Jump to ContentJump to Main Navigation
Multi-level Governance$

Ian Bache and Matthew Flinders

Print publication date: 2004

Print ISBN-13: 9780199259250

Published to Oxford Scholarship Online: April 2004

DOI: 10.1093/0199259259.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 23 October 2019

Conclusions and Implications

Conclusions and Implications

Chapter:
(p.195) 12 Conclusions and Implications
Source:
Multi-level Governance
Author(s):

IAN BACHE

MATTHEW FLINDERS

Publisher:
Oxford University Press
DOI:10.1093/0199259259.003.0012

Abstract and Keywords

Ian Bache and Matthew Flinders reflect on the value of multi‐level governance as both an analytical and normative concept. They suggest that the critical normative challenge is that of legitimacy in the context of multi‐level governance, which demands new means to connect citizens more effectively with the shifting locations of power. In analytical terms, they argue for a clear distinction to be made between multi‐level governance and multi‐level participation, where the latter notion signals greater involvement without effective influence for at least some types of new actors. Pursuing this line of reasoning, they suggest that they highlight the danger of multi‐level governance incorporating neo‐pluralist assumptions in relation to the location and fluidity of power that fail to address the continuing existence of underlying structural inequalities.

Keywords:   accountability, analytical, democracy, fluidity, governance, legitimacy, multi‐level governance, normative, participation, power

One thing is clear from this collection, that multi-level governance is used in different ways and for different purposes. Crucially, there is a need to distinguish between multi-level governance as an analytical model and multi-level governance as a normative concept. Our central task in this volume has been to evaluate multi-level governance as an analytical model: to assess how it captures the changing nature of decision making in the context of complexity. This task, guided by the themes set out in chapter 1, is addressed below. However, such is the attention given by contributors to this volume in relation to multi-level governance as a normative concept, we discuss this issue first.

Multi-level Governance as a Normative Concept

For a number of contributors, multi-level governance is emerging as a normatively superior mode of allocating authority. Moreover, multi-level governance has been assimilated into the discourse of practitioners to describe developments in the European Union, notably by European Commission officials (COM 428 2001). As Hooghe and Marks (chapter 2) suggest, its main advantage lies in its ‘scale flexibility’ in that it ‘allows jurisdictions to be custom-designed in response to externalities, economies of scale, ecological niches, and preferences’ (p.29). However, beyond this appreciation is dispute about how multi-level governance should be organized to maximize its advantages (e.g. efficiency) and minimize its costs (e.g. accountability). Here, the distinction between Type I and Type II developed by Hooghe and Marks is helpful in highlighting the differing qualities of different types of multi-level governance (p.17). Type I multi-level governance creates general-purpose jurisdictions with non-intersecting memberships, with jurisdictions at lower tiers nested into higher ones; it is oriented towards intrinsic communities and suited to the articulation and resolution of conflict. By contrast, Type II consists of special-purpose jurisdictions focused around particular policy problems and is more suited to Pareto-optimality.

Yet, on the value of multi-level governance as a mode of organizing authority there are dissenting voices. Jessop is keen to emphasize that governance, like (p.196) markets and hierarchies, is prone to failure. This is in part because governance theories are closely associated with problem solving and crisis management in relation to specific issues, when issues of metagovernance (the ‘ground rules’ for governance) may be more pertinent. In the context of the European Union, this leads to the relative neglect of ‘questions of the relative compatibility or incompatibility of different governance regimes and their implications for the overall unity of the European project and European statehood’ (p.61). Welch and Kennedy-Pipe (chapter 8) raise a related issue, citing the work of Paul Cammack, which chimes with the notion of metagovernance. This is the role of the World Bank's Comprehensive Development Framework as a ‘management tool’ for monitoring government policies, which Cammack suggests reflects the Bank's aspiration to be ‘the mother of all governments’.

This discussion raises two issues that are worthy of further investigation. The first is whether there is anything intrinsic to multi-level governance as an analytical framework that would exclude a simultaneous focus on both metagovernance and governance issues as Jessop defines them: the case might be made that such a focus is at least an aspiration of the concept. The second, for those concerned with issues of institutional and/or policy efficiency, is how we measure the impact or outcomes of multi-level governance processes. There are claims and counterclaims about the efficiency or otherwise of different governance structures, but little agreement about the appropriate methodological tools for investigation and measurement. The suggestion by Perraton and Wells (chapter 10) that fiscal federalism might be a valuable complement to multi-level governance in this respect is one worth pursuing for those concerned with issues of efficiency. We return to some of these issues in the final section, but now we turn to the seven themes outlined in the opening chapter of this volume that largely addresses the issue of multi-level governance as an analytical concept.

