Multilevel Governance in Europe PDF

9 Multilevel governance in Europe 1 Rainer Eising Introduction In EU Studies, the term ‘multilevel governance’ signiﬁes a move away from classical International Relations (IR) understandings of European integration. Gary Marks (1993) coined the term in his analysis of the European Community’s structural funds policy in the early 1990s, and it has since become a buzzword. It is no coincidence that the concept ﬁrst gained prominence within the multilayered EU institutional context, as the EU is oen considered to be ‘the most active and innovative producer of new types of decision-making arrangements’ worldwide (Bartolini 2011: 5). e concept is also tied into wider developments in the social sciences that drew aention to the increasing complexity of national and international governance arrangements. It is now common practice to aracterize the EU as a multilevel system; however, the multilevel governance concept, whi encompasses broad understandings of two universal elements in politics (i.e. territorial relations and the ways in whi collective courses of action are taken), has also gained currency in a great variety of other contexts. Zürn, Wälti and Enderlein (2010: 5) claim that multilevel governance resear has prompted ‘innovation in at least three main subﬁelds of political science: European integration, comparative federalism and international relations’. Renate Mayntz (2008: 46) locates it within a general shi from an actor- based ‘steering theory’ to the institutional analysis of ‘global governance’. More generally, it forms part of the contemporary solarly eﬀort to move from ‘simple systems’ towards ‘more complex frameworks, theories, and models to understand the diversity of puzzles and problems facing humans interacting in contemporary societies’ (Ostrom 2010: 408). is apter analyses the usage of the multilevel governance theory as it has evolved in EU Studies. First, it elaborates the concept of multilevel governance, situating it among competing perspectives on European Union politics su as liberal intergovernmentalism and neighbouring concepts su as federalism, poly-centric governance, network governance and new modes of governance. Following the reasoning of Hooghe and Marks (2003), two types of multilevel governance will be discussed: one focusing on monopolistic general-purpose jurisdictions, the other on overlapping functional jurisdictions. e second part concentrates on eﬀorts to conceptualize the politics of multilevel governance. It outlines basic models of multilevel decision-making, beginning with Fritz W. Sarpf’s joint- decision trap (Sarpf 1988) and potential escape routes from the institutional logic of the decision-making trap. ereaer, multilevel governance is linked to the variety of governance modes that are in use in the EU. Finally, the relationship between multilevel governance and organized interests is discussed, as the governance concept emphasizes the horizontal interactions among state and private actors as a fundamental aspect of governance in comparison to the hierarical understanding of government. e conclusion evaluates the analytical usefulness of the concept in EU Studies and its impact in other ﬁelds. e concept of multilevel governance e concept of multilevel governance emerged from the empirical study of structural policy-making (Marks 1993). In order to elaborate the elements of multilevel governance, Gary Marks, Liesbet Hooghe and Kermit Blank (1996) juxtaposed this concept to the liberal-intergovernmental understanding of EU politics (Moravcsik 1991, 1998), whi they labelled ‘state-centric’. In their view, the core claim of this earlier model was that ‘policy-making in the EU is determined primarily by state executives constrained by political interests within autonomous state arenas that connect subnational groups to European aﬀairs’ (Marks et al. 1996: 345). e solars took issue with the idea that member state executives control the European integration process and act as gatekeepers between the national arenas and the EU institutions, emphasizing instead the multilayered nature of EU politics. Rejecting a view of the world that reiﬁed the Westphalian state system and separated the sphere of European/international politics from that of national politics, they developed multilevel governance as a model based on the assumption that the state as an aggregate concept must be disaggregated into diﬀerent sets of actors (ibid.: 347–8). e authors put forward three propositions. First, ‘decision-making competences are shared by actors at diﬀerent levels rather than monopolized by state executives’ (ibid.: 346), emphasizing the autonomous political role of EU-level institutions su as the European Commission, the Court of Justice and the European Parliament. Second, collective decision-making in the EU ‘involves a signiﬁcant loss of control for individual state executives’ (ibid.: 346), implying losses or gains for individual member states. ird, it would seem that ‘political arenas are interconnected rather than nested’ (ibid.: 346), including the variety of direct interactions between sub-national and private actors and EU institutions, as well as the formation of transnational associations at the EU level by national groups (ibid.: 346–7). On the basis of these propositions, and drawing on insights from the study of political delegation and historical institutionalism, the authors highlighted the loss of both individual and collective member state control of EU agenda-seing, decision-making and policy implementation. EU decision-making was depicted as aracterized by ‘multiple, intermeshing competences, complementary policy functions, and variable lines of authority’ among levels of government and political institutions (ibid.: 366). ‘[C]ontending, but interloed institutions’ and shiing policy agendas (ibid.: 372) were seen as conducive to the emergence of direct interactions between the European Commission and sub-national actors, interest groups and national experts that could ‘bypass member states and allenge their traditional role as sole intermediary between subnational and supranational levels of government’ (Marks 1993: 402). In short, the initial resear interest was the transformation of statehood in Europe through European integration and the loss of member state sovereignty. Multilevel governance as a concept was intended to repudiate the neat distinction between external relations and national politics that prevails in liberal intergovernmentalism, highlighting the manifold transnational relations that have developed between sub- national actors and the EU institutions. It should also be noted that multilevel governance was in part directed against neofunctional expectations that a European political centre would evolve ‘whose institutions possess or demand jurisdiction over the pre-existing national states’ (Haas 2004 : 16), even though Ernst Haas (1975) had long ago given up on this idea. e concept found immediate resonance beyond the ﬁeld of EU Studies because it addressed not only resear lacunae in EU Studies but also broader developments in the social sciences. In EU Studies, the reinvigoration of European integration in the 1980s and the growing solarly awareness of the role of the European Court of Justice in legal integration had made it clear that the EC/EU was becoming increasingly responsible for the authoritative allocation of values in Western Europe. Prominent commentators came to regard it as a ‘new’ kind of institution (Keohane and Hoﬀmann 1991; Sbragia 1992), observing that the aritecture of territorial rule in Western Europe had substantially anged (Grande 1994). At the same time, there was widespread agreement that the EU did not qualify as a state, as it does not enjoy a monopoly on the legitimate use of force or signiﬁcant taxation rights. Nor could it be reduced to an international organization or confederation, since its rules take precedence over domestic laws and it operates in far more than a single policy area (see Jatenfus 2010). However, the panoply of terms used to describe the EU in the ﬁeld of EU Studies clearly demonstrated that there was mu less agreement on what the EU actually was. In the context of this conceptual ambiguity, aracterizing the EU as a multilevel governance system oﬀered more than just another terminological distinction. e theory provided conceptual leverage, shiing EU Studies away from the sui generis notion inherent in many EU analyses and connecting EU Studies to wider social science developments. Similar to Marks’ claims about the EU, Rosenau had argued with respect to international politics: activity in the global community today is the result not only of nation-states striving for goals, but also of a number of varied transnational collectivities – from multinational corporations to professional societies to international organizations to terrorists, and so on … – engaging in pursuits that are not conﬁned by national boundaries…. Increasingly, it is clear that the breakdown of the old inter-state system is necessitating reformulation of how domestic and international processes sustain ea other. (Rosenau 1993: 5) Although Marks, Hooghe and Blank (1996) did not specify exactly what they understood as ‘governance’ in multilevel governance, beyond the emergence of multilayered interactions and the complex intertwining of competences across diﬀerent levels, their analysis is related to the ‘governance turn’ in Policy Studies and Comparative Politics (CP). Multilevel governance thus became part of a converging resear agenda in International Relations and Comparative Politics. In the laer ﬁeld, the governance concept represented a response to the growing segmentation of both society and state resulting from socio-economic dynamics and institutional evolution. e increasing complexity of actor constellations gave rise to a broad literature on the policy networks, sub-governments and negotiation systems that have supposedly deprived the state of its authoritative position (Rhodes 1997). Despite their diﬀerent roots and emphases, both the IR and the CP literatures emphasize above all the absence of a central authority and the emergence of new processes and means of governing (see Eising and Kohler-Ko 1999a: 4). In that respect, the initial wording of the multilevel governance concept primarily highlighted the emergence of direct interactions among EU political authorities and sub-national entities, as well as the political mobilization of these entities, positing centrifugal tendencies ‘in whi decisionmaking is spun away from member states in two directions: up to supranational institutions, and down to diverse units of subnational government’ (Marks 1993: 402). Subsequent work on the multilevel governance concept was more overtly connected to the general social science debate on governance and the study of federalism: Hooghe and Marks (2003) juxtaposed two ideal types of governance, maintaining their emphasis on the territorial dimension that many other governance studies exclude. Drawing on empirical economics resear on the optimal allocation of jurisdictions in competitive federalism, the authors sought to conceptualize ‘logically coherent types that capture alternative jurisdictional arrangements’ (ibid.: 236). Although competences and resources are distributed and shared across territorial units in both types, these two systems represent diﬀerent ways of organizing political authority. Type I governance features a system-wide aritecture and is based on territorial federalism and its methods of allocating political authority. Its institutions are durable and stiy. Political authority is distributed across general-purpose jurisdictions that exist at a limited number of levels, from the local to the global. Ea jurisdiction is responsible for an extensive bundle of tasks, and there ‘is one and only one relevant jurisdiction at any particular level’ (ibid.: 236). Membership tends to be territorial; the ‘membership boundaries’ of these