Journal of Public Policy

http://journals.cambridge.org/PUP Additional services for Journal Email alerts: Click here Subscriptions: Click here Commercial reprints: Click here Terms of use : Click here

of Public Policy:

The Governance of the European Union: A New Institutionalist Approach

Simon J. Bulmer
Journal of Public Policy / Volume 13 / Issue 04 / October 1993, pp 351 - 380 DOI: 10.1017/S0143814X0000115X, Published online: 28 November 2008

Link to this article: http://journals.cambridge.org/abstract_S0143814X0000115X How to cite this article: Simon J. Bulmer (1993). The Governance of the European Union: A New Institutionalist Approach. Journal of Public Policy, 13, pp 351-380 doi:10.1017/ S0143814X0000115X Request Permissions : Click here

Downloaded from http://journals.cambridge.org/PUP, IP address: 131.174.75.30 on 06 Nov 2013

Jnl Publ. Pol., 13, 4, 351-380

Copyright 1994 Cambridge University Press

The Governance of the European Union: A New Institutionalist Approach*

S I M O N J . B U L M E R , Government, University of Manchester ABSTRACT The analysis of European integration has tended to use a toolkit drawn from international relations. But since the revival of integration in the mid-1980s, the governance of the European Community and European Union has increasingly come to resemble that of a multi-tiered state. Accordingly, this article analyzes the governance of the European Union from a comparative public policy perspective. Using new or historical institutionalism, three levels are considered. In the first part, attention is focused on the EU's institutions and the available instruments of governance. The second part examines the analysis of governance at the policy-specific or sub-system level, and puts forward an approach based on governance regimes. The final part considers the institutional roots of the persistent, regulatory character of governance in the European Union.

How best to analyze the governance of the European Union? What are the characteristics of the policy process associated with the various activities of the European Union? Does its collective public policy have core attributes? Or are individual policies characterised more by the logic of the particular policy or issue rather than by a unifying dynamic of the European Union? These are questions which deserve answers and, in particular, answers which are informed by a methodological framework. The member states have assumed shared policy responsibility over a growing range of areas, something which was again highlighted by the
*The paper is based on research conducted within a project funded by the UK Economic and Social Research Council on regulatory institutions and practices in the single market. The award (no. Wi 13251014) was within the ESRC's Single European Market Initiative. An earlier version was presented at the European Community Studies Association conference, 27-29 May, 1993, Washington D.C. I acknowledge the assistance of Kenneth Armstrong, research officer on the project, in refining the ideas in the article. I am also grateful to John Peterson, Richard Rose and Stephen Wilks, together with Peter Vipond and other panel participants at ESCA for helpful comments. Responsibility for the content rests with the author.

352

Simon J. Bulmer

formal transfer of competencies associated with the Treaty on European Union (TEU). In consequence, collective activities are developing across the whole spectrum of public policy: from policing to foreign and security policy; from citizenship to immigration; from the internal market to the common commercial policy; from the transferability of social security entitlements to labour market regulation; and much else besides. Piecing together this jigsaw of activities, it is striking how the postwar era has seen the transformation of established political and institutional arrangements, with the emergence of a catalogue of collective activities of member states of the European Union (EU). Integration theorists have tried to dissect this transformation but have tended not to do so from the structuralist vantage point of the accumulation of 'state' power at the European level. To some extent this has occurred in the work of historians - most notably Alan Milward (1992) in studies of how the first transfers of policy responsibilities came about. How long will be it before a political scientist sets his or her research agenda as identifying the 'historical origins of modern institutional politics in [the EU], a politics distinguished by incoherence and fragmentation in governmental operations and by the absence of clear lines of authoritative control'? Substituting 'America' for 'the EU', this was part of the task Skowronek set himself in his study, Building a New American State (1982, p. viii). A book entitled Building a New European State would be bound to create controversy, for reference to the European Union as 'a state' does not yet seem to have become acceptable in the social sciences, and certainly not in political discourse amongst practitioners. Nevertheless, is this not what is effectively under construction, even if there is a reluctance to make this explicit? Or even if it is couched in the language of 'fusing' additional layers of authority onto the nation state, as in Wolfgang Wessels' (1992) 'fusion theory' explanation of the dynamics of integration. The point of departure of Skowrenek's study of the United States is illuminating with respect both to this fusion theory and to Milward's thesis of the supranational 'rescue' of the nation state. As Skowronek (1982, p. ix) puts it, '[SJhort of revolutionary change, state building is most basically an exercise in reconstructing an already established organization of state power'. European integration has been concerned centrally with such institutional reconstruction. And both the Single European Act and the Treaty on European Union are major steps in this process. If one seeks parallels from other literatures on nation or statebuilding, the postwar transformation of the European state can be located within the crisis-sequence school of political development, for instance the work of Tilly (1975). Such an exercise makes it striking

The Governance of the European Union how the collective activities of the member states have developed institutionally ahead of a popular, democratic groundswell of support for the process. At a superficial level the development of the European Community's 'state capacity1 is proceeding in advance of the structures for ensuring its democratic legitimation; a pattern which follows the earlier trend for European nation states. However one seeks to interpret this postwar transformation of European state power, the consequence has been that there is a highly developed governance structure; an absence of a clearly fixed state identity to which it is attached, not least because of the possibilities of enlargement of the EU; and a disjuncture between the integration process on the one hand and popular involvement in it on the other. The tension between these three features of the current stage of integration is highlighted particularly by two cases. Firstly, in the context of Economic and Monetary Union (EMU), is it possible to have a currency without a state? And, secondly, is it possible to have a European policing function without a state (den Boer and Walker, 1993)? This paper does not aspire to examine the grand issues of macro-social change associated with the development of governance, state identity and democratic legitimacy. However, it aims to analyze three central issues of the governance of the EU. Firstly, it seeks to locate the EU institutions in a comparative politics context, specifically that of new (or historical) institutionalism. This context, it is argued, allows us to go beyond the detailed configurative studies of the EU institutions that already exist and place them in a context which allows differentiation between formal political institutions, informal conventions and the norms and beliefs embedded within those institutions. This institutionalist focus helps situate analysis of the EU within a central analytical debate prominent in North America, Germany and to a lesser extent in the UK. 1 It is also argued that this approach may help understand the transformation of EU governance, brought about by the Single European Act (SEA) and, subsequently, by the Treaty on European Union. Analyzing governance is not just a matter of the macro or transsectoral level. Rather, it is necessary also to take into account the policyor issue-specific level. Writers such as Heclo and Wildavsky (1974), with their analysis of policy communities in British politics, alerted us to this issue already in the 1970s. It is no less relevant to the governance of the EU. In the intervening period the policy community-policy network literature has developed greatly, especially in the work of British political scientists (for example, Rhodes, 1986; Wright, 1988). In this paper sub-system governance within the EU is analyzed in terms of "governance regimes" this term being adopted to retain a consistent institutionalist focus across the different levels of analysis.

353

354

Simon J. Buhner

T A B L E I : Institutions and Governance in the European Union

Level of Analysis Issue Illuminated

Trans-sectoral Policy Programme Institutional Configuration

Institutions and Instruments of Governance in the EU Discrete Governance Regimes EC/EU as Regulatory State

The paper's third objective is to pick out some of the distinguishing characteristics of the governance of the EU. In some respects, this is an exercise somewhat akin to that undertaken by Richardson (1982) and others, namely the identification of distinctive national 'policy styles'. However, here too we seek to relate the distinctive characteristics of the governance of the EU to institutional characteristics. Accordingly, this paper is divided into three parts; their interrelationship is set out schematically in Table 1. After a brief section contextualising the study of the governance of the EU, the paper first looks at the trans-sectoral level of governance, as seen from a new institutionalist perspective in order to dissect the governance structures which stand between basic political forces and policy outcomes. The second section seeks to reassemble matters at the subsystem level with reference to governance regimes. The third section goes one step further in re-assembling Community governance, arguing that a common thread running through Community governance is its regulatory pattern. Thus, the range of instruments available to the EU - economic, political and legal - results in a peculiarly regulatory mode of governance. And this pattern can be detected across a rang of governance regimes.
Analyzing the Collective Governance of the Member States

