Jump to ContentJump to Main Navigation
The Institutionalization of Europe$

Alec Stone Sweet, Wayne Sandholtz, and Neil Fligstein

Print publication date: 2001

Print ISBN-13: 9780199247967

Published to Oxford Scholarship Online: April 2004

DOI: 10.1093/019924796X.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 30 May 2020

Conclusion: Institutional Logics of European Integration

Conclusion: Institutional Logics of European Integration

Chapter:
(p.221) 11 Conclusion: Institutional Logics of European Integration
Source:
The Institutionalization of Europe
Author(s):

James A. Caporaso

Alec Stone Sweet

Publisher:
Oxford University Press
DOI:10.1093/019924796X.003.0011

Abstract and Keywords

This concluding chapter examines the use that is made in the book of institutionalist theory to address the question of European integration, a use that is described as not typical in the study of the European Union, which is usually approached via international relations and integration theory. Issues that result from taking this institutional approach (i.e. analysing European political integration as institutionalization rather than as international relations) are briefly discussed, before going on to see how the Europe of the last half of the twentieth century has provided social scientists with rich opportunities for evaluating how new political systems evolve and emerge. Some of the aspects examined are: the concept of institutions; the impact of institutions; feedback effects (feedback loops liking actors to organizations, to institutions, and back again); institutional coherence; institutions and society; and the demand for, and supply of, institutions.

Keywords:   European integration, European Union, feedback effects, feedback loops, institutional coherence, institutionalist theory, institutionalization, institutions, institutions and society, integration theory, international relations, political integration, political systems

Changes in basic forms of political rule are measured in the social science equivalent of geological time. In Europe, the transition from the medieval form, based on personal ties and fragmented authority structures, to modernity, in which sovereignty is exercised through impersonal, institutionalized authority, comprises one such deep change. The move to a ‘post‐Westphalian’ order, if indeed this is occurring, provides another example (Caporaso 1996). The focus of this volume—the consolidation of supranational political spaces in Europe—directs our attention to the transformation of the state system from a relatively ‘primitive site of collective governance’ (Stone Sweet, Flikgstein, and Sandholtz, this volume, p. 1) to the densely institutionalized system of interrelationships we witness today.

If it is true that we can gain from crafting methodological lenses for the long view, how much of significance will we miss by downplaying or relativizing the near term? If focus on the shifting and grinding of plate tectonics seems the more metaphorically appropriate strategy to institutionalists, how should we view everyday surface manifestations of these forces? In our view, the appropriate time frame is as long as required and possible. It will inevitably be greater for the development of a stable system of interest‐group representation (Mazey and Richardson, this volume), which began in the 1950s (Haas 1958) than for monetary policy, where the early 1970s provides one convenient place to start (Cameron 1998; McNamara, this volume); and it will be longer for monetary policy than for development of an EU competence in Justice and Home Affairs (Turnbull and Sandholtz, this volume). Further, such episodes can never be properly told in isolation from the more macro ‘integration narratives’ that give our subject‐matter its historical coherence (Fligstein and Stone Sweet, this volume; Héritier, this volume). Indeed, one of the core messages of this book is that, at virtually every point in time, the processes that constitute European integration have been heavily conditioned by existing institutional arrangements. If this project has been right to conceive of the construction of Europe in institutionalist terms, then the future has been, and will continue to be, (p.222) meaningfully organized by the ways in which supranational governance has emerged and become real for actors. In any case, it is a brute fact that an increasing number of actors have, over time, invested in these institutions, reinforcing their centrality as loci for future decision‐making.

In putting things this way, we are not asserting that we understand all that is important about the EU, or that the end point of European integration is known or predictable with certainty. We start by noting that the European state system has been altered in important ways, from a decentralized balance of power system prior to1945 to the current environment characterized by high levels of interdependence, supranational organization, and dense matrices of hierarchically organized rules. This transformation, whatever it might imply for the Westphalian order, provides us with puzzles in search of explanations.1 This project addresses these puzzles as if they were a matter for generic social theory. What we collectively assert is that there are patterns to the changes we observe, that new forms of governance did not simply materialize from the thin air, and that there has been an underlying social logic to how supranational arenas have been constituted.