jurisdictions ‘do not intersect’ (ibid.: 236). As Hooghe and Marks (2003: 236) describe it, ‘every citizen is located in a Russian Doll set of nested jurisdictions’. Democratic Type I jurisdictions usually consist of an elected legislature, an executive and a court system. ese general-purpose jurisdictions tend to foster common identities and voice strategies of their members (ibid.: 240), as they form stable and comprehensive governance arrangements with clearly drawn boundaries. In contrast, Type II governance consists of a set of special-purpose jurisdictions that carry out speciﬁc tasks, su as regulating telecommunications or seing toy safety standards. ere are potentially a vast number of su jurisdictions, and these can operate at various territorial scales across a large number of levels. ey are more ﬂexible than Type I governance arrangements, in that they last as long there is demand for their services (Hooghe and Marks 2003: 236) and as long as they are able to compete with alternative providers of their governance functions. Because they are established to remedy speciﬁc policy problems, their membership depends on utilitarian considerations, su that these jurisdictions prompt exit strategies of their members rather than voice strategies in whi members express their dissatisfaction with these jurisdictions. Frey and Eienberger (1996) have labelled su Type II governance units ‘FOCJ’ – functional, overlapping, competing jurisdictions – and view them as operating within a system of competitive, functional federalism. Real-life examples include the numerous Swiss Zweckverbände, US special districts and, in International Relations, international regimes and treaties. According to Hooghe and Marks (2003: 238), Type I governance is the predominant government type, while Type II arrangements are ‘generally embedded in Type I governance’. eir ﬂexible scope can help exploit economies of scale as well as take into account local preferences for public goods; however, an excessive number of these arrangements may also trigger substantial inter-jurisdictional coordination costs. Although the combination of general-purpose and specialpurpose jurisdictions is said to balance centralization and decentralization costs (see Benz 2009a: 29), there are no universal rules regarding where to draw boundaries between the special jurisdictions and how to adjudicate among them (Piaoni 2009: 171). Principles that are intended to guide the allocation of competences to diﬀerent jurisdictions (su as subsidiarity or proportionality) usually fail to provide clear operating standards from whi the allocation can be unequivocally derived. Nonetheless, advocates of competitive federalism and Type II arrangements suggest that the coordination costs are oﬀset because these units are designed in ways that ‘minimize interjurisdictional spillovers’ (Frey and Eienberger 1996: 319). In the more comprehensive Type I governance (within whi Type II arrangements are embedded), multilevel governance has a statist and institutional core (see Eising 2004: 215). Public actors from at least two levels of government share political authority in formal institutional arrangements. Public actors at the upper level are to some extent autonomous, but lower- level units ‘are not subordinate’ (Mayntz 1999: 101). e EU is usually considered to be a general-purpose jurisdiction, not just a special-purpose jurisdiction (see Jatenfus 2010: 204–5). Notwithstanding its la of statehood, the EU has a mu broader institutional remit than international regimes, and it exerts governing functions within a clearly delineated territory, namely within the borders of its 28 member states. In the EU, several institutions with autonomous powers have been established at the upper (supranational) level: the European Commission, the European Parliament and the European Court of Justice. Its constituent lower-level units, the EU member states, have very strong participatory and decision- making rights, notably through their representation in the Council of the EU and the European Council, and election rights with respect to the College of Commissioners and the judges of the European Court of Justice. e EU Council shares executive functions with the Commission and legislative functions with the European Parliament. In general, a qualiﬁed majority (if not all of the member states) must agree to legislative decisions taken at the EU level. In comparison to most national federal systems, EU multilevel governance is very dynamic and has undergone continuous evolution. Aer 60 years of European integration, it now covers an enormous range of policy areas, largely relying on regulatory policies to create, police and correct the common market. In several of these issue areas, the member states can no longer take autonomous decisions, either on legal or factual grounds. Hence, there is a pronounced need for cooperation and coordination among the EU institutions and national institutions in the EU multilevel system. However, the EU also incorporates important elements of Type II governance. At the most general level, the EU’s authority varies signiﬁcantly across and also within policy areas, as well as along the policy-making cycle. Although the EU has signiﬁcant policy-making capacities, it depends to a large extent on the implementation capacities of its member states. More speciﬁcally, the EU displays some elements of ﬂexible integration. Some member states have opted out of speciﬁc EU policies (see Jatenfus 2010). For example, a subgroup of member states in the Eurozone has adopted a common currency and common institutions (su as the European Central Bank), whereas other member states have kept their own currencies and regimes of monetary policy. In a similar vein, the Sengen system unites those 21 EU member states that have abolished border controls among themselves as well as Switzerland, Norway, Iceland and Lietenstein; however, the United Kingdom and Ireland have decided not to participate in this system. With the Amsterdam Treaty, the EU introduced provisions to enforce the cooperation of EU member states, although they have been infrequently invoked. ese provisions, whi presuppose the failure of EU member states to rea unanimous agreement on legislation beforehand, were ﬁrst used in 2010 when 14 EU member states supported the adoption of a Council Regulation in the area of law applicable to divorce and legal separation. In 2012, this method was used to establish a unitary patent protection system. In February 2013, the Commission suggested the implementation of a ﬁnancial transactions tax to reduce speculative trading and force the ﬁnancial sector to contribute to the cost of the ﬁnancial crisis. is proposal was supported by a group of 11 countries. As these examples illustrate, European governance consists mostly of Type I governance, but includes important elements of Type II governance. Initially employed to study territorial political dynamics in the European Union, the multilevel governance concept has been extended to the general analysis of the territorial structuring of polities. Its understanding of levels is predominantly territorial but also encompasses functional jurisdictions with varying territorial scales. e study of Type I arrangements focuses solarly aention on the vertical interactions among levels, whereas the analysis of Type II arrangements necessitates the study of horizontal interactions among diﬀerent jurisdictions. e typology is abstract and versatile enough to allow not only the study of national federal orders, global governance arrangements and European governance, but also that of sub-national entities and their territorial and functional structures. However, the former are clearly a minority among national modes of governance – Was (2008: Table 2) identiﬁes 25 functioning federations – and the laer are not the focus of aention in the multilevel governance literature. As a concept, multilevel governance is primarily a response to the increasing internationalization of political competences and the ever more complex combinations of (national) territorial and (international) functional jurisdictions. In this sense, the concept represents a response to the paradigmatic limitations of the predominantly single-level models of politics within both IR and CP (Sarpf 2010). Over time, the explanatory focus has shied from the study of power struggles between institutional actors in the EU multilevel system to a more or less functional account of governance structures: ‘Multi-level governance allows jurisdictions to be custom- designed in response to externalities, economies of scale, ecological nies and preferences’ (Hooghe and Marks 2010: 29). In this perspective, multilevel governance is a response to the exigencies of functional problem-solving. e typology of general-purpose and functional jurisdictions highlights the structural ensemble of public institutions and actors involved in governance arrangements as the core of the multilevel governance concept, an aspect that sets it apart from neighbouring concepts and allows it to occupy an important nie in the study of contemporary governance. Its emphasis on the territorial dimension of governance arrangements distinguishes it from understandings of ‘poly-centric governance’ that describe how several formally independent centres of decision-making can perform important governance functions in the same area (e.g. Ostrom 2010). e focus on public actors operating in a multilayered state apparatus diﬀerentiates it from notions of ‘network governance’ (see Kohler-Ko 1999; Kohler-Ko and Eising 1999) that stress the role of private actors and the interactions among private and public actors in these arrangements. e aention devoted to the roles and the interactions of diﬀerent jurisdictional units separates it from the study of ‘new modes of governance’ that investigate the political shi away from the ‘parliament–executive nexus’ and the growing recourse to non-hierarical policy instruments su as self- regulation, benmarking and private dispute resolution (Héritier and Lehmkuhl 2011: 51). Hence, in a broad understanding, it may well be that multilevel governance can imply the study of ‘political mobilization, policy- making, and state-restructuring’, as Simona Piaoni (2009: 175; my emphasis) suggests; however, its more speciﬁc connotation diﬀerentiates it from other theories and concepts in the study of governance. eorizing multilevel governance (in the EU) Marks and Hooghe’s typology focuses on governance structures, outlining the structural aracteristics of multilevel governance, but devotes less aention to the dominant interaction meanisms and governance paerns within multilevel systems. e authors’ work has been criticized on the grounds that ‘it provides a “thi” though compelling, description of contemporary anges in European governance but, in contrast, to standard theories, las a causal motor of integration or a set of testable hypotheses’ (Jordan 2001: 201). is section discusses some eﬀorts to theorize multilevel governance, focusing on German contributions. Fritz W. Sarpf was one of the ﬁrst solars to develop theoretical propositions regarding why and how multilevel governance works and under what conditions it fails (see also Benz 2010: 215). Building on his earlier work on intergovernmental relations in German federalism, he identiﬁed a structural resemblance to the EU multilevel system (Sarpf 1988) – inter alia, with comparisons to the representation of Länder governments in the German Federal Council and its high decision-making hurdles – and has analysed the ‘institutional capacity and the institutional legitimacy’ of the EU multilevel system in several subsequent contributions (Sarpf 2010: 66). In the following analysis, I will consider only aspects of institutional capacity. Beginning with his seminal work on intergovernmental negotiations and the joint-decision trap in EC agricultural policy (Sarpf 1988), Sarpf has persistently emphasized the status of national governments