One of the key features of the study of the European Community hitherto has been the predominance of analytical devices rooted in international relations theory. This has been the case with much of integration theory, particularly functionalism and neo-functionalism, but is also clear in the neo-realist influences in the work of those seeking to explain integrative advances as the product of inter-state bargains, for instance the 'intergovernmental institutionalism' of Moravcsik (1991). Similarly, the literature on policy-making in the EC was centred around neofunctionalism, intergovernmentalism and interdependence: all drawn from international relations frameworks (Wallace, Wallace and Webb,
1983)-

Observation of the day-to-day activities of the European Community suggests that what is going on is less some mutation of an international

The Governance of the European Union organisation and more a multi-tiered system of government. In what other international organisations do we find the involvement of over 500 transnational interest groups, interaction with local government lobbyists, an elected parliament seeking to play its part in interinstitutional decision-making, and a law-based mode of regulating collective governance? Yet this is the substance of the activities of the member states, particularly in the EC 'pillar' of the EU. Economists seem to have come to terms with the core issues of governance, for instance in the attention devoted to distributive issues in Padoa-Schioppa's (1987) study of the economic system of the EC. Similarly, legal scholars have given attention to the legal aspects of governance with their discussions of the 'constitutionalisation' of the EC's legal system. Political scientists, by contrast, seem to a striking degree to have continued working within international relations paradigms. To be sure, there have been ad hoc applications of public policy methodologies to individual case studies (see Holland, 1987; and Peterson, 1991). Even analysis drawn from federal systems of government has had a limited impact thus far on studying the governance of the European Community (see, for instance, Scharpf, 1988; Sbragia, 1992). The approach advocated here is to look at the EC and other collective decision-making arenas of the twelve member states as 'governance structures' familiar in the study of international relations and of national politics. This starting point may offset the 'intellectual apartheid' resulting from divisions within the discipline of political science. In particular, there is something of a continuum between those applying the new institutionalist literature in comparative politics and analysts of the international management of policy issues. An example of this work would be that by Oran Young (1989) on international institutional arrangements for environmental matters. The examination of governance structures places a large amount of emphasis upon institutions, decision-making procedures and rules. In attempting to undertake a pathology of the collective governance of the EU, a core concern is with mapping the institutions, decision-making procedures, rules and norms embedded in these across policy areas. An underlying assumption is that institutions matter: that political struggles are mediated by prevailing institutional arrangements. Hence, this approach follows the broad lines of the 'new institutionalist' literature in political science, viewing institutions as extending beyond the formal organs of government to include standard operating procedures, socalled soft law, norms and conventions of behaviour. All these institutional features may shape the pattern of political behaviour. However, it is recognised that institutions do not generate political behaviour of their own accord; they should not be seen as the determinants of policy.

355

356

Simon J. Bulmer

This 'new institutionalist' approach has been associated particularly with the writings of March and Olsen (1989), but the parameters are quite broad. To take one example, the degree of inclusiveness of the term 'institutions' also varies in the literature (see Hall, 1986, for a particularly inclusive usage). An important distinction within new institutionalism is that between rational choice institutionalists and historical institutionalists. The former group sees institutions imposing constraints upon the rational actions of individuals. By contrast, the latter attributes a larger degree of importance to institutions. The distinction is summarised by Thelen and Steinmo (1992, p. 9) thus: By shaping not just actors' strategies (as in rational choice), but their goals as well, and by mediating their relations of cooperation and conflict, institutions structure political situations and leave their own imprint on political outcomes. The analysis in this paper reflects the historical institutionalist approach. Institutionalist analysis can examine different levels of institutional activity. One recent study identified three different potential tiers at which institutions could affect policy outcomes and governance capacity (Weaver and Rockman, 1993). The first tier compared presidential with parliamentary systems of government; the second entailed comparison within these categories; a third tier looked at the effects of specific institutional features rather than systemic ones, for example the effects of federalism or judicial review. This paper is also concerned with three levels of analysis. The first is trans-sectoral, the overall functioning of the institutions, and how this may be affected systematically by formal constitutional reform, such as through the SEA or the TEU or by informal constitutional conventions, such as the clarifying statements issued on subsidiary since the signing of the TEU or, in an earlier period, the Luxembourg Accords of 1966. It was as a result of the Luxembourg Accords that the European Economic Community did not take up the treaty provision for qualified majority voting at that time but, at French insistence persisted with consensual decision-making. The Accords had a major consequent impact on governance capacity. The second is policy- or issue-specific. The pattern of governance differs between issue areas. This finding was one of the principal conclusions of the UK Economic and Social Research Council's Government-Industry Relations research initiative (see, for instance, Wilks and Wright, 1987). It was argued that earlier comparative studies - for instance, Katzenstein (1978) - had identified different national patterns of governance capacity and policy where, in fact, considerable diversity existed within states. As regards European inte-

The Governance of the European Union

357

gration the existence of different patterns of sub-system governance is well established in academic analyses (Wallace, Wallace and Webb 1983). This paper's emphasis on 'governance regimes' seeks to link together developments at the trans-sectoral, systemic and sub-systemic levels. The third level of institutional analysis addresses the EU's distinctive institutional configuration. Here the concern is with the features that set the European Union apart in a comparative analysis of political systems. Each of these levels of institutionalist analysis is now taken in turn, with a view to establishing the possible insights which can be afforded. Throughout this empirical section, we seek to ensure that the institutional approach is applicable to all three pillars of the EU: - the first pillar, the Founding Treaties of the European Community as amended; the second pillar, the Common Foreign and Security Policy; and the third pillar, Justice and Home Affairs cooperation. The Maastricht Treaty shows that it is no longer sustainable analytically to set aside the non-supranational joint activities of the member states, such as foreign policy cooperation; any viable framework must be applicable to all three pillars of the European Union.

New Institutionalism and the EU: A Trans-Sectoral View

The basic assumption of historical institutionalism is that institutions cannot be regarded as mere arenas within which political action is played out. As March and Olsen put it (1989, p. 17): Without denying the importance of both the social context of politics and the motives of individual actors . . . institutional analysis posits a more independent role for political institutions. The state is not only affected by society but also affects it. Thus formal organisations, bureaucracies, legal systems, together with their patterns of work come to ' . . . provide order and influence change in politics' (March and Olsen, 1989, p. 16). Historical institutionalism is not just concerned with institutions and institutional devices, for embedden within these are ' . . . beliefs, paradigms, codes, cultures and knowledge' (March and Olsen, 1989, p. 26). Taking this skeletal presentation of the historical institutionalism, what light can be thrown upon the collective governance of the member states? If one takes the broad range of Community governance, the basic institutional elements comprise the components set out in Figure 1.

358

Simon J. Bulmer

Figure 1: An institutional pathology of Community governance The Institutions of Governance A. 1. 2. 3. 4. 5. B. 1. 2. 3. 4. 5. 6. 7. The institutions of governance The supranational/intergovernmental institutions Inter-institutional relations Internal institutional organisations Internal institutional procedures Institutional norms The instruments of governance Treaties (supranational or intergovernmental) Constituent agreements International law Secondary EC legislation and decisions ECJ jurisprudence 'Soft' EC or international law Political agreements

The first issue to be addressed is: what are the institutions involved in the governance of the EU? This is not evident from a reading of the treaties. Firstly, not all activities may fall within competences set down in the treaties. This situation applied to foreign policy cooperation before its codification in the Single European Act, or to cooperation on home affairs matters before the TEU. As a result, political rather than constitutional-legal arrangements had to be examined. Even with the codification of Justice and Home Affairs (JHA) cooperation in the TEU, there is little detail provided in that treaty regarding JHA institutions. This economy of detail brings us to the second problem of where governments have deliberately sought to pull a veil over their activities - as in JHA - but it also happens in the well established areas of the EC pillar. Informal meetings of ministers and the European Council before its codification in the SEA are examples of institutional arrangements whose importance could not be adduced from the treaties. The principal legislative institutions of the European Community are well-known.2 Thus the Commission, the European Parliament, the Council of Ministers (including COREPER), the Economic and Social Committee (Ecosoc) are involved in legislation primarily through the formal decision-making process. The European Council's involvement is driven by rather different dynamics, such as developing a more strategic approach to legislation or, at the other extreme, trying to solve specific political problems of principle. The holding of informal minsterial meetings appear to have started with the so-called Gymnich meetings, first

