Democratic Anchorage of Infrastructural Governance Networks: The Case of The Femern Belt Forum

Article
© The Author(s), 2009.

Reprints and permissions:
http://www.sagepub.co.uk/journalsPermissions.nav
Vol 8(3): 282–308
DOI: 10.1177/1473095209104827
http://plt.sagepub.com
DEMOCRATIC ANCHORAGE OF
INFRASTRUCTURAL GOVERNANCE
NETWORKS: THE CASE OF THE
F E M E R N B E LT F O R U M
Jacob Torfing, Eva Sørensen and Trine Fotel

Roskilde University, Denmark
Abstract Governance networks are often praised for their contribution

to making public governance and spatial planning more effective.
However, the democratic performance of governance networks is heavily
disputed. In order to make a more precise assessment of the democratic
quality of governance networks, we need to develop normative criteria
that permit us to measure the democratic quality of governance networks
on different dimensions. Such criteria are developed and brought
together in our model for the democratic anchorage of governance
networks. This article aims to improve the democratic anchorage model
in two different ways: by offering operational definitions of the basic
dimensions of the model and by demonstrating how the assessment
criteria can be applied in an empirical case study of a long-lasting, multi-
level governance network involved in the recent decision to build a bridge
between Denmark and Germany. The democratic anchorage model helps
to assess the democratic performance of specific governance networks
and to gain knowledge about the critical factors determining their degree
of democratic anchorage. Such knowledge is crucial for developing pro-
active strategies for enhancing the democratic performance of specific
governance networks.
Keywords democracy, governance, metagovernance, networks, physical

planning
282
Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Torfing et al. Democratic anchorage of infrastructural governance networks 283
1. Introduction
Citizens and stakeholders are increasingly involved in urban development,
spatial planning and decisions about large transport infrastructures such as
tunnels, airports and bridges. Public engagement is often brief and limited to
the pre-decision phase, where public hearings provide opportunities for people
to contest near-to-final planning proposals before they are endorsed by elected
politicians. Sometimes, however, public and private actors become engaged in
relatively long-lasting governance networks that may contribute to agenda
setting, joint planning, decision-making, or policy implementation (Healey,
2007; Hillier, 2002). Sustained interaction in relational networks is sometimes
preferred to the more episodic forms of citizen involvement, as it helps to
remove policy deadlocks by mobilizing the ideas and resources of private actors
and to reduce implementation resistance by creating joint ownership of public
plans and policies. Among politicians, public managers and planners, there is
growing interest in the formation of collaborative networks. Such networks are
often praised for their contribution to enhancing effective governance and
sustainable planning. While social scientists tend to agree on the positive impact
of governance networks on the effectiveness of public policy and spatial
planning, the democratic quality of governance networks is subject to dispute.
While some researchers claim that governance networks represent a threat to
democracy because they tend to undermine the principles and institutions of
representative democracy by enhancing the influence of unaccountable co-
alitions of self-interested elites (Hansen, 2007; Mayer et al., 2005), others argue
that governance networks help to strengthen democracy because they remedy
crucial shortcomings in representative democracy by enhancing participation,
deliberation and contestation (Dryzek, 2000; Klijn and Skelcher, 2007; Sørensen
and Torfing, 2007).
The current debate about the democratic quality of governance networks is
important, as it places the question of the distribution and exercise of political
power at the heart of the research agenda and helps counter the unfortunate
tendency to describe and analyse governance networks in purely managerial
terms as tools for pragmatic problem solving. However, the debate tends to
build on the assumption that governance networks are either inherently demo-
cratic or inherently undemocratic. This assumption is highly problematic, since
specific, real-life governance networks will tend to be more or less democratic,
depending on the concrete political and institutional conditions and depending
on the standards for measuring their democratic performance. In order to
capture both the theoretical and empirical complexity and ambiguity in the
democratic performance of governance networks, we require a general criteria-
based model that can help us to conduct an empirical assessment of the demo-
cratic quality of governance networks (Agger and Löfgren, 2008).
We have attempted to develop such a model elsewhere (Sørensen and
Torfing, 2005). Our model specifies a number of basic normative criteria for
measuring the democratic anchorage of governance networks. In this article, we
aim to take the model two steps ahead. First, we will translate each of its basic

284 Planning Theory 8(3)
dimensions into a detailed set of assessment criteria. Second, we will show how
these assessment criteria can be applied in an empirical case study of a long-
lasting, multi-level governance network that has been involved in the recent
decision to build a bridge between Denmark and Germany – the so-called
Femern Belt Bridge. After a clarification of what we mean by governance
networks, we briefly present the democratic anchorage model before elabor-
ating a detailed set of assessment criteria in relation to each of its basic
dimensions. In the conclusion, we reflect on our attempt to apply the criteria
for democratic anchorage in an empirical study and flag some of the yet
unanswered questions.
2. Governance networks
Governance networks play an important role in public governance and spatial
planning. Moreover, their role is unlikely to diminish in the near future, as
governance networks are increasingly seen as an effective means of dealing with
cross-cutting and wicked problems while enhancing democratic legitimacy by
linking governments and citizens (European Commission, 2001a, 2001b, 2002,
2005). Hence, there is a persistent growth in joined-up government, where
public agencies and private service providers form collaborative networks; a
surge in networked policy interactions among different levels of governance;
and a proliferation of public–private networks and partnerships. Governance
networks assume many different forms. Some networks are formal and
mandated from above, while others are informal and relatively self-grown.
Some are open and inclusive, while others are closed and exclusive. Some
pursue broad society-wide policy goals, while others have a narrow focus on
single issues. Some are preoccupied with knowledge sharing, agenda setting, or
policy formulation, while others are preoccupied with policy implementation.
The plurality of governance networks is well documented and has led to the
construction of a range of different typologies (Kickert et al., 1997; Marsh and
Rhodes, 1992).
In the face of this diversity of governance networks, one might ask whether
it makes sense to talk about governance networks in a generic sense, or whether
we should rather describe actual practices of networking in their concrete and
specific unfolding. We believe that it does make sense, as there are a number of
distinctive features of network-types of governance that permit us to define
governance networks as a general category that captures different forms of
public-private interaction. Hence, governance networks can be conceived a
stable articulation of interdependent, but operationally autonomous actors who
interact with one another through negotiations that take place within a regu-
lative, normative, cognitive and imaginary framework, which is self-regulating
within limits set by external authorities, and contribute to the production of
public purpose (Rhodes, 1997; Torfing, 2005). As such, the notion of governance
network points to the existence of institutionalized negotiated interaction
among relevant and affected actors from state, market and civil society, who

choose to pool their resources and coordinate their actions in the pursuit of
common understandings and shared goals that are deemed to benefit the
greater public.
Some of the political scientists who have searched for ways of improving
the democratic functioning of governance networks have been driven by the
assertion that governance networks are here to stay and that we should there-
fore seek to minimize the negative impact they might have on democracy
(Peters and Pierre, 2000). Other researchers tend to view governance networks
as a promising cure for the so-called ‘democratic disenchantment’ (Stoker, 2006;
Warren, 2002). As such, it is argued that governance networks can help
strengthen the interaction between politicians and citizens between elections;
supply additional channels of political influence to intensely affected citizens;
mobilize and empower private actors; and facilitate dialogue and collaborative
decision-making among multiple demoi (Bohman, 2005; Dryzek, 2000; Young,
2000). Regardless of whether governance networks are perceived as a potential
threat to democracy that should be mitigated or as a promising tool for re-
invigorating democracy, however, there is a fundamental question that must be
answered: How do we assess the democratic performance of a governance
network? In the attempt to answer this pertinent question, we have developed
a model for measuring the democratic anchorage of governance networks
(Sørensen and Torfing, 2005). Instead of focusing on how governance net-
works might help improve democracy in general, the democratic anchorage
model aims to provide some basic criteria for judging the democratic quality of
specific governance networks.
3. The democratic anchorage of governance networks