within the EU multilevel system, arguing that the EC/EU systematically generates suboptimal policy outcomes. One important baground condition of his joint-decision trap (JDT) model is that competences are shared between EU institutions and the member states in su a way that a unilateral ange in the status quo is not available to the member states. Hence, these actors must participate in a ‘compulsory negotiation system’ in order to resolve the problems they face. Furthermore, it is signiﬁcant that member states’ self- interests enter negotiations in the EU Council basically unﬁltered (because the national governments are directly represented in these negotiations). ird, decisions in the EU Council must be taken unanimously or almost unanimously. According to Sarpf, not only does this constellation preclude unilateral action by the member states and impede policy innovation at the EU level, it is also not amenable to institutional ange, since this would involve major distributive conﬂicts to the detriment of the self-interests of many member states. Even though fully ﬂedged policy bloades are rare, the expectation is that the JDT will systematically generate suboptimal policy outcomes. Connecting these initial ideas on EU decision-making with his analysis of governance modes in policy-making (Sarpf 1997), Sarpf (2001: 4) came to analyse the ‘vertical interactions’ of EU member state governments in terms of four major interaction modes: mutual adjustment, intergovernmental negotiations, hierarical direction and joint decision- making. Mutual adjustment means that national governments make their own policy oices but take into account the policies that other governments have adopted or are likely to adopt. Policies and international order then emerge from the sequence of moves and counter-moves that governments make without direct communication. ese actors enter into intergovernmental negotiations when they feel the need to coordinate or centralize their policies in order to control border-crossing eﬀects but still want to maintain veto rights for themselves and exclude other actors from decision-making. Hierarchical direction centralizes competences under the control of supranational institutions – notably the European Court of Justice and the European Commission – without any further involvement of the member states. Finally, joint decision-making implies the sharing of decision-making powers by supranational institutions and national governments. According to Sarpf (2010: 69–75), a speciﬁc combination of these modes prevails in the EU: mutual adjustment is the default mode, but in order to control border-crossing eﬀects and mitigate the cost of regulatory competition, the member states also enter into intergovernmental negotiations that grant them full control over the collective decision-making process. is mode was important during the integration crisis following the Empty Chair Crisis and also prevailed in the area of freedom, security and justice during the three-pillar regime aer the Maastrit Treaty; currently, it is still predominant in EU foreign and security policy. According to Sarpf, hierarical direction is most important in EU market-making (so-called negative integration), whi forms the babone of the European integration project. It is here that member states fully delegate decision-making powers to EU institutions and subject themselves to the hierarical decisions taken by the European Central Bank (in monetary policy), the European Court of Justice (in legal integration) and the European Commission (in competition policy). is enables the Commission and the Court to unilaterally advance European integration and eliminate national barriers to the free mobility of goods, services, capital and people, as well as other distortions of free competition. In contrast, when market-correcting (positive integration) measures are needed, member states must take joint decisions in enacting EU legislation. When the collective of member states opposes Commission proposals or when national interests diverge widely and are highly salient, reaing agreement in the EU’s joint-decision mode can be diﬃcult, even where qualiﬁed majority voting applies in the EU Council. e governance hurdles in positive integration are thus signiﬁcantly higher than those in negative integration (whi is more ﬁrmly supported by the EU’s primary law). As a consequence, the EU’s multilevel system suﬀers from substantial governance asymmetries: ‘the market-making policies on whi Europe can agree (or whi can be imposed through hierarical direction) will damage the capacity of national governments to adopt those “market-correcting” policies on whi the Union cannot agree’ (Sarpf 2010: 75). According to Sarpf, this problem-solving gap tends to arise exactly in those policy areas ‘where national governing functions are most vulnerable to systems competition’ (ibid.), increasing the existing problems in these areas. Nonetheless, the joint decision-making mode is crucial because ‘its potential to create eﬀective European solutions and binding rules is signiﬁcantly greater than that of stand-alone intergovernmental negotiations, let alone of mutual adjustment’ (Sarpf 2011: 220). Moreover, it seems more susceptible to democratic self-determination than the other modes. Sarpf’s analysis speciﬁes the predominant interaction paerns within the EU’s multilevel systems and connects them to distinct pathologies of EU multilevel governance. Among the most pertinent problems are those resulting from joint decision-making: suboptimal policy solutions and non- decisions (although these are rare, especially in comparison to the legislative process in the United States), the lo-in of prior decisions and the impossibility of policy reversals for elected politicians (Falkner 2011b: 10). In short, joint decision-making makes major policy innovations unlikely. At the same time, it renders major institutional reforms anging the conditions of the compulsory negotiation system almost impossible, as su reforms would involve the substantial redistribution of competences and resources among the member states. Subsequent studies have taken issue with this assessment of decision- making pathologies in EU governance. Several contributions in a systematic analysis of joint decision-making in diﬀerent EU policy areas (Falkner 2011a) have identiﬁed important exit and consensus-building meanisms that are employed to facilitate policy agreement in the EU (Falkner 2011c): the selection of treaty bases that require a less demanding decision rule than unanimity, the shiing of policy arenas to introduce diﬀerent actors that may counteract the politicization of an issue (experts, bureaucrats), as well as the selective inclusion or exclusion of private actors to promote, support or hamper opposition, the strategic use of information, the pressurization of national governments via strategic partners or public discourse, the unseling of the member states through the reinterpretation of the legal status quo, and the ‘hardening’ of so law. Among the consensus-building meanisms allowing an escape from the JDT are the redeﬁnition of issues in terms of their scope (downsizing) or quality (downgrading), issue linkages and side payments, exemptions from EU requirements on a temporary or permanent basis, socialization eﬀects on the basis of long-term cooperation and anticipation eﬀects of future cooperation in a policy area. Although these meanisms are well known to students of EU decision-making (e.g. see Peters 1997; Héritier 1999; Eising 2002), systematically connecting them to the practices and pitfalls of multilevel governance as they are conceived in the JDT model contributes to theory formation in multilevel governance. To this end, Sarpf has suggested that the original JDT model should be extended to include the impact of judgments by the European Court of Justice on the policy status quo and member states’ preferences, as well as the strategic role played by the European Commission in EU policy-making (Sarpf 2011: 232). Accordingly, the evolution of governance modes in the EU is now viewed as the outcome of the interactions and the preference constellations of both supranational institutions and member states. Sarpf’s joint-decision trap accounts particularly well for the stagnation of European integration from the Empty Chair Crisis in the mid-1960s until the late 1970s. However, it was published at a time when the EC was entering a phase of major institutional innovations that reformed EU decision-making – the treaty reforms (beginning with the Single European Act and continuing through the Lisbon Treaty) that moved more and more policy areas from unanimity decisions to qualiﬁed majority voting in the Council and enhanced the decision-making rights of the European Parliament, anging the very conditions of the JDT and indicating the limits of the argument. Nonetheless, Sarpf’s coherent theory systematically connects multilevel structures to decision-making practices, policy outcomes and normative evaluations, highlighting the fact that actors adapt their strategies and interaction modes to the institutional conditions of multilevel governance. Building on this insight, Arthur Benz (2009a) has developed a theoretical understanding of the EU’s multilevel governance system that devotes considerable aention to the strategies the actors employ to prevent policy impasse in the EU. In this perspective, the types of linkages that exist between the institutions and processes at ea level (intra-governmental arena) as well as between these levels (intergovernmental arena) inﬂuence the strategies that actors employ in multilevel systems. In this respect, Benz diﬀerentiates between loosely coupled multilevel systems and tightly coupled systems. He claims that tightly coupled systems (su as German federalism) tend to reduce reform capacities, whereas loosely coupled systems (su as the European Union) allow more eﬀective governance in comparison to the former, since they grant veto players greater leeway to cope with the conﬂicting demands they are exposed to at ea level (Benz 2009a: 82–5). Drawing on Putnam’s (1988: 434) notion of two-level games in international politics, Benz suggests that national governments need to take into account will-formation in national parliaments, parties and constituencies, while also bearing in mind the interests of the other member state governments in EU Council negotiations. Policy solutions that might be agreed upon at the EU level may not ﬁnd domestic support and, vice versa, policy solutions that might be acceptable within a member state may not be supported by other member states. Likewise, the representatives of interest groups in EU-level associations or EU-level expert commiees may agree to positions formulated in these bodies but face resistance from their members. e overaring idea is that multilevel systems are prone to su dilemmas, and actors sear for ways to escape them when addressing policy problems. Benz discusses several institutional conditions facilitating the loose coupling of the EU and the national arenas (Benz 2009a: 140–2). First, there are the agenda-seing rights of the European Commission, whi relies heavily on policy experts when developing its proposals, separating the sear for innovative policy solutions from member state bargaining. Second, there is the mediating role of the Council presidency in EU Council negotiations. ird, the institutional diﬀerentiation of the EU Council’s mainery, ranging from its expert groups to the Commiee of Permanent Representatives (COREPER) of the member states at ambassador level to the level of ministers, allows for level-shiing in the sear for policy solutions. Fourth, there is the mutual adjustment of the EU Council and the European Parliament in EU legislation, particularly during the so-called trilogue with the European Commission. And ﬁnally, there is the decoupling of EU politics from the vagaries of national party politics and party competition. Important