The Governance of the European Union

359

held in the foreign policy cooperation (EPC) arena in 1974. However, such meetings have been held increasingly within the EC pillar. For instance, agriculture ministers have hosted meetings to reflect on policy; these usually are held in the country holding the presidency - and even on the minister's farm, hardly a policy neutral location! At the policy management level similar informal practices may be observed. One example has been the routine management of the European Monetary System by central bankers meeting in Basle under the auspices of the Bank of International Settlements. The European Court ofJustice's (ECJ) activities are confined to those policy areas which are governed by supranational treaty. Moreover, its role in governance is not as an agency having a specific portfolio for a particular policy regime, as does a European Parliament (EP) committee, a Commission directorate-general or a 'technical' Council. Whilst it is not part of a policy network, it can have a major impact on a network's activities. For instance, its jurisprudence might grant the EC institutions a form of regulatory authority which they felt was lacking. This was the case with the Continental Can and Philip Morris judgements which gave the Commission regulatory powers - albeit flawed ones in respect of the control of mergers and acquisitions. The cases were
respectively: Europemballage Corporation and Continental Can Co. Inc. v. Commission, Case 6/72; and British American Tobacco Ltd. v. Commission and

R. J. Reynolds Inc. v. Commission, joined cases 142/84 and 156/84. The ECJ's judgements in the Dassonville and Cassis de Dijon cases were crucial in shaping important aspects of the regulatory arrangements at the heart of the single market programme: respectively Procureur du Roi v. Dassonville, case 8/74; and Rewe-Zentrale v. Bundesmonopolverwaltung fur Brannt-

wein, case 120/78. Whether the court's rulings are fundamental, as in these cases, or relate to the clarification of more detailed points of legislation, its cumulative impact upon the governance of the EC is of great importance. However, outside the EC pillar of EU activity it is virtually powerless to act. Inter-institutional relations are important, for they define the power balance between the bodies concerned. Thus, the power of the European Parliament (EP) is enhanced significantly if the area of governance falls within the ambit of the cooperation procedure rather than within the very limited consultation procedure. With implementation of the Treaty on European Union, its power is enhanced still further in areas governed by the co-decision procedure for, under specified conditions, the EP is able to veto legislation. Thus, the consultation, cooperation and codecision procedures each give a different profile to the EP in the overall institutional balance (Jacobs et ai, 1992). This is why the legal services of the institutions, and especially of the EP, are so exercised by argu-

360

Simon J. Bulmer

ments about the legal base of much EC legislation, for this apparently technical argument defines inter-institutional relations and the means by which interests are articulated. Whilst the EP is the best and most obvious example of an institution whose institutional power is procedurally defined, the situation is by no means exclusive to it. Both the cooperation procedure and the codecision procedure have important implications for the position of the Council of Ministers. For instance, although at a formal level there appears to be a move towards increased majority voting, there is an inter-institutional logic which goes in the opposite direction. Where the cooperation procedure applies, the Parliament has found that it may be able to capitalise upon any division within the Council. Jacobs et al., 1992, p. 186) highlight the case of exhaust emission standards, where the EP was able to influence policy significantly by calculations of interinstitutional interest. However, this has alerted the Council to the fact that it may be advisable to reach decisions consensually to avoid the EP increasing its influence in this way. As for the Commission, its role in inter-institutional relations varies from one area of governance to another. Where the cooperation procedure applies, it may have to act as mediator not only between member governments but also between the EP and the Council. At the other extreme, its role may be relegated to that of observer, for example in JHA cooperation. Like the EP, it too is concerned with constitutional details because these might strengthen its political position. For example, its creativity in basing its proposed directive on the protection of pregnant women at the workplace upon the EEC Treaty's health and safety provisions (Article n 8a) placed it (and the EP) in a stronger institutional position in the negotiations that followed (cf. Cram, 1993), and helped circumvent British opposition to legislation. The ECJ's importance to inter-institutional relations is rather more confined since it is not a participant in the legislative process. Nevertheless, it occasionally has an impact, most strikingly when it issues a judgement concerning inter-institutional obligations. One example of this was in the 1980 Isoglucose judgement, where it annulled a piece of EC legislation on the basis that the EP had not given its opinion on it. This judgement was interpreted as giving some teeth to the EP's involvement in legislation via the consultation procedure (Jacobs et al., '992> PP- 180-182). The distinction between institutional organisation and institutional procedures is rather difficult to identify but the former is perceived here as essentially static in nature, whereas the latter is more dynamic. Indeed, institutional procedures are often devised in order to maximise the effectiveness of the institution's organisation.

The Governance of the European Union

361

The simplest way to illustrate the distinction is by taking a specific example, namely the case of the European Commission. The predominantly vertical nature of the Commission's chain of command is often referred to, and it is suggested that this may be to the detriment of policy coordination across directorates-general (DGs) (see, for instance, Nugent, 1991, pp. 69-72). Thus Commission administrators may see their loyalty as being to 'their' director-general or 'their' commissioner rather than to a more collegiate notion of the Commission. As the vast majority of the 23 DGs are organised by policy portfolio, this has a fragmenting effect for the coordination of Community policy as a whole. Where coordination is more systematically organised, is at the political level, specifically in the College of Commissioners; the Secretariat General also plays a limited coordinating role. As decision-making takes place, institutional procedures come into play. There is a strong reliance in all formal organisations on standard operating procedures. The Commission is no exception to this. In consequence, decision-making tends to start at the middle-ranking level of officials and then be passed upwards through superiors to the Commissioner and his/her cabinet. On the way up, inter-service consultation, i.e. the coordination between staff in different DGs, is sometimes ad hoc in nature rather than systematic and continuing. Moreover, there is the danger that by the time the proposal comes to the formalised collective decisional agency, namely the College of Commissioners, and even though the College may have considerable discretion over the action it can take, thinking will already have been conditioned in a fragmented way within the DGs. Thus the procedural dimension of the Commission's workings may reinforce the organisational tendencies to fragmentation. A similar distinction between organisational and procedural dimensions may be found in the workings of the Council of Ministers. Here, the assignment of legislation to a particular 'technical' Council effectively assigns it to a whole hierarchy. Thus air transport liberalisation was assigned to the Council of Transport Ministers, with the detailed work being conducted in the Transport Questions Working Party (Air Transport), one of the many functioning under the auspices of the Committee of Permanent Representatives, COREPER (on working parties generally, see Nugent, 1991, p. n o ) . Hence, organisationally, the Council may introduce some bias into the process of governance, for other national ministries with an interest in this subject matter have to channel their views indirectly through transport policy mechanisms. This is where the procedural dimension comes into play. Thus the British Department of Trade and Industry, with its interests in the competition aspects of air transport, would have to coordinate its views

362

Simon J. Bulmer

with the Department of Transport within Whitehall. However, whilst such coordination might be assured within the United Kingdom through the role of the Cabinet Office, we know that not all member states have the same level of coordination. In particular, power is divided not only between portfolios, perhaps reinforced by party 'fiefdoms' in a coalition government, but also between different levels of government, as in Germany's federal system (see O Nuallain, 1985, for a comparative picture). Thus, the procedures of member states may exacerbate the isolation within the EC Council of Ministers of the transport policy hierarchy. Moreover, the coordination procedures within the Council itself are known to be deficient, particularly at the ministerial level. The General Council has found itself too overwhelmed with work in its guise as the 'technical' Council responsible for foreign policy and external relations. Similarly the workload of COREPER hampers coordination there. This fragmentation, after all, is how the Council of Agriculture Ministers was able to act for so long with scant respect for the budgetary policy that was being developed by national cabinet colleagues in the Council of Budget Ministers. Reconciling matters was either neglected or left to the European Council under circumstances of impending budgetary crisis. Only with the 1988 budgetary reform measures was an interCouncil reconciliation procedure created to bring both sets of ministers together where their policies were manifestly at odds. A final procedural dimension is that played by the country holding the presidency of the Council. This role, which is given no real importance in the EC treaties, has become an important device for securing agreement on legislation, for instance by means of constructing package deals. Operating at all levels of the COREPER/Council/European Council hierarchy, it is an important procedural device aimed at ensuring that these organisations function effectively. There is scope for explaining other parts of the institutional organisation and procedure in this way, for instance, the range of committees utilised in policy management at the EC level, the 'comitology' issue (Bradley, 1992). Within the Common Foreign and Security Policy (CFSP) pillar of the EU a quite extensive division of labour has also been developed (Nuttall, 1992). Even in its title, the Justice and Home Affairs pillar suggests a fragmentation of responsibilities between interior and justice ministries. As indicated above, March and Olsen's work alerts us to institutional norms. These norms can come into play at all institutional levels. Three examples illustrate this. A first relates to the highest political level, namely the changed climate throughout the EC institutions following agreement on, and ratification of, the Single European Act (SEA). The single market deadline of the end of 1992 had an important effect in