The basic argument underlying the democratic anchorage model is that the
democratic performance of governance networks can be ensured by anchoring
the interactive policy-making in the network in a series of relevant political
constituencies capable of lending democratic legitimacy to the network and in
a democratic grammar of conduct that provides a normative yardstick for
assessing the institutionalized negotiations among the network actors. As such,
we shall claim that governance networks are democratically anchored to the
extent that they:
1) are controlled by democratically elected politicians;
2) represent the membership basis of the participating groups and
organizations;
3) are accountable to a territorially defined citizenry; and
4) facilitate interaction in accordance with a commonly accepted democratic
grammar of conduct.
The first anchorage point expresses the need to establish close linkage
between representative democracy and a specific governance network so as to
ensure that democratically elected politicians are capable of monitoring and

influencing the policy-making that takes place in mandated or self-grown

governance networks. However, as persistently argued by several theorists of
democracy, the authority and regulatory capacity of the institutions of repre-
sentative democracy are unable to provide an undisputed source of democratic
legitimacy, as the link between voters and elected politicians has been system-
atically weakened by the decline in party membership and voter turn-out and
the pervasive role of mass media in organizing political communication
between the people and its representatives (Barber, 1984; Pitkin, 2004; Stoker,
2006). Therefore, we propose that this first anchorage point is supplemented by
the three other anchorage points. In fact, the model builds on the assumption
that none of the four anchorage points alone can ensure the democratic quality
of governance networks. The four anchorage points compensate each others’
shortcomings, and they provide a strong source of democratic legitimacy when
combined.
The four anchorage points re-invoke classical themes in liberal theories of
democracy. However, as we shall see, it is necessary to reformulate the classical
notions of liberal democracy in the context of governance networks and in the
light of new theoretical insights. Let us consider each of the four anchorage
points in turn.
3.1. Anchorage in democratically elected politicians

The basic rationale for anchoring governance networks in democratically elected
politicians is that government control with governance networks helps to ensure
that the structure, procedures and decisions of these networks are in line with
the popular will expressed by the political majority of elected assemblies. Hence,
elected politicians can lend democratic legitimacy to governance networks
insofar as they are able to control the formation, functioning and development
of governance networks.
Now, the problem is that the very idea that elected politicians can control
governance networks is flawed. The attempt to control governance networks
through top-down steering based on rules, commands, narrow targets and tough
sanctions will not only undermine their self-regulatory capacity (Mayntz, 1991),
but also be met by fierce resistance by the network actors, drastically reduce
their motivation to participate, and encourage them to play safe and avoid risks.
Hence, we must rethink the notion of political control in order to understand
how politicians can monitor and influence governance networks through subtle
and indirect forms of regulation that respect the self-regulatory character of
networks.
Following this line of argument, the governance network literature has aimed
to reformulate the notion of political control in terms of metagovernance, defined
as the ‘regulation of self-regulation’ (Jessop, 2002; Klijn and Koppenjan, 2004;
Kooiman, 2003; Sørensen, 2006; Sørensen and Torfing, forthcoming). A reading
of the literature on metagovernance points to four different tools which elected
politicians and other potential metagovernors can use in their attempt to
metagovern governance networks:

1) Network design that endeavours to determine the scope, character,

composition and institutional procedures of networks.
2) Network framing that seeks to define the political goals, fiscal conditions,
legal basis, and discursive storyline of networks.
3) Network management that attempts to reduce tensions, resolve conflicts,
empower particular actors, and lower the transaction costs by providing
different kinds of material and immaterial inputs and resources.
4) Network participation that aims to influence the policy agenda, the range of
feasible options, the decision-making premises, and the negotiated outputs
and outcomes.
With the conceptual displacement from ‘control’ to ‘metagovernance’, the key
question becomes how we measure the extent to which governance networks are
metagoverned by elected politicians. Here it might help to think about when and
how the metagovernance of governance networks can go wrong. First, the
elected politicians might not be aware that a particular governance network
exists, rendering it impossible for them to exercise metagovernance. Second, the
politicians may not want to assume responsibility for metagoverning governance
networks within their jurisdiction, and they may fail to clarify whether the differ-
ent governance network are engaged in ‘high politics’ and therefore need a
careful metagovernance or whether they are engaged in ‘low politics’ bordering
on administrative coordination and, therefore, require less, or even no, meta-
governance. Third, missing or inadequate information about what is going on in
the governance network will prevent adequate metagovernance, and the absence
of clear metagovernance objectives and failure to understand how these objec-
tives can be achieved through a combination of different metagovernance tools
will undermine the effectiveness of actual attempts to metagovern the govern-
ance network. Last but not least, it is a trivial fact that politicians are over-
burdened and often leave the lion’s share of metagovernance to the public
managers. Failure to coordinate with the public managers creates an imminent
risk of inconsistent and conflicting messages from the public metagovernors that
will seriously hamper the effects of metagovernance.
Reversing the arguments about metagovernance failure permits us to draw
up a list of positively formulated norms that can help us measure whether
governance networks are democratically anchored in elected politicians. The
metagovernance norms are summarized in proposition 1.
Proposition 1. Democratic anchorage in elected politicians can be measured by

paying attention to the following norms:
• The elected politicians must be aware of the presence, role and character of the
governance network in question and have access to information about its
processes, outputs and outcomes.
• They must assume the role of metagovernors and decide how important it is to
metagovern the network.
• They must define the objectives of their metagovernance and combine different
metagovernance tools to achieve these objectives.

3.2. Anchorage in the membership basis of participating groups and

organizations
The rationale for anchoring governance networks in the membership basis of
participating groups and organizations is that the members of these groups and
organizations constitute a ‘demos’ of directly affected people that the different
network actors must represent in order for the governance network to obtain
democratic legitimacy (Sørensen and Torfing, 2003). However, the classical
notion of representation is problematic as it erroneously assumes that a pre-
defined interest or preference is reproduced in an undistorted manner at the
level of representation (Laclau, 1993; Saward, 2005). The idea of an unbiased
one-to-one representation of a pre-given interest or preference is not only diffi-
cult to realize in network-based policy processes, it is also theoretically flawed;
it presupposes the presence of a fully constituted object of representation and
assumes that the process of representation does not add anything to the object
of representation. By contrast, we shall contend that the interests and prefer-
ences to be represented are never fully constituted, since they are formed within
a particular discursive context that is constantly destabilized by dislocatory
events that reveal the ultimately undecidable character of all social and politi-
cal identities. As such, the interests and preferences of political actors are
always marked by the ‘non-totalizable’ openness of the policy discourse in
which they are formed (Laclau, 1990). The ultimate failure to construct a fully
constituted preference or interest means that the performative act of represen-
tation in governance networks becomes constitutive of the interests and
preferences that it represents. The interests and preferences of the participat-
ing groups and organizations are defined by the way they are articulated by
their representatives in the governance network.
The represented may not agree with how their representatives construct
their preferences or interests, and they might not even recognize them as repre-
sentatives. However, the constructed representations cannot be seen as a mere
distortion of some primordial object since there was never a fully constituted
interest or preference in the first place; only a vague and ambiguous ensemble
of conceptions, wants and beliefs. Hence, the relationship between the network
representatives and the constituency that they claim to represent should not be
evaluated in terms of correspondence or non-correspondence, but rather in
terms of the degree to which the represented identify with the representatives
(Saward, 2005).
The extent to which the people that the representatives claim to represent
identify with their would-be representatives in a governance network is an
empirical question. However, we should be careful not to perceive identification
as an abstract question of whether people trust their would-be representatives
to ‘do the right thing’. Rather, identification should be evaluated in relation to
the concrete performance of the representatives and by scrutinizing the
conditions of possibility for the represented to critically evaluate this perform-
ance. The most basic condition for identification is that the participants in the
governance network are advancing more or less concrete and direct claims to
represent specific groups that are affected by the negotiated interaction in the

governance network (Saward, 2005). The advancement of such claims is import-

ant in order for a network actor to become recognized as a legitimate player by
the other actors in the network. However, it is not enough that the other network
actors envisage these claims to represent. The represented groups and organiz-
ations must be aware of the claims to represent and must identify with these
claims by constructing a ‘we’-relationship to the representative. The ‘we’-identity
does not construct a completely unified identity, but only a precarious bond
predicated on the construction of a ‘sameness’ that permits disagreements up to
the point where the ‘we’ is replaced by an ‘us and them’ division. The represented
groups and organizations must possess the resources and capacities necessary
for critically evaluating how the representatives construct and pursue particular
interests and preferences, and there should be regular opportunities for the
represented to assess the performance of their representatives and express griev-
ances and criticisms (ex post critique). Identification can be maintained despite
criticisms if the representatives demonstrate willingness to listen, learn and
adjust their performance in the governance network or if they succeed to
convince the critics that, under the circumstances, they have acted in their
interest.
We can now propose a list of positively formulated norms that can help us
measure whether governance networks are anchored in participating groups
and organizations. The identification norms are summarized in proposition 2.
Proposition 2. Democratic anchorage in participating groups and organizations can
be measured by paying attention to the following norms:
• Network actors must advance claims to represent specific groups and/or
organizations.
• The represented groups and/or organizations must be aware of these claims and
accept their validity by constructing a ‘we’-relationship to the representative(s).
• The represented groups and/or organizations must have the capacity and
opportunity to critically evaluate how their interests and preferences are
constructed and pursued by the representative(s).
• The representative(s) must be sensitive to criticism from the represented.
3.3. Anchorage in a territorially defined citizenry

