ARENA Working Papers
WP 97/25

 

 

New Institutionalism, The Single Market and EU Governance


Simon J. Bulmer
University of Manchester


 


Introduction

Analysis of the governance of the European Union has become a growth area in recent times. Amongst the frameworks that have emerged to examine different facets of EU governance have been:

• liberal intergovernmentalism (Moravcsik 1993a);

• multi-level governance (Marks, Hooghe and Blank 1996; Kohler-Koch 1996);

• policy network analysis (Peterson 1995; Richardson 1996);

• new institutionalism, with its various different strands (Bulmer 1993; Jachtenfuchs 1995; Pierson 1996; Pollack 1996);

• the «fusion thesis» explanation of integration (Wessels 1997).

There are doubtless others omitted from this list.

This focus upon the analysis of the governance of the EU has marked a shift away from the grander-scale theorising of earlier eras. With the exception of Wolfgang Wessels' fusion thesis there has been a move away from examining the telos of integration to looking at more closely defined research problems, a trend which developed with earlier moves to look at the substance of policy-making rather than the dynamics of integration as a whole (Wallace, Wallace and Webb 1983). However, the move towards more narrowly defined research problems runs the risk that the field of study becomes very fragmented. We may end up with a bewildering set of policy cases explained by a further array of analytical frameworks so that the «big picture» of integration is lost from view. This could also be one implication of analysis that suggests, for example, that we have different types of theories for different kinds of research problems: e.g. macro-theories for major changes in integration; new institutionalism for institutional change; and policy network analysis for meso-level policy analysis (Peterson 1995: 83-6). Taken to an extreme, this situation could atomise empirical research on the EU, while failing to identify a common methodological strand for analysis of the different levels of research problem.

This paper seeks to suggest a way in which the different levels of research analysis can be linked up through application of a middle-range theory, namely new institutionalism. The paper draws on the methods used in, and findings of, a study of the governance of the Single European Market (1985-96) (Armstrong and Bulmer 1997 forthcoming).

What follows has four parts. The first simply seeks to identify what «the governance of the EU» is, and how to have a complete picture of it. The second part looks at new institutionalism - specifically its historical institutionalist variant - and locates this approach at the intersection of comparative politics, international relations and legal theory: the core areas affected by EU governance. Can historical institutionalism benefit from this location to «capture» the different facets of EU governance? And if so how? In the third part attention is focused on the Single European Market (SEM), with a view to showing how historical institutionalism may encompass the scope of EU governance. Finally, the paper aims to draw some conclusions about whether historical institutionalism is «up to the job».


Identifying the Governance of the EU

«Governance» has become a fashionable term in political science for analysing the pattern of rule generally and in the EU specifically. Although the term has been criticised as imprecise (Rhodes 1996: 652), it has particular value in examining the pattern of rule in the EU. The EU does not resemble, or have, a government, so governance offers some descriptive purchase on the character of the polity. Moreover, within Europe the integration process has been inextricably bound up with the transformation of both the traditional system of «nation states» and of the role of individual Member States. This transformation has not been solely the product of integration but has derived from other developments such as globalisation, new developments in economic management, notably the move towards the regulatory state, and domestic moves towards «new public management» as a way of better administering policy. Beate Kohler-Koch has seen the transformation of governance as affecting: the role of the state; the rules of behaviour; the pattern of interaction; and the level of action (1996: 371). Above all, the task for those utilising the term governance as a research tool is to exploit its greater flexibility (whilst taking care with definitions, of course). The term “governance also problematises the pattern of rule within the EU polity. This exercise need not be post-modern deconstruction, rather it may facilitate the re-connection of: macro- and micro-level studies of the EU; of policy-making and policy-administration; of the political and legal aspects.

The last of this trinity is worth underlining, for neither in the EU nor in individual Member States has the transformed pattern of governance left the role of the judicial system untouched. The regulatory state seems to bring with it an increased juridification of economic management. The EU has not escaped this trend, with a continuing heavy caseload for the European Court of Justice, and a big growth in the workload of the Court of First Instance, established in 1989. The ECJ also stands entrusted with resolving some of the disputes about the level of action.

In this literature on governance it is argued that the state's role has moved from «authoritative allocation and mediation from `above' to the role of partner and mediator» (Kohler-Koch 1996: 371). That perspective underlines the character of supranational governance: of finding joint solutions through multi-levelled partnerships and through mediating the claims of affected interests. This mediating role is precisely the one assigned to institutions by those employing new institutionalist analysis, as will be seen below.

The focus on governance is important also because it highlights the problems of the traditional nation state in managing the policy problems with which it is confronted. The EU, of course, is precisely one available arena for managing such problems. If we are to analyse EU governance, therefore, what empirical dimensions do we need to take into account?

Governance and change

We need first of all to encompass change in governance. Supranational governance has undergone two significant step-changes in the last decade or so, and a third was finalised at the Amsterdam European Council in June 1997 and is currently pending ratification. The impact of these step-changes has been well-charted and, indeed, has been seen as part of a longer-term set of such changes (Wallace 1996a). But these supranational step-changes, which can reverse the integration process (e.g. the Luxembourg Compromise) as well as deepen it, are not simply contained in the bargains typically reached by Member States in the European Council. They also derive from path-breaking legal decisions such as those of the mid-1960s that underlined the special character of EC's legal system (see Weiler 1991), and that were under-estimated by many political scientists who were pre-occupied by the 1966 Luxembourg Compromise. The direct effect of EC law, direct applicability, supremacy, entitlement to damages and similar principles: these all had systemic consequences for the governance of the EU. They form part of the constitutionalisation of the EU and specifically its EC pillar. As Joseph Weiler has put it, «Constitutionalism is the DOS or Windows of the European Community». (1997: 97). Of course, the consequences of these pathbreaking judgements were not confined to the constitutionalisation debate but also had their impact upon the specific patterns of governance in individual policy areas.

The step-changes in EU governance have been given especial attention because of their importance in setting the political tone of integration for medium-term periods. However, the reality of governance at the level of individual policy areas is predominantly one of evolutionary change, punctuated by an occasional system-wide step-change. Accordingly, any attempt to give an analytical explanation of EU governance must take into account the developments in the interstices between politically-negotiated and jurisprudence-led step-changes at the systemic level. The incremental evolution of policy is of key importance and can be illustrated by a number of examples: rolling medium-term policy programmes, such as the Framework programmes in research and technology, environmental action programmes and others in such areas as social affairs or equal opportunities, or the Lomé conventions; institutionalised iteration, whereby negotiators explicitly include a time-specific review clause in legislation in recognition of the fact that the agreed legislation is as far as the negotiating parties could go at the time; and the inevitable trial-and-error aspect of policy making. Each phase of policy normally builds on prior experience. In addition to these features, the close relationship between the judicial process and EC pillar policy-making can also be a source of iteration, as ECJ jurisprudence clarifies the legal situation, creates regulatory «gaps», perhaps offering a better basis for the Commission to propose strengthened policy provision.

Change in the pattern of EU governance may thus be defined along three parameters:

• step-change or incremental change?

• negotiated political change or judicial change?

• system-wide or localised change?

A comprehensive analysis of EU governance needs to be able to encompass all these aspects. In particular it must go behind the institutional forms of change to the very mainsprings, be they matters of national preferences, functional requirements, or institutional self-interest. We also need to be aware of institutional inertia, the extreme version of what Paul Pierson has termed «lock-in», i.e. where particular avenues for policy have been excluded because of bias mobilised, and institutionalised, in the system of governance (Pierson 1996).