strategies employed in multilevel interactions include issue linkages, side payments, budget increases, reliance on fairness criteria and level-and arena-shiing (Benz 2009a: 143–4; Eising 2002). Moreover, in the member states parliamentary or semi-presidential democracies, parliaments usually refrain from exerting their veto rights vis- à-vis their national governments, instead granting them a ﬂexible negotiation mandate or restricting themselves to the ex-post control of their government’s negotiation behaviour, allowing them to hold the respective minister in arge of negotiations accountable (Benz 2009a: 145–7). e work of Sarpf and Benz constitutes important theoretical contributions to the study of multilevel politics in the European Union. Rather than developing holistic theories or concepts of EU multilevel governance, they draw on established theoretical frameworks (su as neo- institutionalism, rational oice, systems and negotiation theories), integrate interaction meanisms and mid-range theorems into these theories (hierary, joint decision-making, loose and tight coupling of levels, etc.) and link them to the institutional conﬁguration of multilevel seings. ese theories of multilevel governance identify speciﬁc coordination problems inherent in multilevel decision-making systems as well as ways to escape these problems. roughout, emphasis is placed on the interactions between institutional structures, interaction strategies and policy outcomes. Varieties of EU governance Further work on multilevel governance in the EU reﬂects the variety of governance modes now present in the European Union. ese studies vary in terms of how the governance modes are deﬁned and what actors are involved. For instance, while Sarpf limits his analysis to the interaction modes of national governments (and the supranational EU institutions), other studies have examined various EU governance structures or governance instruments across the divide of public and private actors. Several researers have found that governance modes diﬀer not just across market-making and market-correcting policies, but also in diﬀerent policy areas within ea of these ﬁelds (see Wallace 2005). It has frequently been observed that the number of governance modes and the mix of these modes within individual policy areas have increased over time in the EU system, although speciﬁc governance modes may continue to dominate in some policy areas, su as voluntary forms of cooperation in education policy and the Bologna process (Tömmel and Verdun 2009: 295). In short, the predominant constellation of EU governance modes is unclear. In the subsequent analysis, I will focus on three governance modes that are frequently referred to in EU governance (and elsewhere): hierary, competition and networks. Tanja Börzel (2007) suggests that EU governance modes generally operate under the shadow of hierary. Her assessment is based on a very broad understanding of hierary that subsumes even majority voting in the EU Council’s joint decision-making under the term (Börzel 2007: 70). Nonetheless, several authors agree that hierary, even when more narrowly deﬁned as unilateral direction, is not limited to the EU’s market-making policies; rather, it extends into a far greater array of EU policy areas. Gerda Falkner (2011c: 254) argues that ‘a policy’s quality of speciﬁcation on the level of EU primary law … is crucial’ when accounting for hierarical decisions by the Commission or Court: ‘Oen, that happens to go hand in hand with market making, but not always and not exclusively’ (ibid.). Moreover, policy implementation in the EU multilevel system relies heavily on hierarical governance instruments: directives, regulations and decisions (Eising and Lensow 2008: 263). An examination of the use of these governance instruments between 1975 and 2012 demonstrates that the EU now employs regulations less than in the mid-1980s and early 1990s (when it implemented the bulk of the Single Market Programme) but continues to produce a steady ﬂow of directives; its use of decisions increased in the mid-1990s (see Figure 9.1). Héritier and Rhodes (2011: 163) point out that ‘the shadow of hierary … looms large’ over the new modes of governance. e ‘credible threat of legislation … usually prompts their emergence as alternatives to traditional forms of decision-making’ as well as securing their eﬀectiveness. However, a caveat is in order: Ingeborg Tömmel (2009: 16) argues that, in comparison to national states, the mode of hierary is important but is not as widely used and its inﬂuence is also weakened in the EU system, the reason being ‘the multilevel and multi-actor structure of the Union. e EU has not only to establish common rules but also to accommodate diversity among the member states.’ us, the implementation of EU directives allows some degree of discretion on the part of the member states, and competition policy decisions are frequently compromises formed with the addressees of these decisions rather than top-down rules. Figure 9.1 e annual number of EU legislative proposals (1975–2012) Source: Häge (2011); European Union Policy-Making Dataset (EUPOL) v04 (1975–2012) (accessed 10 August 2013). e competition mode is more important in the EU than it is in the member states. Not only is it the primary instrument used to govern the common market and foster competitive relations between economic actors, but it is also increasingly employed to foster policy competition and ‘policy convergence among member states’ (Tömmel 2009: 15). However, competition is not the only means to establish the common market; until the mid-1970s, the European Union relied on hierarical standard-seing to establish the common market. When it became evident that this mode was not suited to the task of harmonizing standards throughout Europe, the European Union resorted to two alternative meanisms to advance the common market project. First, the integration of the common market and the interdependence of member states’ economies have been accelerated by the provisions for mutual recognition that were introduced by the Single Market programme in the mid-1980s, obliging member states to accept one another’s regulatory standards as equivalent. Mutual recognition avoids the negotiation and implementation costs of new regulations (Smidt 2009: 124) and is a substitute for harmonization. It has the potential to trigger regulatory competition among the member states that may, in principle, result in lower or higher regulatory standards. To avoid a ‘race to the boom’, mutual recognition is embedded within a regime of minimum harmonization at the EU level and allows for exceptions, should member states have good reasons to stipulate mandatory requirements of domestic speciﬁcations that foreign goods must comply with and that conform to the proportionality principle. Whenever a minimum harmonization of tenical standards is deemed necessary under the so-called New Approa to standard-seing, member states only formulate essential requirements; the task of translating these into speciﬁc standards is delegated to private standard-seing bodies that are umbrella organizations for national private or para-public standards organizations (see Egan 2009). Moreover, in order to assess the equivalence of national standards under mutual recognition, extensive paerns of transgovernmental cooperation have emerged for both goods and services. In contrast, policy competition is intended to foster both mutual learning through communication and benmarking in order to increase the member states’ policy performance and trigger a race towards best policy practices. e most important example here is the Open Method of Coordination (OMC), whi has been introduced in a variety of policy areas concerning public welfare and services (Benz 2009b: 36) as part of the Lisbon strategy to develop Europe ‘into the most competitive and dynamic knowledge-based economy in the world’ (European Council 2000: 2). In the EU multilevel system, the OMC systematically connects non-binding European guidelines and benmarks with national action plans and policy reforms (Eising and Lensow 2008: 268). For the member states, this allows policy problems to be addressed without the surrender of formal competences to the EU (Héritier and Lehmkuhl 2011: 56). e Commission seles for this second-best solution of EU governance from its viewpoint because it anticipates the resistance of the member states to joint decision- making and because it las hierarical governance instruments. Several contributions have emphasized the importance of the meanisms for negotiations among multiple private and public actors (beyond purely intergovernmental negotiations or the joint decision-making of the EU institutions). Some early studies of European governance stressed the signiﬁcance of network governance in the EU (Kohler-Ko and Eising 1999), highlighting the fact that governance in the European Union relies on uniting the relevant state and societal actors and building issue-speciﬁc constituencies. Accordingly, in network governance, state actors and a variety of interest organizations, institutions and experts discuss and negotiate the allocation of speciﬁc values. Given that the EU develops policies for 28 member states, the territorial and functional representatives of national interests come from widely diﬀering political, social and economic systems. Network governance asserts that the functional segmentation of specialized Councils and the Commission’s directorates, regulatory agencies and advisory commiees gives rise to specialized sub-structures within Type I governance that develop their own rationality criteria and operating procedures (see Eising and Kohler-Ko 1999b: 269). e actors involved are autonomous and interdependent. is concept underscores the fact that the Commission proposes EU legislation and that the Council and the Parliament take joint decisions only aer discussion and negotiation with a multitude of private or regional actors: in August 2013, the EU transparency register listed 5,872 interest organizations that seek to inﬂuence EU policies; in comparison, the German Bundestag currently registers 2,034 interest organizations that seek to inﬂuence German legislation. ese actors are consulted and provide information in online consultations; they are also heard in informal meetings and are represented in expert commiees. A few of them, notably the social partners, are accorded rights to participate in the formulation and implementation of EU policies. Depending on their representativeness and their inﬂuence over their members, they may generate support among the target population for the policy measures in question. ey can further contribute to EU governance when the EU institutions delegate powers of self-regulation to them (as in the case of the voluntary agreements of the European paper and PVC industries in environmental policy), when they are authorized to ﬁll in the details of framework directives or standards (as seen in health and safety regulations) or when they act as co-regulators with public institutions (su as the EU’s standard-seing bodies) (Héritier and Lehmkuhl 2011: 60). Some of these actors are also involved in the working groups administering the implementation of EU law. e density of these networks and the amount of trust they generate among the actors involved may be lower than in national politics (because the EU is more active in policy formulation than in policy implementation), but in some cases they can stabilize interaction paerns and contribute to solving EU policy problems. However, it goes without saying that policy networks can also stand in the way of policy anges. Hence, network governance may not be the predominant mode of EU governance (although it is embedded within the EU’s institutional seing), but it is important in the preparation of joint decisions by the EU institutions and also in the implementation of these joint decisions and the resolution of coordination problems during the implementation of