How Should Multi-level Governance be Defined?

In Orwell's (1946) ‘Politics and the English Language’ he lamented the often purposive use of conceptual imprecision in order to justify certain actions, decisions or viewpoints, suggesting that ‘political language has to consist largely of euphemism, question-begging and sheer cloudy vagueness’ (1946: 153). Gallie (1955) developed this line of argument in his work on the ‘essentially contested’ nature of concepts. In this view, concepts become so permeated and surrounded by values that scholars may argue interminably without ever reaching agreement on their meaning and implications (Birch 1993: 8). While it is possible to criticize Gallie's position (Ball 1988), it usefully underpins the nature of the debate on multi-level governance as we seek to assess its value from different perspectives.

(p.197) Throughout the volume, contributors have discussed the need for conceptual clarity in relation to the related notions of governance and multi-level governance. As concepts become more widely applied, they can become devoid of meaning. By bringing together diverse contributions on a common theme, we have sought to identify common understandings and applications of the concept. Four common strands have emerged. First, that decision making at various territorial levels is characterized by the increased participation of non-state actors. Second, that the identification of discrete or nested territorial levels of decision making is becoming more difficult in the context of complex overlapping networks. Third, that in this changing context the role of the state is being transformed as state actors develop new strategies of coordination, steering, and networking to protect and, in some cases, enhance state autonomy. Fourth, that in this changing context, the nature of democratic accountability has been challenged and need to be rethought or at least reviewed. These common strands provide the basis for a parsimonious definition of multi-level governance that raise clear hypotheses for future research.

How is Multi-level Governance utilized by Scholars Working in Different Disciplinary Traditions?

Here we draw specifically on contributions to Part Two of this volume, in which the authors utilize multi-level governance in the traditionally separate academic (sub)disciplines of British Politics, European Studies and International Relations. More general observations on the different interpretations and applications of multi-level governance are contained under the various headings of this chapter.

Given its origins, it is unsurprising that the application and maturity of multi-level governance is most advanced in the field of EU studies. In this field, George suggests, it is an established theory of EU policy making, which generates clear expectations that draw on neo-functionalist antecedents. Crucially, multi-level governance suggests an emerging dynamic in the context of European integration that pulls authority away from national governments and empowers subnational and supranational actors. The consequence of this is that governments will be taken further along the road of diffused authority than they originally intended.

As the concept has travelled to different areas of study, it is been applied in varying ways. In its application to British politics, Bache and Flinders (chapter 6) suggest the concept has potential as an organizing perspective: a map of how things (inter)relate that leads to a set of research questions. The theoretical dimension to multi-level governance here is strengthened in this context by incorporation of hypotheses from the differentiated polity model, which posits aspects of a fragmented versus centralized executive. The multi-level component of the concept (p.198) is useful in this case for highlighting the changing nature of vertical relationships in the context of devolution and, within this new context, the changing state strategies aimed at maximizing control and coordination.

In applying multi-level governance to international relations, Welch and Kennedy-Pipe use it as a contrastive concept alongside the more commonly applied notion of governance. The contrast is made with the notion of government, in a tradition where government refers to ‘the political control of a centralized state’ and (multi-level) governance ‘denotes the coordination of social relations in the absence of a unifying authority’. Alongside this contrastive application, is the use of governance as a generic term similar to authority or coordination. This strategy is necessary in the context of the international realm, where the absence of discrete policy areas suggests that the application of multi-level governance to individual case studies, as is the case with the European Union, is not possible. Instead, the authors take a synoptic view to investigate issue areas of international relations where the question of the distribution of authority arises: governance in the generic sense. While the generic use of governance guides the choice of issue areas, the contrastive concept is used to address a number of questions in these areas relating to the nature of power and authority.

How do the Structures and Processes of Multi-level Governance Differ Across Policy Sectors and how Can these Differences be Explained?