Our use of institutionalist theory to address conventional institutionalist questions is not typical in the study of the EU; counter‐examples include Mattli (1999); Olsen (2000); Sandholtz and Stone Sweet (1998); Scharpf (1999). In doing so, we strongly reject the view that Europe or regional integration are somehow sui generis objects of inquiry. Instead, the Europe of the past half‐century is treated as providing social scientists with rich opportunities for evaluating propositions about how new political systems emerge and evolve. We will return to how the group has done so shortly. Beforehand, we note that the volume also resolutely ignores the standard theoretical reference points in EU studies, which are derived from international relations and integration theory. While we view this orientation as appropriate, given the group's concerns, we think it raises some important issues.

1. Political Integration as International Relations

One implicit claim of this volume is that the processes that have served to institutionalize supranational arenas of governance have also removed Europe from international politics. Generally, institutionalists have good reason to deny essential distinctions between politics within states and politics among states, and then to proceed to the study of the particular rule‐based and organizational forms found in international society. Nonetheless, scholars working in the theoretical traditions of international relations have done the most to develop regional integration theory. We wish to make three summary points about these modes of theorizing and research here.

(p.223) First, although neo‐realist theorists have had something meaningful to say about the conditions leading to the founding of the Communities, they have utterly failed to provide a compelling, even post hoc, explanation for the construction of the European polity. In the classic position spelled out by Kenneth Waltz (1979: 70–1), the emergence of the supranational in Europe is treated as evidence supporting the claim that it is the underlying distribution of power among states in the system that determines outcomes of significance. The post‐World War II shift from multi‐polarity to bi‐polarity placed Europe under the American security umbrella, changing the incentives for institutionalized cooperation. While we can all agree that the outcome of World War II made the treaties of the 1950s thinkable and therefore possible, nothing in neo‐realism is relevant to how supranational governance has developed subsequent to the treaties, or how Europe impacts upon state‐to‐state relations today. Indeed, one of the few straightforward propositions derivable from neo‐realism would have it that, with the decline of the Soviet threat and American hegemony in the West, Europe would unravel (Mearsheimer 1990). In fact, by any measure, integration has continuously deepened.

Of course, realists can always retreat from the more rigorous Waltzian position and simply invoke state power and concern for relative gains. One possible version might view supranational institutions and organizations as largely meaningless, at best glossing first‐order power considerations. The EU is redundant to the expert, who sees that the real action is in the calculation and pursuit of the national interest, and illusory to the non‐expert who is comforted by the rosy cooperative glow in a world that would otherwise expose only naked power. A second possible account emphasizes how international organizations both reflect and accentuate the power of states. Just as the Security Council of the United Nations assigned permanent membership to the victorious powers in World War II, so does the European Monetary Union (EMU) reflect the power and interests of Germany. In this view, institutions and organizations are not irrelevant, but neither do they constitute fair and impartial rules for all. Those who are powerful select the rules they consider to be most advantageous to them.

Our response is that institutionalists, too, must recognize that rule systems are never innocent of power, or of attempts by some groups to dominate others (see the introductory chapter to this volume). Rules are not always and everywhere in the interest of all. One bias of the economic literature on institutions is that the dominant focus is on those institutions associated with collective gain. But institutions are not limited to situations where collective action problems exist. They are just as present in circumstances where distributive concerns dominate, and in circumstances where some actors perceive joint gains through cooperative action while others have strong preferences for going it alone. There is nothing unusual about noticing the conflictual aspect of institutions, and focusing on the EU does not make conflict disappear.