Governance structures

While EU governance entails change, it is also institutionally defined at specific time-points. Moreover, as has already been suggested, institutional definition occurs at multiple levels. The common set of institutions of the EU provides a systemic framework for governance. Nested within this framework is a large array of specific sets of institutional rules and procedures pertaining to individual policy areas. This meso-level phenomenon thus comprises a set of diverse governance regimes. In order to capture EU governance, therefore, it is necessary to encompass the systemic character of EU governance as well as the myriad sub-systems. This requirement is not something specific to the study of the EU, it should be pointed out. Debates in public policy analysis have been grappling with these problems for decades. [1]

In focusing upon the supranational level of governance, it is important to take into account the elements of power-sharing that are involved. In vertical terms, the EU cannot function without sharing power with other levels of governance. The sharing of power with national governments is most obviously expressed through the Council of Ministers, the European Council and the array of intergovernmental committees across all three pillars of the EU. This sharing of power takes place not only at the policy-making stage but also in the informal advisory contacts between national civil servants and the Commission in the pre-legislative phase. It also occurs in various guises of comitology at the policy execution stage. It scarcely needs mentioning that the enormous reliance of the EU on national governments (and their agents) to transpose and implement policy and - where it is the policy instrument employed - EC law, reinforces the interlocking nature of multi-level governance in the EU. Moreover, as scholars working on EU structural funds will attest, multi-level governance extends below national governments to include regions and local authorities (see Marks, Hooghe and Blank 1996). They, too, are involved in the power-sharing, although there is a lively debate as to whether sub-national government has really been empowered in determining structural policy as opposed to merely competing for the available funding (see Allen 1996). [2]

In horizontal terms the EU is also engaged in power-sharing. EU governance is not just about institutions; they are mediating diverse political forces coming from socio-economic interest groups organised at the EU level, from individual firms, and - more diffusely - from wider political sources and public opinion. Political parties, both nationally and at EU level, and public opinion show an ability to seek to engage with the broader constitutional issues of integration, such as monetary union. However, the intermediary role of national opinion-formers exposes such issues to considerable manipulation. Moreover, much of EU decision-making is concerned with the regulatory detail of market integration. In these areas it tends to be the best organised interest groups in Brussels together with interests represented by national governments that shape the policy debates conducted in EU institutions. This regulatory character of policy encourages the development of insider groups. The situation echoes Ted Lowi's portrayal of regulative politics in the USA as disaggregated, decentralised, interest-oriented and localised (Lowi 1972). Elected representatives, i.e. the MEPs, are sucked into this manner of conducting business via a committee structure which to a significant extent mirrors the Commission's organisation. It is in this sense that EU policy-making is pulled towards issue- or policy-related networks, which have both an informal and an institutionalised existence.

Governance, values, norms and ideas

Although they are still young, the institutions of the EU are not free of values. Embedded within them are values and norms which evolve gradually over time through learning-by-doing. Such institutional norms may have a significant impact on how functions allocated to the EU are in fact operationalised. These norms and values may also be subject to stronger, political change. That was the effect of the SEM/SEA package in increasing the presence of neo-liberal values in the work of the EC; or of the Maastricht Treaty in inculcating the notion of subsidiarity in the EU's institutions. New ideas, such as «flexibility» in the context of the 1996/97 IGC, can come to play an important role in shaping the pattern of EU governance. These aspects should also be taken into account in any study of governance, whether of the EU or of a nation-state. They may be difficult to quantify and thus represent a challenge to those whose theory-building tend towards the algebraic. Nevertheless, they represent an important check on «rational-actor» interpretations of politics. We cannot talk of an EU political culture, for that is some way off. However, that does not mean that ideas, norms and values have no explanatory value in analysing the governance of the EU.

Towards new institutionalism

The above considerations can be summarised by re-stating the different dimensions of EU governance that any theoretical framework must encompass:

• its political and legal character;

• the different types of change which are characteristic of a comparatively fragile system of governance like the EU;

• the nesting of policy-level governance structures within an overarching, systemic structure (the EU); and

• the normative dimension of governance.

Most of these considerations would apply to the analysis of policy developments in the national context, so an approach derived from comparative social science appears to have much to offer. Specifically, we will deploy some of the insights offered by new institutionalism. Our attention now turns to outlining that approach.


Historical Institutionalism and EU Governance

New institutionalism is a middle-range rather than a fully-blown grand theory. Unlike neo-functionalist and neo-realist (or intergovernmentalist) theories its institutionalist focus does not entail a teleology of integration. Rather, new institutionalism is agnostic on the end-goal of the integration process. The most modest sales pitch for new institutionalism is that it offers a methodology for research. This methodology generates research questions and orientations rather than mapping out a macro-social model of integration. The core assumption of this approach is that institutions matter.

This assumption, it should be noted, is shared with the “institutionalists” in international relations theory. Particularly linked to the analysis of international regimes, this view of international relations argues that institutions represent “persistent and connected sets of rules (formal and informal) that prescribe behavioural roles, constrain states, and shape expectations” (Keohane 1989: 3). Although the international relations literature on institutions is not engaged with systematically in this paper, we do emphasise the existence of this intellectual link. By adopting a comparative approach we do not wish to contribute to the balkanisation of social science, for we believe that, in the past, the analysis of European integration has suffered from such fragmentation. The sui generis assumptions of some political integration theory and the lack of interdisciplinary dialogue have risked confining European integration to an intellectual “ghetto” within the social sciences.

New institutionalism

New institutionalism is one of the principal methodological approaches to have emerged in the recent comparative social science literature. There are different variants of new institutionalism, so it should be regarded as something of an umbrella term. The most obvious question is: what is “new” about the approach? Two aspects are particularly distinctive to new institutionalism. [3]

One is a wider interpretation of what constitutes institutions. Thus, there is a shift away from formal constitutional-legal approaches to government, with their tendency to be configurative. It is possible to take into account some of the less formalised arenas of politics. A new institutionalist concern, therefore, encompasses these broader aspects of governance: a wider remit than the formal institutions of state or government. In including the less formal arenas of politics new institutionalism can be sensitive to the valuable findings of the “policy community” and “policy network” literature that has exposed the interconnectedness between formal state organisations and communities and networks of actors with an expertise and interest in a given policy area (see Peterson 1995).

A second distinction is a concern with the “beliefs, paradigms, codes, cultures and knowledge” embedded within the institutions (March and Olsen 1989: 26). This concern with institutional values is important, for the machinery of government is steeped in norms and codes of conduct and it is difficult to isolate formal institutional rules from the normative context. Of course, nobody would read the EU treaties and expect to gain an accurate impression of the operation of the institutions: witness the divergence of the practice of decision-making in the Council from treaty rules following the Luxembourg Compromise. To be sure, we did not need the invention of new institutionalism to provide that insight. On the other hand, it helps organise analysis of the evolution of ideas within institutions; of differing institutional cultures embedded within different parts of, say, the European Commission; of the change in institutional values brought about by the commitment to complete the SEM by the end of 1992 (which, it will be recalled, had no automatic legal effect); and similar normative changes. March and Olsen's work, it will be recalled, was inter alia an attempt to “correct” overly rationalist and functionalist interpretations in political science.

New institutionalism places the analytical focus on the polity. We can understand politics as comprising three separate components: politics, polity and policy. The presumption is that the polity structures the inputs of social, economic and political forces and has a consequential impact on the policy outcome. [4]


Political Forces ----> Political Structures (Institutions) ----> Policy

In such an approach institutions play a key mediating role. What do we consider to be “institutions” for the purposes of this study? They are taken as meaning formal institutions; informal institutions and conventions; the norms and symbols embedded in them; and policy instruments and procedures. This definition helps us to incorporate the traditional constitutional-legal notions of governance. But it also enables us to bring in the culture of political institutions; the informal decisional arenas - the “smoke-filled rooms” of politics; the accumulation of jurisprudence and the development of legal norms as factors contributing to institutional and policy norms; and finally, it allows us to incorporate the role of “soft law” and political declarations as further influences upon policy outcomes.

An additional issue upon which we must comment is that of institutional autonomy: of how much autonomy institutions possess in the political process. Most analysts see institutions as playing a mediating role. That is why the simple model of politics outlined above sees political (and socio-economic) forces as the starting point of governance. We certainly are not proposing that institutions provide the fundamental dynamics of politics. However, we do not see institutions as neutral arenas within which political forces are played out. Firstly, institutions structure the access of political forces to the political process, creating a kind of bias. Thus institutional rules, norms, resources or symbols shape actors' behaviour. Secondly, institutions can themselves develop endogenous institutional impetus for policy change that exceeds mere institutional mediation. This second aspect enables us to take into account the active contributions of the ECJ to the governance of the EU or of the Commission as an agenda-setter (see, respectively, Weiler 1991 and Peters 1994).

