild resurgence in the 1990s. Beyond challenging conventional
wisdom about how much multilateral treaty-making there is and how it is
distributed over time, Figure 1 also casts doubt on four common conjec-
tures about multilateral treaty-making: that it should be related to crises,
demographic shifts in the international system, alterations in (mainly eco-
nomic) connectivity and hegemonic influence.
First, there is no evidence that multilateral treaty-making is a crisis-
driven activity or that as crises fade into historical memory our propensity
to make treaties declines. Certainly, the two world wars were followed by
increased treaty activity, but what is remarkable is how the post-war pat-
terns return to the trajectories of the pre-war eras. Multilateral treaty-
making does not appear crisis-driven so much as crisis-interrupted. Even with
the slate wiped clean by the events of 1918 and 1945, and the opportunity to
remake the international system, states returned to the practice and prior
trajectory of multilateral treaty-making.
Second, we might be concerned that the number of multilateral treaties
would increase as the number of sovereign states rises. This relationship is
not a certainty. It could be that states that are new to the system, especially
the relatively small ones that emerged from decolonization from the 1950s
to the 1970s, and again after 1989, would tend to join existing agreements or
follow more powerful states into new agreements, as opposed to sponsoring
their own. Assuming that additional states would nonetheless tend to
increase the number of agreements, this might provide an explanation for
some of the increase noted in Figure 1. Figure 2 casts doubt on the efficacy
of using demographic shifts to explain treaty-making over time, showing
that there is a complex relationship between the number of states in the sys-
tem and the number of treaties signed. We may see a threshold effect or
even an inversion, but the signing of multilateral treaties is no simple func-
tion of the addition of new actors to the global system. Bivariate regression
results (Model 1 in Table 1) appear to support a more direct relationship,
but when the post-war period is controlled for, as in Model 2 in Table 1, the
relationship between the number of states in the system and treaties signed
disappears.2
A third argument concerns the increase in treaty-making that might
emerge from an increase in the interaction of nation-states. Assuming that
treaties emerge at least in part from the discovery of collective problems,

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


180

160

140

120

100

80

60

Number of Treaties Signed


40

20

0
1596 1616 1636 1656 1676 1701 1 721 1 74 1 1 761 1781 1801 1821 1 841 1 861 1 881 1901 1 921 1941 1961 1981

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING

FIGURE 1
Multilateral Treaties Signed Per Year 1596–1995
189
180
190

160

140

120

100

80

60

Number of Treaties Signed


40

20

0
COOPERATION AND CONFLICT 43(2)

0 50 100 150 200 250

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


Number of Independent States in International Community

FIGURE 2
Size of International Community and Treaty Signings
Note: Figure 2 provides data for this relationship from 1816 to 1995, the period within which we could obtain reliable data on the number of independ-
ent states in the international system on an annual basis. We obtained the data at: http://weber.ucsd.edu/~kgledits/statelist.html
(accessed January 2006). See Gleditsch and Ward (1999).
DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 191

and the search for collection solutions, the more interdependent states
become, the more incentives for treaty-making we might expect. Increasing
trade, important both in its own right and as the most generally agreed-
upon surrogate measure of interdependence, does not appear to explain the
rise of multilateral treaties. Figure 3 contrasts the growth of international
trade with multilateral treaty signings over a period of 40 years that
includes both increases and decreases in treaty behaviour. The figure casts
doubt on the ability of (trade-based) interdependence to explain treaty-
making patterns. Models 3, 4 and 5 of Table 1 also question the efficacy of
explaining the growth of treaty-signing with traditional markers of interde-
pendence. While trade is a significant variable in all three models, it has the
wrong sign, indicating that increasing trade volume is correlated with a
decrease in the number of treaties signed per year. There is no simple rela-
tionship between interdependence and treaty-making.
Finally, those who portray multilateralism as a post-Second World War
phenomenon driven by the hegemony of the US are at best only partially cor-
rect (Martin, 1993; Ruggie, 1993b; Ikenberry, 2001). While multilateral treaty-
making surged in the 20 years following the Second World War, this is
remarkable not for its novelty but rather for how it fits within a pattern of
accelerating multilateral treaty-making that began a century prior. Figure 1
casts suspicion on any suggestion that the incidence of multilateral treaties
ought to rise or decline with any given distribution of power. Treaty-making
did not suffer with the decline of the British (circa 1885 to 1910), nor enjoy
unusual increases with the rise of the US (circa 1945 to 1955).With the excep-
tion of the largest and most disruptive wars, treaty-making continues apace.3
This lack of apparent impact between hegemony and treaty-making is
interesting given the prevalence of suggestions about such a link. Not all of
the arguments are consistent. Dominant powers, especially those dedicated
to representative forms of rule, are suggested to facilitate more agreements
(Keohane, 1984; Ikenberry, 2001). Alternatively, such powers may eschew
agreements given problems of group size and their lack of a need for col-
lective support (Oye, 1986; Taylor, 1987). Or their decline might provide
incentives for others to engage in increased cooperation (Krasner, 1982;
Keohane, 1984; Hasencleaver, 1997). Certainly, as Ruggie (1993b) contends,
‘American hegemony’ was important, but the existence and promulgation
of multilateral treaties from 1850 to 1945 appears to have had a profound
effect on the manner in which the emergent US sought to order the global
system.4 What the data force us to consider is that hegemony may affect
multilateral treaty-making less than the acceptance of the practice of
treaty-making influences hegemonic behaviour. In the case of hegemony
and the establishment of a multilateral treaty system, the traditional causal
arrow may have to be reversed (Jönsson and Hall, 2005).

Explaining Multilateral Treaty-Making

Much of what we have come to expect about multilateral treaty-making does


not fit with how multilateral treaty signings are distributed through time.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


192

TABLE 1
Regression Results
Model 1 Model 2 Model 3 Model 4 Model 5
Parameters (1816–1995) (1815–1995) (1950–1995) (1950–95) (1950–95)

Constant 23.1 14.7 102.3 75 35.7


States 0.72 (15.3)*** 0.04 (0.55) 0.28 (1.1)
Post-war 69.3 (11.6)***
Trade 0.03 (2.5)** 0.63 (1.94)* 0.16 (2.2)**
R2 0.56 0.75 0.12 0.14 0.10
N 179 179 45 45 45

The dependent variable for Models 1–4 is the number of treaties signed per year. For Model 5 the dependent variable is the number of trade
COOPERATION AND CONFLICT 43(2)

and economics related treaties signed per year. The independent variables are operationalized as follows: States  number of states in the
international system in each year; Post-war  Dichotomous dummy variable (1  1945–1995; 0  1816–1945); Trade  World export/import vol-

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


ume measured as a percent of 1995 trade. (*p  0.1; ** p  0.05; *** p  0.001)
100 180

90 160

80
140

70
120

60
100
50
80
40

60
30

Global Export Levels (1995 = 100)


Number of Multilateral Treaties Signed

40
20

10 20

0 0
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995

Total Exports (Volume)* Total Multilateral Treaties W+M*

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING

FIGURE 3
Multilateral Treaties and Export Levels
*Exports (total world merchandise export volume) are measured as a percentage of 1995 levels. The source data are drawn from the World Trade
Organization at: http://www.wto.org/English/res_e/statis_e/its2004_e/its04_appendix_e.htm (accessed January 2006).
193
194 COOPERATION AND CONFLICT 43(2)