(p.224) Of course, just how much institutions reinforce or mitigate power asymmetries among actors, or channel politics down certain kinds of paths while closing off others, are empirical questions. This volume is partly a response to these questions, and we think that realists are wrong. But these issues also raise tricky conceptual problems that we have not resolved. The neo‐rationalists in this project tend to define power and resources instrumentally, as finite commodities that actors deploy in conflict or cooperation with others. At a bare minimum, institutions constitute the rules of the game that serve to stabilize interactions among players, that is, actors conceived as bearers of interests. For others in the group, power is conceived more constructively as ‘a dynamic field that enables individuals to define themselves, existentially and in community with others. Normative systems constitute and animate these fields, and therefore also constitute and animate individuals as political actors’ (Stone Sweet 2000: 10). In any event, either view takes institutions more seriously than do realists.

Second, neo‐liberalism (Moravcsik 1999) and its variant, liberal intergovernmentalism (Moravcsik 1993; 1995; 1998), has proved to be a flexible framework for explaining episodes of intergovernmental bargaining and institutional design. But it has not been very good at identifying or explaining the flow, or the larger patterns, of integration across time. For Andrew Moravcsik, supranational institutions are the codified outcomes of discrete intergovernmental deliberations, which are treated partly as state‐to‐state bargaining over relative gains. He systematically downplays the extent to which European rule structures, and the spaces they organize, generate institutional innovation within the interstices of the grand bargains with which he is concerned. As with the European Integration and Supranational Governance volume, this project recognizes that the decision‐making of actors who comprise or represent Member‐State governments is often crucial to certain kinds of outcomes. But we also think that larger processes of integration construct the contexts for such decision‐making, and that a great deal of importance occurs around and between what governments do. Last, we reject the view that Member‐State governments control the pace and scope of integration in any definitive way.

Third, the present status of neo‐functionalism is quite ambiguous. If the seminal works of Ernst Haas are still ritually cited, they appear to be less read than invoked and dismissed. Few scholars are willing to bear the neo‐functionalist standard, and even fewer engage the actual arguments Haas and his followers made. As with the previous project, some but not all members of our group are quite comfortable being called (modified) neo‐functionalists, if labels we must have. The editors of both volumes, for example, think that Haas got most things right, and that the longer the time‐frame we adopt, the more prescient his ideas appear. Haas argued that economic interdependence and the growth of transnational society would push supranational organizations, like the Commission, to work creatively to facilitate further integration, at the same time raising the costs of intergovernmental inaction. As important, Haas's conceptualization of positive feedback, through loops of (p.225) institutionalization connecting actors, organizations, and institutions, was far ahead of his time. Critics (Moravcsik 1993) may rightly insist that neo‐functionalists never developed a rigorous theory of positive integration. This problem deserves discussion, not least because it cuts across the intergovernmentalist‐neo‐functionalist divide. We take it up again in the next section.

2. Political Integration as Institutionalization

The central problematic of the book is how European political space—‘supranational policy arenas or sites of governance, structured by EU rules, procedures, and the activities of the EU's organizations’ (introductory chapter, this volume)—emerges and evolves. Institutionalization constitutes the outcome to be explained, and it partially provides part of the explanation. In short, institutions are both exogenous and endogenous. This becomes particularly clear in the more ‘sociological’ accounts of institutional change where social structure and patterns of agency are more or less co‐constituted, or where the actors who have helped to build institutions are then induced to behave in ways that lead to further institutional change (McNamara, Fligstein and Stone Sweet, and Smith, this volume). But even in more economistic accounts, the analyst typically stipulates, for a given institution‐building moment, the relevant rules of the game (institutions) in order to focus on the interplay of interests and agency (strategy). When the game being played is itself, at least partly, an institutional‐design game, distinctions between sociological and rational‐choice perspectives on change necessarily weaken. Several chapters of this volume, especially those of Cichowski, Héritier, Le Galès, and Shapiro, demonstrate this point, if only implicitly