Historical institutionalism

Following these general remarks on new institutionalism, we now adopt a specific variant from within this literature, and examine the kinds of issues which it can help analyse. The approach adopted is historical institutionalism, which has a particularly encompassing interpretation of the role of institutions, as will be seen below. The particular distinction made here is with so-called rational choice institutionalism, which takes into account institutions and their rules but only in so far as they modify an essentially rational choice model of politics (see Thelen and Steinmo 1992: 7-10). [5] That essentially rationalist approach may be described as being at the “thin” end of institutionalism. By contrast our historical institutionalism is at a “thicker” end of institutional analysis by virtue of including the normative and cultural dimensions which go beyond rationalist calculations. We argue that historical institutionalism can highlight some important analytical issues in the governance of the SEM.

In order to show the potential insights of historical institutionalism, we identify four particular dimensions of EU governance which it can help “capture”: systemic change, governance structures, policy evolution, and the role of values and norms.

Historical institutionalism and systemic change

An important research terrain of new and historical institutionalism has been concerned with the state's role in macro-social change. Reacting to the society-centred, behavioralist analysis of the 1950s and 1960s, some political scientists moved, in the 1980s, to “bring the state back in” as the focus of analytical attention (see Evans, Rueschemeyer and Skocpol 1985). These analysts were particularly concerned with the evolution of political systems, and with the question of the degree of state autonomy in such developments.

The European integration literature has concentrated in particular on the issue of macro-level change: on the patterns of overall development. However, the attention paid to the role of the supranational institutions or to the embryonic European “state” has been skewed by the two predominant narratives. Neo-functionalism tended to underplay state dynamics in the integration process. One of the first studies in this vein to attribute importance to systemic transformation and the institutional capacity of the EC institutions was Lindberg and Scheingold's Europe's Would-Be Polity (Lindberg and Scheingold 1970: especially Chapter 4). In general, however, neo-functionalists saw a reconstruction of state capabilities as the outcome of what they were examining rather as an explanatory focus of the integration process.

A concern with the state is much more at the heart of realist or neo-realist interpretations of the integration process (see, for example, Hoffmann 1966; 1982). In such analyses, however, it is the resilience of the nation states that is the focus of attention. In other words, European integration is seen as a dependent variable of state development at the national level. In terms of comparative politics this kind of interpretation is unusual since the phenomenon being explained (supranational integration) is on a different level from that of the analysis (the nation state). Nevertheless, this approach has enjoyed something of a resurgence in the last decade, notably through the historical research of Alan Milward (1992) and Andrew Moravcsik (1991; 1993b). Their work, with its emphasis upon the “rescue” and “strengthening” of the nation state respectively, argues a particular line about where state power lies.

What does historical institutionalism have to offer in this context of systemic transformation? Essentially it places the focus upon those formal and informal institutions where systemic change is negotiated. Moreover, it pre-supposes no particular teleology of development. Hence, whilst the research agenda may be compatible with those of Milward and Moravcsik, the prescriptions of historical institutionalism are not concentrated on whether nation states are the winners or losers of the process but more neutrally on the way in which the negotiating fora shape the outcome of negotiations in a process of state reconstruction (see Bulmer 1996 forthcoming). The process of state reconstruction is a multi-level phenomenon. Hence there is not a zero-sum game between nation states and the EC/EU. Indeed, the process of integration involves a reconstruction of state responsibilities and activities at multiple levels, as officials and politicians from British local government or the German Länder will attest.

In the context of systemic change an historical institutionalist perspective comes closest to the work of analysts who examine the reconstruction of state authority over time. The work of Wolfgang Wessels, with its strong rooting in continental European writing on the state, displays consistency with historical institutionalism. The connection is formed by the emphasis which Wessels places on how the EU is the product of the “logic” of the European state system, with supranational solutions being sought for the problems of managing welfare states in an increasingly interdependent Europe (Wessels 1992; 1997). Although his work uses the terminology of “fusion”, which implies a convergence of state power on one level, he does take into account multiple levels of governance. His approach aims to capture the dynamics of change, while recognising that the integration process is open-ended. Marks, Hooghe and Blank (1996) advocate a multi-level governance explanation of integration, and this too displays strong consistency with an historical institutionalist approach, for they look not only at the intergovernmental institutions but also examine the role of supranational and subnational institutions and actors. Paul Pierson (1996) explicitly employs historical institutionalism to explain characteristics of the integration process. He offers a corrective to intergovernmental analysis and does so by explaining the importance of evolutionary change at the systemic level.

The principal value of historical institutionalism in respect of systemic change is twofold. It can explain, through institutionalist lenses and without a teleology of integration, the involvement of key institutions and actors in the transfers of competence at particular junctures of the integration process. It can also explain systemic change between those critical episodes of integration, such as the SEM/SEA, the TEU and so on, for much systemic change is evolutionary in character and judicial-normative in form (Pierson 1996; Weiler 1982; Dehousse and Weiler 1990). Consequently this form of change may be omitted from the research focus of those who only examine so-called history-making decisions. Analysts focusing upon such decisions also bias their research data away from the constitution-building that derives from the judicial process.

Historical institutionalism and governance structures

Historical institutionalism is not only an analytical approach which is helpful to examining systemic change, it can also assist analysis of sub-systemic governance. In this case the emphasis is upon how different institutional configurations - between political systems or between policy sub-systems - can impact upon governance capacity. [6] In the context of the governance of the SEM, two particular research issues are highlighted in this way.

The first is the governance capability of the EC/EU. Weaver and Rockman (1993a: 5-7) have identified various ways of measuring governance capability, and how the structural properties of the decision-making system may affect that capability. The SEM/SEA package-deal was a specific attempt to increase the EC's governance capability, both generally and with a view to ensuring the necessary legislation for the SEM programme. It is important, therefore, if we are seeking to analyse the governance of the SEM, to review performance in achieving that goal. Did the SEM/SEA package enhance the governance capability of the EC; and how? Thus, an institutionalist agenda is concerned with the impact of the polity upon putting the SEM into operation.

The second research issue is one which is well established in the literature on policy-making in the EC, namely the variable performance of the EC/EU in executing its various policies (Wallace, Wallace and Webb 1983; Wallace and Wallace 1996). The interest in differing sectoral policy dynamics is thus well developed but the particular contribution of historical institutionalism is to attribute varying policy outcomes to the different institutional arrangements. In order to highlight the different character of these arrangements, we term them “governance regimes”.

Historical institutionalism and policy evolution

Historical institutionalism emphasises the cumulative nature of policy-making, a feature explored by Pierson (1996). Thus, initial policy choices may restrict subsequent evolution so that a kind of path-dependency influences the course of policy. The result may be a policy which outlives its usefulness, or which does not correspond to the requirements of a new era. Nevertheless, it may have its own internal “logic”. The Common Agricultural Policy (CAP) may be seen in these terms. With an emphasis upon institutions, an interpretation of the CAP's character would be that those engaged in agricultural policy-making were able to isolate themselves institutionally from broader issues of public policy - including the financial aspects - and thus exploited supranational policy-making to enhance their own power resources. Only with the 1988 reform package were agriculture ministers, and the farm lobby influencing them, formally required to reconcile their policy decisions with those of EC budget ministers. According to such analysis, the result, both before 1988 and since, has been a sub-optimal CAP which none the less corresponds to the internal logic of the prevailing governance regime.

Historical institutionalism offers a number of valuable insights into the dynamics of the policy process.

First, it helps us to organise an exercise in process-tracing in policy case studies. Much policy-making is iterative and incremental. Thus, most of the legislation associated with the internal market White Paper and with many of the flanking measures did not simply emerge in the mid-1980s. Rather, new strategies were adopted to achieve goals already set out in the Treaty of Rome. And a new “bargain” gave renewed political impetus. Thus, whilst historical institutionalism often reveals path-dependency and policy solutions outliving their usefulness, there is the more positive possibility that lessons may be drawn from past experience, resulting in new strategies being put forward. This situation combined with the enhanced governance capability, brought about by the SEA, to create the new policy dynamics of the SEM. New strategies were attached to achieve existing goals. [7] Thus, the rationality of policy changes is a bounded rationality based on the endogenous construction of experience: “learning by doing”.

At the policy level, we are able to encapsulate this iterative process by examining the development of “governance regimes”. [8] From the modest beginnings of the first Commission initiative in an illustrative (EC pillar) policy area, perhaps reflected in a “mere” recommendation and itself preceded by a phase of agenda-setting, interest groups establish their claims to consultation. [9] National officials are also engaged in many meetings before the first piece of legislation is agreed. Gradually a governance regime is established. Then comes the learning process of how implementation by the national authorities corresponds to intentions. Further, ECJ jurisprudence develops legal norms. New legislation may be proposed, reflecting pressure from affected interests and incorporating the benefits of institutional learning. And so policy evolves in a manner which is structured by the institutional capacity of the EU generally and of the specific governance regime as well.