When assessing conjectures empirically, we note that extant understandings


of treaty-making cannot account for the most striking feature of Figure 1,
the century-long exponential increase. One explanation for this failure is that
most aspects of conventional wisdom about multilateral treaties assume
they are essentially reactive. They assume that multilateral treaty-making
responds to politics. ‘Politics as usual’ is not multilateral. Something happens
(i.e. crisis, change in connectivity, hegemonic transition, demographic shift)
and we suddenly need multilateral treaties. These data suggest a radically dif-
ferent conception. ‘Politics as usual’ is multilateral, as it has come to include
the practice of multilateral treaty-making. Multilateral treaty-making has a
long-term historical dynamic with which other forces must contend.
We posit that once multilateral treaty-making emerges as an accepted
practice, increasing returns to this activity alter expectations and behav-
iours. The increasing returns are both material (in that states engaging in
multilateral treaty-making are successful even if the organizational form is
not necessarily optimal) and ideational (in that states come to see multilat-
eral treaty-making as appropriate), and these returns enhance the attract-
iveness of future multilateral treaty-making. This dynamic leads to the
spread of multilateral treaty-making geographically and across issue areas.
The result is that treaty-making begets treaty-making, leading to exponen-
tial growth. Multilateral treaties are not mere ‘scraps of paper’ because
states have internalized this means of interaction (Jönsson and Hall, 2005).
In the following sections we elaborate on this proposed explanation for the
exponential rise in treaty-making and provide initial tests of it with treaty
data. We first explore the notion of social construction through path depend-
ence and increasing returns, detailing the mechanisms that could explain
exponential growth. We then examine the plausibility of this explanation by
reviewing key literature on multilateralism and multilateral treaties. Finally,
we interrogate our explanation empirically with the MATRS dataset.

Constructivism, Path Dependence and Multilateral Treaties

The central notion of social constructivism is that actors and their social con-
texts are mutually constitutive. The behaviours of actors help create their
social context (rules, meanings, institutions), that social context shapes actor
identities, perceptions and wants, which are then translated into subsequent
behaviours. We posit that the accelerating adoption of multilateral treaties
as a mode of international cooperation in the early to mid-nineteenth cen-
tury altered the global social context. Multilateral treaty-making was intro-
duced as a way to deal with a growing set of transnational issues. Increasing
use of this instrument reinforced its appropriateness and states began to ori-
ent their domestic structures, behaviour and expectations to this practice.
Multilateral treaty-making became the accepted way to deal with trans-
national issues, and came to constitute how states conceived of appropriate
interactions.
This broad understanding is incomplete. It fails to provide insight into
why and how states might come to conceive of multilateral treaty-making

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 195

as appropriate, and thus how and why it came to be a popular response to


transnational issues. We argue that the relevant mechanism of social con-
struction was path dependence through increasing returns.
Path dependence arises when a confluence of actions, intended or not,
provides a given mechanism with the advantage of being considered ‘best
practice’.5 Once a ‘best practice’ emerges, this advantages one mechanism
at the cost of others, some of which might even be technically superior.6 The
dynamics of path dependence naturally complement social constructivist
ideas about mutual constitution. Many path dependency arguments rely on
rational choice to drive the processes by which practices are adopted or
locked-in around an institution, but rational cost-benefit calculation is
not the only mechanism that leads to path dependence. At its core, path
dependence is a probability mechanism whereby the chance of future use
of a technology or institution increases every time it is used in the present
(North, 1990; Arthur, 1994; Thelen, 2000; Couch and Farrell, 2004; Boas,
2007). While utilitarian considerations can drive the ‘self-reinforcing
sequences’ inherent in path-dependent processes, an institution can also
exhibit path dependency when ‘an initial precedent about what is appropri-
ate forms a basis for making future decisions about what is appropriate’
(Mahoney, 2000: 523). There is thus no reason that the self-reinforcing
mechanisms of path dependence cannot work within the logic of appropri-
ateness highlighted by constructivists or through some hybrid that con-
siders both costs and ideas of appropriateness in reinforcing a practice like
multilateral treaty-making.
Social constructivism recognizes the limits of flexibility in social life. While
behaviours shape social context, the historical accumulation of behaviours
serves to reify/alter structures, putting history on a certain path, closing off
some avenues and opening others. Path dependence is a crucial mechanism
of social construction. With regard to multilateral treaties, we agree with
Reus-Smit’s (1997: 569) claim that:

It matters little whether, in an abstract rational sense, arbitration or multilat-


eralism constitute a more efficient response to coordination and collaboration
problems; what matters is that at particular historical moments states have
deemed them the right responses.

Once multilateral treaty-making was deemed the right response in the late
nineteenth century, a certain path was chosen. Increasing returns to this
institution solidified the choice and the practice moved, in the parlance of
normative dynamics, from emergence to internalization (Finnemore and
Sikkink, 1998). Once chosen, continued use of a given practice alters expect-
ations such that it eventually achieves the enviable state of being simply
taken for granted.
The increasing returns that drive path dependence are expected in sys-
tems exhibiting four characteristics (Pierson, 2000). First, start-up costs and
the costs of switching institutions are high. It is expensive or difficult to
establish institutions in the system, or to change them once established.
Second, learning effects are significant. There are steep learning curves

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


196 COOPERATION AND CONFLICT 43(2)

allowing for higher returns with continued use of existing institutions.


Third, coordination effects are significant. Benefits from the use of an insti-
tution or practice increase as more actors adopt the same mechanisms.
Fourth, adaptive expectations dominate actor calculations. In this sense,
‘projections about future aggregate use patterns lead individuals to adapt
their actions in ways that help make those expectations come true’
(Pierson, 2000: 254). Our claim is that the practice of multilateral treaty-
making that emerged significantly in the nineteenth century exhibits these
characteristics and can fruitfully be analysed through the lens of increasing
returns with a combination of what Mahoney (2000) calls legitimation and
functional processes of self-reinforcement whereby the efficacy and appro-
priateness of an institutional choice is reinforced over time.
Combining notions of social construction with path dependence through
increasing returns presents a unique mechanism for understanding the per-
sistence and evolution of global practices. In addition, such an analytic
framework can bring ideational and material factors into a single mech-
anism, a type of explanation for which both constructivist and rationalist
scholars have been calling. Very simply, multilateral treaties emerge in the
1850s as a potentially dominant practice for a series of material and ide-
ational reasons. The emergence and early success of multilateral treaty-
making alters the social context and states respond with changes in their
expectations about international interactions. The new expectations make
further treaty-making more likely, reinforcing these early actions and reify-
ing the appropriateness and utility of multilateral treaty-making. Eventually,
this positive feedback lends multilateral treaty-making a taken-for-granted
quality across large swaths of the global system. In the sections that follow
we examine this argument in greater detail and challenge it with MATRS
data.

Emergence of Multilateral Treaty-Making

Treaties, multilateral and otherwise, have been in use for millennia (Beckman,
1996; Cohen and Westbrook, 2000). Multilateral treaty-making exploded in
the latter half of the nineteenth century and grew exponentially through the
1960s. The puzzle we address is why this form of global interaction flowered
at this time, and with this pattern. The growth and dominance of multilate-
ral treaty-making as a solution to the problems generated by transnational
interactions began with a specific confluence of changes in the material
environment, ideational context and distribution of power resources of the
global system.
Regarding the material context, Murphy (1994) claims that the emergence
of multilateral practice is inextricably tied to the development of the global
capitalist system. As capitalist industry outgrows the physical boundaries of
the nineteenth and early twentieth century state, capitalist interests express
a preference for republican forms of government where they might have a
chance to counteract both the power of entrenched privilege and the pol-
icies of sovereigns bent on territorial aggrandizement rather than prosperity.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 197