The rationale for anchoring governance networks in a territorially defined
citizenry derives from the idea that democratic legitimacy can also be obtained
from the network being accountable to the citizens who are indirectly affected
by the decisions made by a governance network, for example, because the
decisions affect the overall allocation of resources or have repercussions for
other policy sectors. As such, the citizens living within the local, regional,
national or transnational territories in which a certain governance network is
making binding decisions should be able to hold that particular governance
network accountable.
The classical notion of public accountability assumes that the responsible
political decision-makers can be identified and held accountable for un-
acceptable policy outcomes (Behn, 2001). A fundamental problem with this
assumption is that complex social causalities often render it impossible to

determine the exact reason why a particular policy or project failed. There is
always the possibility that contingent circumstances beyond the control of the
political decision-makers caused the policy or planning failure. This problem is
acerbated in relation to governance networks that often share the responsibility
for policy implementation with a variety of administrative government agencies
that may affect the outcome of particular policy outputs in a negative manner.
Another problem is that the causal link between a particular policy output
and the decisions that brought it about can be difficult to establish, as political
decisions are often ambiguous, inconsistent and subject to competing interpret-
ations. Moreover, this problem is aggravated in governance networks, since it is
often impossible to identify clear-cut decisions in networks, as decisions are
results of complex and informal interactions in which problem definitions and
policy solutions are subject to ongoing revisions. Moreover, it is often difficult
to identify the responsible decision-makers, as governance networks rarely
resort to voting, aiming instead to establish a ‘rough consensus’, according to
which a certain decision is tacitly accepted, despite varying degrees of griev-
ance, protest and dissent.
Finally, there is the problem of sanctioning ill-conceived policy decisions.
Blundering politicians can be ousted from office in the next election. However,
since network actors are not elected, but most often (self-)appointed, these
options are unavailable in relation to governance networks. There might be
other ways of excluding failing network actors, but many private organizations
tend to monopolize the representation of particular groups of people, which
makes it difficult to find relevant and qualified substitutes for those who are
excluded. Alternatively, discontented citizens may press the politicians to cut
back funding and reduce the political influence of a governance network that
is responsible for clear and discernable policy failures. However, this will merely
tend to dismantle the governance network and replace it with more hierarchical
forms of top-down government that will be unable to deal with cross-cutting
and wicked planning problems.
As we have seen, the idea of a one-to-one correspondence among policy
outcomes, policy outputs and policy decisions is particularly problematic in
relation to governance networks, as is the idea of imposing sanctions. Neverthe-
less, ensuring the public accountability of governance networks is crucial if we
want to prevent governance networks from degenerating into closed and secret
clubs operating in the dark and without public approval (Dryzek, 2007; Fox and
Miller, 1995; Newman, 2005). Instead of tracing unacceptable policy outcomes
back to the responsible decision-makers in governance networks in order to
penalize them, we should therefore aim to ensure public contestation in terms
of critical public scrutiny of how governance networks define policy problems;
how they frame and select policy options; how they evaluate and make choices
among different solutions; and how they perceive the direct and indirect results
of the implemented policy. This introduces a new kind of narrative account-
ability whereby governance networks produce public accounts that are
contested by critical counter-accounts (March and Olsen, 1995).
There are three crucial requirements that must be fulfilled in order to
facilitate public contestation of narrative accounts produced by governance

networks. The first requirement is transparency. Governance networks and their

tasks, remit and composition should be visible to concerned publics, and they
must produce regular public accounts explaining why and how they arrived at
particular decisions and what the results have been. The public accounts ought
to be comprehensive, informative and accessible for laypeople. Overly brief
policy statements, misleading figures and arguments, and technical jargon will
leave the citizens in the dark as to what goes on in a governance network.
The second requirement is access to public dialogue with governance
networks. The narrative accounts of the deeds of various governance networks
can be made publicly available on the Internet, or through hearings, press
meetings, mass media coverage, or the presence of key network actors in
accessible public spaces where they can meet laypeople face-to-face. The
citizens should be able to ask questions, raise concerns, advance criticisms and
organize protests, and the members of the governance network should respond
by giving additional accounts addressing the issues raised by the citizens.
The third requirement is responsiveness on the part of the governance
network. Fostering a critical dialogue between the members of a governance
network and a territorially defined citizenry does not help much if the former
are unwilling to take eventual criticisms into account by adjusting the course of
action. Public contestation is reducible neither to the repressive tolerance of
critical voices nor to a polite dialogue with no effect. Public contestation must
be played out in a manner that promotes responsiveness on the part of the
governance network. Just as the critical citizens must be prepared to learn that
their criticisms are ill-founded or misguided, the network actors must be
prepared to respond positively to constructive proposals, to make concessions,
or at least to compensate the losers.
On the basis of these three requirements, we can now draw up a list of norms
for measuring the democratic anchorage of governance networks in the
territorially defined citizenry. The public contestation norms are summarized in
proposition 3.
Proposition 3. Anchorage in a territorially defined citizenry can be measured by
paying attention to the following norms:
• The tasks, remit and composition of a governance network must be fully visible to
concerned publics.
• The governance network must produce regular narrative accounts that seek to
justify its decisions, actions and results in the eyes of the broader citizenry.
• The network actors must engage in a constructive dialogue with those who are
publicly contesting their decisions, actions and results.
• The governance network must display some level of responsiveness towards
criticisms and alternative proposals raised in the public debate
3.4. Anchorage in democratic rules and norms

The rationale for anchoring governance networks in democratic rules and
norms is to ensure that governance networks are not only democratically
anchored in external political constituencies, but also in a set of democratic
standards regulating the processes and interactions proceeding within the
network. The democratic quality of a governance network also depends upon

the degree to which the network actors follow the rules and norms inherent
to a democratic grammar of conduct (Mouffe, 1993). The interweaving of
theoretical discussions of how to define democracy and political discussions of
how to institutionalize democratic forms of governance means that democratic
rules and norms are subject to endless dispute. Hence, the attempt to draw up
a complete and unbiased list of democratic norms and rules is doomed to fail.
At best we can draft an open-ended list of relevant and commonly accepted –
but ultimately contingent – rules and norms that can help us assess the demo-
cratic performance of governance networks.
Admitting the incomplete and ultimately political character of the demo-
cratic rules and norms constituting the grammar of democratic conduct in most
Western societies leaves us with another and even more delicate problem about
what it means to follow democratic rules and norms. We might be able to show
that some rules and norms are relevant, consistent and necessary for ensuring
democratic decision-making in governance networks, but we cannot expect the
governance networks and actors participating in these networks to follow such
rules in a mechanical and unproblematic way assuming that their form and
content are self-evident. Ludwig Wittgenstein (1986) has taught us that rule-
following is not as simple as it may seem. When the network actors have
succeeded to match a given situation with a certain rule or norm (March and
Olsen, 1989), the real problems have only just begun. For as Wittgenstein (1986)
demonstrates, rules are structurally ambiguous and there are always numerous
ways to interpret and follow a particular rule. The search for another rule that
can help us clarify the content of the first rule leads to an infinite regress, which
we can only escape by making our own constitutive reformulation of the rules
we encounter on our way. In that case, rules merely become an instance of their
usage, although there are clearly some outer limits to this usage since the actors
must be able to demonstrate that they act in accordance with broadly recog-
nized and recognizable rules (Staten, 1985). This means that the democratic
rules and norms can only be followed in and through a situational and con-
textual re-enactment, which is adjusting the rule to particular circumstances
while retaining a family resemblance with other usages of the rule.
The grammar of conduct that we claim should govern the pragmatic re-
enactment of democratic rules and norms includes three sets of normative
regulations that relate to the formation, functioning and development of
governance networks. The first set of normative regulations relates to the demo-
cratic demand for the inclusion of the affected actors. The norm that the citizens
should be able to influence the decisions that affect them is inherent to the basic
democratic idea of the self-government of the people (Dahl, 1989; Pateman,
1970). However, since governance networks are always predicated on the
inclusion and exclusion of more or less affected actors and since formal
inclusion is not a guarantee political influence, we must further qualify the
generally accepted norm regarding the inclusion of the affected actors (Dryzek,
2007; Fung and Wright, 2003; Saward, 2005; Young, 2000). The first rule concerns
the need to ensure that there is an ongoing debate about who should be
included and excluded. The second rule is that the negotiated criteria of