One of the book's core messages is that institutional design and change do not take place in an institutional void, or only through the sway of actors' preferences and material resources. A rendering of prior structural facts must always accompany an attempt to explain structural change. The economist's dream of devising an explanation of institutions based solely on relative preferences and relative scarcities must be abandoned from the start (Field 1981; 1984). Economic theory can never be complete—all variables included—or closed—no stochastic or error term—as presently formulated, since by definition institutions are outside the basic microeconomic model.2

Although the project achieved broad consensus on a wide range of concepts and first‐order principles, an eclectic mix of methodological approaches is (p.226) represented in the volume. Chapters draw from economics, political science, law, and sociology, seizing opportunities from comparative politics as well as international relations, and resolutely avoiding paradigm wars in favour of devising successful explanations of institutional change in specific areas. To some extent, this eclecticism is prompted by the very nature of the European Union (EU). William Wallace's (1983) much quoted phrase, ‘less than a federation, more than a regime’, instructs us as to the difficulties political leaders and scholars have experienced in categorizing the EU. Further, as EU studies have grown, so have opposed disciplinary claims. In political science, early studies—and many contemporary US scholars—have used tools provided by international relations theory, while Europeans have often relied on approaches rooted in policy analysis, public administration, and federalism (Jupille and Caporaso 1999). In law, European scholars have not interacted much with lawyers who focus on parts of national legal orders, even when their respective subject‐matters overlap. The European Court, in its jurisprudence on supremacy, has declared that EC law constitutes ‘an autonomous legal order’, and legal scholarship is still debating whether or not this order is more ‘international’ than ‘constitutional’: see the response of Weiler (1999: Ch. 9) to Schilling (1996).

This research project constitutes a denial that the EU's uniqueness precludes the use of more general theoretical materials to identify and explain the central features of the polity and of integration processes. Some may still find it useful to distinguish the ‘anarchic’ international system, where state power and interests are all that count, from domestic regimes, where rulers govern by law. Such characterizations are not only empirically wrong, they reify disciplinary boundaries, polarizing comparative and international politics, for example. Institutionalists typically bring common elements into the foreground. Contract enforcement is universally problematic, whether in the domestic or the international system, and the problem is basic to the social logic of courts everywhere. The binding nature of agreements, incentives to defect, the ability of parties to signal commitment, and the organizational capacity to implement decisions, monitor compliance, and use threats are variables not fixed parameters of a social setting. It should always be less important to know whether political phenomena occur within or among nation states than to know the institutional properties of the situation within which action occurs.

Of course, at some level of description, every polity is unique; but as we move on to higher levels of abstraction—the domain of theorizing—the system and its parts appear as specific instances of things familiar. This project not only collectively uses previously existing, relatively generic, institutionalist theory to guide its work; but contributors relate their findings to other approaches that are relevant to their respective chapters. The Brussels complex today possesses a system of interest‐group representation, for example, so the social science on how interests organize, and with what effects, is salient to Mazey and Richardson's account in this volume. The European legal system (p.227) may look new and special from some perspectives, but hypotheses derived from the study of other legal systems can generate a productive research programme on the Court and its impact on other policy‐makers (Shapiro, this volume; Stone Sweet and Brunell 1998a).

2.1. The Concept of Institutions

The project has sought to draw out some of the consequences of Douglass North's (1990: Ch. 1) earlier, influential formulation of institutions as ‘humanly devised constraints that shape human interaction’, which include ‘rules of the game’, ‘customs and traditions’, ‘conventions, codes of conduct, norms of behavior, [and] . . . law’ (see also March and Olsen 1989: Chs 1–3). In the introductory chapter to this volume, the authors begin by noting that any rule or rule system will vary along three dimensions, each of which is a defining characteristic of institutions more generally. Rules vary by degrees of precision, formality, and bindingness—also called authority.3