What results is an ideology of liberal internationalism, supported by a coali-


tion of social forces and devoted to the creation of international institutions,
that might facilitate peaceful interaction, further integration and the result-
ing prosperity. Far from the blind choice of simple and immediate pecuniary
interests, this system offers us a variety of institutions with which to over-
come different short, medium and longer-term barriers to profitable inter-
actions. Those practices that garner sufficient backing, generating sufficient
resources to attract viable support, are ‘selected’ in the evolution of global
organization. Successful organizational forms then mould behaviour by
shaping the understandings and motivations that confront actors.
What Murphy seeks to understand are not multilateral treaties but the
great conferences that led to the development of various public unions in
the last half of the nineteenth and the earliest years of the twentieth cen-
turies. Some 105 multipurpose conferences gave birth to over 30 world
organizations designed to foster industry, manage social conflict and streng-
then states, societies and the state system. The great conferences had sev-
eral advantages, being easy to call, efficient in the breadth of their coverage
and popular with both political elites and a wide range of technical and
bureaucratic professionals. They also had advantages relative to treaties,
Murphy argues, because they allowed for the exposition of a problem,
its consideration by experts and the development of agreements without
the need to submit to a formal process of ratification. Though Murphy con-
siders conferences and public unions the more important form of multi-
lateralism in the early period, his explanation of the forces that led to the
emergence of multilateral practice in the 1850s is entirely relevant for our
purposes.
Reus-Smit (1997, 1999) provides a primarily ideational counterpoint to
Murphy’s material analysis, explaining the emergence of multilateral prac-
tices and contractual international law (especially permanent universal
conferences and international judiciaries, which are analogues to domestic
legislatures and courts), by referring to two deep constitutional structures:
the moral purpose of the state and norms of procedural justice. Reus-Smit,
like Murphy, finds the origins of multilateralism in the nineteenth century,
but rather than looking to the needs of capital, he focuses on ideational
changes. In particular, he claims that the ideological revolution in the late
eighteenth century changed how people and politics were considered. He
does not ignore the material context, and claims that ideational shifts, com-
bined with the industrial revolution, ‘profoundly altered the nature and
terms of intraterritorial governance, generating distinctively modern stan-
dards of legitimate statehood and rightful state action’. After the shift, the
‘moral purpose of the state’ became ‘increasingly identified with augment-
ing individuals’ purposes and potentialities’ generating ‘a new legislative
norm of procedural justice’ (1997: 577).
Changes toward ‘reciprocally binding social rules’ and ‘legislative norms
of justice’ filtered outward to the international sphere. The domestic values
of dominant states diffused, and ‘the principle that social rules should be
authored by those subject to them’ … ‘came to license multilateral forms of
rule determination …’ (ibid., pp. 568, 578). Such an expansion of domestic

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


198 COOPERATION AND CONFLICT 43(2)

norms into the global sphere was the catalyst for the multilateral order as
evidenced by the initiation of universal conferences, the proliferation of
treaties, and the formation of international courts. Keene (2007) observes a
similar ideational shift in the dominant legal discourses of the eighteenth
and nineteenth centuries that also facilitated increasing use of international
legal mechanisms.
The hallmarks of the multilateral practice that Reus-Smit analyses are
the most crucial multilateral agreements that set the parameters of inter-
national law and its adjudication. He (1997, 1999) provides an in-depth
analysis of The Hague Conferences of the late nineteenth and early twentieth
century, the Treaty of Versailles in 1919 and the founding conference for the
United Nations. In tying the emergence of multilateralism to ideational
concerns, Reus-Smit provides additional elements to Murphy’s primarily
material account and significant evidence for the existence of a multilateral
order that extends beyond the temporal boundaries of material incentives.
He claims that the norms of procedural justice that flow from ‘the moral
purpose of the state’ constrained institutional choices by shaping both cog-
nitive horizons and the range of legitimate actions. In the modern case, legis-
lative norms led the international community (or at least the dominant
states within it) to advance multilateral forms of interaction.
Ikenberry (2001) reminds us that powerful actors have a significant role
to play in the emergence of multilateral treaty-making as a potentially dom-
inant practice, and his account of its emergence diverges sharply from both
Reus-Smit and Murphy. For Ikenberry, multilateral order, or ‘constitutional
order’ as he describes it, is a potential solution for the problem that hege-
mons face at the end of major wars: how to institutionalize their advantages.
He (2001: 4) argues: ‘Major postwar junctures are rare strategic moments
when leading or hegemonic states face choices about how to use their newly
acquired power.’ A constitutional order is one of the choices, and it has sig-
nificant advantages in that it allows the new hegemon to show restraint and
cope with the mistrust that accompanies asymmetric power. By institutionally
binding itself into a multilateral order, the hegemon is able to achieve ‘buy-
in’ by smaller powers and institutionalize a durable, advantageous inter-
national order.
Different hegemons may institute different post-war orders, and
Ikenberry attributes the choice to the extent of power disparities and
fidelity to democracy (the more unilateral, the more readily a hegemon can
cement its preferred order, while more democratic hegemons can more
readily lock-in a constitutional order). While Ikenberry sees ‘hegemonic
establishment of constitutional orders’ and we observe a relatively un-
broken dynamic of interaction through multilateral treaty-making, his
reminder of the role of material power in reinforcing multilateralism is
well-taken.
The emergence in the 1850s of multilateral treaty-making as a potentially
dominant practice resulted from a confluence of material, ideational and
distribution of power factors. Industrialization altered the needs of power-
ful actors, providing a drive toward cooperative international interaction
at the same time that a new set of values favouring legislative justice and

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 199

reciprocity within dominant states was being externalized into the inter-
national community. These forces combined to build powerful momentum
behind the practice of multilateral treaty-making.

Acceptance and Spread of Multilateral Treaty-Making

What we see in the mid to late nineteenth century is the critical juncture
that Mahoney (2000: 513) urges those pursuing path-dependent arguments
to seek. A number of contingent forces led to the emergence of multiple
multilateral institutional possibilities — most notably public unions, multi-
lateral conferences, systems of arbitration and multilateral treaty-making.7
What we observe in subsequent decades is the growing dominance of mul-
tilateral treaty-making as the increasing returns to this institutional practice
reinforce its legitimacy and generate positive feedback. States orient them-
selves toward making multilateral treaties, altering their organizational
structures and shifting domestic resources. Their actions reify the appropri-
ateness of multilateral treaty-making and the practice spreads over time,
across space and to varied issues.
The secondary literature on multilateral treaty-making confirms the
plausibility of the social construction through the increasing returns argu-
ment. The hallmarks of increasing returns — high start-up and switching
costs, learning effects, coordination effects and adaptive expectations — are
evident in observers’ accounts of the multilateral treaty system. A brief dis-
cussion of each characteristic of increasing returns reveals how multilateral
treaty-making spread and came to be a dominant institutional feature of
world politics.
Both the start-up and switching costs of cooperation via multilateral treaty
are high. Murphy (1994) notes that it was only under the extreme pressure
applied by those defining the new economic system (both its dominant fig-
ures and popular sectors) that sovereigns altered their tendency to seek
aggrandizement through territorial conquest and moved toward a system
that facilitated increased wealth.This required extensive and carefully crafted
cooperative agreements. The social movements that emerged behind these
changes were costly to establish and difficult to decommission. Ikenberry
(2001) makes similar claims about the establishment of constitutional orders
and highlights the ‘stickiness’ of a system founded on multilateral treaties and
organizations. In addition, as multilateral treaty-making grew in legitimacy,
the costs of switching to a different interaction mechanism grow as well.
Once multilateral interactions became more frequent, it grew difficult to
alter the form or nature of that behaviour. Conferences may have had sig-
nificant advantages, but the agreements that emerged and lasted the longest
were those posed as contracts negotiated by experts, ratified by legislatures
and/or signed by sovereigns. Eventually the need to expand and update
cooperative agreements increased beyond the ability, willingness or neces-
sity of leaders and their vast entourages to travel. Once this happened, it
was difficult to even conceive of an alternative form for the updating and
further codification of binding agreements. These first movements changed