inclusion and exclusion should be transparent and consistently applied in

relation to all potential network actors; the excluded actors should be able to
understand and contest the reasons why they are not included. The third rule
is that all of the intensely affected actors that can and will contribute to the
networked policy process are included in the network. The ‘intensity’ of
people’s affectedness is a matter of degrees, and the same goes for their
‘inclusion’, as it is often possible to distinguish a core group of participants from
more peripheral groups of participants. Therefore, the rule might be that the
more intensely affected an actor is, the closer to the core group of participants
the actor should be. The final rule should be that none of the formally included
actors must be marginalized in a way that systematically prevents them from
influencing the actual decisions of the network (Young, 2000). Hence, the
political agenda and storylines defining the policy problems and the range of
appropriate solutions should be defined in broad terms that are open for
interpretation and reformulation, thereby permitting actors with different views
and preferences to put forth claims and ideas that are not excluded a priori from
the negotiating table (Hajer, 1995; Torfing, 2007).
The second set of normative regulations concerns the demand for demo-
cratic deliberation that is based on openness; agonistic respect for other
people’s opinions; commitment to reach a rough consensus; and a relatively
transparent decision-making process. All of the actors in the network should
have opportunity to express their views without being censored by strict
demands that one should be reasonable, articulate, moderate, civilized, etc.
(Young, 2000), and if they cannot find any support for their views, they should
be free to leave the governance network and form, or become part of, another
network. Democratic deliberation thrives on political disagreement, but
network actors with conflicting views and interests should not regard each other
as enemies to be eliminated by all possible means, but rather see each other as
adversaries who might strongly disagree with one another, but who respect each
others’ right to voice their opinion and fight for it (Mouffe, 2005). The network
actors may attempt to influence one another by means of controlling the
agenda setting, persuading and confronting each other, or manipulating their
opponents’ subjective perception of their interests (Lukes, 1974). But no matter
how much they bargain, negotiate and deliberate with each other, they should
be committed to attempt to reach an inclusive compromise in the form of a
‘rough consensus’, where conflicts and grievances remain, but are deemed
acceptable by those who disagree. Finally, decisions about public policy and
governance should be reached through a relatively transparent process where
everybody knows the terms of the debate, the available options and the final
decisions. Hence, standard procedures for ensuring decision-making trans-
parency through written agendas and minutes and a clear and visible leadership
should be observed.
The third set of normative regulations concerns the demand for a democratic
improvement of the interactive system of governance. Such an improvement
can be obtained by taking steps to ensure that the interactions within the
governance network contribute to developing the democratic identities of the

participating actors and to augmenting their political empowerment (Sørensen

and Torfing, 2003); however, it can also be obtained through an active search
for new institutional designs capable of contributing to a further democratiz-
ation of the overall system of participatory governance. Democracy never
reaches a final form; neither as a perfect set of institutions nor as a regulative
idea. Democracy must be constantly developed through a plurality of creative,
partial and experimental methods (Dewey, 1927), and governance networks
should be judged on their ability to spur such a development.
The three sets of normative regulations help us derive a list of norms for
measuring the democratic anchorage of governance networks in a democratic
grammar of conduct. The regulatory norms are summarized in proposition 4.
Proposition 4. Democratic anchorage in democratic rules and norms can be
measured by paying attention to the following norms:
• Inclusion and exclusion must be subject to ongoing negotiations, and these

negotiations should result in explicit criteria for admittance that should be
consistently applied.
• The degree of inclusion in governance networks should be a function of the
intensity of the actors’ affectedness, and the included actors should be able to
influence the decisions made by the governance network.
• The deliberation within governance networks must be governed by a democratic
ethos that ensures openness, agonistic respect, commitment to reach a rough
consensus, and a relatively transparent decision-making process.
• Participation in the governance network must contribute to the enhancement of
the political empowerment of the participating actors.
• The governance network must stimulate democratic innovation through self-
reflexive and experimental processes.
4. Applying the democratic anchorage model

The four propositions outlined above aim to translate the basic norms in the
original model of democratic anchorage into a detailed set of assessment
criteria that can be used in the empirical measurement of the democratic
performance of governance networks. In principle, a specific governance
network can be evaluated on the basis of all of the dimensions and criteria
explicated above. As a general rule, the more politicized a governance network
is, the more important it becomes for it to perform well on all of the different
dimensions and norms. However, we cannot expect all governance networks to
perform equally well on all four dimensions. Not only have we set the bar high,
but there are also structural conditions pertaining to the form and character of
governance networks that will prevent them from doing equally well in relation
to all four anchorage points. Hence, when evaluating and comparing the demo-
cratic anchorage of different governance networks, we must bear in mind that
certain types of governance networks are likely to fare quite well in relation to
certain anchorage points while doing less well in relation to other anchorage
points. In other words, we should conduct a pragmatic and contextualized
assessment of the democratic anchorage of governance networks that focuses
on the most relevant and pertinent anchorage points.

The fourth anchorage point demanding that the negotiated interaction within
a governance network should be anchored in a democratic grammar of conduct
is relevant and important in relation to all kinds of governance networks.
Hence, a governance network that fails to re-enact commonly accepted demo-
cratic rules and therefore tends to make undemocratic decisions is highly
problematic and calls for some kind of political intervention. However, the
relevance and importance of the three other anchorage points depend on the
form and character of the governance network in question.
The first anchorage point that emphasizes the need to anchor governance
networks in metagoverning politicians is more relevant for governance
networks that are mandated from above than for governance networks that are
self-grown from below. As such, it will often be quite illegitimate, and also diffi-
cult, for elected politicians and public managers to metagovern self-grown
governance networks formed by individual citizens, civil society organizations
and business firms aiming to enhance the sustainability of a small village in the
countryside or to settle disputes over multiple uses of land in a rural area. From
a democratic point of view, metagovernance strategies aiming to ensure the
three other anchorage points might be in high demand, but public authorities
will have great difficulties legitimizing even the most indirect and subtle forms
of political interference with a private, self-grown governance network. The
exceptions to this general rule are those cases where an overly private, self-
grown governance network is publicly funded, has been delegated competence
to solve public tasks, or tends to make political decisions of great importance
to the wider community. In these cases, public metagovernance is important and
warranted, but private, self-grown governance networks that are merely
engaged in a self-regulated governance of civil society should enjoy an
outstretched autonomy and, therefore, cannot be expected to perform well in
relation to the first anchorage point.
The second anchorage point that aims to anchor governance networks in the
membership basis of the participating groups and organizations is more
relevant for governance networks engaged in agenda setting or policy formu-
lation than for governance networks that are merely engaged in knowledge
sharing or policy implementation. Hence, if governance networks pave the way
for or make authoritative decisions that are crucial for the groups and organiz-
ations that are represented, or if these groups and organizations are expected
to deliver active support to such decisions, then the anchorage of the network
actors in their hinterland is extremely important. By contrast, if the governance
networks are merely preoccupied with exchanging relevant information or
coordinating the actions of different groups and organizations in order to
prevent implementation problems, then a firm anchorage in their membership
basis might not be as important, since there are few decisions and activities that
will interest the members and call for their approval or disapproval.
The third anchorage point that insists on anchoring the output and
outcome of governance networks in a territorially defined citizenry is highly
relevant for governance networks that are involved in decision-making with
repercussions for other policy areas and for large sections of the population
within a particular jurisdiction. By contrast, governance networks dealing with

relatively isolated problems of little concern for the broader citizenry cannot
be expected to be firmly anchored in a critical public scrutinizing their policy
decisions.
These reflections merely serve to illustrate the need to contextualize the
empirical application of the democratic anchorage model. It is impossible to
make iron-cast rules for the relevance of the different anchorage points, since
everything ultimately depends on a concrete assessment of the form and
character of the governance network in question. Contextualization is also
needed when interpreting and acting upon the results of the assessment of the
democratic anchorage of a specific governance network. The results of the
measurement of the strength of each of the four anchorage points will not be a
question of either/or, but a matter of degrees (in the following measured as
‘weak’, ‘moderate’, or ‘strong’), and the results will only make sense when the
political and institutional conditions for the democratic performance of the
governance network in question are taken into account.
The democratic anchorage of the Femern Belt Forum 1