Rules provide guidelines for action, of more or less clarity and precision. The rules of traditional diplomatic practice fixed vague standards for dealing with particular conflicts, but provided more detail in terms of the role and treatment of ambassadors and diplomats. As with all modern European constitutions (see Stone Sweet 2000: 41–4), the EU's treaties are paradigmatic examples of what organizational economists refer to as ‘relational contracts’. Such agreements seek to frame complex, long‐term relationships, broadly rather than in precise detail, by (1) stipulating general purposes and objectives, and (2) establishing procedures and organizational capacity to ‘complete’ the contract over time (see Milgrom and Roberts 1992: 127). The Treaty of Rome announced certain goals—to guarantee the free movement of productive factors throughout EC territory, and to produce common European policies to regulate the common market—but said less about how to attain these goals. Indeed, when the Treaty entered into force, operational distinctions between what looked to be binding rules, programmatic statements, and policy recommendations were underdeveloped. Today it may seem obvious that ex‐Art. 119 EEC was specific and precise, and that women must be paid the same as men for equal work, and that maternity benefits constituted pay (Cichowski, this volume). But before the Court's announcements of the doctrines of direct effect and supremacy, Art. 119 EEC had little operational content. That Art. 119 EEC came to provide the basis for a policy of gender equality owes everything to developments that were unpredictable in 1960 or 1970. Just as important to (p.228) the history of the Community, parts of the Treaty have not been interpreted as requiring action on the part of the EC legislator and are, unsurprisingly, of little consequence. Ex‐Arts 2 EEC, 117 EEC, and 118 EEC, for example, set out particular social objectives, but the European Court of Justice has not interpreted them as legally necessary to achieve in order to create the common market (Ball 1996: 338–9).

However important, normative precision can not stand on its own. Some rules may be vague, but higher degrees of formality may enable or require actors to interpret them, authoritatively. It matters that the core rules governing the free movement of goods (ex‐Arts 30–6 EEC) are there to be found, as black letter law, in a formal treaty solemnly expressing the will of the contracting parties to be bound. Ex‐Art. 36 EEC provides that the Member States can hinder the flow of goods across borders through national regulatory instruments if a legitimate ‘public policy’ or ‘public morality’ argument can be mustered. In 1970, no one knew what would constitute a compelling ‘public policy’ exception to free‐movement rules. But traders, whose rights the Court located in ex‐Art. 30 EEC, litigated, provoking a long line of ECJ judgements that gradually clarified, by making more precise over time, the terms of ex‐Art. 36 EEC and its relationship to ex‐Art. 30 EEC (see Poiares Maduro 1998). The high degree of formality of a relatively vague rule undergirded the process through which greater degrees of precision have been achieved. The dynamic is well‐known to those who study litigation as public policy, and it features in several chapters of this book—those of Cichowski, Héritier, Le Galès, and Shapiro, for example.

Authority speaks to the issue of obligation. The more any rule is obligatory, or compulsory, the more likely threats and sanctions will come into play. Rules with high levels of precision, formality, and authority may be internalized by the actors, making threats and sanctions less pressing; that is, the rules are more likely to be self‐enforcing. Still, the more any rule possesses higher degrees of all three elements, the less difficult it becomes to monitor compliance. Further, organs with the ability to sanction in the name of a rule of law typically have more authority than organs that do not.

There are other dimensions that could have been singled out but were not, such as the coherence of an institutional construct—the degree to which different parts fit together. (In)coherence can be a crucial problem in the legal arena, since legal reasoning commonly values consistency, precedent, and normative hierarchies. It can create tensions among political entrepreneurs, as when the jurisdiction of the various Directorate Generals in the Commission overlap enough on a given issue to bring then into open conflict with one another. The EU is typically viewed as a relatively incoherent polity in institutional terms. The US Constitution propagates the vision of a Greek temple on the hill, with all of its structures balanced in harmonious symmetry; of course, the reality of American governance is far messier. On paper, the European treaty system, however constitutionalized, looks like the hoary Gothic cathedral, slapped together somewhat haphazardly by different architects, at different times, using different materials (for more of this imagery, see the lovely piece by De Witte (p.229) 1999). But just as normative imprecision begs for clarification, incoherence has led would‐be constitution‐makers—the Court, skilled social actors operating in Brussels, and academics (Ehlermann and Mény 2000; Weiler 1999: Part II)—to seek to enhance the aesthetics of Europe's institutional architecture. But this issue is not just a cosmetic one, as we will argue shortly.