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


200 COOPERATION AND CONFLICT 43(2)

the cost of future choices. Adopted early on, it would have been quite diffi-
cult to prospect for, prove the value of and convince others to adopt differ-
ent norms. Norm-destroying wars of the sort that gave birth to modern
Europe after 1815 were few in number, and even before 1918 the tendency
was to settle them with the same mechanisms (conferences and treaties)
that had come to dominate international relations in the century prior.
Learning effects emerge quickly in the cooperative environment of con-
ferences and multilateral treaty-making. Not only are sovereigns satisfied
with what they have created, but also we see the almost immediate growth
of an international civil service staffed by what Murphy (1994) calls ‘public
systems builders’ with generic skill sets. Additional officials and bureaucracies
emerge to cope with the new activities. Specialists, and the agreements they
negotiate, eventually eclipse the conferences themselves.
What is generally observed is the emergence of a stratum of experts co-
evolving with the practice of multilateral treaty-making. As these public
system builders become more expert in the handling of a given issue area
at a conference or in the drafting of a treaty, expertise and cosmopolitan
experience become associated with the ability to navigate the great confer-
ences and create or service the important treaties that emerged. Murphy
(1994) traces this process through the nineteenth century and into the twen-
tieth. Sacriste and Vauchez (2007) illuminate how a heterogeneous popula-
tion of international lawyers came to define international law and alter the
orientation of states towards international law and treaty-making in the
late nineteenth and early twentieth centuries. Finnemore (2003) observed
similar dynamics in the politics of military intervention, where the disciplin-
ary growth of international lawyers led to a reinforcement of ideas about
the rule of law in the international system. Ikenberry (2001: 66) records the
same kind of support for the constitutional order, noting that:

[I]nstitutions are not just agreements, they are also interstate processes that
require state officials to engage in ongoing interaction with other states. This
requires state bureaucracies be organized in particular ways — with mandates,
missions and routines.

This trend tracks nicely with how Boli and Thomas (1999) and Barnett and
Finnemore (2004) describe a longer-term trend toward the acceptance of
rational-legal norms. Indeed, what emerges around the institution of multi-
lateral treaty-making is what Adler (2005) describes as a ‘community of
practice’. States and their agents come to expect multilateral interaction
and are geared toward multilateral treaty-making. In this way, multilateral
treaty-making structures state practices and is in turn the result of those
practices.
The end result of learning effects is that multilateral treaty-making gains
its own momentum as participation comes to be understood and advocated
by state functionaries as the composition of acceptable practice. Other com-
munities of experts follow suit. Murphy (1994: 112) notes that once great
powers began to interact in characteristic ways, ‘Governments and private
associations regularized their own work to match …’ Experts in solving the

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 201

problems of the state were defined as experts in dealing with conferences


and treaty-making.
The reorientation of states toward multilateral treaty-making has effects
beyond altering the costs of institutional change. As more and more states
begin to engage in the same sorts of endeavours, there are growing advan-
tages and a growing sense of appropriateness attributed to engaging in
more and more multilateral treaty-making, both into new issues-areas and
in the engagement of new actors. As these coordination effects emerge,
states adapt their practices so as to be able to participate in global life.
Multilateral interaction came to be perceived as the key to advancement,
given the needs of the new industrial markets that Murphy identifies. The
exact form of that interaction was not as important as was its existence. Any
organizational form that offered access and generated results would have
been seen as superior to lack of access to the growing global system. And,
once adopted, whatever advantages multilateral treaties possessed, their
continued and expanded use by actors in the system would significantly
magnify.
This is not to suggest that multilateral treaty-making was a wholly arbi-
trary formulation that could provide no more real advantage than any other
conceivable system. The co-evolution of industrial markets and sovereign
states posed no small difficulty, but emerged as the dominant political form
given the benefits the two systems offered one another (Chase-Dunn,
1981). The treaty system, which recognized the sovereignty of the actors
involved and facilitated cooperative interaction, was well suited to this con-
text. Given the material returns that states (and individuals and capitalists)
were receiving, and the ease with which this cooperation gained acceptance
with leaders so jealous of their sovereignty, the incentive to coordinate with
the dominant treaty-making system was maximized.
The final attributes of the process of increasing returns are adaptive
expectations. Like learning effects, not only do new actors attempt to co-
ordinate with the new ‘realities’ of the system, but also actors come to
expect that this will be both the current and future norm in international
relations. Individuals, states and organizations came to expect multilateral
activities and adjust their strategizing and actions to fit this expectation
(Jönsson and Hall, 2005). When states come to expect multilateral treaty-
making, this changes not only how they view interstate interactions, but also
how they view the global system in general. The multilateral treaty-making
lens helps them to define transnational problems, in some sense creating or
constituting these problems themselves.
The result of these material, ideational and institutional features is
uniformly positive feedback. Issues came to be viewed in terms of their
ability to be handled with some form of treaty. As states sought to deal
with them, treaties were not surprisingly the tools used. Other states
adapted, populations adapted and the future was defined in terms of the
application of more treaties to more issues. Material, ideational and institu-
tional factors served to reinforce the place of multilateral forms of inter-
action in the global system, and an exponential rise in multilateral treaties
was the result.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


202 COOPERATION AND CONFLICT 43(2)

Interrogating Social Construction through Path Dependence

Path dependence through increasing returns is a novel way to explain the


process of social construction — how an emergent practice comes to be taken
for granted and becomes a regular feature of the global social context — and
it plausibly captures the observed exponential rise in treaty-making. It is
novel because it both contains a specified mechanism that provides testable
implications and because it allows for the integration of ideational and mater-
ial factors that should be evident in a constructivist explanation. The social
construction/path dependence/increasing returns approach also provides a
number of additional conjectures about multilateral treaty-making that can
be explored with the MATRS dataset.
In this section, we suggest three non-obvious and verifiable indicators of
this explanation. These include the diffusion of treaty-making across issue-
areas, across geo-political contexts, and the continued importance of initial
locations/procedures. The relatively recent peaking and apparent decline of
treaty-making over the past 30 years may also emerge from our argument.
In each case, one or more of the four elements of path dependence (high
start-up and switching costs, learning effects, coordination effects and adap-
tive expectations) are implicated.
Our large-n examination of hypotheses about treaty-making departs from
traditional constructivist empirical analyses that call upon rich historical
description. This methodological move is necessary precisely because the
appropriateness of treaty-making is so ingrained that the forces shaping the
choice of this instrument in any particular instance are invisible. Once a
practice is so well ingrained that it is taken for granted, it becomes difficult
to apprehend empirically if analysis is restricted to individual treaties or
even sectoral treaty-making. Historical accounts and recollections are
inevitably concerned with the particulars of constructing agreements to
address the issues at hand and lend little insight into the influence of larger
patterns of practice. With a broader view of treaty-making practices, path-
dependent processes become more visible.

Diffusion across Issues — Switching Costs and Waves of Treaty-Making

One hallmark of the constructivist explanation is that practices come to be


taken for granted. For a number of ideational (framing how actors conceive
their world) and material (distributing the resources that actors use to
respond to their world) reasons, multilateral treaty-making becomes a lens
through which actors look upon and define transnational problems and
issues. If the argument is correct, we should observe the diffusion of multi-
lateral treaty-making practices to different transnational issue areas as they
arise over time. Disaggregating the treaty data into specific categories
reveals waves of treaty-making, just as the constructivist explanation would
predict. Consider the composite Figure 4. What we see is the exponential
pattern repeated in five of the six major substantive treaty categories with
different dates of initiation and peaking. As the international community

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


States and Relations
40

35

30

25

20

15

10

Number of Treaties Signed


5

0
1596 1616 1636 1656 1676 1696 1716 1736 1756 1776 1796 1816 1836 1856 1876 1896 1916 1936 1956 1976

Trade and Economy

100

90

80

70

60

50

40

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


30
DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING

Number of Treaties Signed


20

10

0
1596 1616 1636 1656 1676 1696 1716 1736 1756 1776 1796 1816 1836 1856 1876 1896 1916 1936 1956 1976
203
Communication and Tr anspor tation
204