The proof of the pudding is in the eating. We shall therefore aim to show how
the normative criteria elaborated above can be applied in an empirical
measurement of the democratic anchorage of a specific governance network.
The governance network that we have chosen for this purpose is a long-lasting
infrastructural network that has played a major role in promoting the recent
decision to build a bridge over the Femern Belt separating the Danish island
Lolland from the northern shores of Germany. The Danish governance network
was initially called the ‘Idea Committee for Infrastructure’ (1994–2003) but was
later renamed the ‘Femern Belt Forum’ (2003), which is the label we shall
apply here.
The first attempt to reach an agreement about the building of a bridge over
the Femern Belt was made at a summit meeting in 2006 between Danish Prime
Minister Anders Fogh Rasmussen and German Chancellor Angela Merkel.
Despite the high hopes of the Femern Belt Forum, no final decision was made
on that occasion. This generated much frustration among the network actors
who felt that they had been very close to reaching the goal that they worked
for since the early 1990s. However, frustration was turned into applause on
29 June 2007 when a bilateral government agreement to build a bridge was
finally signed. The agreement states that the Femern Belt Bridge is to be
financed by user fees (bridge toll) and that the Danish government will cover
all costs in case of unforeseen fiscal problems. The EU has agreed to cover
20 percent of the total costs.
The idea of building a bridge between Germany and Denmark goes a long
way back. The first mention of a bridge over the Femern Belt dates back to the
14th century, and concrete plans to establish a road connection from the north
of Scandinavia all the way south to Germany were advanced as early as 1914
(Bech, 1914). These plans were further developed in the 1940s and 1950s
(Danish Railroad Company, 1940; Erngaard, 1952), and after a relatively quiet

period in the 1970s and 1980s, the idea of a bridge to Germany resurfaced in
connection with the building of the Great Belt Bridge and the Øresund Bridge
in the 1980s and 1990s.
The construction of the Øresund Bridge connecting Denmark and Sweden
meant that the Femern Belt Bridge could be re-launched as the ‘missing link’
that could help realize the old dream about a beeline road connection between
Scandinavia and Germany. In 1994, local mayors from the southern part of
Zealand, together with a group of public administrators from the Regional
County, therefore took the initiative to form a governance network with the
declared intention of gaining political support for the building of a bridge
between the small Danish town of Rødby and the Femern peninsula on the
northern shores of Germany. A bridge was seen as a central means for promot-
ing economic growth in the southern region of Zealand, which had long suffered
from economic decline and mass unemployment.
The Femern Belt network aimed to influence the political agenda, build a
strong multi-level coalition, and prepare the southern region of Zealand for the
building and future exploitation of the Femern Belt Bridge. One of its central
tasks was to lobby the Danish government and parliament, both of which had
displayed limited interest in the project. The reluctance of the political parties
in the Danish parliament was due to the predominance of MPs from Jutland
who saw highway construction in Jutland as the top priority. Since its formation
in 1994, the governance network has undergone several reorganizations,2 but
has continued its activities.
The Femern Belt Forum has brought together a large number of public and
private actors in a multi-level governance network that was highly successful in
lobbying and preparing for the construction of a Femern Belt Bridge. In that
sense we might judge the governance network to be rather effective. However,
the question remains: to what extent has the governance network been demo-
cratically anchored?
Building a 28-kilometre-long bridge over the Femern Belt is a controversial
decision due to the high costs and uncertain consequences. The Femern Belt
Forum was initiated by local and regional government officials, and it has had
a considerable influence on the political agenda and the actual policy making
in both Denmark and Germany. The policy output will affect traffic patterns,
environmental sustainability and regional development, and it will have a
significant impact on citizens and corporate actors on both sides of the Femern
Belt. As such, we can conclude that all of the four anchorage points are relevant
in the assessment of the democratic anchorage of the Femern Belt Forum.
First anchorage point: democratic anchorage in elected politicians

The elected politicians at different levels of government were undoubtedly
aware of the presence and role of the governance network. Not only was the
Femern Belt Forum a formal network initiated by local and regional politicians
and administrators, it was also highly visible in the sense that it contacted
elected politicians at all levels and at both sides of the border through various
kinds of lobbying activities such as petitions, meetings, conferences, etc.

However, the elected politicians in the national government and Danish

parliament did not assume the role of metagovernors vis-à-vis the governance
network. Due to the lobbying role of the governance network, it was the
governance network that contacted the national politicians rather than the
other way around. Former Minister of Industry Nils Wilhjelm was appointed
chair of the Femern Belt Forum in 1994, but he was selected by the network
because of his connections to the government and his ability to negotiate and
obtain results; he did not act as a government representative. Nevertheless, the
political authorities at the local and regional levels metagoverned the govern-
ance network in several ways. A number of elected politicians from the local
municipalities and the Regional County participated in the governance
network; the County created an administrative support unit; and the new
regional body entitled Growth Forum aimed to frame the activities of the
governance network, both politically and economically.
The politicians and administrators at the local and regional levels worked
closely together in order to support and develop the Femern Belt Forum. They
conceived the network as an important tool for mobilizing local actors, aligning
interests and gaining national support for a Femern Belt Bridge. The local and
regional authorities helped the network obtain funding from the EU structural
funds programme (INTERREG). This added a transnational dimension to the
metagovernance of the network. Hence, part of the EU funding was used to
finance a series of expert reports on the environmental (VVM) and economic
consequences of the bridge.
The multi-level metagovernance exercised by political authorities at the local,
regional and transnational levels is important for ensuring democratic anchor-
age, but it is noteworthy that the national politicians, who were going to make
the authoritative decision about whether or not to build the bridge, failed to
metagovern the governance network. Hence, whereas the local and regional poli-
ticians supporting the Femern Belt Bridge joined the network and actively
supported its development, the national politicians, who were much more reluc-
tant and constituted the primary target of the lobbying activities of the network,
did not attempt to influence the proceedings in the governance network. The
national politicians did not formulate any clear metagovernance objectives and
did not combine hands-on and hands-off metagovernance tools. The Ministry of
Traffic issued a few, but rather comprehensive, reports with calculations of the
total costs and estimations of the effects on the transport sector, but it did not
really aim to metagovern the Femern Belt Forum. Few individuals within the
Ministry took an interest in the Femern Belt project, and although they had some
influence through their participation in an informal, national network referred to
as the ‘Friends of the Femern Belt Bridge’, the Ministry of Traffic did not attempt
to frame or direct the activities in the Femern Belt Forum. At some point, the
Ministry predicted that a final decision on the Femern Belt Bridge would be
made in 2000, but that prediction later proved to be completely misguided and
it was ridiculed by the participants in the local and regional governance network.
In sum, the Femern Belt Forum has been democratically legitimized by the
multi-level metagovernance exercised by the local and regional politicians and

administrators, and to some extent also by the transnational political

authorities in the EU, although we should bear in mind that the link between
the structural fund programmes and democratically elected politicians in the
EU is extremely weak. However, democratic legitimacy could have been
further enhanced if the national politicians – who were one of the key targets
of the lobby activities of the network – had assumed responsibility for meta-
governing the Femern Belt Forum. The failure to do so is a serious drawback,
and we shall therefore conclude that the democratic anchorage of the Femern
Belt Forum in elected politicians is: MODERATE.
Second anchorage point: democratic anchorage in participating

groups and organizations
In the negotiated interaction within the Femern Belt Forum, it neither seems to
be important who the participating actors represent nor whether they represent
anybody at all. Representation is apparently not an issue in the network. The
only thing that counts in the governance network is the knowledge, resources
and energy that the different network actors bring to the table. What really
matters is how the actors can contribute to achieve the common objective.
However, the absence of explicit references to ‘representation’ in the network
does not mean that the individual network actors do not claim to represent
particular groups or organizations. Rather, it means that the network actors have
free hands to construct themselves in terms of whom or what they represent.
As a result, there are huge differences in how the network actors perceive
their role as representatives. For example, there is a newspaper editor who
claims that he both represents the supporters and the critics of a bridge, and
two representatives from environmental organizations claiming to represent
the public interest in protecting the environment. There is a local trade union
leader who claims that he only represents himself, and another trade union
leader who states that he represents both the regional labour market and the
Regional Labour Market Council. There is also a local business leader claiming
to represent a number of specific business and employer organizations. He does
not rely on their active support, reckoning instead that they would tell him if
they disagreed with what he is doing in the governance network. Finally, there
is a large group of politicians from different local municipalities and the new
Regional Council, all of whom claim to represent the elected political body they
come from. The claim to represent a particular level of government is very
important to the politicians, because there is an escalating struggle between
the municipalities and the region about the future control over the Femern
Belt project.
Unfortunately, our data do not permit us to conduct a thorough evaluation
of whether the represented groups and organizations accept the validity of the
network actors’ claim to represent them. Thus far, however, there have not been
any protests against the participants in the governance network from the key
stakeholder groups; only a few minor disputes. Some of the local businessmen
have contacted the business leader in the network about their growing concern
for the emergence of bottlenecks in the local labour market, but that merely