In this way an historical institutionalist perspective rejects the idea that European integration is just about the SEA and the Maastricht Treaty. Those negotiations represent the tip of an iceberg; we need to look at the nine-tenths of the iceberg which are below the waterline. Moreover, an historical institutionalist perspective focuses on the evolution of policy. It rejects the notion that politics can be separated from public administration, for the putting of policy into practice is an essential part of the whole: the experience of administration may start a new cycle of policy development.

Secondly, historical institutionalism helps forge a link between jurisprudence of the European Court and the legislative process. Unlike some areas of EU activity, notably the second (Common Foreign and Security Policy) and third (Justice and Home Affairs) pillars established under the TEU, the first (EC) pillar is regulated by EC law. In consequence, disputes between private parties, or disputes involving the supranational institutions or the member governments can lead to the establishment of important policy principles. Since the Commission's role is to act as conscience of “Europe”, it follows that its staff closely monitor the work of the European Courts (the ECJ and the CFI). Indeed, they may have intervened in litigation brought before the Court and thus be well-placed to draw lessons from consequent jurisprudence. Historical institutionalism offers a framework for understanding the judicial and legislative processes as complementary to one another. This situation is illustrated by the Commission seizing upon mutual recognition as a regulatory strategy for market integration in the wake of the Cassis de Dijon judgement.

A third contribution of historical institutionalism is through illustrating the ways in which institutions structure the policy process. Here we expand on the point made earlier on new institutionalism generally: that institutions are not mere neutral arenas.

In their analysis of the governance of the American economy, Lindberg and Campbell alert us to three ways in which the institutional structure shapes the policy process through a kind of mobilisation of bias (Lindberg and Campbell 1991: 357-361). [10] The first means of influence derives from the fact that the supranational institutions do not provide equal access for influencing the policy process. Thus the strong position of the Council of Ministers privileges national governments and their civil servants in the policy process, although increased power-sharing with the EP and varying rules on voting in the Council qualify this statement. Another instance of bias deriving from the institutional structure can be shown by a further reference to the CAP. The weakness of consumer responsibilities within the Commission and, for that matter, in national governments, has been another factor explaining why broad public interests, as well as those of taxpayers, were for a long time scarcely heard in the CAP governance regime.

A second means of institutional influence derives from the fact that the supranational institutions are not mere arbitrators but are key players in their own right. The Commission is an obvious example, for it is charged with finding supranational solutions to policy problems. As the initiator of policy, it has useful cards to play in setting the policy agenda (Peters 1994). So, too, do other institutions: the European Council (Bulmer 1996 forthcoming), but also the presidency of the Council of Ministers.

A third means of institutional influence derives from the fact that “the state” has its own distinctive configuration which predisposes it to certain types of activity. Within the EC pillar that distinctive configuration is its regulatory character. Hence the whole character of the White Paper was to legislate or regulate within the context of a small EC budget. Regulatory costs are borne by the Member States and the regulated. They are hardly borne by the supranational institutions at all, since implementation is left overwhelmingly to national arrangements, apart from in competition policy. As Guy Peters puts it:

Regulatory policy may ... minimize (although not eliminate) national, regional, and even class conflicts over Community policy. The choice [of policy instrument] also enhances the relative powers of the Commission and the bureaucracy (Peters 1992: 93).

The configuration of the EU institutions and the rather untypical set of policy instruments available predispose the Union to certain types of policy action but limit the scope for others, especially involving large financial transfers. This situation provides a specific illustration of Schattschneider's general observation that organisation is the mobilisation of bias (Schattschneider 1960: 71).

The three ways in which institutions influence the policy process - as identified by Lindberg and Campbell (1991) - represent elements of institutional autonomy within the EU. They comprise the third and final contribution of historical institutionalism to studying policy evolution.

Historical institutionalism and the norms of governance

A major intellectual input into new institutionalism derives from the work of James March and Johan Olsen (March and Olsen 1984; 1989). Their work has rather different origins from that described thus far, for its roots are in organisation and decision theory. For example, they criticise rational choice assumptions that political and administrative action is the product of calculated self-interest. In short, their criticism is of the utilitarianism implicit in much of the existing literature rooted in neo-classical economics. March and Olsen, like other analysts considered here, place emphasis upon institutions. Their view of institutional analysis is perhaps best encapsulated thus:

political actors are driven by institutional duties and roles as well as, or instead of, by calculated self-interest; politics is organized around the construction and interpretation of meaning [our italics] as well as, or instead of, the making of choices; routines, rules, and forms evolve through history-dependent processes that do not reliably and quickly reach unique equilibria; the institutions of politics are not simple echoes of social forces; and the polity is something different from, or more than, an arena for competition among rival interests (March and Olsen 1989: 159).

We find several new institutionalist pre-occupations in this passage. However, what is particularly distinctive to their work is the importance of norms, values and routines embedded within institutions; and what they term a “logic of appropriateness” that shapes individuals' actions within institutions (March and Olsen 1989: 160-162). What, then, can historical institutionalism offer by highlighting the normative dimension of EU governance?

Three contributions are particularly worthy of note. The first is to go beyond the idea of the institutional structure of the EC/EU being important and to look at the internal organisational features of individual institutions. This “micro-institutional” (or organisational) analysis allows us to focus on the routines, norms and symbols within individual institutions. Thus, the internal processes of decision-making in the European Commission, the committee structure of the EP, comitology and a range of other specific organisational features may help explain policy outcomes. These features, or what Burch and Holliday have termed the “disposition of an institution”, also privilege certain policy actors over others because they define a pattern or distribution of power potential (Burch and Holliday 1995). In one sense this is simply giving a micro-level focus to institutionalism. However, the addition of March and Olsen's second contribution - the role of norms and values - goes beyond mere structure to include such aspects as administrative culture.

March and Olsen highlight the norms and values held within organisations. Cultural explanations of the politics of the EU are problematic because we cannot identify a European culture. [11] However, within individual institutions norms and values accumulate and create a kind of institutional culture. These institutional norms, codes of conduct and values provide some kind of stability to a political system which is very fluid in character (see Mazey and Richardson 1996: 54). We can identify systemic norms associated with the EU, for example efforts by all institutions to respect the subsidiarity principle. Equally, norms and values may be attached to individual institutions. For example, they explain the informal voting practices in the Council of Ministers whereby attempts are made to reach a consensus and not overrule “significant minorities”, even where qualified majority voting is specified in the treaty rules. Taking another example, the Commission has a pro-integration mission which is partly inculcated by its rules and partly by its institutional culture. Lord Cockfield, responsible in the first Delors Commission for the SEM programme adopted such a missionary approach that he was deemed by the Thatcher government to have “gone native”. He adopted the institutional role and culture of the Commission to such an extent that he was not re-nominated for a further term. It is by linking together the micro-institutional and normative dimensions of March and Olsen's work that one can also analyse the different values held in different parts of the same institution (see Bulmer 1993: 363-364; Cini 1996).

Thus, to take an illustration relating to air transport, we find that state aids in this sector of economic activity are handled by Directorate-General (DG) VII (transport policy). This situation is quite different from what applies to state aids in general, for they are dealt with by DG IV (competition policy). This division of responsibility means that airlines seeking approval of state aids are dealing with a DG and a set of officials whose task is to oversee a functioning transport infrastructure. Their institutional remit is thus different from that of DG IV, whose concern is with competition rules. DG VII has traditionally been seen as more industry-friendly and more likely to be sympathetic on state aids than DG IV. Suffice it to note here that the organisational role - the organisational logic and administrative culture - of DG VII staff may affect the character of their decisions. Organisational roles are reinforced by the different norms and values embedded in the DGs. Organisational roles, norms and values affect the access points open to lobbyists and other political forces.