35

30

25

20

15

10

Number of Tr eaties Signed


5

0
1596 1616 1636 1656 1676 1696 1716 1736 1756 1776 1796 1816 1836 1856 1876 1896 1916 1936 1956 1976

Environment
25

20
COOPERATION AND CONFLICT 43(2)

15

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


10

Number of Tr eaties Signed


0
1596 1616 1636 1656 1676 1696 1716 1736 1756 1776 1796 1816 1836 1856 1876 1896 1916 1936 1956 1976
Social Affairs
25

20

15

10

Number of Treaties Signed


0
1596 1616 1636 1656 1676 1696 1716 1736 1756 1776 1796 1816 1836 1856 1876 1896 1916 1936 1956 1976

War and Peace


50

45

40

35

30

25

20

15

Number of Treaties Signed


10

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


5
DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING

0
1596 1616 1636 1656 1676 1696 1716 1736 1756 1776 1796 1816 1836 1856 1876 1896 1916 1936 1956 1976

FIGURE 4
205

Multilateral Treaty-Making across Issue-Areas


This figure presents the number of treaties signed in each of 6 issue areas, ordered from earlier exponential take-off and peak to later, along with the
exponential trend line (except for War and Peace). Please contact the authors for a breakdown of treaty titles included in each graph.
206 COOPERATION AND CONFLICT 43(2)

confronted issues, it turned to multilateral treaty-making as the means for


addressing them over and over again.8 These waves produce the overall
exponential curve and they are evidence that the international community
identifies multilateral treaty-making as the mechanism for dealing with
global issues.
Furthermore, the figures support the argument that start-up/switching
costs inherent in systems with increasing returns constrain institutional
choices. When confronted with a new transnational issue, state leaders are
not likely to spend time thinking about new forms of interaction with which
to address the problem. Conceiving a new form of interaction to address
each transnational issue-area as it arises would be inherently inefficient.
There are also costs involved in switching to a new and untried institutional
form.9 The high cost of switching engenders a taken-for-granted quality
such that state leaders do not actively weigh the costs of continuing with a
given form of interaction. Thus, once certain forms of interaction emerge
and begin to lock-in, states continue to use these tools to approach prob-
lems in new issue-areas.
An extreme, though telling, example of how diffusion arises from path-
dependent lock-in is the fact that multilateral treaty-making migrated so
easily to issues of space exploration in the latter half of the twentieth cen-
tury. In this entirely new environment, one where the prerogatives of sov-
ereignty and international cooperation are ambiguous at best, multilateral
treaty-making is a standard way for states to interact. In the late 1980s,
March (1988: 328) noted that ‘for over two decades, multilateral treaties
have governed the spaceborne activities of nations in a variety of situa-
tions’. Prasad’s (2005) report on international efforts to deal with space
debris further demonstrates that multilateral treaty-making structures how
states deal with issues of ambiguous transnational character. The possibili-
ties for dealing with space debris appear to be bounded by amending exist-
ing treaties or creating new ones (Prasad, 2005: 247).

Diffusion across Space — Learning/Coordination Effects and Socializing


New Members

Demonstrating that multilateral treaty-making has migrated to most, if not


all, transnational issues is one hallmark of the effects of increasing returns.
Another is the diffusion of this practice from the community that adopted
it in the modern era (the West) to the remainder of the international sys-
tem. If there are increasing returns to multilateral treaty-making, not only
should existing states use it, but those that newly emerge should adopt (or
be made to adopt) the practice as well. Increasing returns facilitate the
socialization of new actors. The treaty data demonstrate this effect. In
Figure 5 we observe the spread of the locus of treaty-signing beyond the
West with the decolonization movements of the 1950s and 1960s. As new
states join the international community, even when their independence was
bitterly contested, they still chose to adopt multilateral treaty-making in
patterns similar to those of the more established members. Not only did

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 207

these states start signing treaties devised in the West, they began to negoti-
ate treaties on their own and host the negotiations and signings.10 There
were serious concerns in much of the Third World about Western imperial-
ism and neo-colonialism, but this did not appear to impact on the adoption
of (typically Western) multilateral treaty-making as a generic form of
interaction.
In a report on how recently independent states viewed the instruments of
international law, Abi-Saab (1962: 100) pointed out that such states ‘do not
easily forget that the same body of international law that they are now
asked to abide by, sanctioned their previous subjugation and exploitation’.
Even so, these states readily accepted treaty-making as a means of inter-
action with the global community (Abi-Saab, 1962; Sinha, 1965; Akintoba,
1996). International law and multilateral treaty-making served a purpose
for newly independent states because they define what it is to act like a
state. As Akintoba (1996: 21) notes ‘this need to communicate and trade
invariably makes necessary resort to the system of law originally created by
Western states for regulating interstate relations’.
This adoption or socialization demonstrates either learning effects or
coordination effects, depending on one’s perspective on the autonomy of
non-Western states. If one takes seriously the de jure sovereignty of states
in the periphery, and considers them to be autonomous actors with inde-
pendent initiative, then Figure 5 demonstrates the learning effects of
increasing returns. As states enter the international community, they find a
system of international interaction in place with its experts, diplomats and
the other technologies of multilateralism. States adopt this practice as part
of the process of becoming ‘modern’ (Finnemore, 1996).
On the other hand, if one questions the autonomy of peripheral states,
focusing on the lack of de facto sovereignty in some areas, coordination
effects appear to be driving this extension of multilateral treaty-making.
This interpretation recalls that increasing returns are evident when benefits
from a given practice increase to the extent that more participants adopt
them. From this perspective, it is in the interests of the West to incorporate
emerging actors into the extant multilateral order. Of course, both perspec-
tives are likely implicated in explaining Figure 5. States outside the West
have had varying levels of autonomy and thus varying abilities to ‘choose’.
In either case, our argument anticipates the spread of multilateral treaty-
making evident in Figure 5.

Stability in Centrality — The Continued Importance of Initial


Locations/Procedures

While evidence for increasing returns is found in the diffusion of treaty-making,


additional support is revealed in the stability of the organizational form. It is not
only the functional form that gets taken for granted; the procedures for opera-
tionalizing the practice are also internalized. Thus, we expect and observe
remarkable stability in the multilateral treaty system, here viewed in terms of
where treaties are negotiated and signed. While multilateral treaty-making has

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


208

120

100

80

60

40

Number of Tr eaties Signed


20

0
COOPERATION AND CONFLICT 43(2)

1648 1668 1688 1708 1728 1748 1768 1788 1808 1828 1848 1868 1888 1908 1928 1948 1968 1988

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


West Non-West

FIGURE 5
Number of Treaties Signed in Western States and Non-Western States in Each Year
DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 209

diffused around the globe, the sites where most of the treaty-making takes place
have altered only slightly. Figure 6 presents data on the top five locations for
treaty-making (where the most treaties are signed) in six different periods of
global politics.The identities of the top locations for multilateral activity remain
notably consistent over 400 years.
This geographical consistence demonstrates the functioning of adaptive
expectations. Not only do states adapt their actions to fit the expectation of
multilateral treaty-making, but they also adopt specific expectations of
where treaty-making will take place. Table 2 reinforces this finding by fur-
ther illustrating the continuity. Thirty cities could conceivably fill our six dif-
ferent ‘top five’ lists. We find only fifteen. Eight of those fifteen appear only
once, and seven of those eight appear only during the two short wartime
periods. This leaves a list of only seven cities that serve as the top treaty-
signing locations over a period of four centuries.
Going a step further in collaboration with geographer Herman van der
Wusten, we analysed the development and persistence of urban specializa-
tion in treaty negotiations. As early as the period from 1782 to 1849, we note
the focus of ‘Vienna in state relations…. London in others’. Although some
specializations emerge by issue-area, like Berne in the area of communica-
tion treaties, ‘those places that lack any specialization receive conferences
concerned with a wide variety of issues’ like London, Paris and Brussels
(van der Wusten et al., 2007). Once a particular location becomes an estab-
lished site of multilateral activity, this alters the way actors view appropri-
ate places to negotiate. These adaptive expectations are driven by a number
of factors, including learning and coordination effects. The cities in the top
five thus remain pre-eminent centres for treaty-making for long stretches of
time, beyond any rationale they may have had for being an early primary
location for negotiations in the first place.