The Governance of the European Union

363

mobilising the institutions beyond the legislation in the White Paper. For example, it was a contributory factor - along with the changed corporate context and ECJ judgements - to the Council's agreement on the EC Merger Control Regulation in 1989. This legislation had first been tabled in 1973 but effectively shelved by the end of the decade. In the context of the single market deadline, however, it became clear that resistance by the Council (or, more accurately, by some of the member governments) was no longer feasible, given the normative context within which the EC was functioning. Thus, although it was not in the White Paper, the need for merger control legislation was recognised by all institutions as indispensable, given the agreement on completion of the single market. In consequence, the normative change was also perceived within a much narrower context too. A second example relates to norms embedded at the institutional level; the clearest example is the Commission. This institution has to take on the role of'conscience of the European Community'. Its actions must reflect the collective interests of the Twelve. This is a strong motivating value in the Commission's work. Nonetheless, it is clear that it can advance this conscience more boldly when supranational integration is in the ascendant. Thus, there is an interaction between two 'levels' of norms - this helps explain how the Commission, under the presidency of Jacques Delors, was able to play a more pro-active role in the EC in the post-SEA period (see Ludlow, 1991). It should also be mentioned that the Court of Justice has been held to have taken a broadly expansionary interpretation of the treaties and of EC law (Rasmussen, 1986). This situation has come about as part of the development of a doctrine which Weiler sees as 'self-referential, legal, internal to a possible logic of the Treaty itself (1993, p. 419)However, in the same article Weiler argues that in the future the ECJ will have to take account of a changed political environment. These developments in the doctrine of the ECJ have a strong normative component. The third example concerns the norms on display at the intrainstitutional level. Again, the Commission may serve as an example. Different DGs have different implicit mission statements. These may reinforce the elements of organisational and procedural fragmentation that have already been referred to. DG IV (competition) is a case in point. Air transport liberalisation within the EC falls under the responsibility of DG VII (transport) but also necessitates interaction with DG IV. This is because block exemptions from the competition rules have to be negotiated or renewed for a range of activities, so as to permit interairline discussions on route capacity or to facilitate joint ownership of computer reservation systems. DG VII is regarded as taking a rounded

364

Simon J. Bulmer

view of the EC air transport sector, including its relatively fragmented size in the face of American mega-carriers. Thus, it has a certain industry sensitivity. By contrast, DG IV is more concerned with competitionbased comparators. There is also an underlying view that non-exposure to competition within the EC market-place undermines international competitiveness. Not surprisingly, this results in a different perspective on what regulatory arrangements are desirable for EC air transport. An interviewee at the European flag-carrying airlines' interest group, the Association of European Airlines, highlighted this in dubbing DG IV as the 'ayatollahs of competition'. Similar tensions may arise along these lines in other areas of the Commission's work. For instance, there can be tensions between DG III (Internal Market and Industrial Affairs) and DG IV in reviewing concentrations under the Merger Control Regulation. DG III may take a more favourable view of larger-scale enterprises on the grounds that these may facilitate increased European industrial competitiveness in the global context. DG IV, by contrast, is more concerned with the effects upon competition. This tension was highlighted most graphically in the 1991 consideration of the Aerospatiale/Alenia bid to buy De Havilland Canada from Boeing. By the narrowest of majorities within the College of Commissioners, Sir Leon Brittan, the Commissioner for competition, gained support for his proposed decision to block the acquisition, against the wishes of his counterpart presiding over DG III, Martin Bangemann. Institutional roles develop that combine norms with obligations that are rooted in EC legislation and the treaties. There are elements of this in the Council too, where agriculture ministers may perceive their role as defending the interests of farmers rather than those of consumers or taxpayers. But the picture in the Council is more complex still. Its organisation is not just by hierarchies of portfolios but also by member state. Thus, British ministers may consistently present a more procompetition line in a range of Councils than, say, their French or Italian counterparts. This aspect, too, has a normative component.
The Instruments of Governance

The most fundamental basis of governance is by means of treaties, especially the EC treaties themselves (i.e. the founding treaties, the Single European Act and other amending treaties). Their importance is threefold. Firstly, the treaties set out the decision-making arrangements and formal institutions. However, the European Council was not set down in any treaty until it had been in regular operation for a decade or so.

The Governance of the European Union

365

Even within the context of EC activity, not all institutions need be set out in the treaties. Of course, this assumes that one defines the European Council as an institution in the period prior to 1987. That is the position taken here, as it is consistent with the new institutionalist framework. There is very little that has changed, as far as the organisation and functioning of the European Council is concerned, in the period since 1987. Secondly, the treaties set out rights and duties for individuals. This aspect of the treaties, where nation states would normally set out a bill of rights, is not especially developed therein because many national governments would jib at such overt recognition of the EU's state-like rank. Nevertheless, the TEU's provisions on citizenship could be seen as part of the development of this aspect. Thirdly, the treaties set out policy responsibilities. The Treaty of Paris sets out these responsibilities in considerable detail in the coal and steel sectors, whereas the EEC Treaty is more concerned with setting out broad policy principles. The supranational treaties have a significant influence upon the nature of governance in individual areas of activity. For instance, they may determine the level of competence possessed by the EC. Thus, explicit reference to a policy area in a treaty will give greater authority to the Commission in proposing legislation. By contrast, resort to Article 235 of the Treaty of Rome, which gives a basis for proposing action beyond what is set out in the EEC Treaty if this is necessary as a means to achieving Community goals, puts the Commission in a weaker role in proposing legislation. We have also noted that the treaty base can have a major impact on inter-institutional relations and thus on the strategies of other policy actors. Similar broad functions to the three identified above may also apply to the intergovernmental pillars. However, the character of the law thereby created is of a quite different nature from that deriving from the supranational treaties. Politically, there is a clear interaction between the three pillars but this is hardly the case in terms of constitutional law. Finally, there is the possibility of geographically wider treaties with an impact upon governance. These could include treaties between the EC and third parties, i.e. where the EC is an international legal actor. They could also include international treaties or Council of Europe conventions which shape the collective governance of the Twelve in some particular issue-area, particularly in the JHA pillar. The second instrument of governance concerns constituent agreements. These lack the formal status of treaties and indeed it is usually precisely because of their non-binding character that they come into being. Nevertheless, they may result in the member states undertaking new activities, developing new institutional roles and so on. Within the EC context we

366

Simon J. Bulmer

could locate the 1983 Solemn Declaration on European Union in this category or the various reports which set down the operation of foreign policy cooperation until it was formalised in the SEA. A third instrument of governance is international law. This largely relates to areas outside the supranational treaties but not exclusively so. International law has a limited role to play within the domain of the supranational treaties when the ECJ uses principles of international law in arriving at its judgements on cases it is examining. Further, where EC legislation is not compatible with fundamental human rights, the ECJ will regard it as invalid (Steiner, 1990, pp. 45-47). The EC itself may become subject to international law where it becomes party to a treaty in its own right. In addition to such cases, law comes into play when the EU interacts with activities of the Council of Europe and its conventions, or other such organisations. Such interaction could apply to all three pillars of the EU. With the fourth instrument of governance we encounter the more detailed territory of EC legislative acts. They chiefly comprise the secondary legislation adopted by the EC institutions, as well as the consequential implementing legislation by member states. The form of EC legislation is well known to students of integration: regulations and directives and their equivalents under the terminology of the Treaty of Paris, (see, for instance, Nugent, 1991, pp. 168-171). The decision to employ a Regulation rather than a Directive, or vice versa, may be related to the mode of governance prevailing in the particular policy situation. Thus, in competition policy the preferred policy instrument is the Regulation. This reflects a range of factors: the fact that the principles of the policy are set down in the EEC Treaty, so only their implementation is required; the fact that the Commission's competence is strong in this policy area; the need to have uniform application of the law; the fact that the law is addressed to private economic actors rather than to the member states; and so on. Directives, by contrast, introduce some limited discretion as to how the EC's wishes are administered and operationalised in national courts. Thus, the choice between using a Directive or a Regulation is likely to be closely linked to the prevailing form of governance for the particular subject matter. A final form of EC legislative act is the Decision. Typically, this is the way in which the EC institutions - principally the Commission implement their own legislation. To take the competition policy area again, it is by means of legally binding Decisions that the Commission may find companies guilty of transgressions of competition law and, where appropriate, impose a fine. Generally, Decisions are employed by the Commission where it enjoys a regulatory role, i.e. where it has some discretionary power of its own.