proves that they consider him to be their representative. The absence of protests
directed against the individual network actors might reflect the stakeholder
groups’ ignorance of the governance network, but a more likely explanation is
that the stakeholders feel that, until now, the members of the governance
network have represented what all the stakeholders perceive to be their
common interest, namely the building of a bridge to Germany that will spur
regional growth and development.
One thing that we can say something about is the degree to which the repre-
sentatives have had regular contact with the groups and organizations they
claim to represent. Regular contact is important, as it creates opportunities for
the represented groups and organizations to critically evaluate the way that
their interests and preferences are pursued by their would-be representatives.
Again, we have a varied picture. The newspaper editor has received letters from
the readers, but they were not directed to him as a network participant. The
environmental representatives did not discuss the Femern Belt Bridge within
their respective organizations, relying instead on their own expert judgement.
One of the trade union leaders reports that he has had close and regular contact
with his hinterland throughout. He regularly discusses the agenda and
upcoming issues and decisions with the other members of the Regional Labour
Market Council. The business leader has not consulted with the organizations
that he claims to represent, merely anticipating their reactions instead. The poli-
ticians have participated in regular meetings in their respective political bodies,
but these meetings do not provide opportunities for critical feedback.
As stated above, there have been very few cases of dispute between the
representatives and their hinterlands; however, our data indicate that the repre-
sentatives have been sensitive to the criticism that has been raised and
attempted to justify their actions and ideas vis-à-vis their constituencies.
In sum, the democratic legitimacy of the Femern Belt Forum has been
enhanced by the relatively well-developed relations of representation between
the individual network actors and the groups and organizations they claim to
represent. However, the political significance of the anchorage of the network
actors in the represented groups and organizations is mitigated by the fact that
the representation of different groups and interests hardly plays any role in the
governance network, because everybody agrees with its overall objective.
Nonetheless, it is important for the democratic anchorage of the network that,
with a few exceptions, there is a relatively close link between the representa-
tives and the represented, and that the latter seems to identify with the former.
Hence, we can conclude that the democratic anchorage of the Femern Belt
Forum in the participating groups and organizations is: STRONG.
Third anchorage point: democratic anchorage in a territorially

defined citizenry
The meetings in the Femern Belt Forum are not open to the general public. The
meetings were open in the beginning, when the local press was invited to the
meetings as part of the network attempts to create consensus and mobilize local
resources behind the project. Later, however, when the network thought that

the political battle was won, the number of participants was reduced and the
local press was no longer part of the network. While the close relations to the
overly positive local press persisted, that did not necessarily enhance trans-
parency. There have been at least two incidents in which the governance
network attempted to get the local press to scale down its coverage of the oppo-
sition to the bridge. Nevertheless, the local citizens continued to enjoy access to
information about the network through the local press. The rest of the Danish
population has not had the same opportunity, since national television and
newspapers did not cover the Femern Bridge project very well. This can be
explained as a result of a public fatigue after the completion of two other major
bridge-building projects. There has been some coverage in national professional
magazines, such as The Engineer, but the debate has been of a technical nature
and has not found its way to more broadly circulated mass media.
The interested citizens could obtain information about the Femern Belt
Forum by visiting its website, where the members and organizational structure
of the network were displayed. They could also obtain access to numerous
highly relevant documents and reports about the Femern Bridge project as well
as an overview of the activities organized by the network. However, critical to
the possibility for public contestation, the agendas and minutes from the
meetings in the Femern Belt Forum were not available on the website.
The governance network produced a number of written and oral reports
about the Femern Bridge project. The original storyline was that the Femern
Belt Bridge would transform the island of Lolland from an economic back-
water to a prosperous economic centre without harming the environment and
without any costs to the taxpayers, as the bridge would be financed by user fees.
The storyline later focused on how the cross-border region would cope with the
bridge-building project and reap the fruits of the bridge. The reports and
discussions were relatively open in the sense that they also addressed critical
issues as long as they did not cast doubt on the relevance of the bridge. All this
changed in 2000, when the governance network believed that the Danish
parliament would formally endorse the Femern Bridge project. When this did
not happen and the political process stalled, the governance network became
extremely tense and less tolerant vis-à-vis critical voices.
This change seriously affected the ability of the governance network to
engage in a critical dialogue with people contesting their storyline and those
aiming to broaden the project by introducing new elements to the storyline.
While the storyline and dialogue became more rigid, closed and totalizing, a
certain willingness to discuss with opponents remained. For example, a group
of German critics from Femern were invited to discuss the project at a con-
ference organized by the Femern Belt Forum. Hence, the dialogue with the
territorially defined citizens on both sides of the border remained, although the
network actors found it increasingly difficult to respond to critical voices and
alternative suggestions, as too much was at stake. For example, the location of
the Femern Bridge was not open to discussion, despite good arguments for
other placements. The willingness to discuss with the opponents differed among
the network actors. Some network actors continued to engage in a constructive

and respectful dialogue with people outside the network; others developed a
much more condescending attitude, claiming that they had ‘wasted too much
time on the critics’, and that these ‘spoke against better judgement’. There was
also an incident in which some of the network actors declined an invitation to
attend a meeting in Germany to discuss the project.
In sum, we can conclude that democratic legitimacy has been enhanced by
the visibility and transparency of the governance network that permitted public
scrutiny of its storyline and lobbying activities. The Femern Belt Forum has not
been a hidden and secluded network, where decisions are taken in smoke-filled
rooms. There are reasonably good opportunities for the citizens to learn about
the network through the local press and the website. The more indirectly
affected citizens have been able to raise criticism, and the governance network
has engaged to a certain point in a constructive debate with opponents. At the
same time, there is considerable room for improvement on at least three differ-
ent counts. First, there has been little coverage and debate about the Femern
Belt Bridge in the national press. Second, agendas and minutes from the
meetings in the Femern Belt Forum have not been available on the website.
Third, the governance network has deliberately aimed to close down the public
debate in the years preceding the final decision out of fear of ‘unnecessary
politicization’. Hence, we can conclude that the democratic anchorage of the
Femern Belt Forum in a territorially defined citizenry is: MODERATE.
Fourth anchorage point: democratic anchorage in democratic

rules and norms
The criteria of inclusion and exclusion of network actors changed over time, but
there was hardly any discussion of the criteria in the governance network, and
they were never made explicit. This has made it difficult for the outside world
to understand who could become a part of the network and who not; however,
we have not heard of any actors that have been denied access.
The implicit criteria of inclusion were formulated by Nils Wilhjelm, chair-
person of the Femern Belt Forum, who favoured a broad inclusion of all of the
actors interested in improving the local infrastructure and building a bridge to
Germany. Hence, the governance network initially included people from the
municipalities, the County, the central labour market organizations, financial
institutions, educational institutions, tourism, agriculture, local business, etc. The
critics of the bridge project, including environmental groups, the local ferry
companies, the national railway company (DSB), and the local press were also
included in the network. As the administrative leader of the Femern Belt Forum
secretariat remarks: ‘We believed that it was better to include them in the
discussions than to have them standing on the sidelines and shouting at us.’ The
broad inclusion of interested actors was not motivated by democratic concerns;
instead, it represented an attempt at building a strong supportive alliance. Later
on, in 2000, when the network actors thought that the final approval of the
Femern Belt Bridge was near, the number of network actors was reduced and
the governance network became less inclusive. This change was largely
accepted by the network participants, who believed that the job was done and

could now be left to a few lead actors – organized in an executive committee –

to see the project through. After the disappointment with the failure to obtain
the final approval, the governance network again became more inclusive, as it
now incorporated a number of local and regional politicians who could resume
the lobbying activities. Recently, after the final decision has been made, some
of the network actors have expressed interest in transforming the Femern Belt
Forum into a regional implementation network consisting of experts and public
administrators, while others seek to replace the Femern Belt Forum with a local
implementation network bringing together the most intensely affected actors
from the local municipalities. The latter have stirred up conflicts with the poli-
ticians and administrators from the Zealand Region, who claim that all citizens
in Zealand are intensely affected by the Femern Bridge project.
The high, but varying, degrees of inclusiveness should not make us forget
that three affected groups have been systematically excluded from the Femern
Belt Forum. First, there have been no representatives of the Danish taxpayers
who are going to pay for the economic deficit that might arise if the optimistic
prognosis of the numbers of cars and trucks crossing the bridge does not
hold. Second, the business firms and transport companies from the other
Scandinavian countries that stand to benefit tremendously from a Femern
bridge are not represented either, although both the Scandinavian Round Table
of Industrialists and the EU transport authorities have given the Femern Belt
Bridge their seal of approval (The Idea Committee for Infrastructure, 1998).
Last but not least, the highly critical actors on the German side of the border
were initially invited to some of the meetings in the network, but cultural differ-
ences and political controversies made it difficult to sustain their participation.
Thus, they have also been excluded. Of these three exclusions, the latter is
by far the most problematic, since the German citizens on the Femern penin-
sula are more intensely affected than both the Danish taxpayers and the
Scandinavian businesspeople, since their daily lives and business opportunities
will be affected by the environmental consequences of an infrastructure
development they have no interest in.
To these external exclusions should be added the internal exclusion of the
actors who are formally included but tend to be marginalized in the decision-
making process. The public administrators and the labour market organizations
have clearly had the upper hand in the network, whereas the elected politicians
and some of the other private actors have had a hard time working their way
through the piles of paper circulated in the network.
However, the power relations in the network have not led to a silencing of
the weaker part. The negotiated interaction among the included actors has been
governed by a set of commonly accepted rules that tend to ensure that every-
body gets to talk and that other people listen and respond in a civilized manner
and with proper respect. There has never been any voting. Decisions are made
on the basis of deliberation and consensus formation is easy, as most of the
actors agree on the basic idea of building a bridge across the Femern Belt.
Written agendas and minutes tend to ensure transparency in the internal
decision-making process.