2.2. The Impact of Institutions

The contributors to this volume quickly go beyond functional transaction cost and efficiency approaches to institutions. All of the contributors accept that institutions make collective governance possible and more efficient, but emphasize others roles and functions. Fligstein and Stone Sweet demonstrate that institutionalization both constitutes actors and shapes the integration process by forging causal linkages between the activities that take place in otherwise distinct supranational arenas. Since the book is about the emergence and evolution of policy spaces in Europe, some description of the process by which new arenas come into existence, the way older ones change, and how relatively autonomous spaces interact with each other would seem to be critical. The transition from a loosely coordinated balance‐of‐power system operating with bilateral measures and in an ad hoc way to a dense network of actors and rules in multiple issue areas turns our attention to the formation of issues and arenas. Mark Pollack (1994) has made an important attempt to generalize about how and why new issues and arenas emerged, from the point of view of the Commission. In this volume, Fligstein and Stone Sweet offer a rather different causal account, as well as comprehensive data on economic, interest group, legislative, and judicial activity across the life of the Rome Treaty.

A second way in which institutions have an impact is through their effect on the choice of strategies. Game theorists have demonstrated conclusively that, given a stable set of pre‐strategic preferences, altering the institutional context will alter behaviour by inducing a shift in strategy. Under prisoners' dilemma incentives, for example, the dominant strategy is defection, given an environment of anarchy, lack of effective communication, and inability to enforce agreements. But the more an institution allows communication and encourages binding commitments—such as an international organization that carries sanctions, or simply one that increases the value of reputation—the more it will shape preferences in a different direction. Notice that this is true even though, or just because, the preferences are derived and not fundamental. The desire of a country for a trade agreement or arms control agreement may not change at all, but the institutional context could allow for a shift in strategy. In one sense, this is the whole idea behind liberal theories of government, that is, that a proper specification of institutions enables individuals to do what they would want to do in any case. All of the chapters in this volume provide evidence that actors partly seek, in building or seeking to alter institutions, to induce certain kinds of outcomes by altering the rules of game (see especially the chapters of Cichowski, Héritier, and Le Galès).

(p.230) Third, institutions can change the presumptions of actors concerning the proper arenas within which to solve problems. Certainly Turnbull and Sandholtz, as well as Smith, are correct when they point out that immigration, policing, and national security were dealt with presumptively within the national political arena. The definition of crime, of criminal behaviour, of standards used in prosecuting criminals, had, and continue to have, a strong national stamp on them. Military security and foreign policy were, and continue to be, within the control of national decision‐makers and institutions. Nevertheless, in many important areas, including the two mentioned here, significant aspects of problem‐solving have migrated to the European level. This is true for the domains of agriculture, competition, trade, aspects of transport, and monetary policy, among other issue areas. While it cannot be denied that most progress has been made in regulatory areas—making laws that regulate or control, set standards, provide guidelines—rather than in the areas of stabilization or redistribution, such regulation is important in its own right. The EU should not be judged only with reference to an image of the state bent on producing the sinews of war, a strong centralized administration, and powerful tax‐collecting institutions. It is not a question of which level of government is more powerful, or who would win in a showdown over national sovereignty. The EU is a political construction that is being built on foundations which include the existing nation states; and the division of labour between the EU and these states reflects that elementary fact.