Here, we consider the application of multi-level governance to the policy sectors in Part three of the book: environmental policy, EU regional policy and economic policy. Each of these policies has a strong EU dimension, which reflects both the advanced use of multi-level governance in this area of study and the development of discrete EU policy competences in these sectors. The three sectoral studies provided a mix of ‘easy’ and ‘hard’ tests for multi-level governance. Conventionally, environmental and regional policies are considered ‘low politics’ areas for national governments, while economic policy is generally recognized as ‘high politics’ (Hoffmann 1966, 1982). The structures and processes of multi-level governance differed across the case studies. The case of regional policy highlights the effective formalization and promotion of multi-level governance relations through the partnership principle, the case of environmental policy suggests multi-level interactions that were less formalized but nonetheless evident, while the case of economic policy highlights more formal layering of authority and less evidence of non-state and multi-level interaction. However, despite the differences in structures and processes of multi-level governance, there were some commonalities in terms of outcomes.

(p.199) Fairbrass and Jordan's study (chapter 9) of EU environmental (biodiversity and land use planning) policy making and the implementation of these policies in the United Kingdom presents a ‘critical test’ for multi-level governance. This study takes a policy area characterized by dispersed decision-making competences and the involvement of state and non-state actors, that would seem to make it ‘an excellent vehicle’ for analysing multi-level governance. However, the United Kingdom provides a ‘hard case’, in terms of its likely response to multi-level governance pressures emerging from the European Union, because of its history of centralized authority and its sceptical approach to the European Union. The study revealed mixed evidence for multi-level governance. Fairbrass and Jordan argue that the proliferation of actors with competing policy objectives led to unintended consequences. There was ‘clear evidence’ of the national executive struggling to predict, let alone fully control the course of policy making. Despite this, the study showed that ‘the state can still be an important shaper of post-decisional politics and a powerful gatekeeper’.

Similar conclusions emerged from Bache's discussion of EU regional policy, another policy area where multi-level governance could be expected to be prominent. This chapter suggested that there was some evidence of multi-level governance in some territories and in relation to some issues, but also evidence of effective gatekeeping powers at the implementation stage in the United Kingdom.

Perraton and Wells looked at economic policy, an area where expectations for evidence of multi-level governance would not necessarily be high. They found that structures and processes of multi-level governance varied across different areas of economic policy making. In their cases, European Monetary Union (EMU) and the transition of former communist countries, tended to demonstrate decisional authority at a number of levels, but little evidence of wide non-governmental participation. They conclude that ‘this would appear to conflict with much of the literature on multi-level governance which downplays the role of powerful state and institutional actors’.

It would be inappropriate to draw too many conclusions from three studies about how and why multi-level governance structures, processes and outcomes vary across policy sectors. However, we can sketch out a number of findings that at least point the direction for future research. First, the distinction between ‘high’ and ‘low’ politics issue areas is important. The argument from this follows that multi-level governance is likely to be more prominent in areas that state actors deem less important to their interests. This distinction is a crude one and we do not suggest here that policies and issues cannot travel from one end of the high-low spectrum to another over time. However, the studies confirm that multi-level governance is most likely to be prominent in sectors or issues deemed ‘low politics’ by national governments. Second, multi-level governance has overstated the prospects for diminishing central control at the implementation and post-decisional stage. The evidence from environmental and regional policies is that determined state executives can be effective gatekeepers in (p.200) relation to the impact of EU policies within the domestic arena. However, we note here that the studies looked at implementation in the ‘hard’ case of the United Kingdom (see above). Third, while state executives remain pivotal actors, there is evidence of unintended consequences arising from multi-level governance that are beyond state control and this is true even in the case of the United Kingdom.

Does Multi-level Governance Connote Hierarchy? If so, is This a Weakness in the Approach?

Our contributors suggest that the answer to this question is highly dependent on the conceptual lens through which individuals view the notion of multi- level governance. Some read the concept as clearly hierarchical, suggesting vertically layered tiers of authority, making it ‘misleading and imprisoning’ (Rosenau, chapter 3). It is for this reason that Rosenau prefers the term ‘spheres of authority’, while others suggest that the ‘multi-level’ aspect of the concept is unnecessary and the notion of ‘governance’ is itself enough (Welch and Kennedy-Pipe, chapter 8). The counterview is that multi-level governance ‘sets up an alternative hypothesis that the hierarchy of levels of governance is being eroded’ (George, chapter 7). These contending views on multi-level governance reinforce our earlier point that, to date, the concept has lacked clarity. However, Hooghe and Marks's development of Types I and II multi-level governance provides more clarity in this respect by suggesting that these different typologies describe different circumstances captured by either orderly or disorderly processes and either the presence or absence of vertical layering of authority. Rosenau's argument reflects an understanding of Type I multi-level governance; Type II is explicitly characterized by ‘specialized, territorially overlapping jurisdictions in a relatively flexible, non-tiered system with a large number of jurisdictions’ (emphasis added).