A third normative issue which historical institutionalism can highlight extends March and Olsen's work into the domain of ideas, an area which has been examined by Jachtenfuchs (1995). Particular institutional configurations can facilitate the spread of policy ideas if they possess suitable resources. [12] One pertinent example of that was the decision taken in early 1986 by internal market commissioner, Lord Cockfield, to ask Paolo Cecchini, a former deputy director-general of DG III, to undertake a report on the benefits of creating the SEM (Cockfield 1994: 90). Cecchini's findings, The European Challenge 1992. The Benefits of a Single Market (Cecchini et. al. 1988), succeeded in constructing positive values around the SEM programme at a very crucial stage in the White Paper's operationalisation. Not only did this report have a wide impact on economic actors but it proved to be a major factor in giving purpose to DG III itself. Together with the White Paper and other studies on “the costs of non-Europe”, it became a crucial factor shaping the administrative culture of DG III (which held responsibility for the SEM at that time).

------------------------------------------------------------------------------------------------------------------

· How the SEM/SEA package came to be agreed (systemic change).

· How changes to the EC's institutional structure contributed to the realisation of the SEM (governance structures).

· How institutional structures shape the governance of individual policy areas and issues into governance regimes (governance structures)

· How EC/EU policy is shaped over time as the product of both rule-making and judicial methods of regulation (policy evolution).

· How the institutional structure of the EU is not purely and simply a neutral arena but structures the policy process (policy evolution).

· How organisational disposition as well as institutional norms, values and ideas shape policy outcomes (normative dimension).

(Box 1: Historical institutionalist insights into the SEM/SEA package )

------------------------------------------------------------------------------------------------------------------

Historical institutionalism - a summary

There is an analytical core which runs through the four dimensions of EU governance discussed above. Centrally, the analysis is institution-centred rather than being actor-centred and behavioural in character. For March and Olsen a more organisational focus is adopted, with the endogenous organisational features structuring politics (1984; 1989). Further, informal and formal institutions are seen as structuring actors' political behaviour. Another recurrent feature is an historical focus; past choices restrict subsequent policy action. Finally, the values and norms embedded within institutions are ascribed explanatory value. Emphatically, none of this means that our empirical concern is just with institutions.

Historical institutionalism is not some kind of grand theory. That is not considered to be a deficiency in the context of European integration and specifically of the governance of the SEM. Attempts at grand theory within European integration have been shown to be highly problematic (Caporaso and Keeler 1995). Moreover, the EU is not yet a state with a stable institutional system, so the development of a convincing grand-theoretical explanation of its governance seems to be highly unlikely. Nevertheless, historical institutionalism offers a method for deriving analytical insights. And this method allows us to focus on different aspects of the SEM/SEA, summarised in Box 1. Above all, historical institutionalism offers a method which helps bring together the high-profile politics of the SEA and the TEU, on the one hand, and the day-to-day policy-making, policy-administration and judicial dynamics of the EC pillar, on the other.

Finally, how does historical institutionalism help pull together the “social science” of the governance of the EC/EU? Firstly, at the cost of parsimony, it emphasises the multi-levelled and multi-faceted nature of governance. In referring to the reconstruction of state authority, and by dropping the teleological presumptions that the trend is for that authority to flow in one direction (or not flow at all), we necessarily are describing a phenomenon of multi-level governance. Our perception of the governance of the EU is broadly consistent with that of Marks, Hooghe and Blank (1996), although we take a different route to reach our conclusions.

The importance of seeing the governance of the EU as multi-level governance is important for the social sciences. By contrast with the narratives of integration offered by neo-functionalism and intergovernmentalism, multi-level governance facilitates a discourse with those social scientists whose primary focus is the national political arena into which the supranational level has intruded. This multi-levelled character of governance is also important for linking up the political science and legal analysis of the EU. The matrix of relationships between national courts, national legislatures, supranational courts and supranational legislatures can also be examined. And these relationships are central to “bringing the legal dimension in”. Thus, just as we can consider the issue of the autonomy of the EU's political institutions, so we can examine the way in which the institutionalisation of legal norms has created some autonomy for the EC legal system. The interaction of the legal and political systems is facilitated in a way that is not possible with the predominant approaches (see the review in Armstrong 1997).


The Single Market: An Historical Institutionalist Analysis

In examining the development of the governance of the Single European Market (SEM), concentrating on the period 1985-96, we sought to examine both the step-change brought about by the decision taken at the Milan European Council in June 1985 together with the institutional changes brought about by the Single Act, and the operationalisation of the programme. As regards the latter, we selected six illustrative case studies.13 Space precludes a detailed account of our findings. However, we can summarise them in respect of the four dimensions of EU governance where we argued that historical institutionalism has analytical purchase.

Systemic change and the single market

Our starting-point was to identify the single market as not being confined to the Milan European Council agreement on the Cockfield White Paper. This package of measures was a somewhat arbitrary collection and by no means sufficient to complete a process of market liberalisation in the EC. Rather, it is necessary also to take into account the Single European Act, with its flanking measures and institutional reforms. This might seem self-obvious but it can highlight two particular institutionalist points. The first of these is revealed by the need for negotiators to make accurate judgements on the boundaries of reform proposals. This fate befell the British negotiating team at the Milan summit. It was of the view that the White Paper was a free-standing programme which could be accepted without the need to embrace other reform proposals that were afoot. In part this view derived from a persistent British wish to avoid constitutionalised reform. Moreover, where the UK government was motivated by neo-liberal economic ideas, other Member States had other norms: some sought to forge a strong linkage with side-payments through the structural funds; some sought to press ahead with institutional reforms already broached in 1983 at Stuttgart and then explored by the Dooge Committee. It is important to take care here because we may fall into the rationalists' trap of seeing the SEM and SEA as the tidy reconciliation of Member State preferences over the period from Milan to the signing of the SEA in February 1986. Such rational-actor analysis is dangerous, however, for the calling of the inter-governmental conference at Milan was achieved by a deft manipulation of European Council procedures by the Italian presidency in a way that surprised several states.14 This point reminds us, then, that institutional rules can amount to important contingencies. Without that vote, who knows how successful the White Paper would have been? Member State preferences may not suffice in accounting for systemic change.

A second point to derive from defining the shape of the package is that it is crucial for analysts, just as for negotiators, correctly to identify the boundaries of developments. Garrett and Weingast (1993) are correct to emphasise the importance of the Cassis de Dijon decision to the development of the internal market programme, in part perhaps as a corrective to Moravcsik's under-emphasis of this aspect in his original analysis (1991). However, they see the EC's legal arrangements as the guarantee for the SEM and neglect the log-rolling and linkages with other policy areas. The latter were also part of the SEM's guarantee.

A further key finding was that neo-functionalist and realist accounts of the SEM/SEA package amount to incomplete narratives of what happened. Member States' preferences were important but so were: the ECJ's Cassis ruling; Commission entrepreneurship (only partially covered by Sandholtz and Zysman 1990); and the role of transnational interests (see Cowles 1995). The key to getting a balance for us was to focus on the institution where the key decisions were taken: the European Council. The Commission has “insider status” in this institution and availed itself of all the available opportunities to project its agenda. Hence, while Moravcsik is formally correct to interpret the Cockfield White Paper as merely a response to the wishes of government heads (1991: 66), the question is whether they all supported market integration. The skill of Delors and Cockfield was to turn their ritualistic declarations back on the government heads; the latter could not then reject the internal market programme since “they” - and it was some more than others - had called for its completion at a series of summits, and they could scarcely turn down a re-working of the objectives of the Treaty of Rome. The missing item in integrating the various factors behind agreement on the White Paper is the way in which the underlying ideas gained a foothold in the European Council and could be exploited there by an astute Commission president. The institutional framework mattered.

Finally, our analysis went forward from agreement on the White Paper to see what happened. How was it put into operation? Again, much detailed evidence must be condensed. The vast majority of White Paper measures were agreed before the end of 1992. The thrust of market liberalisation adjusted over time. Concerns with subsidiarity, competitiveness and the regulatory burden on business led to new pre-occupations. The first of these emerged from the Maastricht process but the others reflected concerns, such as with the competitiveness of the European economy, which found expression at various subsequent sessions of the European Council. Concerns about compliance were one factor behind the UK government's uncharacteristically supranational contribution to the Maastricht Treaty, namely the provision to fine states flouting ECJ rulings. As these overarching policy concerns have evolved, so has the EU's understanding of what the single market is. In its review of its achievements, the Commission comes round to including social regulation as well as a set of sectoral liberalisation programmes - e.g. electricity, telecommunications - and a general understanding of the scale of the SEM which has grown considerably since 1985 (CEC 1996). However, our analysis went further than this generalised overview of the SEM because it looked at the transformation of the SEM into action at the policy level. In other words, our examination went beyond the grand bargains to include the more routine stuff of the EC: regulative politics.