2.3. Feedback Effects

Approaches such as the one adopted by this project notice and emphasize feedback loops connecting actors to organizations, to institutions, and back again. These loops are conceived as generic mechanisms of rule innovation and institutionalization (March and Olsen 1989; North 1990; Stone Sweet and Sandholtz 1998). In this volume, a collective concern for positive feedback gives rise to themes that we believe merit summary discussion.

Despite intense debates about how best to explain or assess myriad aspects of European integration, few would deny that there is an overall direction to the EU's development. Competencies have expanded, new members have been added, new systems of rules and procedures have proliferated, controversial moments—and whole domains—of law‐making are increasingly judicialized, and so on. Why? The European polity that millions of people now take for granted as providing the basic social context for their activities was not preordained, and no other international organization has experienced the same evolution or impact.

One straightforward way to address this question is through the logic of path dependence (Arthur 1994; North 1990; Pierson 2000). While this term is too often diluted to describe any process affected by its own past, we refer here only to the path dependence generated by increasing returns, through network externalities and lock‐in. Since the lock‐in mechanism was discussed by (p.231) Paul Pierson (1998) and James Caporaso (1998) in European Integration and Supranational Governance, we will ignore it here and focus instead on increasing returns in the presence of network externalities.

Once an institutional infrastructure conducive to productive exchange is in place, cooperation obviously becomes easier. The marginal net benefits of cooperation rise, and the marginal costs of additional cooperative efforts will be smaller. The infrastructure in which EU Member States and so many other actors have invested now includes not only its legislative, executive, administrative, and judicial organizations, but a system of property rights and human rights, financial bodies, and a dense and elaborate corpus of hard and soft laws and procedures. Throughout the history of the EC/EU, the infrastructure in place has never been depleted with continued usage; on the contrary, it has been expanded and reinforced through use, more like the growth of muscle mass than the wasting of a finite natural resource such as oil. The chapters in this volume of Cichowski, Fligstein and Stone Sweet, and Shapiro, among others, make these points well.

Nearly all important studies of path dependence made by economic historians (see the reviews in David 1992 and Foray 1997) link increasing returns to positive network externalities through feedback. It is often the case in economics that the utility from consuming a product or service is enhanced as the number of others similarly consuming goes up, and that this networking effect depends critically on the adoption of common standards—of production, behaviour, and so on. This is true for users of a particular type of PC, or for producers of telecommunications equipment, or even for buyers and sellers of a particular type of car. Performances related to compatibility, repair, spare parts, and communication immediately raise issues of how many others are out there in the same situation. A similar logic exists inside institutions which are, after all, standards of behaviour. Institutions are important only in a social context. Variables such as trust, transparency, signalling and communication, credibility, and reputation are salient within institutional settings, although we would not imply that they are non‐existent outside of institutions. As a working hypothesis, we would argue that the more an institutional context is used, the more others will be encouraged to use it to, ceteris paribus. To the extent that one set of actors—say, Member State governments, or an association of interest groups—exhibits trust, transparency, and commitment within an arena, by adapting to a particular institution, others will be similarly encouraged; such dynamics are more akin to an assurance than a prisoners' dilemma game. Of course, institutions have to show that they facilitate the ongoing resolution of coordination problems, be it in trade, health and safety, or whatever. But often exploitable gains do not occur for purely institutional reasons, even when the objective situation allows profitable transactions. ‘Big bills’ are indeed left on the sidewalk.

In sum, actors within institutional settings benefit to the extent that rules, understandings, and standard operating procedures are shared. If EU actors, facing a crisis in the environment, have 15 different ways of responding, all (p.232) unrelated to the others, the exploitation of positive network externalities will not be possible. In this case, we would say that for all practical purposes the EU does not exist in some areas. As the early integration theorists understood, only if a subset of actors shares knowledge regarding definition of problem, expectations regarding other actors, and presumptions about how to deal with the problem within a common institution, are the beginnings of network externalities in place.