Where multi-level governance is described as connoting hierarchy, this is usually intended as a criticism, suggesting that it overemphasizes vertical relationships to the detriment of horizontal coordination. The distinction between Types I and II multi-level governance provides a useful heuristic tool for deconstructing and refining arguments in this area. While Type I eschews tiering of authority, Type II directs attention to relationships between actors organized at distinct and different territorial levels, which in some circumstances may be appropriate and advantageous. chapter 6, on Britain, provides an example of the heuristic value of Type I multi-level governance, in which attention is placed on the newly devolved governmental framework around which broader networks and processes in the United Kingdom are largely structured. The case of the United Kingdom is, to paraphrase Jessop, an illustration of multi-level governance in the shadow of governmental hierarchy: but a hierarchy where authority is being vertically layered.

(p.201) What Implications does Multi-level Governance Have for the Power, Position and Role of the Nation state?

This question has been central to the debate around multi-level governance and is addressed throughout the contributions to this volume. We did not expect an easy consensus in response to this question, and we did not find one. State power remains important, but how it is exercised is increasingly complex.

Theories of governance and multi-level governance are closely associated with predictions of the demise of the nation state. Yet, this appears to be a parody of multi-level governance. None of the ‘governance’ theorists contributing to this collection deny the continued importance of nation states at various territorial levels and throughout the policy process. There is broad recognition that state executives possess both legitimacy and nodality, which ensures that they are at least ‘first among equals’ in the context of governance.

There is also a growing recognition of the role of states in shaping and regulating governance: what Jessop (chapter 4) describes as ‘metagovernance’. It is possible that much of the misunderstanding in relation to governance and the nation state stems from the work of hyperglobalists (e.g. Ohmae 1990, 1996) who, while disagreeing about the risks and opportunities of globalization, all emphasize the demise of the nation state. As Welch and Kennedy-Pipe argued, the relationship between globalization and governance has frequently been opaque. While the two concepts raise a large number of related issues, the nature of the relationship between these processes is often assumed rather than demonstrated. Moreover, the variance between the two concepts is arguably stark in relation to the nation state. A common theme running throughout the chapters in this book has been the transformation or reconfiguration of the state to protect and, some would argue enhance, its continuing centrality within evolving structures of multi-level governance.

Consequently, a number of arguments emerge from this collection in relation to how states respond to the changing context of governance to retain or enhance state power. First is the role of states in metagovernance, that is in providing the ground rules for governance. Second, state executives have some control over what powers are transferred upwards, downwards and sideways and may exercise this control to pass on some responsibilities while concentrating resources on issues and projects deemed more important (see Jessop). Third, state executives may mobilize and draw on the resources of supportive non-state actors to achieve specific objectives and outcomes. However, as George's discussion of the additionality dispute demonstrates, this strategy is also available to supranational organizations as well as national state executives. Fourth, state executives have illustrated the importance of the implementation stage in shaping the outcomes of policy decisions taken in a more pluralistic arena (see Bache, chapter 10; Fairbrass and Jordan, chapter 9). Fifth, states can introduce institutional reforms with the aim of increasing their vertical and horizontal (p.202) strategic capacity while also augmenting their ‘gatekeeping’ capacity. The creation of the Regional Coordination Unit within the Cabinet Office in the United Kingdom can be identified as an example of this strategy. Finally, states can re-scale state powers as a response to subnational and supranational pressures in ways that may increase state capacity in some areas at least (see Bache and Flinders, chapter 1).

Of course, listing the possibilities for state action in the context of multi-level governance does not imply success in each instance. The collection has demonstrated evidence of limited state executive rationality and of the unintended consequences of governance processes. In context of complexity and rapid change, states will face increasing pressures to reinvent themselves and develop new strategies: popular legitimacy secured through free elections will continue to provide states with the right to exercise authority, but not with the means.

What are the Implications of Multi-level Governance for Democratic Accountability?

Until now the implications of multi-level governance for democratic accountability have been relatively neglected. Yet this normative debate emerged as a theme concerning a number of contributors to this collection. The issue is discussed in some detail by Peters and Pierre (chapter 5), but is a feature of other chapters also. Peters and Pierre ask whether problem-solving capacity and outcomes has taken precedence over democratic input and accountability. Similarly, Perraton and Wells raise the question of a ‘trade-off between democratic notions, such as accountability and participation, and more managerial concerns, such as efficiency and control’.