Establishing Plausibility

The initial plausibility of our social constructivist/path dependence/increasing


returns argument was founded on its ability to explain the exponential rise
in multilateral treaty-making over the long term, the conjecture being that
treaty-making came to beget treaty-making. The test of this conjecture con-
sisted of empirically assessing a number of expectations that flow from it,
namely that if increasing returns are in place we should see a spread of
treaty-making across issues and actors, while also reinforcing the initial con-
ditions (viewed here as the matter of where treaty-making originally took
hold) of the multilateral system. All three of these expectations are borne
out by the MATRS data. We conclude that the practice of multilateral
treaty-making has emerged, grown and evolved through a process of social
construction by increasing returns — a process that is independent of the
content of individual treaties, that is resilient to change in the distribution
of power, and that shapes (rather than responds to) changes in the socio-
economic characteristics of the global community.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


210

600
1596–1849 1850–1914 1915–1918 1919–1938 1939–1945 1946–1995
500

400

300

200

Number of Treaties Signed


100

0
t
ue on na ris urg aris gue don sels ton don vsk res vre olm eva aris don ton gue don deo ton rlin cow eva sels aris don nna
ag nd ien Pa rsb P Ha on us ing on P on ing Ha on evi ing Be os en us P on e
e
H Lo V te e L r h L t -L ito ha Ha ckh en
c e o G L nt h M G Br L Vi
e h B a s s B u L t L sh e
o as
Th P T W re S W
a Th
M W
. B
St
COOPERATION AND CONFLICT 43(2)

City of Signature

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


FIGURE 6
Top Five Places for Signing Multilateral Treaties
DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 211

TABLE 2
Top Places for Treaty-Making
Number of
Place of Signature Eras in Top 5

London 6
Paris 4
The Hague 3
Washington 3
Vienna 2
Brussels 2
Geneva 2
St. Petersburg 1
Montevideo 1
Moscow 1
Berlin 1
Stockholm 1
Bucharest 1
Le Havre 1
Brest-Litovsk 1

The Future of Multilateral Treaties

In Figure 1, multilateral treaty-making grows exponentially from the 1850s


through the 1960s, peaks and declines. The decline predates the current pre-
occupation with the erosion of multilateral norms and institutions and pro-
vides some perspective on the current anxiety. This raises the question of
why a socially constructed, path-dependent process that enjoys increasing
returns should evidence decline. There are three possible conjectures that
appear from within the context of our research that would help us under-
stand this apparent anomaly and shed light on current concerns about the
fate of multilateralism.
First, this decline may be an artifact of the data. Treaties may take as long
as a decade to make it into the official record, and some may not be entered
at all. Linguistic and geographic biases are more prevalent in the sources we
used in the later years than earlier on. The resurgence of the Cold War in
the 1980s may have been a crisis large enough to interrupt the series, and
new data for the last decade, once gathered, may reveal a continuation of
the exponential pattern.
A second conjecture would find multilateral treaty-making in real decline.
Perhaps there are forces at work that are making the turn to this practice less
automatic and states are escaping from the path (Couch and Farrell, 2004).
These could consist of the familiar arguments about hegemonic unilateral-
ism, or the intriguing possibility that the forces of globalization are making
states less central relative to private regulatory regimes that manifest struc-
tures with historically different modes of interaction (Hall and Biersteker,
2002; Haufler, 2003).

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


212 COOPERATION AND CONFLICT 43(2)

Finally, a growth in treaty signings by international organizations may be


signalling another evolution in this multilateral practice. Ruggie (1993b: 9)
argues that the ‘adaptive and reproductive capacities’ of multilateral forms
are key to their effectiveness and persistence. Figure 7 provides evidence
for this adaptiveness, demonstrating the growing number of non-state sig-
natures on multilateral treaties. At first glance, the increase in treaties
signed by (and among!) international organizations serves to support our
argument, just as the adoption of treaty-making by peripheral states did.
But as international organizations grow more autonomous it is possible
that they, rather than treaties, will become the locus of multilateral diplo-
macy. Murphy described a similar transition in global processes during the
second half of the nineteenth century. Barnett and Finnemore’s (2004)
recent work on the autonomy of international organizations is suggestive
along these lines. Adding treaty data beyond 1995, an ongoing project, will
enable us to begin to adjudicate these various possibilities.

Conclusion

Treaties have long been a staple of global interaction, and this has been even
truer for the last century and one half. Their exponential increase since 1850
has been interrupted, but not generated, by global crises. The exponential
increase cannot be attributed to the growing number of actors, interdepend-
ence or hegemony. A plausible explanation rests with constructivism/path
dependence/increasing returns. Such an explanation is consistent with major
material, ideational and power distribution treatments of international
organization and cooperation. The constructivist explanation is supported
empirically by the replication of patterns across issue-areas, across political
and geographic contexts, and in terms of the uncanny consistency with which
certain cities serve as the locus of treaty signings.
These insights differ from those that could be generated by more trad-
itional, small-n, treatments of treaty-making. Substantive concerns regarding
particular issue areas and the effects of different designs provide important
information, but do not speak to the overarching role of treaty-making in
the global system as does this project. From this perspective, treaties are
more prevalent, and their exponential growth through the 1960s and 1970s
suggests a far more important role for cooperative interaction in the evolu-
tion of international relations than other perspectives can offer.
Specifically, the results of this study provoke us to consider the connec-
tions between the practice of multilateral treaty-making and the broader
norms of multilateralism. Ruggie (1993b: 7) argues that the generic multilat-
eral form entails more than arrangements of three or more states and
includes the generalized principles of indivisibility (participants see them-
selves as a unit), and diffuse reciprocity (participants expect roughly equiva-
lent benefits over time). Multilateralism is thus more than a quantitative
description; it has to do with the quality of an institutional arrangement
(ibid., pp. 7–12). Individual treaties can be signed for a number of reasons. It
would be folly to suggest that all treaties are consistent with the principles

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


100

90

80

70

60

50

40

30

20

10

Number of Treaties Signed by Non-State Actors


0
1840 1860 1880 1900 1920 1940 1960 1980

FIGURE 7
Non-State Treaty Signators. Number of International Organization Signators of Multilateral Treaties in Each Year

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING
213
214 COOPERATION AND CONFLICT 43(2)

of indivisibility and diffuse reciprocity. Some treaties (especially alliances)


are discriminatory and others do little to promote reciprocity. But multilat-
eral treaties are also consistently employed to serve the purpose that Ruggie
defines for multilateralism, namely to ‘define and stabilize international
property rights of states, to manage coordination problems, and to resolve
collaboration problems’ (ibid., p. 8).
While there is not an exact match between multilateral treaties and mul-
tilateralism, studying multilateral treaty-signing dynamics may have much
to offer us in studying the dynamics of multilateralism. It may contribute to
our understanding of where the global order has been and where it is going.
If the dynamics of multilateral treaty-making track the dynamics of multi-
lateral norms and institutions more broadly, we observe a profoundly multi-
lateral world in the MATRS data. Both banal and weighty matters are
addressed through multilateral treaty-making, whose role has expanded
and evolved over time. Sustained understanding of the global system may
depend on treating multilateral treaty-making and multilateralism as con-
stitutive elements of the global system and on seeking to understand the
path-dependent treaty-making dynamics that continue to unfold.