Some of the marginal network actors have probably been empowered by

their participation in the network, but the central actors were already highly
skilled and very competent when they joined the network. Technical skills and
political competences have been important due to the complex character of the
infrastructure project and the task of securing national political support for the
project. Some of the long-lasting members of the governance network complain
that some of the less experienced actors have too little knowledge of what is
going on, which might indicate that the empowerment effects have been limited.
The point where the governance network fails completely is in relation to
the demand for a self-reflexive contribution to democratic innovation. The
Femern Belt Forum does not perceive itself as being part of a broader demo-
cratic process nor does it address the issue of how its internal democracy is func-
tioning and how it could be improved. The governance network conceives itself
mainly as a lobbying organization participating in a political game that it aims
to win. The forging of a strong and political alliance is more important than
concerns for democratic innovation.
In sum, democratic legitimacy has been enhanced by the re-enactment of
some commonly accepted democratic rules and norms, while the failure to
comply with other democratic demands has pulled in the opposite direction.
The network has included most of the relevant and affected actors on the
Danish side of the border – even some of the critical voices were included in
the initial phase in an attempt at neutralizing them and turning them into
supporters. The internal interactions within the Femern Belt Forum have also
been governed by democratic rules ensuring agonistic respect among the
participants. However, there has been a problematic external exclusion of the
critical actors on the German side and an internal marginalization of the elected
politicians and some of the private actors. Last but not least, the empowerment
effects have been limited and democratic innovations totally absent. As such,
we shall conclude that the democratic anchorage of the Femern Belt Forum in
democratic rules and norms is: WEAK.
The overall verdict is that the democratic anchorage of the Femern Belt
Forum is: MODERATE. There is a strong anchorage in the membership basis
of the participating groups and organizations and a moderate anchorage in both
elected politicians and the territorially defined citizenry. However, the anchor-
age in democratic rules and norms is weak and leaves considerable room for
improvement.
6. Conclusion
In this article we have combined public administration research and political
theory in an analysis of participatory governance in the field of spatial planning.
The new interactive forms of network governance are often reduced to non-
political, managerial mechanisms for pragmatic problem-solving and evaluated
merely in terms of their contribution to an effective governing of society. By
contrast, we have claimed that governance networks are arenas of political
power struggles and partake in the authoritative definition and allocation of

values. This calls for an assessment of their democratic performance, which we

have proposed should be analysed in terms of their democratic anchorage.
The argument presented in this article has elaborated on the original model
of democratic anchorage in three crucial ways. First, it has operationalized and
contextualized the key dimensions of the model in order to facilitate a context-
sensitive empirical analysis. Second, it has demonstrated how the democratic
anchorage model can be used in an empirical case study of a governance
network that has been formed in relation to a major infrastructure project.
Finally, it has shown how the qualitative measurement of the democratic
performance of the Femern Belt Forum in relation to the four anchorage points
can lead to a relatively clear verdict.
The empirical application of the democratic anchorage model has
revealed three crucial merits. As such, the model is helpful in structuring a
multi-dimensional assessment of the democratic performance of specific
governance networks on the basis of an analysis of qualitative data. In addition,
the model can provide a nuanced understanding of the strengths and weak-
nesses in the democratic performance of particular governance networks. Last
but not least, it can help identify shortcomings that can be addressed by pro-
active strategies aiming to enhance the democratic performance of governance
networks.
The empirical assessment of the Femern Belt Forum generates three crucial
tasks for future research. The first task is to further address the problems that
arise from applying the democratic anchorage model in democratic audits of
multi-level and cross-border networks. As such, it is difficult to evaluate the
questions about metagovernance and inclusion in relation to multi-level and
cross-border networks. The second task is to reflect on how to handle the
dynamic development of governance networks in the assessment of their demo-
cratic anchorage. As such, the results of the evaluation of the basic dimensions
of the anchorage model will tend to differ over time, and it is not clear how to
draw clear conclusions in that situation. The third task is to develop empirical
methods for measuring the degree to which the represented groups and
organizations identify with the network actors who claim to represent them.
Two final tasks of a more general nature should be mentioned. The first
concerns the development of a comparative approach to democratic auditing
of governance networks that can help identify the structural and contingent
factors conditioning the democratic quality of governance networks. The
second task is to develop a systematic means of translating a critical diagnosis
of democratic anchorage into proactive strategies for democratic meta-
governors and into reflexive learning processes within governance networks.
Completing these tasks may help establish an important and fruitful interface
between researchers and practitioners.
Notes
1. The empirical data on which this analysis is based has been collected in 2007–08. It
includes qualitative interviews, archival material and observation of meetings and
events.

2. The first reorganization took place in 2000, when a small Executive Committee was
formed and took over most of the tasks from the ‘Idea Committee for
Infrastructure’. The reorganization turned the relatively broad and open network
into a more professionalized and elitist network. The second reorganization took
place in 2003, when the network was re-named ‘Femern Belt Forum’ and acquired
its own administrative secretariat (Raahavegaard). The last reorganization took
place in 2007, where the long-term chairperson was replaced by a new former
minister who aimed to involve more regional and national actors in the network.
This reorganization coincided with a huge amalgamation reform that reduced the
number of Danish municipalities from 275 to 98 and replaced the 14 old counties
with five new regions. This reform has triggered the formation of a new competing
network named ‘Ad hoc-committee Femern Belt’, which was initiated by the new
and larger municipality in Lolland, where the bridge will begin.
References
Agger, A. and Löfgren, K. (2008) ‘Democratic Assessment of Collaborative Planning
Processes’, Planning Theory 7(2): 145–64.
Barber, B. (1984) Strong Democracy: Participatory Politics for a New Age. Los Angeles:
University of California Press.
Bech, C. (1914) ‘Det danske Landbrugs Stilling til Etableringen af Rødby-Femern
ruten’, Tidsskrift for Landøkonomi, special issue.
Behn, R.D. (2001) Rethinking Democratic Accountability. Washington, PA: The
Brookings Institution.
Bohman, J. (2005) ‘From Demos to Demoi: Democracy across Borders’, Ratio Juris
18(3): 293–314.
Dahl, R.A. (1989) Democracy and Its Critics. New Haven, CT: Yale University Press.
Danish Railroad Company (1940) ‘Forslag til anlæg af Jernbane i forbindelse med
Rødby-Femern Ruten’, Bilag 1.
Dewey, J. (1927) The Public and Its Problems. Chicago, IL: Swallow Press.
Dryzek, J.S. (2000) Deliberative Democracy and Beyond. Oxford: Oxford University
Press.
Dryzek, J.S. (2007) ‘Networks and Democratic Ideals: Equality, Freedom and
Communication’, in E. Sørensenand and J. Torfing (eds) Theories of
Democratic Network Governance, pp. 262–73. Basingstoke: Palgrave-Macmillan.
Erngaard, E. (1952) ‘En bys aarhundredgamle drøm om fugleflugtslinien’, Berlingske
Aftenavis, 14 August.
European Commission (2001a) European Governance: A White Paper, COM(2001) 428
final.
European Commission (2001b) Networking People for Good Governance in Europe,
Brussels.
European Commission (2002) General Principles and Minimum Standards for
Consultation of Interested Parties, COM(2002) 704 final.
European Commission (2005) Plan D for Democracy, Dialogue and Debate,
COM(2005) 494 final.
Fox, C.J. and Miller, H.T. (1995) Postmodern Public Administration. Thousand Oaks,
CA: SAGE.
Fung, A. and Wright, E.O. (eds) (2003) Deepening Democracy: Institutional Innovations,
Empowered Participatory Governance. London: Verso.
Hajer, M. (1995) The Politics of Environmental Discourse: Ecological Modernization
and the Policy Process. Oxford: Clarendon Press.
Hansen, A.D. (2007) ‘Governance Networks and Participation’, in E. Sørensen and