The single market, the SEA and governance structures

What was the impact of the SEM/SEA upon the governance of the EU? In short it was profound, for economic governance has been transformed over the period since 1985. Amongst the factors assisting the dynamising of the market integration were: a clear programme (the White Paper); the 1992 deadline; a “philosophical framework”; an emphasis upon removing barriers without confronting sovereignty; and the Cecchini Report as a guide to the anticipated benefits.15 These factors reinforced institutional “missions”, especially on the part of DG III (originally containing the internal market portfolio) and of successive Council presidencies, which sought to post a good score on the board. For policy transposition, the use of league tables ensured an openness of information in pursuit of putting the SEM into practice. An associated development was the ECJ's handing down of key decisions on SEM matters. All our case studies were affected to some degree, facilitating major regulatory reform, for instance in air transport liberalisation (the Nouvelles Frontières and Ahmed Saeed judgements). Policy and judicial norms were transformed.

The formal institutional changes brought about by the SEA complemented these factors. Of particular importance were: the significant shift to provision for qualified majority voting in the Council; and the designation of clearer legal competences, something important for the Commission's authority.16 The EP was also mobilised through its stronger powers in those parts of the SEM that were subject to the co-operation procedure. These developments also reinforced dynamism in the institutions.

Finally, the supporters of market liberalisation were swimming with the tide. These economic ideas were already establishing themselves in the global political economy. Hence the direction of the EC was in conformity with these developments. Larger-scale enterprises and their interest groups were already becoming exposed to liberalisation; they were amenable to persuasion of the benefits of supranational regulation. In short, the norms of informal integration were in step with those of the formal integration structure.17

If the governance structure and capability of the EC/EU was transformed over time, the pattern was not identical across all the cases we examined. The exact balance of forces in policy-making varied considerably according to policy issue and the prevailing procedures and rules, from which national governments and interest groups took their cues. For example, the Merger Control Regulation of 1989 required unanimous agreement but some national governments were reluctant to give up their national regulatory powers (see Bulmer 1994). Business interest groups were vociferous in pressing for the Commission to have maximum authority so that cross-border mergers and acquisitions would be subject to one regulatory authority. Finding a balance was not easy and national preferences were of considerable importance. By contrast, in air transport liberalisation, subject to QMV, the ECJ ruled existing bilateral inter-state, inter-airline regulatory arrangements to be illegal, whereupon the Commission used its new-found competition policy powers to threaten legal proceedings against national flag-carrying airlines until liberalising legislation was in place. The Commission had as its principal allies only the governments (and, to a lesser extent, the airlines) of the United Kingdom and the Netherlands. A rolling three-package programme of legislation offered some side-payments to other states and some time for their industries to adapt.

In short, there was a broad picture of the dynamising of the policy process, but the form was shaped by the governance regime prevailing for the issue concerned.

The single market and policy evolution

A first point to make regarding policy evolution was that neither the SEM nor the majority of the case studies represented new policy issues. The common market was reinvented as the single market. The Commission's proposed merger legislation, which had languished in the Council for over a decade was given greater new pertinence by industrial re-structuring (and the ECJ's Philip Morris ruling). The free movement of goods benefited from a new SEM route-map: mutual recognition (following the Cassis ruling), harmonised minimum legal requirements; the hiving-off of standard-setting to independent European agencies and so on. Of the six cases examined, it was only with the transport of toxic waste that a substantially new issue come onto the policy agenda: and that because the internal market's removal of frontier controls opened up the possibility for the uncontrolled transport of hazardous materials across borders. The overwhelming character of the SEM was not one of starting from scratch but of “dusting down” pre-existing provisions and proposals, up-dating them and judging their feasibility in the changed institutional and normative context. The single market was largely a story of a step-change affecting multiple policy issues. However, a path-dependency of policy was maintained.

As has already been identified, each of our case studies proved to entail a close interaction between evolving judicial norms emanating from the ECJ and the negotiated policy process. Our cases were not selected to yield such results, and we have no reason to think they are unrepresentative of the broad range of SEM issues (including flanking measures). This interaction was of considerable importance to air transport liberalisation, the movement of goods and merger control.

Policy evolution was found to be shaped by the prevailing institutional structure. We already highlighted the contrasting circumstances of merger control and air transport liberalisation. In the former case the institutional rules privileged those states with pre-existing national merger control regimes. Under unanimity their demands, based on domestic experience, had to be met to secure agreement. In consequence, Germany, the UK and France were the decisive governmental players. In air transport, the EC had played next to no regulatory role until the 1980s. The ECJ's ruling that existing bilateral regulatory arrangements were illegal, combined with the QMV introduced by the SEA, made liberalisation inevitable and thus skewed things in favour of states advocating such a policy (the UK and the Netherlands). It was fortunate for these states that they held the presidency of the Council in 1986, the year of the Nouvelles Frontières ruling. The institutional circumstances of the EC could scarcely have been more helpful for the realising of a priority goal of the UK's Euro-sceptic Transport Secretary, Nicholas Ridley!

In the development of policy proposals, the Commission was in many cases not a mere arbitrator but a key player. In the case of the Maternity Directive the Commission used an expansive reading of Article 118a, EC Treaty to include the regulation of maternity pay levels on the basis of its SEA-created competence in health and safety matters. Some of its ideas in fact derived from the work of the European Parliament, whose Women's Rights Committee has acted as a ginger group on equal opportunities matters. This legislation, subject to QMV and the co-operation procedure, entailed a real power struggle between the EP, the Commission, and resistant governments, chiefly the UK. All this had to be presided over by the President of the Social Affairs Council, the UK minister, Gillian Shephard. The British government reputedly sought as few meetings as possible of this formation of the Council. It was perhaps symbolic, therefore, that the legislation was adopted by Fisheries Ministers. The EC's Maternity Directive, it is worth noting, directly prompted the establishment of a new interest group in this policy area: the European Women's Lobby.

Finally, in considering policy evolution, attention is drawn to the institutional configuration of the EC/EU. The SEM had a big impact on the character of governance. It underlined the notion of a supranational regulatory state with a rather unbalanced set of policy instruments. The predominance of interest-driven, elite-level politics has been reinforced. And it was a kind of reaction to this type of supranational governance that contributed to the popular unease during the Maastricht ratification exercises. Governance capability was strengthened by the SEM/SEA in terms of efficiency measures. In terms of democratic rule this episode also highlighted problems such as subsidiarity and transparency that plague the EU today.

The single market and the normative dimension

In this story of the SEM the role of norms, values and ideas should not be under-estimated. The SEM programme infused a new set of values amongst policy-makers and business elites alike. This was most clear in the supranational institutions. Illustrative of this is Lord Cockfield's note that, Fernand Braun (his director-general for the internal market) considered that “at long last and for the first time in many years...[the staff] knew exactly what was expected of them” (Cockfield 1994: 42).

The SEM dispersed throughout the EC the ideas of economic liberalism, even if they were reluctantly received in some Member States. However, the member government most keen on neo-liberal economic ideas, i.e. the UK, had not expected to find the presence of countervailing ideas. Yet that is what occurred with the 1992 Maternity Directive where the Conservative Government, having negotiated an opt-out of the Social Chapter at Maastricht in 1991, found itself having to introduce stricter labour market regulation. New norms were thus generated in flanking policy areas. In different policy sub-systems of EC governance, “logics of appropriateness” were re-defined by socio-economic and institutional actors alike.


Conclusion

This paper has sought to advance historical institutionalism as a method of analysing the multi-faceted nature of EC/EU governance. The analytical focus is placed upon the institutional aspect of governance but not to the neglect of wider political forces. We have sought to summarise some of the findings from an application of the method to the governance of the single market.

Of course, we have not discovered the holy grail. Historical institutionalism cannot get to the very mainsprings of integration. It offers an explanation based on intermediating factors rather than going to the underlying sources of macro-social change. In more modestly doing this, however, we have argued that it offers a balanced and inclusive `reading' of European integration. We argue that it is “up to the job” defined at the outset of the paper.