Notes

We acknowledge the research assistance of April Collins, Charlotte Freeman, Tom


Isherwood, Markus Lang, Lauren Twist, Stephen Walls and, especially, Hasan
Yonten. We also thank Mark Boyer, Yale Ferguson, Robert Johansen, Stuart
Kaufman, Phil Triadafilopolous and the three anonymous reviewers from Co-
operation and Conflict for insightful comments on previous versions — all remain
blameless for inadequacies that remain. This research was supported by the Center
for International Studies and the Undergraduate Research Program at the
University of Delaware.
1. We, of course, realize that signing treaties is only one dimension of the treaty-
making process; ratification, duration and compliance being other crucial dimen-
sions. However, treaty-signing is a significant endeavour and tracking the signing of
treaties over time provides an illuminating if less than complete picture of the treaty-
making practice.
2. The bulk of any linear relationship between the number of states and treaty
signings is actually an artifact of the post-1945 period, while the exponential expan-
sion of treaty-making begins well before 1945.
3. The suggestion that treaties of special importance emerge during periods of
hegemony is not supported by the preliminary data, and will be the focus of forth-
coming work.
4. This is contraposed to the literature that finds modern multilateralism result-
ing from an externalization of United States’ domestic preferences. See Martin,
1993; Ruggie, 1993a.
5. See Arthur, 1994; Pierson, 2000. Mahoney (2000) stresses the necessity of contin-
gency in the emergence of the ‘best practice’. As we discuss below, because significant
multilateral treaty-making was a response to industrialization, a highly contingent
event, this condition is satisfied in this case.
6. The classic example of this from economics is the QWERTY keyboard. This
configuration arose for rational reasons (the need to slow typists to stop typewriters

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 215

from jamming), yet its emergence as best practice closed off avenues of history that
might have led to the development of more ergonomic letter placement more
rational for the computer age. See Arthur, 1994.
7. In line with Couch and Farrell’s (2004) analysis of changes in paths, alterna-
tives to multilateral treaty-making do not necessarily disappear even when multi-
lateral treaty-making grows in prominence. Public Unions survive into the twentieth
century and large multilateral conferences saw resurgence in the late twentieth cen-
tury. This retention of latent institutional forms is entirely consistent with the path-
dependence argument and is, according to Couch and Farrell, a means for change in
paths.
8. It has not escaped our notice that treaties of war and peace depart from this pat-
tern. The politics of war and peace may be more crisis-driven than the bulk of inter-
national interaction.We take up the implications of this departure in forthcoming work.
9. Ikenberry (2001) considers these costs to be so high that nothing short of a
major war provides the opportunity to restructure the system.
10. Initial social network analyses of treaty signings indicate the emergence of
distinct, non-Western regional treaty systems (Hoffmann et al., 2007).

References

Abi-Saab, Georges (1962) ‘The Newly Independent States and the Rules of
International Law’, Howard Law Journal 8: 95–121.
Adler, Emanuel (2005) Communitarian International Relations: The Epistemic
Foundations of International Relations. London: Routledge.
Akintoba, Tayo (1996) African States and Contemporary International Law: A Case
Study of the 1982 Law of the Sea Convention and the Exclusive Economic Zone.
The Hague: Martinus Nijhoff.
Arthur, Brian (1994) Increasing Returns and Path Dependence in the Economy. Ann
Arbor: University of Michigan Press.
Barnett, Michael and Finnemore, Martha (2004) Rules for the World: International
Organizations in Global Politics. Ithaca, NY: Cornell University Press.
Beckman, Gary (1996) Hittite Diplomatic Texts. In: Harry Hoffner (ed.) Writings
from the Ancient World, Volume 7. Atlanta, GA: Scholars Press.
Beilenson, Laurence (1969) The Treaty Trap: A History of the Performance of
Political Treaties by the United States and European Nations. Washington, DC:
Public Affairs Press.
Boas, Taylor (2007) ‘Conceptualizing Continuity and Change: The Composite-
Standard Model of Path Dependence’, Journal of Theoretical Politics 19: 33–54.
Boli, John and Thomas, George, eds (1999) Constructing World Culture:
International Non-Governmental Organizations Since 1875. Stanford, CA:
Stanford University Press.
Bowman, Michael and Harris, David (1984) Multilateral Treaties: Index and Current
Status. London: Butterworths.
Chase-Dunn, Christopher (1981) ‘Interstate System and Capitalist World-Economy:
One Logic or Two?’ International Studies Quarterly 25: 19–42.
Cohen, Raymond and Westbrook, Raymond, eds (2000) Amarna Diplomacy: The
Beginnings of International Relations. Baltimore, MD: Johns Hopkins University
Press.
Couch, Colin and Farrell, Henry (2004) ‘Breaking the Path of Institutional
Development? Alternative to the New Determinism’, Rationality and Society 16:
5–43.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


216 COOPERATION AND CONFLICT 43(2)

Finnemore, Martha (1996) National Interests in International Society. Ithaca, NY:


Cornell University Press.
Finnemore, Martha (2003) The Purpose of Intervention: Changing Beliefs about the
Use of Force. Ithaca, NY: Cornell University Press.
Finnemore, Martha and Sikkink, Kathryn (1998) ‘International Norm Dynamics
and Political Change’, International Organization 52(4): 887–917.
Gleditsch, Kristian S. and Ward, Michael D. (1999) ‘Interstate System Membership:
A Revised List of the Independent States since 1816’, International Interactions
25: 393–413.
Goldstein, Judith, Kahler, Miles, Keohane, Robert and Slaughter, Anne-Marie, eds
(2000) ‘Legalization and World Politics’, International Organization 54.
Grieco, Joseph (1990) Cooperation Among Nations. Ithaca, NY: Cornell University
Press.
Hall, Rodney Bruce and Biersteker, Thomas, eds (2002) The Emergence of Private
Authority in Global Governance. Cambridge: Cambridge University Press.
Hasencleaver, Andreas (1997) International Regimes. Cambridge: Cambridge
University Press.
Haufler, Virginia (2003) ‘Globalization and Industry Self-Regulation’, in Miles
Kahler and David Lake (eds) Governance in a Global Economy: Political
Authority in Transition. Princeton, NJ: Princeton University Press.
Hoffmann, Matthew, Denemark, Robert and Isherwood, Thomas (2007) ‘It’s the
Relationships’. Paper presented at the Annual Meeting of the International
Studies Association, Chicago, 28 February.
Horne, Charles (1923) The Great Events of the Great War. New York: National
Alumni.
Ikenberry, G. John (2001) After Victory: Institutions, Strategic Restraint, and the
Rebuilding of Order After Major Wars. Princeton, NJ: Princeton University Press.
Jönsson, Christer and Hall, Martin (2005) Essence of Diplomacy. Basingstoke:
Palgrave Macmillan.
Keene, Edward (2007) ‘A Case Study of the Construction of International
Hierarchy: British Treaty-Making Against the Slave Trade in the Early
Nineteenth Century’, International Organization 61: 311–39.
Keohane, Robert (1984) After Hegemony. Princeton, NJ: Princeton University
Press.
Koremenos, Barbara, Lipson, Charles and Snidal, Duncan, eds (2001) ‘The Rational
Design of International Institutions’, International Organization, Special Issue 55.
Krasner, Stephen, ed. (1982) International Regimes. Cambridge, MA: MIT Press.
Leeds, Brett A., Long, Andrew G. and McLaughlin Mitchell, Sara (2000)
‘Reevaluating Alliance Reliability: Specific Threats, Specific Promises’, Journal of
Conflict Resolution 44: 686–99.
Leeds, Brett A. (2003) ‘Alliance Reliability in Times of War: Explaining State
Decisions to Violate Treaties’, International Organization 57: 801–27.
Mahoney, James (2000) ‘Path Dependence in Historical Sociology’, Theory and
Society 29: 507–48.
March, Scott (1988) ‘Law Aboard the Space Station’, Space Policy 4: 328–35.
Martin, Lisa (1993) ‘The Rational State Choice of Multilateralism’, in John Ruggie
(ed.) Multilateralism Matters, pp. 91–121. New York: Columbia University Press.
Mearsheimer, John (1994/5) ‘The False Promise of International Institutions’,
International Security 19: 5–49.
Mitchell, Ron (2003) ‘International Environmental Agreements: A Survey of Their
Features, Formation, and Effects’, Annual Review of Environment and Resources
28: 429–61.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 217