The Governance of the European Union

367

A fifth instrument of governance is the jurisprudence of the ECJ. Court decisions can be seen as having different kinds of effect upon governance, but the unifying feature, as with EC legislative acts, is that they are confined to those areas subject to the supranational treaties. The first kind of effect is broad in nature: where an ECJ judgement has a broad impact. The kinds of decision referred to here are those of a transectoral nature that clarify legal principles across a range of policy areas. Examples would include the landmark decisions on direct effect, supremacy, proportionality or the ERTA principle: 'once the Community, in implementing a common policy, lays down common rules, member States no longer have the right, individually or collectively, to contract obligations towards non-member States affecting these common rules' (Steiner, 1990, p. 37). This derives from case 22/70
Commission v. Council (Re European Road Transport Agreement). A final

example would be individual entitlement to damages where a Directive is not properly implemented, as established in the Francovich case: C-6/
90 and C-90-90, Andrea Francovich and Daniela Bonifaci v. Italian Republic.

The second kind of decision is quite restricted in effect, i.e. the decision simply clarifies a point of law, what the treaty intended, whether an institution has acted within its powers, whether a member state is in breach of its obligations, or something similar, with an impact confined to a particular policy area. Sometimes such judgements have a rhetorical effect upon their immediate legal circumstances. Examples of this might be the Cassis de Dijon or Dassonville cases. In narrow terms these simply resolved specific legal disputes. However, their impact was not confined to the specifics of the two cases. Rather they established principles which the Commission was able to employ in seeking agreement on its new approach to standards and mutual recognition. Similarly, in the case of merger control, the Philip Morris judgement made a specific ruling on a case affecting an agreement between enterprises in the tobacco industry. However, in effect it also gave the Commission regulatory powers. The Commission's threats to use these helped to get the member governments to legislate a more clearly defined Merger Control Regulation that offered greater certainty to the corporate actors. Whilst EC legislative acts are the principal means of purposeful transposition into practice of the principles in the EC treaty, supported by the jurisprudence of the ECJ, there may remain areas of discretion that the negotiating parties wish to clarify. These areas of discretion are often clarified by reference to instruments of soft law. Soft law is somewhat elusive and not easy to delineate from hard law, on the one hand, and purely political agreements, on the other (see Wellens and Borchardt, 1989). Soft law, in simple terms, relates to rules of conduct that are not legally enforceable but nonetheless have a legal scope in that they guide

368

Simon J. Bulmer

the conduct of the institutions, the member states, individuals and undertakings (Wellens and Borchardt, 1989, p. 285). Examples of soft law include Commission practice notes and memoranda, declarations attached to the treaties, and statements recorded in the minutes of the Council of Ministers but not normally published (Nicoll, 1993). Two specific examples of the impact of soft law can be shown in the case of merger control. Both relate to how the legislation agreed in 1989 has been put into practice. Firstly, the principal means by which the Commission has regulated mergers has been through the attachment of conditions, such as the disposal of a division of a company involved in a merger, to its approval of a concentration. Typically, this would occur if a merger would lead to a perceived problem of market dominance in one particular area of activity. Such conditions have legal effect, for they are set down in a Commission Decision. Yet there is no provision in the Merger Control Regulation (MCR) for making this kind of conditional approval. The origin of the practice lies in fact in a 'Commission notice regarding restrictions ancillary to concentrations' (Official Journal, 1990, C2O3/5ff.). This is a clear case of soft law in operation. Secondly, the minutes of the December 1989 meeting of the Council, at which the MCR was agreed, have become public (see Cook and Kerse, iggr, pp. 226-230). They are quite revealing as to the function that statements in the minutes can play, ranging from individual member states/the Commission indicating their preferences when the MCR is reviewed through a statement that the Commission does not normally intend to apply its powers under Articles 85 and 86 to concentrations (i.e. relying exclusively on the MCR instead) to interpretative statements as to how the MCR shall be implemented. The statements have had an important impact upon the governance of this particular policy area. The concept of soft law was developed in public international law before being applied to the specific circumstances of EC law. It thus follows that it may come into play in policy areas covered by the two intergovernmental pillars of the EU. Thus, codes of conduct and similar instruments may come into play in policy areas governed by the intergovernmental pillars of the EU, or where other international treaties are employed, such as conventions of the Council of Europe. Indeed, soft law is perhaps the principal means of operationalising policy principles outside the domain of the supranational treaties. Political agreements are not easily distinguished from soft law. The inventory of EC soft law, as drawn up by Wellens and Borchardt (1989), is so extensive as to encompass political agreements. Where the exact divide is drawn, therefore, may be somewhat arbitrary. Here, political agreements are seen as trans-sectoral in nature and set out, in a manner analagous to soft law, as clarification of political conventions. Specific

The Governance of the European Union

369

examples would include the 1966 Luxembourg Accords or, more recently, the Edinburgh European Council's declaration on subsidiarity. Drawn up to provide further clarification of the formal treaty provision on the subject (in the TEU), the provisions set out a number of tests, such as whether European action is desirable and whether it is proportionate to policy goals. This detailed presentation of the institutional components of government aimed not only to illustrate some of the insights that can be obtained from working within a historical institutionalist framework. It was also designed to illustrate the range of institutions and policy instruments that can be employed in the context of Community governance, both within and outside a supranational treaty framework. However, dissection of this kind cannot offer a full explanation of the governance of the European Union. Firstly, it fails to present a dynamic picture of governance. Secondly, it omits detailed analysis of the role of political forces, namely the role of interest groups, national interests arising from domestic politics and so on. It is possible to hypothesise, along the lines of historical institutionalism, that strategies and goals will be influenced by the institutional arrangements which political forces are confronted with. To test this hypothesis, however, it is necessary to undertake case studies, something which cannot be attempted within the confines of this paper. Obtaining a dynamic picture of governance is also something best achieved within a case study context. However, there is a second aspect to institutional dynamics that not only helps link together this transsectoral view of EU governance with case studies conducted at the governance regime level but also is a potential addition to the historical institutionalist literature. It has been regarded as more effective at explaining: the 'stickiness' of institutional arrangements; the constraints placed upon radical policy change; and the distinctive institutional and policy circumstances across states (see, for instance, Katzenstein, 1978; Hall, 1986; Ikenberry, 1988; and the discussion in Thelen and Steinmo, What is striking about European integration is the dynamic nature of governance since the mid-1980s. To be sure, supporters of a federal Europe have bemoaned some of the reform measures as limited in nature but, in international comparison, there is a striking degree of institutional dynamism. Krasner's (1984) characterization of institutional change, as 'punctuated equilibrium' suggests that, at times of crisis, institutional explanations of politics are suspended. However, it would be very difficult to sustain the argument that the EC and its member states were in crisis when the SEA and the TEU were negotiated. There were, of course, major economic and political challenges but the condi-

370

Simon J. Bulmer

tion of crisis would suggest that the negotiation of EC treaty reforms occurred in circumstances other than the fairly widespread public apathy or ignorance that prevailed at the time. Of course, the ratification of the TEU encountered public opposition in some member states but this was after institutional reform had been agreed. Indeed, the idea of the institutional dynamic racing ahead of change in political forces appears to turn on the head the assumptions of historical institutionalism. One possible explanation for this dynamic may be attributable to the institutional role of the European Council (cf Cameron, 1992). He sees the roots of the SEA going as far back as the European Monetary System (EMS), another creation of the European Council. Moreover, under Delors, the Commission developed its 'policy entreneurship' and capitalised on its presence inside the European Council as a way of contributing to that institution's role of developing the strategy of integration. Thus, the self-perceived role of the European Council might be a potential explanation for the comparative dynamism of the supranational institutions. Setting aside that research question, the fact remains that the SEA and the TEU brought about major change which can be understood using the analytical tools of historical institutionalism. They created increased competences for the EC/EU; they brought about changes in institutional and inter-institutional behaviour; they brought about changes in conventions and, above all, the norms embedded in the institutions. And the explosive growth of lobbyists in Brussels, according to one survey, a quadrupling between 1985 and 1992, indicates that interest groups have responded to the institutional dynamics (Andersen and Eliassen, 1993, p. 37). These were the changes which occurred at the trans-sectoral level. But there is a close interaction with changes occurring within individual policy areas. The internal market policy area was dynamised by the SEA and the normative commitment to completion of the single market. The flow was not always top down. Thus the dynamics in EC environmental policy and technology policy ahead of the SEA were such that treaty provision had to catch up with these developments but gave stronger treaty powers in doing so. This macro-meso institutional interaction is another aspect of EU governance which historical institutionalism can encompass.
Governance Regimes and Policy Sub-Systems within the EU

There has been an increasing realisation in policy analysis that generalised statements about national patterns of governance are deficient.