EU policies. It purportedly has the potential to strengthen democratic control ‘by involving stakeholders in policy-making and by mobilizing their commitment to speciﬁc policies, thereby increasing output legitimacy’ (Héritier and Rhodes 2011: 164). Private actors and multilevel governance Many studies of the European Union emphasize the fact that multilevel governance is also multi-actor governance. Private actors are assigned an important role in the governance of the European Union; in fact, the move from hierarical direction towards more horizontal relationships among state and private actors is oen considered to be the deﬁning aracteristic of the shi from government to governance. e concept of network governance reﬂects this development, and the paerns of private-actor participation in EU politics underscore their relevance to EU governance. Given the signiﬁcance aaed to private actors in modern governance, this section more closely examines the impact of multilevel governance systems on interest mediation. In this respect, a major puzzle involves how interest organizations adjust to the opportunities and constraints imposed by multilevel political systems (see Beyers et al. 2008). Adjustment can refer to the ability to adapt the organizational format, the adoption of speciﬁc political strategies, the sear for new allies or the modiﬁcation of the policy agenda. Importantly, an organization’s embeddedness within institutional contexts or its privileged consultation by a national or international institution can ease or stimulate adjustment. Some groups ﬁnd it easier to take advantage of multilevel politics than others (Eising 2004). Multilevel systems have important consequences for interest groups. e fundamental institutional aracteristic of Type I governance arrangements is the distribution and sharing of power between the upper level and its constituent units. I limit my discussion to these arrangements because they apply to the EU and because many aracteristics of network governance correspond to interest representation in Type II governance. While the speciﬁc form and distribution of authority across and within levels varies signiﬁcantly across multilevel systems (see Was 2008: 83 on territorial federations), in general, Type I multilevel seings tend to oﬀer interest groups a variety of points of access and options for ‘venue shopping’. In multilevel governance, ‘interest groups at any territorial level are free to lobby government at any number of levels’ (Constantelos 1996: 30). Political representation is no longer conﬁned to a single jurisdiction. erefore, David Coen (2007: 339) suggests that ‘it is logical and responsible’ for interest organizations ‘to develop a mix of political annels to inﬂuence policy’ in the EU multilevel system. However, interest representation at diﬀerent levels is costly. Even if interest representation is the raison d’être that forces them to respond to anges in the political environment, many interest organizations are tied to their members and constituencies, as well as to the national or regional contexts in whi they emerged. ey are embedded in social relationships (see Granoveer 1992) and are dependent on routine exanges with established partners (see Wilson 1973). eir speciﬁc location in the EU multilevel system shapes their political activities as well as their access to the political institutions. An important response to the formation of political multilevel systems is organizational isomorphism, in the sense that interest group systems are also structured in several layers, with the territorial scope of ea interest group layer corresponding to the boundaries of the political units at that level. Accordingly, in the EU’s multilevel system a large number of EU-level interest organizations are federations of national interest organizations whose organizational costs can be shared by their members. e presumption, then, is that in European Union governance national associations concentrate their activities on domestic institutions, whereas EU associations focus their political activities on EU institutions. Similarly, in German federalism it is mostly nationwide interest organizations that enter into political exanges with federal institutions, whereas their Länder organizations are in arge of relations with Länder governments and parliaments. However, there are various reasons why national groups may not limit themselves to representing interests through their EU-level federations, instead becoming active themselves vis-à-vis the EU institutions: when EU regulation has a major impact on these groups or on their members, when the division of labour among them and the EU associations that are supposed to represent them vis-à-vis EU institutions are unsatisfactory, or when the terms of EU policy implementation must be worked out at the EU level, it is likely that these groups will extend their activities to the EU level. It is generally accepted that interest organizations require substantial governance capacities to represent their interests at diﬀerent levels in the EU multilevel seing. e abilities to recognize the needs of state actors in decision-making processes, to mediate between the competing demands of state institutions and their own members, and to contribute to compliance with and implementation of public policies (by means of interest group self- regulation or co-regulation) are crucial (see Smier and Stree 1981). In part, these abilities hinge on an organization’s type. Empirical studies indicate that, in line with Mancur Olson’s logic of collective action, speciﬁc interest organizations are more heavily involved in EU policy-making than diﬀuse interests, even though various annels are open to diﬀuse interest organizations. Diﬀuse interest groups ‘la a well-delineated and concentrated constituency’; they defend ‘interests that are linked to broad and general segments of society’ that are not necessarily identical to the self- interests of their members (Hassel 2010: 161). In contrast, speciﬁc interest organizations have a well-delineated and concentrated constituency. ey defend the material interests of their members, whi tend to have clear-cut commercial or professional concerns. Moreover, in part, the governance capacities of groups are shaped by the ways in whi they are incorporated in public policy-making and implementation. e study of national interest group systems diﬀerentiates between two major types of interest group systems featuring groups with rather diﬀerent governance capacities: pluralism and corporatism. In pluralist seings, a multitude of interest organizations compete in the same domains and strive for political inﬂuence, without any public authority for self-regulation or co-regulation. In corporatist seings, a limited number of centralized interest organizations that are highly representative of their domains also compete for inﬂuence, but are authorized to formulate public policies and/or implement these measures. In exange, state institutions expect them to moderate their demands on public policies (Hassel 2010). In sum, interest organizations in corporatist seings are likely to have broader governance capacities than those rooted in pluralistic seings. What type of interest-group system and what governance capacities of interest groups can we expect in multilevel systems? Studies on the relationship between federal systems and interest-group systems suggest three reasons why multilevel systems should be more conducive to the formation of interest groups than unitary systems (see Armingeon 2001: 214). First, multilevel systems allow greater disparities across regions in interest-group organization than unitary states. Second, cultural, social and economic diﬀerences are more pronounced in multilevel systems than in unitary states, giving rise to a greater variety of interest organizations. Finally, the dispersion of political authority in multilevel systems results in greater diﬀerentiation within the associational landscape than the centralization of political authority in centralized political systems does. e expectation for Type I multilevel systems is thus that the greater dispersion of political authority across diﬀerent levels will result in greater interest-group pluralism. Figure 9.2 displays the bivariate association between the extent of federalism and interest-group pluralism in 36 established democracies. Data are taken from Arendt Lijphart’s (2012) revised and updated study on paerns of democracy in 36 countries for the period from 1945 to 2010. e graph includes two regression lines. e negative association between federalism and interest-group pluralism is based on the analysis of all 36 democracies, including those that cannot be regarded as multilevel systems in the strict sense, as they display only a very limited dispersion of political authority across the diﬀerent levels. ese highly centralized countries exhibit great variation in terms of interest-group systems, ranging from strong corporatism in Luxembourg to pronounced pluralism in Greece. Evidently, in highly centralized countries the structure of the interest-group systems strongly depends, inter alia, on the ways in whi groups are incorporated into public policy-making and implementation. Although this is also true in multilevel systems, of course, the greater dispersion of political authority and the larger number of access points in these countries should nonetheless give rise to greater interest- group pluralism than is the case in unitary states. Figure 9.2 Multilevel governance and interest-group pluralism in 36 democracies (1945–2010) Source: Author’s calculation, data based on Lijphart (2012). Note: Regression 1 Intercept 2.337 – 0.151 index of federalism (SE 0.92, F 2.215 on 1 and 34 df, p 0.146, adj. r 2 0.034). Regression 2 Intercept 0.009 + 0.400 index of federalism (SE 0.81, F 6.435 on 1 and 15 df, p 0.023, adj. r 2 0.254). When we exclude unitary countries from the analysis and focus only on those countries that display at least some competence dispersion and sharing across diﬀerent levels (i.e. countries with a value greater than 1.5 on the federalism index), the relationship between state structures and interest- group systems anges: the more political authority is dispersed (and shared) across diﬀerent layers of government, the more pluralistic the national systems of interest representation turn out to be. Germany, Austria and Switzerland represent important exceptions to this paern. Not mu resear has been done on this question, but one underlying reason might be that particularly the former two represent an ‘extreme form’ of cooperative or interloing federalism, ‘reducing the opportunities for ﬂexibility and variety of policy through autonomous decision-making by diﬀerent governments’ (Was 2008: 84). Hence, tight coupling in multilevel systems with policy-making authority allocated to the upper level and policy- implementation authority to the constituent units may promote greater interest group centralization. In itself, this is insuﬃcient to account for the emergence of interest-group corporatism. From Stree and Smier’s (1991) seminal study of interest representation in the European Union, further requirements can be inferred: upper-level control of (ﬁnancial) policy instruments, the annelling of access to the central institutions, a power balance between the diﬀerent interest groups involved in corporatist arrangements, high salience of the policy area in whi corporatist arrangements are formulated and a manageable divergence of interests based on more or less homogeneous policy contexts in the constituent units. Conclusions e concept of multilevel governance is a response to the internationalization of politics and the increasing interactions of territorial and functional jurisdictions. It highlights the structural ensemble of the public institutions and actors involved in the governance arrangements at the core of the multilevel governance concept. is focus sets it apart from neighbouring concepts