As the nation state has become more enmeshed in complex networks, responsibility for certain functions is increasingly blurred (Flinders 2001). In this context, the important question is whether governance can be democratic. The inevitable answer is this depends on how democracy is defined and understood. For those attached to liberal democratic notions of democracy, the rise of multi-level governance should be accompanied by the extension of the basic framework of representative democracy to the relevant territorial levels. An alternative view suggests the creation of new mechanisms of control and accountability that draw on the principles of self-governing communities. What both these top-down and bottom-up approaches to democracy seek is a way to ‘save our souls from the Faustian bargain’ to ensure frameworks of accountability that do not diminish the independence and operational efficiency of these bodies. However, as the work of Bellone and Goerl (1992), Moe (1994), Oughton (1995) and Roth (1996) suggests, achieving a synthesis between accountability and efficiency is difficult. We return to this issue in the final section.

(p.203) What are the Strengths and Weaknesses of Multi-level Governance as an Analytical Framework?

Jessop (p.61) suggests that work on governance-centred approaches ‘largely remains at a pre-theoretical stage’. Similarly, Fairbrass and Jordan prefer to avoid calling multi-level governance a ‘fully-fledged theory’, but suggest that it provides a ‘compelling description’ of what happens to decisions once they escape the domain of intergovernmental bargaining in the European Union. Peters and Pierre (p.88) also suggest that ‘multi-level governance appears incapable of providing clear predictions or even explanations (other than the most general) of outcomes in the governance process’. George rejects this position, suggesting that multi-level governance is a theory that generates the hypothesis that the European Union is a system not dominated by national governments. Further hypotheses come from this, relating to the role of other actors, particularly supranational and subnational institutions. As discussed above, we suggest that whether multi-level governance is accepted as a theory, an organizing perspective or a contrastive concept, it can be used in a number of ways that contribute to our understanding of the changing nature of policy making. The use of multi-level governance as a contrastive concept by Welch and Kennedy-Pipe illustrates this well.

Welch and Kennedy-Pipe suggest that analytically, the emphasis of multi-level governance on the effect of international civil society upon states appears unproblematic, but the empirical evidence suggests considerable variation of this effect. More firmly, they conclude that this is an ‘exceptionally unpropitious time to be suggesting, as the concept of multi-level governance does, a general trend of the eclipse of the government of the nation state’. Yet, the advantage of multi-level governance here is in acting as a counterpoint to the state-centric approaches that have dominated analysis of international relations: in this case, multi-level governance raises questions about the role of non-state actors and highlights variation in different patterns of participation and influence in different cases that state-centric approaches may well overlook.

Final Thoughts and Future Directions

In this book we have assembled and analysed critical perspectives on multi-level governance in order to clarify some conceptual confusion and to assess the value of the concept. It is clear that, whatever perspective is taken, multi-level governance (and related notions of governance) is a concept with an appeal that demands attention. In other words, it may not always be liked, but it is very difficult to ignore. We hope the book has contributed to understanding on multi-level governance—its uses, misuses, and limitations—but recognize that there is much more that ought to be done if the potential of the concept is to be fully (p.204) developed. We would suggest that the concept deserves further testing and analyses in relation to both the outstanding analytical and normative issues, which we discuss here in turn.

It is both a strength and a weakness of multi-level governance as an analytical framework that it can be applied in a wide range of circumstances: it is careful to avoid strong claims about the demise of the state and the influence of non-state actors, but this means that its theoretical content is weak. As Jessop suggests, there are ‘marked ambiguities in the referents of multi-level governance’. A particular problem is the emphasis on the increased mobilization and participation of ‘new’ (subnational and non-state) actors in policy processes rather than a focus on how this change affects policy outcomes. Put another way, a distinction must be drawn between multi-level governance and multi-level participation, where the latter notion signals greater involvement without effective influence for at least some types of new actors (the role of UK trade unions in EU regional policy provides an example of this, see Bache and George 1999). Pursuing this line of reasoning, it is possible to suggest that multi-level governance incorporates a weakness of neo-pluralist assumptions in relation to the location and fluidity of power that fail to address the continuing existence of structural inequalities (Marsh et al. 2003).