Mostecky, Vaclav, ed. (1965) Index to Multilateral Treaties. Cambridge, MA: Harvard
Law School Library Publications.
Murphy, Craig (1994) International Organization and Industrial Change: Global
Governance Since 1850. New York: Oxford University Press.
North, Douglass (1990) Institutions, Institutional Change, and Economic Performance.
Cambridge: Cambridge University Press.
Oye, Kenneth, ed. (1986) Cooperation Under Anarchy. Princeton, NJ: Princeton
University Press.
Pahre, Robert (2001) ‘Most-Favored-Nation Clauses and Clustered Negotiations’,
International Organization 55: 859–90.
Pierson, Paul (2000) ‘Path Dependence, Increasing Returns, and the Study of
Politics’, American Political Science Review 94: 251–67.
Prasad, M. Y. S. (2005) ‘Technical and Legal Issues Surround Space Debris — India’s
Position in the UN’, Space Policy 21: 243–9.
Reus-Smit, Christian (1997) ‘The Constitutional Structure of International Society
and the Nature of Fundamental Institutions’, International Organization 51:
555–89.
Reus-Smit, Christian (1999) The Moral Purpose of the State: Culture, Social Identity,
and Institutional Rationality in International Relations. Princeton, NJ: Princeton
University Press.
Ruggie, John G., ed. (1993a) Multilateralism Matters: The Theory and Praxis of an
Institutional Form. New York: Columbia University Press.
Ruggie, John G. (1993b) ‘Multilateralism: The Anatomy of an Institution’, in John
Ruggie (ed.) Multilateralism Matters, pp. 3–47. New York: Columbia University
Press.
Sacriste, Guillaume and Vauchez, Antoine (2007) ‘The Force of International Law:
Lawyers’ Diplomacy on the International Scene in the 1920s’, Law and Social
Inquiry 32: 83–107.
Sinha, S. Prakash (1965) ‘Perspective of the Newly Independent States on the
Binding Quality of International Law’, International and Comparative Law
Quarterly 14: 121–31.
Taylor, Michael (1987) The Possibility of Cooperation. Cambridge: Cambridge
University Press.
Thelen, Kathleen (2000) ‘Timing and Temporality in the Analysis of Institutional
Evolution and Change’, Studies in American Political Development 14: 101–8.
van der Wusten, Herman, Denemark, Robert, Hoffmann, Matthew J. and Yonten,
Hasan (2007) ‘The Production Sites of Multilateral Treaties: Political Center-
Formation and the World-System (1600–2000)’, in N. van Nuffel (ed.) Van
Christaller tot Wallerstein. Liber Amicorum Prof. Dr. Piet Saey, pp. 131–46. Nautilus
Academic Books Zelzate (Belgium).
Wiktor, Christian (1998) Multilateral Treaty Calendar. The Hague: Martinus Nijhoff.

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


218 COOPERATION AND CONFLICT 43(2)

Appendix I

There exists a great deal of information on multilateral treaties. The chal-


lenge, from the perspective of those who wish to understand general trends
in treaty-making behaviour, is that this information is usually presented
in formats designed to serve very specific purposes. The majority of indi-
viduals with a professional interest in treaties want to know the current
status of agreements in a given issue-area, or the current status of a given
state’s obligations. Hence there are several places to find the most current
versions of treaty texts, the most current set of signators and the most
current sets of agreements to which a given state may be party. Very few
students of multilateral treaties have been interested in cumulating infor-
mation on overall trends in treaty-making behaviour. This dataset was built
on those few sources, and supplemented with information garnered from
national or special-focus treatments.
The three largest contemporary sources of general information on multi-
lateral treaties are Christian Wiktor’s Multilateral Treaty Calendar, Vaclav
Mostecky’s Index to Multilateral Treaties and Multilateral Treaties: Index
and Current Status by M. Bowman and D. Harris (1984). The two most use-
ful older sources, both with something of a national orientation, include
William Malloy (for the first two volumes) and G. Charles (for volume 3),
Treaties, Conventions, International Acts, Protocols, and Agreements
Between the United States of America and Other Powers and Clive Parry,
Consolidated Treaty Series, 1648–1919.
We reviewed the material in each of these sources, noting every inter-
national agreement concluded between three or more states (or interna-
tional organizations) that emerged in written form and were understood to
be governed by international law. This is essentially the definition of the
term ‘treaty’ under the Vienna Convention on the Law of Treaties. We
recorded the title, date of signing, number of signators, identity of signators
and place of signing. We also recorded information on links to other treaties
and dates of coming into or going out of force. When this information
was not available, we searched other sources, including the United Nations
Treaty Series, national collections and specialty treatments, for missing
material. In all, we identified 6976 treaties, although certain elements are
still incomplete. For example, we only have a full list of signators for 6033
treaties, and complete information on place of signing is only available in
5957 cases. We continue to work on completing these elements.
We categorized the treaties into both specific and more general group-
ings. Our more specific categorizations were adopted from those used by
Wiktor in the largest of the contemporary compendia. He uses 282 catego-
rizations that were assigned ‘… following those used by the U.S. State
Department, Treaty Office, slightly modified, by subjects employed by the
United Nations in its treaty publications’ (Wiktor, 1998: xxii). We familiar-
ized ourselves with those categories so that we could categorize treaties not
included in Wiktor’s work. Categorization was based on consensus among
two faculty members, the graduate student in charge of the dataset, and

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010


DENEMARK AND HOFFMANN: MULTILATERAL TREATY-MAKING 219

either the graduate or undergraduate coders who were involved.A set of six
broader categorizations was created into which all sub-categories were
placed. Appropriate placement of about two-thirds of the sub-categories
into their broader categorizations was apparent, although the graduate stu-
dent in charge of the dataset, and both faculty members involved, reviewed
these choices. Those categories that were more difficult to fit were broken
down by treaty title, and all three of us discussed proper placement. All
decisions were based on review and consensus, and as a result there are no
concerns with inter-coder reliability.
In this article, we use version 1.2 of the dataset. It is not perfect. The
largest source is self-consciously biased toward treaties that were published
in either English or French (Wiktor, 1998: xxi). This is justified by Wiktor in
part given the requirements of the League of Nations and the UN to sub-
mit all treaties in one of those languages to its depository. Our second
largest sources (Mostecky, 1965) is less constrained by language, but ends in
1963. Version 1.3 of this dataset is in process, and will include treaties from
updates of the Mostecky dataset, as well as material from treaty calendars
of the USSR and China. It will also include materials from specialty sources
in areas like the environment and human rights. While the dataset is not
perfect, we believe it is sufficiently representative of multilateral treaty-
making to serve as the foundation of the large-n study presented here. We
look forward to future modifications and we welcome input.

ROBERT A. DENEMARK is an Associate Professor in the Department


of Political Science and International Relations at the University of
Delaware. His research and teaching interests include international politi-
cal economy, multilateralism and the historical evolution of the global
system.
Address: Department of Political Science and International Relations,
University of Delaware, Newark, DE 19716, USA.
[email: denemark@udel.edu]

MATTHEW J. HOFFMANN is Assistant Professor of Political Science


at the University of Toronto. His research and teaching interests include
global governance, social constructivism, multilateralism and environ-
mental politics.
Address: Department of Political Science, University of Toronto, 100 St.
George St, Toronto, Ontario M5S2K3, Canada.
[email: mjhoff@utsc.utoronto.ca]

Downloaded from http://cac.sagepub.com at Stockholms Universitet on April 7, 2010