The one question surrounding historical institutionalism is this: is it a theory with predictive qualities? Whilst it certainly cannot predict the destiny of the integration process, it can make more localised predictions. In particular, it can predict that institutions take on their own dynamics, norms and values. With such a diversity of regulatory arrangements for the SEM - executive and judicial, supranational and national - we can predict the emergence of a highly fragmented and opaque set of arrangements for regulating economic governance in the EU, with implications for democratic accountability. The “logic” of global trade regulation by the World Trade Organisation intrudes in this area, too. This subject-matter looks ripe for further applications of historical institutionalism as the SEM evolves.


Footnotes

1. The emphasis in some literature on comparative, national characteristics in policy-making (cf. Katzenstein 1978, Richardson 1982) led to an adverse reaction on the part of some analysts, who resisted the idea that there were “strong” states and “weak” states. Examination of different policy cases revealed examples which did not fit with the supposed national profile. This kind of approach was one source of the burgeoning literature on policy networks and communities (Atkinson and Coleman 1989; Freeman 1985; Wright 1988). This literature was a new “take” on Lowi's argument that policy shapes politics: that the character of politics is shaped by the policy concerned (Lowi 1964).

2. This paper is concerned with EU governance as mediated through supranational institutions. There is a separate institutionalist research agenda which seeks to explain national patterns of articulating European policy, and why they differ, see for example Rometsch and Wessels (1996).

3. On new institutionalism in general see the surveys by Hall and Taylor (1996); Lowndes (1996); and March and Olsen (1996). For applications of new institutionalism to different aspects of European integration, see Armstrong (1995; 1997); Armstrong and Bulmer (1996); Bulmer (1993; 1994; 1996); Kerremans (1996); Pierson (1996); Wincott, (1995).

[4]. This view corresponds also to German institutional theory: see Göhler, 1987, pp. 15-47. For a more elaborate development of this chain in connection with the issue of governance capabilities, see Weaver and Rockman 1993a: 7-11.

[5]. This distinction between rational choice institutionalism and a “thicker” variant which sees institutions embedded in norms, values and even identity is also present in the international relations literature. The former approach is illustrated by Keohane (1989); the latter, for example, by Ruggie (1993b) or Young (1989).

[6]. For an application of this approach, comparing the American political system with others, see Weaver and Rockman (1993b).

[7]. The situation was not quite as straightforward as suggested here, since the notion of an “internal market' was not identical to that of a “common market”. In truth, there were elements of all of the explanations for change put forward by Thelen and Steinmo (1992: 16-18).

[8]. See also Campbell and Lindberg (1991). Their analysis of the governance of the US economy has contributed to our thinking here.

[9]. This illustration is just that: it does not claim to incorporate all the possible variations in the development of governance regimes. The pattern would differ in each of the other two pillars of the EU.

[10]. We substitute the term “institutional structure” for what, in an American context, they term “the state”

[11]. Different national cultures do meet in the Council of Ministers, of course. See Shackleton (1991) for an application of cultural analysis to the EU.

12. For an illustration of this in British economic policy - namely the shift from Keynesianism to monetarism - see Hall (1992). Also see Finnemore (1993) for discussion of how international organisations may “teach norms” to states.13. The case studies included two from the Cockfield White Paper (the free movement of goods and public procurement in the utilities sector). We also chose a policy sector, namely air transport liberalisation, which was mentioned in very modest terms in the White Paper but has now become emblematic of other important sectoral exercises in liberalisation, such as telecommunications. A fourth case examined the merger control regulation, which was not in the White Paper but amounted in 1985 to an important gap in the EC's instruments for regulating corporate behaviour, especially in light of the encouragement given to industrial re-structuring. Our final two cases considered aspects of social regulation which flanked market integration: the regulation of toxic waste (was it to be considered as a “good”, for which free movement should be assured, or as an environmental hazard, whose cross-border movement should be regulated?); and the regulation of pregnant women at the workplace, considered to be a matter falling under the health and safety competence introduced by the SEA.14. See, for example, the account of this episode offered by Lord Howe, the then British Foreign Secretary, in his memoirs (Howe 1994: 409). The Italian presidency called a vote in the European Council, using the argument that the calling of an IGC was a procedural matter.15. See also the account by Lord Cockfield himself (Cockfield 1994).16. Significantly, the principal remaining problem areas of the White Paper retain unanimous voting: free movement of persons and indirect taxation matters.17. On informal and formal integration, see Wallace (1990).

REFERENCES

Allen, D. (1996), “Cohesion and Structural Adjustment”, in H. Wallace and W. Wallace (eds.), Policy-Making in the European Union, Oxford, Oxford University Press, 209-33.

Armstrong, K. (1995), “Regulating the Free Movement of Goods: Institutions and Institutional Change” in J. Shaw and G. More (eds.), New Legal Dynamics of European Union, Oxford, Oxford University Press, 165-91.

Armstrong, K. (1997 forthcoming) “New institutionalism and EU Legal Studies”, in C. Harlow and P. Craig (eds.), Law-Making in the European Union, Dublin, Round Hall.

Armstrong, K. and Bulmer, S. (1996), “United Kingdom”, in D. Rometsch and W. Wessels (eds.), The European Union and Member States, Manchester, Manchester University Press, 253-90.

Armstrong, K. and Bulmer, S. (1997 forthcoming), The Governance of the Single European Market, Manchester, Manchester University Press; New York, St. Martin's.

Atkinson, M. and Coleman, W. (1989), “Strong States and Weak States: Sectoral Policy Networks in Advanced Capitalist Economies”, British Journal of Political Science, 19:1, 47-67

Bulmer, S. (1993), “The Governance of the European Union: A New Institutionalist Approach”, Journal of Public Policy, 13:4, 351-80.

Bulmer, S. (1994), “Institutions and Policy Change in the European Communities: the Case of Merger Control”, Public Administration, 72:3, 423-44.

Bulmer, S. (1996 forthcoming), “The European Council and the Council of the European Union: Gatekeepers of a European Federal Order?”, Publius: The Journal of Federalism, 26:4, 000-000.

Burch, M. and Holliday, I. (1995), The British Cabinet System, Hemel Hempstead, Harvester Wheatsheaf.

Campbell J. and Lindberg, L. (1991), “The Evolution of Governance Regimes”, in Campbell, J., Hollingsworth J.R., and Lindberg L. (eds.) (1991), Governance of the American Economy, Cambridge, Cambridge University Press., 319-55.

Caporaso, J. and Keeler, J. (1995), “The European Union and Regional Integration Theory”, in C. Rhodes and S. Mazey (eds.), The State of the European Union, Harlow, Longman.CEC (1996), “The Impact and Effectiveness of the Single Market”, COM(96) 520 final.

Cecchini, P. et. al. (1988), The European Challenge 1992. The Benefits of a Single Market, Aldershot, Wildwood House.

Cini, M. (1996), The European Commission, Manchester, Manchester University Press.

Cockfield, Lord A. (1994), The European Union: Creating the Single Market, Chichester, Chancery Law Publishing.

Cowles, M. G. (1995), “Setting the Agenda for a New Europe: the European Round Table and EC 1992”, Journal of Common Market Studies, 33:4, 501-526.

Dehousse, R. and Weiler, J. (1990), “The Legal Dimension”, in W. Wallace (ed.), The Dynamics of European Integration, London, Pinter, 242-60.

Evans, P. Rueschemeyer, D. and Skocpol, T. (eds.) (1985), Bringing the State Back In, Cambridge, Cambridge University Press.

Finnemore, M. (1993), “International Organizations as Teachers of Norms: the United Nations Educational, Scientific and Cultural Organization and Science Policy”, International Organization, 47:4, 565-97.

Freeman, G. (1985), “National Styles and Policy Sectors: Explaining Structured Variation”, Journal of Public Policy, 5:4, 467-96.

Garrett, G and Weingast, B. (1993), “Ideas, Interests and Institutions: Constructing the European Community's Internal market”, in J. Goldstein and R. Keohane (eds.), Ideas and Foreign Policy: Beliefs, Institutions and Political Change, Ithaca, NY, Cornell University Press, 173-206.

Göhler, G. (1987), “Politische Institutionen und Politikwissenschaft”, in G. Göhler (ed.), Grundfragen der Theorie politischer Institutionen: Forschungsstand - Probleme - Perspektiven, Opladen, Westdeutscher Verlag, 15-47.