The Governance of the European Union

371

Instead greater analytical attention has been paid to policy subsystem approaches, a rubric comprising the literatures on policy networks and policy communities (cf. Van Waarden, 1992). Empirically, a disaggregated policy sub-system approach was established in the bench-mark analysis of the EC written in the 1970s and revised in the 1980s (Wallace, Wallace and Webb, 1983). Thus, studies of Community governance have tended to focus on, or include, the policy level, paralleling the public policy literature's concentration on the policy or issue level (Sbragia, 1992). Understanding EU governance requires an analytical framework for the sub-system level of analysis. But, what criteria must it fulfil? It must be as clear and parsimonious as possible. It must here be consistent with an historical institutionalist approach, so that systemic contextualisation is possible. It must be able to identify the boundaries of the particular sub-system. It must be able to map the institutions and actors involved. It must facilitate comparison across EU sub-systems in both static and dynamic terms. It must be able to encompass both the formulation of policy and its implementation. Finally, it must enable conclusions to be drawn about the key political issues of power and the distribution of benefits. The framework employed is that of the governance regime. The term governance encapsulates some key aspects of the empirical reality of the EU and is thus appropriate because the EU represents governance without a formal government and is not just concerned with formal institutions but can encompass procedures, norms, conventions and policy instruments as a core around which interest groups and other actors cluster. The term 'regime' can encompass the looser institutional arrangements in some areas of EU activity, for example in JHA policymaking. Regimes, too, have a substantive and normative content (cf. Krasner, 1983). Finally, regimes can be seen as purposive arrangements, formal or informal, which govern the interface between the private and public domains. The term governance regime derives from work undertaken by Campbell, Hollingsworth and Lindberg (1991) on sectoral governance in the American economy. Their use of governance regimes is in a slightly different context, for they also seek to identify different methods of economic governance, from the market to promotional networks, across several sectors of the American economy. However, that does not disqualify application to a more limited scope of enquiry. Indeed, it makes Campbell et a/'s explanation more parsimonious for the objectives here. Campbell et al. work within the broad framework of new institutionalism. They particularly develop three points of institutionalist explanations of interest to studying governance of the EU. First is their

372

Simon J. Bulmer

threefold identification of the role of the state (Campbell et ai, 1991, pp. 357-361). The state is conceived as an institutional arena which provides unequal access to societal groups for influencing the policy process. Secondly, the state is not just an arbitrator but also a 'player' in governance itself. And, thirdly, the state has its own distinctive configuration which predisposes it to certain types of activity. Of course, we must note that in their work 'the state' means American institutions in the present context so this must be understood to mean EU institutions. The first role of the state is the well established basis of new and historical institutionalism. The role of the state as a player within the institutional arena is important for understanding governance transformation in the EU. But, as was noted earlier in the paper, the Court and the Commission are key players in the transformation of EU governance. Taking the case of the ECJ, this applies both to trans-sectoral governance or to the governance regime level yet neither the ECJ nor the Commission are institutions designed to reflect political forces, unlike national governments or the EP. Thus the analytical dimension of the work of Campbell et al. alerts us to the elements of institutional autonomy within the EU. This issue addresses one of the key tensions within new institutionalism: how much autonomy do institutions possess. Historical institutionalists agree that institutions are not mere arenas but structure the way in which socioeconomic forces act. The suggestion here goes beyond that somewhat, for it posits that the ECJ and the Commission may generate endogenous institutional impetuses for policy change that go beyond the usual representation of institutional mediation. Their second contribution underscores the role of culture and norms embedded in the state as shaping the selection of policy. In the EU context, culture is likely to be about shared procedural norms rather than shared views towards, say, the conduct of economic governance. The explanation is simple. Culture develops over a long period; European integration is still a recent phenomenon. Thus, the EU has quite different cultural bases: Rhenish, Mediterranean and Anglo-Saxon capitalism, for example; or Scandinavian, Anglo-Saxon, Mediterranean and Bismarckian systems of welfare capitalism. The cultural context embedded within the national policies is one of the factors that makes the development of supranational policy so fraught: indeed, so much more fraught than in comparable federal systems, such as Germany or the USA. Their third contribution is to give attention to compliance mechanisms in economic governance; any study of sub-system governance that fails to take this dimension into account will be too parsimonious. Com-

The Governance of the European Union pliance mechanisms are quite diverse across the policy portfolio of the EU, from legal sanctions in many EC activities to a kind of peer pressure amongst state elites in respect of CFSP activities. Compliance systems are also important in another respect, for they are central to ensuring whether policy works. Because trial and error are an integral component to policy evaluation, they emphasise the process element of the term governance. If policy is perceived to be deficient in its implementation, this may provide new impulses for reform. The perceived failure of foreign policy cooperation during the 1990-91 Gulf Crisis is one factor explaining the TEU's commitment to a Common Foreign and Security Policy. How, then, to operationalise governance regimes as a tool for analyzing sub-system governance in the EU? Below we set out five different components.
Identifying the boundaries of the sub-system. Campbell et al. do not offer

373

guidance here because the level of their analysis is sectoral governance in the American economy. However, EU governance is overwhelming non-sectoral in nature, and much is outside the economic policy area, most obviously in the JHA and CFSP pillars. Agriculture, coal and steel are the exceptions which prove the rule! The solution to this lies in adopting the policy programme as defining the boundaries of governance (Rose, 1985). This definition may necessitate some flexibility in application because policy may in some cases be camouflaged by the way it is handled institutionally. Nevertheless, the policy programme approach generally facilitates identification of policy in the key executive agencies within the EU, namely the Commission and the Council hierarchies. Mapping the policy participants. This identifies the institutions and the socioeconomic interests, political parties etc. that are involved. Firstly, a focus on the EU-level institutions is of primary importance but their interaction with the domestic level in the member states is important. Institutional traditions and their cultural dimension are more embedded at that level than at the EU level. Secondly, non-governmental actors continue to use the national level as one of their channels of lobbying, reflecting the Council's continued importance. Hence this potential source of power must be examined. Thirdly, the institutions engaged in policy implementation must be included in this mapping exercise. Fourthly, it must be borne in mind that the institutional configuration of governance regimes, and the norms associated with them, will place some policy participants in a privileged position. As said in a policy network context, 'access to a policy network is normally controlled by the dominant government agency' (Wright 1988,

374

Simon J. Bulmer

T A B L E 2: Sources of Governance Transformation - The Single Market

Source of change Single market examples

Economic efficiency Technological development Actors seeking power and control Cultural factors

Move to exploit EC-wide markets to achieve economic efficiency in single market Perceived lagging behind Japanese and US rate of development European multinational companies and business leaders seeking to strengthen competitiveness Complex mixture of support for greater integration; increased support for neo-liberalism; trade-off with economically weaker states in return for a strengthening of cohesion National: deliberate transfer of power to supranational level as a way of trying to achieve more successful policy management and greater economic growth. Supranational: wish to increase power and fulfil original goals of EEC Treaty