and allows it to occupy an important nie in the study of contemporary governance. Multilevel governance must not be equated with a speciﬁc mode of governance. e governance literature has identiﬁed a variety of governance modes – e.g. hierary, networks, competition, negotiations – that are now combined in the EU multilevel system. eories of multilevel governance draw on well-established theoretical frameworks, integrating causal meanisms into them that stress the operating logic and institutional conﬁguration of multilevel seings. As su, they are far more general than sui generis theories of European integration. Emphasis is placed on the interactions between institutional structures, interaction strategies and policy outcomes. Finally, multilevel governance tends to promote interest-group pluralism and works to the advantage of groups with substantial governance capacities. Note 1 e author anowledges support of the German Science Foundation for this resear under the DFG grant EI 461/6–1. Bibliography Armingeon, Klaus (2001), ‘Verbändesysteme und Föderalismus. Eine vergleiende Analyse’. In: Arthur Benz and Gerhard Lehmbru (eds), Föderalismus. Analysen in entwicklungsgeschichtlicher und vergleichender Perspektive, special issue of Politische Vierteljahresschrift 32: 213–33. Wiesbaden: VS Verlag. Bartolini, Stefano (2011), ‘New Modes of European Governance: An Introduction’. In: Adrienne Héritier and Martin Rhodes (eds), New Modes of Governance in Europe: Governing in the Shadow of Hierarchy. New York: Palgrave Macmillan, pp. 1–18. Benz, Arthur (2009a), Politik in Mehrebenensystemen. Wiesbaden: VS Verlag für Sozialwissensaen. –––– (2009b), ‘Combined Modes of Governance in EU Policy-making’. In: Ingeborg Tömmel and Amy Verdun (eds), Innovative Governance in the European Union. Boulder, CO: Lynne Rienner, pp. 27–44. –––– (2010), ‘e European Union as a Loosely Coupled Multi-level System’. In: Henrik Enderlein, Sonja Wälti and Miael Zürn (eds), 10 e EU as a citizens’ joint venture Multilevel constitutionalism and open democracy in Europe Ingolf Pernice Introduction e year 2014 is the ﬁh year of operation under the Treaty of Lisbon, and the Union thus reformed has gone through a diﬃcult period. e ﬁnancial crisis has led to an economic depression, but also to new trends of scepticism and even nationalism, both in the southern countries that were so dramatically aﬀected by the crisis and the austerity policies imposed upon them and in the northern countries whose citizens have found it diﬃcult to accept the demands of solidarity (Pernice 2013b: 25–56). With a view to the embarassing results of the European elections, this seems to be reason enough for a re-theorization of Europe – a contemplation and explanation of what Europe is and what basically constitutes the European Union. Is it a Union of states alone, or is it a Union of citizens? Further developing the concept of ‘multilevel constitution alism’, the present contribution seeks to raise awareness of the role of the citizen as the real source of power and legitimacy in the European Union. A Union of states and citizens In terms of political philosophy, the EU is commonly described as an international or supranational organization (Polla 2005: 357–98; Risse- Kappen 1996: 53–80; Stone Sweet and Sandholz 1998: 1–26), an organization sui generis, if not a federal state (Mancini 1998: 29–42; Sa 2005: 67–98) or an unidentiﬁed beast – a monstro simile (Pufendorf 1994 : 198–9) impossible to deﬁne, eluding any aempt at description.1 Some degree of common understanding is required to describe the type of venture the citizens of the European Union are engaging in ever more deeply as European politics become increasing relevant for their daily life. However, this is not the place to rehash all the aempts to qualify or categorize this speciﬁc type of political organization that have failed to clarify what we observe taking shape step by step: a federation of states, a compound of states, an association of sovereign states (Staatenverbund) or a supranational union (Wiener and Diez 2009). Most of these terms make reference to states. But is the European Union really only a maer of states? At least legally, ever since the very early case law of the European Court of Justice (ECJ) (ECJ case 26/62 – Van Gend & Loos 1963 ECR 1; case 4/64 – Costa/ENEL, 1964 ECR 585), individuals seem to play a signiﬁcant role in the process of European integration. is role is hidden behind all the state- oriented concepts, but debates on the proper protection of fundamental rights, democratic legitimacy and the principles of subsidiarity and solidarity in the EU clearly demonstrate that this organization diﬀers in nature from all other traditional forms of transnational cooperation. Multilevel constitutionalism Here is where the concept of ‘multilevel constitutionalism’ can serve a beneﬁcial purpose. e term is not used to describe the European Union, nor is it intended to give this political system a categorial name. e issue is, rather, to develop an underlying concept for theorizing the European Union in a constitutional perspective. Since its inception, the concept of multilevel constitutionalism has undergone many aempts by various authors to determine its scope and interpret its meaning, receiving both acclaim and criticism (for a survey of both aﬃrmative and critical assessments of the concept of multilevel constitutionalism, see Pernice 2009: 352). It is primarily the proponents of the traditional theory of the state (Staatslehre) who critically target the underlying functional concept of the constitution and its application or extension beyond the state (Kirhof 2006: 768–76). Others (Bogdandy and Sill 2010: 702–7) view this as part of the federal tradition, but criticize its ‘uncertain aitude toward sovereignty’ (Walker 2003: 14). On an analytical note, the conceptual terminology (i.e. the use of ‘levels’ and ‘multilevel’) has been criticized as evoking a hierarical structure or remaining ambiguous, whi diminishes its descriptive value (Cananea 2010: 83–317). In this vein, the author further submits that the concept regreably over emphasizes the vertical dimension (see, however, Pernice 2006). In a recent publication, René Barents (2012: 159–83) has harshly criticized the four pivotal premises upon whi the concept of multilevel constitutionalism rests (for a ﬁrst reply, see Pernice 2013a): ﬁrst, the ‘unity in substance’ thesis, whi holds that the EU constitution and the national constitutions of the member states form a coherent and substantive whole; second, the ‘European citizenship’ thesis, whi argues that the EU’s legitimacy can and must be traced ba to the collective will of its citizens (voiced, mediated and executed through their respective national governments); third, the ‘autonomy’ thesis, whi presents the EU’s legal order as autonomous with regard to the national legal orders, a concept that lies at the very heart of the idea of a non-hierarical constitutional composite in a pluralistic seing; and, ﬁnally, the ‘divided sovereignty’ thesis, whi states that the EU and the member states jointly bear and exercise sovereign public power. While some commentators (Mayer and Wendel 2012: 127) underscore and expound the fundamental link between multilevel constitutionalism and constitutional pluralism, others (Jestaedt 2004: 638, 662, 664), in the Kelsenian tradition of legal theory, question the possibility of a pluralistic framework in general. More speciﬁcally, Neil Walker properly distinguishes between ‘narrower’ and ‘wider’ notions of multilevel constitutionalism. As a ‘narrower’ concept, it focuses on the EU context and the vertical relationship between the Union and its member states (understood as non-hierarical) as well as the laer’s horizontal relationships; in addition, it substitutes the concept of the constitution for the concept of the state, a notion that is more concerned with abstract quality (constitutionalism) than concrete entities and presents itself as centred on the citizens rather than the polity. is narrower notion, however, may well be (and, in fact, has been) explored in a ‘wider’ sense, as Walker does, expanding multilevel constitutionalism beyond the conﬁnes of the EU seing to investigate the application of constitutional ideals, institutions and practices beyond the state at large (Pernice 2006b: 973–1005; Walker 2010: 143–68). Developing the citizen’s perspective On the basis of ‘multilevel constitutionalism’, the present proposition is to submit and further develop a comprehensive understanding of the progressive construction of the Union as a divided power system – or, beer, as a process of ‘constituting’ the EU multilevel structure in the original sense of the term ‘constitution’, a word derived from the Latin constituere, meaning ‘puing together, constructing, establishing’, and giving this process a name. Who are the authors and actors in this process? Who is at the origin of the EU? And who is able – and has the legitimacy – to drive this process forward? In terms of multilevel constitutionalism, the answer is: ideally the citizens alone, the citizens of the EU member states acting through their national governments, thereby – directly by referendum or indirectly via representation in their parliaments – eﬀectuating the Treaties establishing the European Union. Could there be anybody else in any democratic system equipped to do this? To be absolutely clear: in modern democracies, nothing ‘earthly divine’ (Buwalter 2008: 495–509) or absolute remains in the state. Given the interdependence of states in the age of globalization (or in the ‘postnational constellation’, as Habermas [2001: 58 et seq.] puts it), there is similarly no room le for ideas su as absolutism of states or sovereignty. If there is any sovereignty at all, it is the sovereignty of the people. ‘People’ here does not refer to an abstract entity, a Volk or nation; rather, it has a political meaning as the individuals who have decided to unite and constitute themselves as the subjects of a legitimate power by organizing themselves in the form of a political community that we typically call a ‘state’ and assuming citizenship in the resulting body. e instrument used to accomplish this is the constitution of that state. We should consider the process of the constituting of Europe in the same fashion. e same people, citizens of their respective member states, through their national governments and parliaments, have commonly agreed upon treaties by means of whi they have constituted the EU as a supranational political entity to serve their common purposes and interests through common institutions acting on their common behalf. us, these citizens of the member states are mutually granting ea other a new additional identity by establishing through the EU Treaties a complementary legal status: citizenship in the European Union. e constitution of the European Union and its further development can therefore be called a citizens’ joint venture (Pernice 1999: 727; 2001: 166–8).2 Here, emphasis is placed on the citizens as the true authors and owners of the EU, no less than they are authors and owners of their respective national legal-political orders. Both the member states and the EU are serving the citizens’ interests, according to the competences conferred to ea level of action. In this regard, reference can be made to the famous description of the federal system provided by James Madison in the ‘Federalist No. 46’: e federal and state governments are in fact but diﬀerent agents and trustees of the people, instituted with diﬀerent powers, and designated for diﬀerent purposes. (Hamilton et al. 1787/88: No. 46) Article 10 of the Treaty of European Union (TEU), whi sets out the principle of representative democracy in the Union, refers to the dual aracter of legitimacy, underlining its two diﬀerent strands of accountability: to the European Parliament and to the national parliaments – or to the citizens directly. e provisions on democratic principles in the Treaty clearly indicate that the citizens are the source of legitimacy: Article 9 TEU requires the Union to ‘observe the principle of equality of its citizens’, Article 10(4) TEU anowledges the role of political parties at the Union level to ‘contribute to … expressing the will of citizens of the Union’ and Article 11 TEU deﬁnes the participatory rights of citizens, including the citizens’ initiative. Conceptualizing the European Union from the citizens’ perspective allows us to detect and correct certain misunderstandings underlying arguments that question the democratic legitimacy (regarding the alleged democratic deﬁcit of the EU, see, for instance, Moravcsik 2002) and even the desirability of the Union as su: 1. e conferral of competences upon the European Union will progressively extract powers from national parliaments, to the point that general elections at the national level will become meaningless. 