If multi-level governance is going to overcome this weakness, and in general terms move from being generally viewed as a ‘proto-theory’ to one that is ‘fully fledged’, it needs to generate clearer expectations in relation to the influence of subnational and non-state actors, as well as highlighting their mobilization and participation. Similarly, it might more clearly hypothesize the behaviour of state actors in response to the different challenges posed by different types of multi-level governance in different spatial, temporal, and policy context and, in particular, on how central control is maintained in some states and sectors rather than in others. We note here the potential dangers highlighted by Sartori's distinction between ‘conceptual travelling’ and ‘conceptual stretching’. The first of these notions refers to the effective application of a concept to new cases, while the second involves the distortion of a concept where it does not fit new cases. Travelling is valid and proper, stretching is problematic in that ‘gains in extensional coverage tend to be matched by losses in connotative precision’ and that ‘we can cover more—in travelling terms—only by saying less, and by saying less in a far less precise manner’ (Sartori 1970: 1035). Only when multi-level governance has travelled further in hostile academic and policy territories will we know if the concept is being overstretched.

In normative terms, multi-level governance is seen by its advocates to present a number of advantages. However, as Jessop suggests, it is wrong to think that multi-level governance can solve old problems without creating new ones. Here, we suggest, the critical problem is that of legitimacy in the context of multi-level governance. While multi-level governance in its different forms may add to the legitimacy of public policy making through increased efficiency, it (p.205) may reduce legitimacy in the form of democratic accountability unless new means are found to connect citizens more effectively with the shifting locations of power. The electoral legitimacy of national governments ensures them a pivotal role in this changing context, but the diffusion of competencies and the changing patterns of participation demand additional mechanisms of accountability beyond those provided by representative institutions. This does not only mean democratizing supranational and global processes, but also revisiting and revising the mechanisms of democracy within the state, at both national and subnational level. The increasingly fragmented nature of governing processes is a feature of politics within and without the nation state driven by related processes (hence ‘fragmegration’), and, as such, demands a response that takes this as its starting point. Our view is that the evolving structures of multi-level governance are likely to necessitate new forms and models of accountability that seek to build new and innovative conduits between the public and the institutions involved in complex networks. In essence, this may involve a fundamental reappraisal of the meaning of democracy and the role of representative institutions within nation states.

However, the very developments that some see as potential threats to legitimate authority, others suggest might be part of the democratic solution. Thus, while on the one hand, complex multi-level governance may be seen to reduce transparency and accountability, on the other, the proliferation of actors involved may be seen to provide ‘safety in numbers’ by limiting the power of any single actor or type of actor (see Rosenau, chapter 3). Yet, while ‘safety in numbers’ through increased participation may provide part of the answer to the challenges of democracy, this cannot be taken for granted. Participation does not equate to power and the emergence of multi-level governance does not necessarily enhance the position of weaker social groups and may indeed concentrate power more in the hands of those groups and actors with the necessary resources to operate most effectively in the context of complexity. In the words of Peters and Pierre (p.89), multi-level governance that is ‘embedded in a regulatory setting which enables weaker actors to define a legal basis for their action might be the best strategy to escape the Faustian bargain and to cheat the darker powers.’ In one view, this may involve granting ‘special rights’ to groups who are disadvantaged within this new context (Young 1997).

Following this argument, we would suggest that the next phase of research on multi-level governance pays particular attention to the implications for democracy of empirical developments and, related to this, to the design of frameworks of accountability that adopt a positive-sum gain in relation to the accountability versus efficiency debate (see Jessop 1998: p. 42). We would not expect a ‘one size fits all’ solution to this challenge: it is likely that in the context of different types of multi-level governance, different types of democratic arrangements will be needed to ensure popular consent.

Already, there are indications that this research agenda is being taken forward. Olsson (2003) has recently considered three models of democracy in (p.206) the context of theorizing multi-level governance and the EU regional policy: the parliamentary model, the pluralistic model, and the elite-democratic model. In a similar vein, Pollack (2003: 14) has characterized the three options for enhancing democracy in the EU's complex polity as parliamentarization, deliberation and constitutionalization. Another alternative would be to increase state regulation of governance: an echo of Jessop's work on metagovernance. However, as Olsson and others acknowledge, democratic renewal and efficient policy making in the context of emerging multi-level governance cannot be easily legislated for. Public cynicism towards politics and politicians has increased, rather than diminished, as elected leaders have experimented with new forms of governance to meet heightened public demands characterized by the desire for better public policies alongside lower taxes. As this pressure and other processes accelerate complexity and obscure accountability, meeting the challenges presented by multi-level governance in both analytical and normative terms will become ever more urgent.