State policy

p. 608). In empirical terms, part of this mapping exercise entails establishing the particular 'mix' of the institutions and instruments of governance from those discussed earlier in the paper. Sub-system comparison. Comparative statics entail comparing the findings established in the mapping of different governance regimes. More interesting is the issue of'comparative dynamics'. Here Campbell et al provide useful categorisation for the transformation of governance regimes, identifying five sources of transformation: economic efficiency; technological development; actors seeking power and control; cultural factors; and state policy. In Table 2 we apply these explanations to the single market programme for illustrative purposes. The categories are applicable at various levels of analysis: from the constitutional change of the SEA down to the details of regulating public procurement. Implementation is incorporated already by virtue of including the compliance arrangements when mapping the participants. Since governance is an iterative process, it is important to incorporate examination of compliance mechanisms, for any shortcomings are likely to feed back into the governance regime. Including implementation in governance regimes incorporates the judicial process, in particular the important role of the ECJ in the development of governance. The policy network literature, by contrast, has difficulties in including the ECJ. It also has difficulties in incorporating the compliance arrangements, for these may comprise separate networks in each member state.
The impact of policy outcomes on the distribution of power. March and Olsen's

(1989) work sought to explain sub-optimal policy outcomes and unintended consequences by means of bounded rationality within institu-

The Governance of the European Union tions. There is little difficulty, therefore, in extending this kind of analysis to the policy sub-system level. It is also possible, assisted by process tracing, to seek the roots of policy outcomes in the institutional mediation of bargaining positions. The question of compliance reminds us that policy outcomes may be subverted at that stage, either deliberately or accidently, for example by bureaucratic forces. The governance regime approach to sub-system governance in the EU retains a clear methodological link with the wider context of EU governance. For example, policy programmes become subject to horsetrading in the Council of Ministers, so it is crucial to encapsulate the embeddedness of governance regimes in the institutional superstructure of the EU. Moreover, the inclusion of judicial input and the important role of compliance ensures that these important components of governance are not simply residualised or unexplained. Above all, the governance regime concept remains as yet exempt from what might be termed the 'Kama Sutra' problem, namely the need to have an illustrated guide to the numerous positions adopted by political scientists within the policy networks literature. This brings the analysis back to the wider canvas and to the distinctive organisational configuration of the European Union.

375

The Regulatory Characteristics of EU Governance

In recent years, as social scientists have begun to reflect on Community governance one particular theme that has emerged from empirical observation is the regulatory pattern of governance (Majone, 1991; 1993; Dehousse, 1992). From an economic perspective the governance of the EU is striking because, compared to other multi-tiered systems of governance, the European framework has strikingly modest resources of its own. Consequently, there are major constraints upon the nature of activity that can be pursued. Spending programmes are frequently merely supportive of national measures, such as most structural fund expenditure. Alternatively, as in technology programmes, they normally require co-funding from the private sector. Thus, in EC activities expenditure lacks the measure of autonomy possessed in other multi-tiered or federal systems. Outside the EC pillar, activity is further constrained because budgetary provision at the international level is more limited or even non-existent. If this is translated into a public finance interpretation, there is scarcely any scope for the pursuit of two of the main functions of public budgets, namely stabilisation policy or redistributive measures. Instead expenditure is much more confined to an allocation function (cf. Musgrave,

376

Simon J. Bulmer

1959). Critics of the Maastricht road to EMU argue that the proposals are not viable because of this (Hughes-Hallet, 1994). In a different context, Majone (1993) explains that this bias in the function of the EC budget accounts for the emphasis upon social regulation rather than welfare policy in the Community's social dimension. But the effect is more far-reaching. As Majone (1991, p. 96) puts it elsewhere: an important characteristic of regulatory policy-making is the limited influence of the budget on the activities of regulators. The size of non-regulatory, directexpenditure programmes is constrained by budgetary appropriations and, ultimately, by the size of government tax revenues. In contrast, the real costs of most regulatory programmes are borne directly by the firms and the individuals who have to comply with them. Compared with these costs, the resources needed to produce the regulations are trivial. Thus, given the small size of the EC budget, and the large share taken by agricultural spending, 'the only way for the Commission to increase its role is to expand the scope of its regulatory activities'. Outside the supranational treaties a similar principle applies. Since economic or public finance instruments of the EC are very constrained, this necessitates budgetary activity through the allocation function of spending or it requires legislative regulations which place their costs upon private economic actors or upon the administrative apparatuses of the member states. A final point on economic aspects concerns regulation and the statemarket relationship. The EU does not intervene in the economy through ownership: it has no such holdings. Moreover, the size of the budget restricts subsidies and restricts using the awarding of contracts as a means of intervention in the market. Thus, traditional avenues of state intervention in the market are closed off, and the EU's actions are limited largely to determining the framework conditions within which the market operates. This takes us to the political dimension of regulation. In his analysis of different types of policy decision in the United States, Lowi (1972) examined the key characteristics of regulatory politics, identifying group bargaining as predominating, as opposed to party and electoral politics. He also saw regulatory politics as disaggregated, decentralised, interest-oriented and localised. Diametrically opposed to regulatory policy in his schema was constituent policy, a category which would include, in the European context, the Maastricht negotiations and their ratification. Strikingly he saw such policy issues as characterised by electoral and party politics, and centralised, ideological and 'systems' level in focus. This distinction could scarcely be

The Governance of the European Union more apposite to the European context. Regulatory politics is typically group bargaining and interest-oriented politics (see Wallace, 1983, pp. 410-415). This bias is one of the features tending to discourage popular interest in the European integration. It again vindicates historical institutionalists1 view that institutions shape the strategies of political forces. Finally, there is the administrative-legal aspect of Community governance. The central administration of the EC is small. It is not a full blown federal civil service, for it lacks the resources and manpower. Thus, the predominant method of regulation relies heavily on the member states for the lion's share of implementation, with all the problems that this raises for the legal issue of compliance. And it relies upon the member states to play a major role in the committees assisting the Commission in its executive role (Dehousse, 1992, p. 390). There are some elements of long chains of implementation in other multi-tiered political systems. In the Federal Republic of Germany particularly, the administration of federal law is in many cases devolved to the Lander. However, the skeletal nature of the EC's institutions for policy implementation is unique in multi-tiered governance. Hence the regulatory characteristic of the EC/EU is exaggerated markedly in the administrative-legal domain. By virtue of these economic, political and administrative-legal characteristics, a centripetal feature of Community governance is its tendency towards a regulatory approach to governance. The best match with these features is in those issue-areas falling under the supranational treaties, this unifying trend is a counterpart to the disaggregative findings of policy sub-system analysis.
Conclusion

377

This paper set out with the question: How best to analyze the collective governance of the European Union? In seeking to provide an answer to this question, it has considered the 'tool-kit' of EU governance. Hood (1983, p. 115) has argued that thinking of government as a tool-kit can help in three ways. Firstly, it can help simplify, and make sense of, a complicated subject-matter. Secondly, and conversely, it can help reveal the different ways in which government might tackle a problem. Thirdly, it offers a way of making comparisons: between government systems, policy areas and so on. In the first part of the paper attention focused on the range of institutions and tools available for EU governance, illustrating the different ways the EU can address particular policy areas. In the second part, the paper offered a framework for comparing the sub-systems of EU

378

Simon J. Bulmer

governance, namely the notion of governance regimes. Finally, the paper identified a regulatory pattern of politics as a feature lending coherence to the collective policies of the EU. Above all, this paper has emphasized institutions in the trans-sectoral, governance regime and institutional configuration perspectives on the EU, for institutions play a central role in structuring the governance of the EU. Institutional analysis is thus the central element of an EU tool-kit. NOTES
1. The analytical debate within Germany is less well-known in English language studies but derives from a research programme of the Deutsche Forschungsgemeinschaft. The roots of German institutional theory differ from those of new institutionalism. On this, see Gohler, 1987. 2. These institutions are not limited to legislative roles but our exposition is confined in this way for illustrative purposes. Outside the legislative arena, other institutions come into consideration, such as the Court of Auditors, or issue-specific agencies, such as the European Monetary Institute. 3. On the three functions performed by public budgets, see for instance Musgrave, 1959

REFERENCES
Andersen, S. and Eliassen, K. (1993) 'Complex Policy-Making: Lobbying the E C , in S. Andersen
and K. Eliassen (eds.), Making Policy in Europe: The Europeification of National Policy-making