2. Further European integration will place the national sovereignty of member states at risk, as the national parliaments and their governments will be compelled to implement policies in concreto, occasionally even without their prior consent. 3. Democracy and collective political self-determination are endangered in the member states due to the remoteness of European institutions from the citizens and a la of democratic accountability at the European level. If we understand the citizens to be the source of any legitimate aribution and exercise of public authority in a political system, answers to these allenges can be summarized by three principles: the principle of additionality, the principle of voluntary participation and the principle of open democracy. e ﬁrst principle concerns powers and shared sovereignty, the second is about exercising sovereignty and the third focuses on legitimacy. All three address (from their respective perspective) the issue of the ‘democratic deﬁcit’ in the Union. ese principles shall ﬁrst be explained before some conclusions are drawn regarding the upcoming reform of the EU. Powers: the principle of additionality To illustrate what is meant by this principle, the metaphor of a troubled apartment house may help. Having experienced several conﬂicts among the families living in the house, some tenants felt the need to convene regularly to discuss and resolve issues of common interest. However, there was no room big enough for su a meeting. So the tenants came up with the idea of constructing a meeting room in an upper ﬂoor for these purposes. e room and the meetings would be open to all the other families to join, subject to their acceptance of common rules. e joint venture proved to be successful. e tenants found that it was beneﬁcial for ea of them when issues were commonly discussed and decided under their established rules. is immediate success encouraged other families to join. We could further develop this metaphor to mirror the European Union. What it suggests is that there are maers of importance for the peoples of the member states of the European Union that can be solved in common, at a supranational level, more eﬀectively than by ea state individually. Supra – that is, the meeting on the upper ﬂoor – does not necessitate hierary,3 although we are talking about a multilevel structure. Peaceful coexistence in Europe was the ﬁrst focus of European integration, as countries had proven themselves incapable of ensuring this on their own. Other issues followed. Matters beyond national reach and democratic self-determination e principle of additionality means that the supranational structure adds to the member states and their respective power. e European Union was designed to address allenges that the individual member states could not handle on their own. In eﬀect, the powers conferred upon the EU are not powers that the member states previously possessed; rather, they represent new competences added to those of the member states in the form of collective action through common institutions. States would not give away their power voluntarily. Instead, their citizens, by common agreement, have found it useful to establish new institutions with powers that are additional to those of their member states. If it is true that in democratic societies people confer powers upon institutions by means of their constitutions, the origin of the new powers conferred to the European Union cannot be states, but instead only citizens. e citizens of the European member states, through the European Treaties, have constituted and further developed a new instrument, in addition and complementary to their respective nation- states, in order to aain the objectives the several states in isolation were themselves unable to aieve. Would people or politicians accept the idea that certain issues will be decided at the regional level, when they can be dealt with eﬃciently by local authorities; or at the national level, when they can easily be seled by regional authorities; or at the European level, when member states could take care of them as eﬀectively as the EU? Clearly, the answer must be: no, they would not. is refusal reﬂects the desire to ensure that decisions are taken as closely as possible to those aﬀected by them. It is a question of optimizing democratic self- determination, cognizant of the diﬀerences in responsiveness of the various levels of authority. e relative inﬂuence of ea citizen on what is ﬁnally decided diminishes as the number of participants increases. us, if democracy signiﬁes self-determination, the level of relative self- determination decreases with the increase in the size of the group – that is, with the level of political organization: local, regional, national, European. On the other hand, maers decided on a level of authority that is closer to the citizen may well engender profound external eﬀects, impacting citizens in other polities who did not have a say in the maer, su that a democratic asm gapes open, as Jürgen Neyer recently explained: Under conditions of interdependence, and in the absence of a supranational regulatory body, all democratic nation-states suﬀer from the structural problem that the policies of one nation impinge on the policies of others, with no country having the ability to systematically internalize these repercussions. (Neyer 2012: 4 et seq.)4 is relates to the complementary nature of the EU system of dual legitimacy as the ﬂip side of the principle of additionality. At the Union level, the structural democratic deﬁcit emerging at the member state level can be addressed and at least partially remedied as other constituencies gain a voice in the decision-making process in order to internalize pertinent negative externalities, su that the EU is best understood as a corrective meanism enhancing the democratic legitimacy of governance in Europe as a whole (Neyer 2012: 68–70). The democratic meaning of the principle of subsidiarity e principle of subsidiarity reﬂects this fundamental insight. In conjunction with the principle of proportionality in Article 5(4) TEU, it is not only a criterion for the legitimate use of competences conferred upon the Union (Article 5(3) TEU), but also the guiding principle of the aritecture of competences within the European Union (on the idea of subsidiarity in the constitutional context of the EU, see Pernice 1996d). is is what citizens as the authors of the Treaties (should) consider when deciding upon the conferral of powers to the Union. is principle has been included in the integration clauses of national constitutions, su as Article 23(1) of the German Basic Law; it corresponds to the principle stated in Article 1(2) TEU that ‘decisions are taken as closely as possible to the citizen’ (for a comprehensive and comparative legal analysis of integration clauses of both the member states and the EU, see Wendel 2011: 144 et seq., 525 et seq.). us, from the perspective of the citizens, the principle of subsidiarity can be understood as a general rule ensuring the highest possible degree of political self-determination in a multilevel political system (Barber 2005: 305–25). If, as the German Federal Constitutional Court (GFCC) recognized in the Lisbon Judgement, democratic self-determination is related to human dignity (GFCC 2009: para. 211), it therefore obtains a prestigious position among the founding values of the European Union enshrined in Article 2 TEU and in Article 1 of the EU Charter of Fundamental Rights (for a comparative legal analysis of the two conceptions of human dignity pursuant to Art. 1 of the Basic Law and Art. 1 of the EU Charter, see Swarz 2011). Sovereignty lost – or new powers gained? If the principle of subsidiarity is faithfully applied, the claim that member states or national parliaments have lost and continue to lose their powers is ill founded. Clearly, the ban on barriers to trade, the prohibition of discrimination on grounds of nationality, the requirement to respect the common principles and values of the Union and many other rules agreed upon in the Treaties restrain political options at the national level and even the constitutional autonomy of the member states. In turn, however, citizens gain freedoms and rights they never had before and that a member state could not grant individually. is beneﬁt could not be secured and the Union could not function properly without functioning national democratic institutions, administrative bodies and judiciaries based upon the rule of law implementing and ensuring the proper application of European law. All this means that national authorities are now subject to new constraints and loyalty obligations, and that they have – at least in part – altered their function (Hufeld 2011: 118, 121–3).5 is, however, does not necessarily take away powers from national institutions. It is the ﬂip side of the newly established possibility of actively participating in collective decision-making that reaes far beyond national borders, an increasingly important opportunity to eﬀectively extend rights and secure adequate living conditions to all European citizens in the age of globalization. A constitutional system designed for a multilevel political entity us, the proposal is to understand the European Union as a political entity that is not separate from the member states but instead comprises them, an organization that is composed of the member states and the supranational institutions. e constitution of the Union, consequently, does not allenge the national constitutions; rather, it is based upon and can be considered a complementary part of them. It is a sort of extension, adding new capacities for action of common interest to the beneﬁt of the citizens – even overcoming democratic deﬁcits at the national level. is is why citizenship in the Union can be said to be ‘additional to national citizenship’, as Articles 9 TEU and 20(1) TFEU emphasize. Clearly, this does not mean that Europe’s citizens are sizophrenic. Rather, to paraphrase Goethe, two souls are dwelling in their ests, as a second legal status has been added to the status that citizens of the member states already had (Habermas 2012: 28–36). Union citizenship reﬂects a speciﬁc belonging or constitutional relationship to the European Union and its institutions. It means ownership of and adherence to the Union, in the same sense that national citizenship is the expression of ownership of and adherence to the respective member state, the component and basis of the Union. Participation: the principle of voluntariness e preceding argument has shown that it would be diﬃcult to understand European integration as posing a real threat to sovereignty. In fact, the opposite is true: from the perspective of the citizens, it is an expression of their voluntary and sovereign decision that creates new opportunities for self-determination at the supranational level. e principle of voluntariness applies to both membership in the Union and the implementation of its legislation. Membership in the European Union No country or people is forced to accede to the Union, nor were