Hall, P. (1992), “The Movement from Keynesianism to Monetarism: Institutional Analysis and British Economic Policy in the 1970s”, in S. Steinmo, K. Thelen and F. Longstreth (eds.), Structuring Politics: Historical Institutionalism in Comparative Analysis, Cambridge: Cambridge University Press, 90-113.

Hall, P. and Taylor, R. (1996), “Political Science and the Three New Institutionalisms”, Political Studies, 44:5, 936-57.

Hoffmann, S. (1966), “Obstinate or Obsolete: the Fate of the Nation State and the Case of Western Europe”, Daedalus, 95, 862-915.

Hoffmann, S. (1982), “Reflections on the Nation-State in Western Europe Today”, Journal of Common Market Studies, 21:1/2, 21-37.

Howe, G. (1994), Conflict of Loyalty, London, Macmillan.

Jachtenfuchs, M. (1995), “Theoretical Perspectives on European Governance”, European Law Journal, 1:2, 115-33.

Katzenstein, P. (ed.) (1978), Between Power and Plenty: Foreign Economic Policies of Advanced Industrial States, Madison, WI, University of Wisconsin Press.

Keohane, R. (1989), “Neoliberal Institutionalism: A Perspective on World Politics”, in R. Keohane (ed.), International Institutions and State Power, Boulder, Co., Westview Press, 1-20.

Kerremans, B. (1996), “Do Institutions Make a Difference? Non-Institutionalism, Neo-Institutionalism, and the Logic of Common Decision-Making in the European Union”, Governance, 9:2, 217-40.

Kohler-Koch, B. (1996), “Catching up with Change: the Transformation of Governance in the European Union”, Journal of European Public Policy, 3:3, 359-80.

Lindberg, L. and Campbell, J. (1991), “The State and Economic Governance” in J. Campbell, J. Hollingsworth and L. Lindberg (eds.), Governance of the American Economy, Cambridge, Cambridge University Press, 356-95.

Lindberg, L. and Scheingold, S. (1970), Europe's Would-Be Polity, Englewood Cliffs, Prentice-Hall.

Lowi, T. (1964), “American Business, Public Policy, Case Studies and Political Theory”, World Politics, 16, 677-715.

Lowi, T. (1972), “Four Systems of Policy, Politics and Choice”, Public Administration Review, 32:4, 298-310.

Lowndes, V. (1996), “Varieties of New Institutionalism: A Critical Appraisal”, Public Administration, 74:2, 181-97.

March, J. and Olsen, J. (1984), “The New Institutionalism: Organizational Factors in Political Life, American Political Science Review, 78, 734-49.

March, J. and Olsen, J. (1989), Rediscovering Institutions: the Organizational Basis of Politics, New York, The Free Press.

March, J. and Olsen, J. (1996),”Institutional Perspectives on Political Institutions”, Governance, 9:3, 247-64.

Marks, G., Hooghe, L. & Blank, K. (1996), “European Integration From the 1980s: State-Centric v. Multi-Level Governance”, Journal of Common Market Studies, 34:3, 341-78.

Mazey, S. and Richardson, J. (1996), “EU Policy-Making: a garbage can or an anticipatory and consensual policy style?”, in Y. Mény, P. Muller and J-L. Quermonne (eds.), Adjusting to Europe, London, Routledge, 41-58.

Milward, A. (1992), The European Rescue of the Nation-State, London, Routledge.

Moravcsik, A. (1991), “Negotiating the Single European Act: National Interests and Conventional Statecraft in the European Community”, International Organisation, 45:1, 651-88.

Moravcsik, A. (1993a), “Preferences and Power in the European Community”, Journal of Common Market Studies, 31:4, 473-524.

Moravcsik, A. (1993b), “Why European Integration Strengthens the Nation-State”, Cambridge, Ma, Harvard University Center for International Studies Working Paper.

Peters, B. G. (1992), “Bureaucratic Politics and the Institutions of the European Community”, in A. Sbragia (ed.), Euro-Politics, Washington DC, Brookings Institution, 75-122.

Peters, B. G. (1994), “Agenda-Setting in the European Community”, Journal of European Public Policy, 1:1, 9-26.

Peterson, J. (1995), “Decision-Making in the European Union: Towards a Framework for Analysis”, Journal of European Public Policy, 2:1, 69-93.

Pierson, P. (1996), “The Path to European Integration: A Historical Institutionalist Analysis”, Comparative Political Studies, 29:2, 123-63.

Pollack, M. (1996), “The New Institutionalism and EC Governance: The Promise and Limits of Institutional Analysis”, Governance, 9:4, 429-58.

Rhodes. R.A.W. (1996), “The New Governance: Governing without Government”, Political Studies, 44:5, 652-67.

Richardson, J. (ed.) (1982), Policy Styles in Western Europe, London, Allen & Unwin.

Richardson, J (1996), “Policy-making in the EU”, in J. Richardson (ed.), European Union: Power and Policy-Making, London, Routledge, 3-23.

Ruggie, J. G. (1993), “Multilateralism: The Anatomy of an Institution”, in J. G. Ruggie (ed.), Multilateralism Matters: The Theory and Praxis of an Institution, New York, Columbia University Press, 3-47.

Sandholtz, W and Zysman, J. (1989), “1992: Recasting the European Bargain”, World Politics, 42:1, 95-128.

Schattschneider, E. E. (1960), The Semisovereign People: A Realist's View of Democracy in America, New York, Holt, Rinehart and Winston.

Shackleton, M. (1991), “The European Community Between Three Ways of Life: A Cultural Analysis”, Journal of Common Market Studies, 29:6, 575-601.

Thelen, K. and Steinmo, S. (1992) “Historical Institutionalism in Comparative Politics”, in S. Steinmo, K. Thelen and F. Longstreth (eds.), Structuring Politics: Historical Institutionalism in Comparative Analysis, Cambridge: Cambridge University Press, 1-32.

Wallace, H. (1996a) “The Institutions of the EU: Experience and Experiments”, in H. Wallace and W. Wallace (eds.), Policy-Making in the European Union, Oxford, Oxford University Press, (1996), 37-68.

Wallace, W. (1990), “Introduction: the Dynamics of European Integration”, in W. Wallace (ed.), The Dynamics of European Integration, London, Pinter/RIIA, 1-24.

Wallace, H. and Wallace, W. (eds.) (1996), Policy-making in the European Union, Oxford, Oxford University Press.

Wallace, H. Wallace, W. and Webb C. (eds.) (1983), Policy-Making in the European Community, 2nd edn., Chichester,

John Wiley Weaver, R. Kent and Rockman, Bert A. (1993a), “Assessing the Effects of Institutions”, in Weaver, R. Kent and Rockman, Bert A. (eds.) Do Institutions Matter? Government Capabilities in the United States and Abroad, Washington, D.C., Brookings Institution,1-41.

Weaver, R. Kent and Rockman, Bert A. (eds.) (1993b), Do Institutions Matter? Government Capabilities in the United States and Abroad, Washington, D.C., Brookings Institution.

Weiler, J.H.H. (1982), “Community, Member States and European Integration: Is the Law Relevant?”, Journal of Common Market Studies, 21:1/2, 39-56.

Weiler, J.H.H. (1991), “The Transformation of Europe”, The Yale Law Journal, 100:8, 2405-83.

Weiler, J.H.H. (1997), “The Reformation of European Constitutionalism”, Journal of Common Market Studies, 35:1, 97-131.

Wessels, W. (1992), “Staat und (westeuropäische) Integration: die Fusionsthese”, Politische Vierteljahresschrift, Sonderheft 23 (Die Integration Europas, ed. M. Kreile), 36-61.

Wessels, W. (1997), “An Ever Closer Fusion? A Dynamic Macropolitical View on Integration Processes”, Journal of Common Market Studies, 35:1, 267-99.

Wincott, D. (1995), “The role of the law or the rule of the Court of Justice? An 'institutional' account of judicial politics in the European Community”, Journal of European Public Policy, 2:4, 583-602.

Wright, M. (1988), “Policy Community, Policy Network and Comparative Industrial Policies”, Political Studies, 36:4, 593-612.

Young, O. (1989), International Cooperation: Building Regimes for Natural Resources and the Environment, Ithaca, NY, Cornell University Press.






[Date of publication in the ARENA Working Paper series:15.10.1997]