(London: Sage Publications), pp. 35-53. Bradley, K. ST. C. (1992) 'Comitology and the Law: Through a Glass Darkly', Common Market Law Review, Vol. 29, No. 4, pp. 693-721. Cameron, D. (1992) 'The 1992 Initiative: Causes and Consequences', in Sbragia, 1992, pp. 23-74. Campbell, J., Hollingsworth, J. R. and Lindberg, L. (eds.) (1991), Governance of the American Economy (Cambridge: Cambridge University Press). Cook, J. and Kerse, C. (1991) EEC Merger Control: Regulation 4064/8$. (London: Sweet and Maxwell). Cram, L. (1993), 'Calling the Tune without Paying the Piper? Social Policy Regulation: the role of the Commission in European Community Social Policy', Policy and Politics, Vol. 21, No. 2, PP- '35-'4 6 Den Boer, M. and Walker, N. (1993) 'European Policing After 1992'', Journal of Common Market Studies, Vol. 31, No. 1, pp. 3-28. Dehousse, R. (1992) 'Integration v. Regulation? On the Dynamics of Regulation in the European Community', Journal of Common Market Studies, Vol. XXX, No. 4, pp. 383-402. Gohler, G. (1987) 'Politische Institutionen und Politikwissenschaft', in G. Gohler (ed), Grundfragen
der Theoriepolitischer Institutionen: Forschungsstand- Probleme - Perspektiven ( O p l a n d e n : Westdeutscher

Verlag), pp. 15-47. Hall, P. (1986) Governing the Economy (Cambridge: Polity Press). Heclo, H. and Wildavsky, A. (1974) The Private Government of Public Money: Community and Policy Inside British Politics (London: Macmillan). Hood, C. (1983) The Tools of Government (London: Macmillan). Holland, M. (1987) 'Three Approaches for Understanding European Political Co-operation: A Case-Study of EC-South African Policy', Journal of Common Market Studies, Vol. XXV No. 4, pp. 295-3'3Hughes-Hallett, A. (1994) 'Comment on Thygesen', in S. Bulmer and A. Scott (eds.), Economic
and Political Integration in Europe: Internal Dynamics and Global Context (Oxford: Blackwells), p p .

259-264. Ikenberry, G. J. (1988) 'Conclusion: An Institutional Approach to American Foreign Policy', in G. J. Ikenberry, D. Lake and M. Mastanduno (eds.), The State and American Foreign Policy (Ithaca, NY: Cornell University Press).

The Governance of the European Union

379

Jacobs, F., Corbett, R. and Schackleton, M. (1992) The European Parliament, 2nd edn. (Harlow: Longman).
Katzenstein, P. (1978) Between Power and Plenty, Foreign Economic Policies of Advanced Industrial States

(Madison: University of Wisconsin Press). Krasner, S. (ed.) (1983) International Regimes (Ithaca, N.Y.: Cornell University Press), pp. 1-21. Krasner, S. (1984) 'Approaches to the State: Alternative Conceptions and Historical Dynamics', Comparative Politics, Vol. 16, No. 2, pp. 223-246. Lowi, T. (1972) 'Four Systems of Policy, Politics and Choice', Public Administration Review, Vol. 32, No. 4, pp. 298-310. Ludlow, P. (1991) 'The European Commission', in R. Keohane and S. Hoffmann (eds.), The New
European Community. Decisionmaking and Institutional Change (Boulder, Co.: Westview Press), pp.

85-132. Majone, G. (1991) 'Cross-National Sources of Regulatory Policymaking in Europe and the United States', Journal of Public Policy, Vol. 11, No. 1, pp. 79-106. Majone, G. (1993) 'The European Community between Social Policy and Social Regulation',
Journal of Common Market Studies, Vol. 31, No. 2, pp. 153-169.

March, J. and Olsen, J. (1989) Rediscovering Institutions, (New York: Free Press). Milward, A. S. (1992) The European Rescue of the Nation State (London: Routledge). Moravcsik, A. (1991) 'Negotiating the Single European Act', in R. Keohane and S. Hoffmann
(eds.), The New European Community. Decisionmaking and Institutional Change (Boulder, Co.: Westview

Press), pp. 41-84.

Musgrave, R. (1959) The Theory of Public Finance: a Study in Political Economy (New York:

McGraw-Hill). Nicoll, W. (1993) 'Note the Hour and File the Minute', Journal of Common Market Studies, Vol. 31, No - 4. PP- 559~5 66 Nugent, N. (1991) The Government and Politics of the European Community, 2nd edn. (Basingstoke:

Macmillan). Nuttall, S. (1992) European Political Cooperation (Oxford: Clarendon Press).

O Nuallain, C. (ed.) (1985) The Presidency of the European Council of Ministers (Beckenham: Croom

Helm).
Padoa-Schioppa, T. (1987) Efficiency, Stability and Equity. A Strategy for the Evolution of the Economic

System of the European Community (Oxford: Oxford University Press). Peterson, J. (1991) 'Technology Policy in Europe: Explaining the Framework Programme in Theory and Practice', Journal of Common Market Studies, Vol. XXIX, No. 3, pp. 269-290.
Rasmussen, H. (1986) On Law and Policy in the European Court of Justice: a Comparative Study in Judicial

Policymaking (Dordrecht: Martinus Nijhoff). Rhodes, R. (1986) The National World of Local Government (London: Allen & Unwin). Richardson, J. (ed.) (1982) Policy Styles in Western Europe (London: Allen & Unwin). Rose, R. (1985) 'The Programme Approach to the Growth of Government', Britishjoumal of Political Science, Vol. 15, No. 1, pp. 1-28. Sbragia, A. (1992) 'Thinking About the European Future: the Uses of Comparison', in A. Sbragia
(ed.), Euro-politics: Institutions and Policymaking in the 'New' European Community (Washington:

Brookings Institution), pp. 257-291. Scharpf, F. (1988) 'The Joint-Decision Trap: Lessons From German Federalism and European Integration', Public Administration, Vol. 66, pp. 229-278.
Skowronek, S. (1982) Building a New American State. The Expansion of National Administrative Capacities,

i8yy-i()20 (Cambridge: Cambridge University Press). Steiner, J. (1990) Textbook on EEC Law, 2nd edn. (London: Blackstone Press Ltd.) Thelen, K. and Steinmo, S. (1992) 'Historical Institutionalism in Comparative Polities', in
S. Steinmo, K. Thelen and F. Longstreth (eds.), Structuring Politics: Historical Institutionalism in

Comparative Analysis (Cambridge: Cambridge University Press), pp. 1-32. Tilly, C. (ed) (1975) The Formation of National States in Western Europe (Princeton, N.J.: Princeton Univ. Press). Van Waarden, A. (1992) 'Dimensions and Types of Policy Networks', European Journal of Political Research, Vol. 21, No. 1/2, pp. 29-52. Wallace, H., Wallace, W. and Webb, C. (eds.) (1983) Policy-Making in the European Community, 2nd edn. (Chichester: John Wiley). Wallace, W. (1983) 'Less than a Federation, More than a Regime: the Community as a Political System", in Wallace, Wallace and Webb (1983) pp. 403-436.

380

Simon J. Bulmer

Weaver, R., Kent and Rockman, Bert, A. (1993) 'Assessing the Effects of Institutions', in R. Kent
Weaver and Bert A. Rockman, Do Institutions Matter? Government Capabilities in the United States

and Abroad (Washington, D.C.: Brookings Institution), pp. 1-41. Weiler, J. H. H. (1993) 'Journey to an Unknown Destination: A Retrospective and Prospective of the European Court of Justice in the Arena of Political Integration', Journal of Common Market Studies, Vol. 3t, No. 4, pp. 417446. Wellens, K. and Borchardt, G. (1980) 'Soft Law in European Community Law', European Law Review, Vol. 14, No. 5, pp 267321. Wessels, W. (1992) 'Staat und (westeuropaische) Integration: die Fusionsthese', Politische Vierteljahreschrifl, Sonderheft (Die Integration Europas, ed. M. Kreile), pp. 36-61.
Wilks, S. and Wright, M . (eds) (1987) Comparative Government-Industry Relations: Western Europe, the

United States and Japan (Oxford: Clarendon Press). Wright, M. (1988) 'Policy Community, Policy Network and Comparative Industrial Policies', Political Studies, Vol. XXXVI, No. 4, pp. 593-612.
Young, O . (1989) International Cooperation: Building Regimes for Natural Resources and the Environment

(Ithaca, N.Y.: Cornell University Press).