any of the original member states forced to participate in this joint venture. Likewise, none of the existing member states is legally bound to stay. e new provisions of Article 50 TEU introduced by the Treaty of Lisbon make this voluntary nature explicit by stating the option for unilateral withdrawal from the Union – an option unknown to federal states. Politically, though, any withdrawal of a member state would be contrary to the idea of European integration and the common objective ‘of creating an ever closer union among the peoples of Europe’ (Art. 1(2) TEU). It is particularly diﬃcult to imagine Germany withdrawing from the Union,6 for historical and political reasons, in particular due to its existential interest in being embedded in a political union that oﬀers its citizens a hospitable environment, an enduring peace, economic and social welfare, and the opportunity to maintain inﬂuence at the global level. Implementation and the rule of law Membership in the European Union is thus a voluntary decision taken by the people of ea member state, beginning with a state’s accession and extending to its continued membership. However, the ‘principle of voluntariness’ has a broader meaning that gives the European Union a unique aracter, distinct from any other model of political organization. e Union is founded upon the binding force of law instead of physical coercion. ere is no European army, nor does the EU dispose of troops or deploy police forces to enforce obligations under the Treaties or secondary legislation. It is based on the rule of law only, as well as the common consensus that the Union serves the common interest of all its citizens best when the commonly established rules are observed. Union law is not imposed from the outside, but rather built into the national systems as it works ‘from the inside’ through the national authorities that enforce it. Voluntariness and disobedience Voluntariness includes the option of disobedience and exit. However, empirical evidence has shown that the system, based upon the rule of law, generally functions well. It is by conviction and the force of law, not by physical coercion or the threat of force, that member states (including their judges and administrative bodies) obey the law of the Union and give it preference even over national constitutional law. e worst cases of disobedience seen thus far arose in an area in whi the ‘Community Method’ (including the jurisdiction of the Court of Justice of the European Union, CJEU) does not apply: economic and ﬁscal policies. Under the Treaties, these policies still remain ‘Member States’ economic policies’ (Art. 119(1) of Treaty on the Functioning of the European Union [TFEU]). e damage done by breaes of ﬁscal discipline under the Treaty – in particular by France and Germany, unfortunately followed by others – cannot yet be measured. From a legal point of view, the present crisis must be aributed not to the absence of physical enforcement of the rules, but rather to an unrealistic trust in cooperation among states and the la of eﬀective meanisms for judicial decisions identifying breaes of law and requiring correction. Limits of primacy and the role of the courts in a pluralist system e rule of law and the aracteristic of the European Union as a union based upon the rule of law instead of physical force are oen what convince prospective member states and their citizens to join the EU as a civilized political entity. is implies limits to submission, as even obedience to Union law remains voluntary. In concreto, as national constitutional courts have already made clear, cases may arise where a Union measure clearly violates the national identity of a member state as described in Article 4(2) TEU (ECJ 2009), is evidently ultra vires (Article 5(2) TEU) or otherwise violates the substance of the fundamental rights of the individual (Article 6 TEU) to an extent that the values common to the Union and its member states (Article 2 TEU) are called into question, thereby threatening the very basis of the EU legal order. For a national court to deny the application of su a measure to the citizens of a member state is not in contradiction with the principles of primacy and direct eﬀect, as established by the ECJ, but rather the expression of a common responsibility typical of a non-hierarical, pluralist system su as the EU (Mayer and Wendel 2012: 105–27; Pernice 2014a; Walker 2002: 317–59). Safeguarding respect for these common values – in particular human dignity and the fundamental rights of the individual – is a shared responsibility of European and national authorities, speciﬁcally the CJEU and the national constitutional courts, for the beneﬁt of the citizens of the Union (Pernice 2006; Voßkuhle 2010a: 108; 2010b: 175–98). Mutual constitutional stabilization is respect and the shared responsibility of the courts at both levels ensures it can be understood as a condition for the citizens of ea of the member states to agree upon the common exercise of sovereign rights by Union institutions at a supranational level and to accept the binding force of their actions. Article 23(1) of the Basic Law clearly expresses this conditionality with regard to the operation of European Union institutions as a basic requirement for German participation. However, it reﬂects also the conditions for accession and continued membership to the Union. Article 2 TEU summarizes the common values, Article 49(1) TEU states that only a European state that respects the values referred to in Article 2 TEU ‘and is commied to promoting them’ may be accepted as a new member state, and Articles 7 TEU and 354 TFEU set up a procedure of supervision and sanctioning in cases of ‘serious and persistent brea by a Member State of the values referred to in Article 2’7 (see also Chapter 1). Respect for these values is not only a condition for the proper functioning of the Union;8 the corresponding provisions at national and European levels also play an important role in the protection of the rights of individuals. ey form a system of mutual constitutional stabilization (Pernice 1995: 225–64) established by the citizens of the member states with a view to ensuring the respect of their fundamental rights, in parallel with the European Convention of Human Rights, for all cases in whi a member state might fail to observe its duties towards the individual. Voluntariness and national sovereignty e principle of voluntariness is thus supported by a vested interest of the citizens, as the state may be enjoined to protect these rights and values in the case of a serious violation. Arguing that this represents a threat to national sovereignty would mean that the state is sovereign, not the people. e same holds for other provisions on powers conferred to the Union for purposes beyond the rea of national authority. ese provisions may subject national authority to rules and limits, but it would be a misconception of democratic sovereignty if su constraints resulting from the common exercise of sovereign rights by supranational institutions were understood as a limit on the self-determination of the citizens in ea of the member states. In fact, the opposite is true. Legitimacy: the principle of open democracy Democracy means collective self-government: those who are aﬀected by the actions of the public authority must have equal rights to participate in the process of determining its policies in order to accept them as legitimate. Practice has illustrated many ways in whi democracy can be organized, but one common denominator seems to be that the system ought to be self- referential, insofar as democracy seems to be equivalent to popular sovereignty (Grimm 2010: 35–41). Democratic legitimacy, or the recognition and acceptance of decisions by those aﬀected by them, thus depends on the perception that the decisions are in some way one’s own oice, a concept related to the Rousseauean ideal of self-authorship. As there are varying views and interests in ea society, the decisions taken by a majority are accepted as legitimate, but only when the competent institutions observe certain conditions, procedures and fundamental rights guaranteed in the constitution. The democratic deficit and special EU standards of democracy All this seems to be the case for the European Union as mu as it applies – ideally – in the member states. Nevertheless, there is a general complaint that the EU suﬀers from a democratic deﬁcit. People seem to feel that Brussels is ‘remote’ or unresponsive, that people do not have any inﬂuence on politics in Brussels and that nobody can be held accountable for the decisions taken there (Weiler 2013: 111, 116). Public opinion on the issue is still split among member states – there is no common language and basically no European-wide public sphere (Grimm 1995: 590). On the other hand, it is important to note that this political analysis is not reﬂected in the legal analysis of the German Federal Constitutional Court. e Court has accepted that the EU system of governance meets the requirements of democracy, at least those laid down in the German Basic Law. However, this view is based upon the assumption that democratic legitimacy for European policies ultimately relies upon the national parliaments, as the European Parliament is not considered suﬃciently democratic to assume this role; rather, it plays a supplementary part. e reason for this statement is the institution’s la of equality resulting from the principle of degressive proportionality (Article 14(2) TEU). e weight of a vote of a citizen of Malta or Cyprus counts 12 times as mu as that of a German citizen. As long as legitimacy can be considered to be derived from the national parliaments, however, and as long as the Union is not a federal state to whi the criteria traditionally applicable to states would apply, the GFCC does not see any reason to consider the EU to be undemocratic (GFCC 2009: paras. 263–72, 278–97; for the circularity of the argument, see Halberstam and Möllers 2009: 1241 et seq.). is judgement eﬀectively claims that the European Parliament would not be a democratic body capable of providing legitimacy to the Union’s policies. e GFCC has conﬁrmed its critical aitude in its judgement on the 5 per cent threshold for parties competing in federal German elections (GFCC 2011: para. 118). However, the Court’s denial of the parliamentary quality of this institution has been widely criticized (Sönberger 2009: 535– 58; ym 2009: 559–68). is is not the place for further comment on the jurisprudence regarding the speciﬁc democratic powers of the European Parliament (Neesheim 2010: 119). As correctly stated by the German Constitutional Court, the European Union is not a state (GFCC 2009: para. 277). Consequently, European legislation and policies may follow functionally equivalent democratic principles that correspond to its speciﬁc structure. Democracy within the member states would not be aﬀected, because even if the principles applied at the European level did not meet the standards for national policies, it is not possible to argue a democratic deﬁcit as long as the purposes of the decisions taken could not eﬀectively be aieved at the national level. As intimated above, this is guaranteed by the principle of subsidiarity. If the principle of subsidiarity is systematically applied, the maers that are decided at the European level are only those that cannot (or cannot eﬀectively) be dealt with at the member state level. If a maer is beyond the scope of national measures – and dealing with su issues is precisely what the EU was established to do – other rules for democratic legitimacy must be accepted if non-action is not the desired outcome. Enhancing democracy in the European Union e question is therefore how to organize the institutional framework and the decision-making processes at the Union level in order to meet the fundamental democratic requirement of

Multilevel Governance in Europe PDF

Document Details

Tags

Related

Summary

Full Transcript

Upgrade to continue