J. Torfing (eds) Theories of Democratic Network Governance, pp. 247–61.

Basingstoke: Palgrave Macmillan.
Healey, P. (2007) Urban Complexity and Spatial Strategies. London: Routledge.
Hillier, J. (2002) Shadows of Power. London: Routledge.
The Idea Committee for Infrastructure (1998) ‘The Fixed Connection of the Femern
Belt’, TFK Transportforschung.
Jessop, B. (2002) The Future of the Capitalist State. Cambridge: Polity Press.
Kickert, W.J.M., Klijn, E.H. and Koppenjan, J.F.M. (eds) (1997) Managing Complex
Networks: Strategies for the Public Sector. London: SAGE.
Klijn, E.H. and Koppenjan, J.F.M. (2004) Managing Uncertainties in Networks. London:
Routledge.
Klijn, E-H. and Skelcher, C. (2007) ‘Democracy and Governance Networks: Compatible
or Not?’, Public Administration 85(3): 587–608.
Kooiman, J. (2003) Governing as Governance. London: SAGE.
Laclau, E. (1990) New Reflections on the Revolution of Our Time. London: Verso.
Laclau, E. (1993) ‘Power and Representation’, in M. Poster (ed.) Politics, Theory and
Contemporary Culture, pp. 277–96. New York: Columbia University Press.
Lukes, S. (1974) Power: A Radical View. Basingstoke: Macmillan.
March, J.G. and Olsen, J.P. (1989) Rediscovering Institutions. New York: The Free Press.
March, J.G. and Olsen, J.P. (1995) Democratic Governance. New York: The Free Press.
Marsh, D. and Rhodes, R.A.W. (eds) (1992) Policy Networks in British Government.
Oxford: Oxford University Press.
Mayer, I., Edelenbos, J. and Monnikhof, R. (2005) ‘Interactive Policy Development:
Undermining or Sustaining Democracy?’, Public Administration 83(1): 179–99.
Mayntz, R. (1991) ‘Modernization and the Logic of Inter-organisational Networks’,
Working Paper, Max-Planck Institut für Gesellschaftsforschung, pp. 1–22.
Mouffe, C. (1993) The Return of the Political. London: Verso.
Mouffe, C. (2005) On the Political. London: Routledge.
Newman, J. (2005) ‘Participative Governance and the Remaking of the Public Sphere’,
in J. Newman (ed.) Remaking Governance: Peoples, Politics and the Public Sphere,
pp. 119–38. Cambridge: Polity Press.
Peters, B.G. and Pierre, J. (2000) Governance, Politics and the State. Basingstoke:
Macmillan.
Pitkin, H.F. (2004) ‘Representation and Democracy: Uneasy Alliance’, Scandinavian
Political Studies 27(3): 335–42.
Pateman, C. (1970) Participation and Democratic Theory. Cambridge: Cambridge
University Press.
Rhodes, R.A.W. (1997) Understanding Governance: Policy Networks, Governance,
Reflexivity and Accountability. Buckingham: Open University Press.
Saward, M. (2005) ‘Governance and the Transformation of Political Representation’, in
J. Newman (ed.) Remaking Governance: Peoples, Politics and the Public Sphere,
pp. 179–93. Cambridge: Policy Press.
Staten, H. (1985) Wittgenstein and Derrida. Oxford: Basil Blackwell.
Stoker, G. (2006) Why Politics Matter. Basingstoke: Palgrave Macmillan.
Sørensen, E. (2006) ‘Metagovernance: The Changing Role of Politicians in Processes
of Democratic Governance’, American Review of Public Administration 36(1):
79–97.
Sørensen, E. and Torfing, J. (2003) ‘Network Politics, Political Capital, and Democracy’,
International Journal of Public Administration 26(6): 609–34
Sørensen, E. and Torfing, J. (2005) ‘The Democratic Anchorage of Governance
Networks’, Scandinavian Political Studies 28(3): 195–218.

Sørensen, E. and Torfing, J. (eds) (2007) Theories of Democratic Network Governance.

Basingstoke: Palgrave Macmillan.
Sørensen, E. and Torfing, J. (forthcoming) ‘Making Governance Networks Effective and
Democratic through Metagovernance’, Public Administration.
Torfing, J. (2005) ‘Governance Network Theory: Towards a Second Generation’,
European Political Science 4(3): 305–15.
Torfing, J. (2007) ‘Discursive Governance Networks in Danish Activation Policy’, in
M. Marcussen and J. Torfing (eds) Democratic Network Governance in Europe,
pp. 111–29. Basingstoke: Palgrave Macmillan.
Warren, M.E. (2002) ‘What Can Democratic Participation Mean Today?’, Political
Theory 30(5): 677–701.
Wittgenstein, L. (1986) Philosophical Investigations. Oxford: Basil Blackwell.
Young, I.M. (2000) Inclusion and Democracy. Oxford: Oxford University Press.
Jacob Torfing is Professor in Politics and Institutions at Roskilde University,

Denmark. He is a political scientist and his research interests include discourse
theory, employment policy, democratic participation, and governance networks.
He has published widely on discourse theory and democratic network governance.
His books include Theories of Democratic Network Governance (2007) – co-edited
with Eva Sørensen, Discourse Theory in European Politics (2005) – co-edited with
David Howarth, New Theories of Discourse (1999) and Politics, Regulation and the
Modern Welfare State (1995). He is director of the Centre for Democratic Network
Governance.
Address: Department of Society and Globalisation, Roskilde University, Building

25.2, PO Box 260, DK-4000 Roskilde, Denmark. [email: jtor@ruc.dk]
Eva Sørensen is Professor in Public Administration at Roskilde University,

Denmark. She has written several books and articles on democracy, public adminis-
tration, and new forms of governance. Her books include Theories of Democratic
Network Governance (2007) – co-edited with Jacob Torfing – and Public Adminis-
tration in Transition (2007) – co-edited with Gunnar Gjelstrup. She is co-founder of
the Centre for Democratic Network Governance and directs a large-scale research
project on ‘Collaborative Innovation in the Public Sector’. She is director of the
PhD programme on ‘Governance, Welfare and Citizenship’ at Roskilde University
and leader of the national Political Science PhD network Polforsk.

25.2, PO Box 260, DK-4000 Roskilde, Denmark. [email: eva@ruc.dk]
Trine Fotel is a post-doctoral researcher at Roskilde University, Denmark. She

has published several articles on urban planning, mobility and local democracy.
She is currently participating in a large research project on network governance
in the new Danish regions. She is a member of Centre for Democratic Network
Governance.

25.2, PO Box 260, DK-4000 Roskilde, Denmark. [email: trinen@ruc.dk]

Democratic Anchorage of Infrastructural Governance Networks: The Case of The Femern Belt Forum

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

Democratic Anchorage of Infrastructural Governance Networks: The Case of The Femern Belt Forum

Uploaded by

Copyright:

Available Formats

Article

© The Author(s), 2009.

Jacob Torfing, Eva Sørensen and Trine Fotel

Abstract Governance networks are often praised for their contribution

Keywords democracy, governance, metagovernance, networks, physical

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

3. The democratic anchorage of governance networks

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

influencing the policy-making that takes place in mandated or self-grown

3.1. Anchorage in democratically elected politicians

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

1) Network design that endeavours to determine the scope, character,

Proposition 1. Democratic anchorage in elected politicians can be measured by

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

3.2. Anchorage in the membership basis of participating groups and

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

governance network (Saward, 2005). The advancement of such claims is import-

3.3. Anchorage in a territorially defined citizenry

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

networks. The first requirement is transparency. Governance networks and their

3.4. Anchorage in democratic rules and norms

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

inclusion and exclusion should be transparent and consistently applied in

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

participating actors and to augmenting their political empowerment (Sørensen

• Inclusion and exclusion must be subject to ongoing negotiations, and these

4. Applying the democratic anchorage model

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

The democratic anchorage of the Femern Belt Forum 1

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

First anchorage point: democratic anchorage in elected politicians

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

However, the elected politicians in the national government and Danish

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

administrators, and to some extent also by the transnational political

Second anchorage point: democratic anchorage in participating

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Third anchorage point: democratic anchorage in a territorially

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Fourth anchorage point: democratic anchorage in democratic

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

could now be left to a few lead actors – organized in an executive committee –

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Some of the marginal network actors have probably been empowered by

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

values. This calls for an assessment of their democratic performance, which we

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

J. Torfing (eds) Theories of Democratic Network Governance, pp. 247–61.

Downloaded from plt.sagepub.com at UNIV NEBRASKA LIBRARIES on April 9, 2015

Sørensen, E. and Torfing, J. (eds) (2007) Theories of Democratic Network Governance.

Jacob Torfing is Professor in Politics and Institutions at Roskilde University,

Address: Department of Society and Globalisation, Roskilde University, Building

Eva Sørensen is Professor in Public Administration at Roskilde University,

Address: Department of Society and Globalisation, Roskilde University, Building

Trine Fotel is a post-doctoral researcher at Roskilde University, Denmark. She

Address: Department of Society and Globalisation, Roskilde University, Building