We want to stress that path dependence is not a magic wand. It does not logically imply that the EU will forever expand and continue to integrate. Such a prediction could be made only on the basis of observed behaviour, over time, in light of how variables have been specified and operationalized. Increasing returns can stabilize or reverse. Lock‐in can work so as to produce gridlock as well as to ratchet up gains. And network externalities can be negative as well as positive. Nothing we have said here necessarily implies an ever‐upward, ever‐onward view of the EU, whatever that can mean. Instead, given impressive progress in the EU, it is difficult not to entertain the notion that this progress might be profitably viewed through the lens of path‐dependence theory. And when we do, we see paths taking shape and pressing forward; we see increasing returns to European institutions; and we see that logics of networking infect more and more domains of action. Not everyone will agree with their conclusions, but the issues of reversibility and deinstitutionalization are address explicitly in this volume in the introductory chapter and in the chapter by Fligstein and Stone Sweet.

2.4. Institutional Coherence

While coherence is not treated as a definitional property of institutions, it is treated as a variable worthy of our attention. As we move away from a system based on the material power and interests towards one based on normative or ideational power and rules, the need to include consistency increases. Regarding a particular outcome—say, an environmental agreement—a country cannot simply say ‘this should be the outcome because we are more powerful and this is what we want’. Material power obviously still counts, but it works through and is constrained by rules. Interests are always present, but they are more effectively pursued through principled arguments and counterarguments. There is pressure to apply rules evenly, to look upon similar events as requiring similar responses, to think in terms of equity, precedent, and consistency. Shapiro's contribution to this volume shows just how much rules and delegated discretion are intimately linked. Indeed, this tension makes room for judges. As soon as discretion is granted, there is pressure to rein it in, to provide interpretations that limit discretion, to progressively specify the conditions under which such and such will be allowed. All contracts may be incomplete and disputes over the meaning of these contracts may help to fill in the lacunae, but ‘the relentless particularity of experience’ (Eckstein 1988: 795–6) always produces new political challenges that will renew the cycle.

(p.233) As critics have noted for some time (Kratochwil 1989; Stone 1994), international‐relations scholars have not much dealt with the idea of normative hierarchies, analogical reasoning, and normative consistency. In the international realm, to what would consistency apply? To power and interest? These things are expected to shift. The only kind of consistency expected has to do with a conformity of leaders' actions to underlying national interest. Now we are confronted with a world, a European one, that includes formal and informal systems of rule, doctrine, and precedent, and all of these things have to be taken seriously not because of their intrinsic interest but because they are important for understanding variation in outcomes. Shapiro identifies a process that requires EU administrators first to ‘give reasons’ for having generated a rule, then to ‘give good reasons’, and finally to ‘give better reasons for the rule than against it’. In this world, material power can be expressed only in the form of a rule‐governed, heavily judicialized, discursive politics. Outcomes of this process, of course, are no less political for being generated by normative deliberation. Thus, the idea of institutional coherence is important not only for doctrine, it also plays a role in the politics of institutional innovation and change.

2.5. Institutions and Society

Another important theme of this volume concerns the relations between institutions and society, both domestic and international. Political institutions possess some level of autonomy, at least in the sense of being abstract social constructs, but they are also embedded in society or implicated in concrete situations. Some institutions, such as those that sustain highly technocratic or expert modes of governance, may be less generally embedded, but some degree of linkage to ongoing decision‐making or problem‐solving is required. In this volume, we see this very clearly in the chapters on gender equality policy (Cichowski) and on policing and immigration (Turnbull and Sandholtz). These issue areas provide a strong demand‐side component to propel institutional change. Feminist organizations, labour unions, and pressure groups were all quite active in the judicial expansion of gender equality policy, a conclusion shared by everyone who has studied the issue (for example, Alter and Vargas 2000; Caporaso and Jupille 2001; Tesoka 1999). The Rome Treaty and the ECJ's expansive interpretations empowered domestic groups, conferred legal status on individuals, held out the prospect of judicial remedy in light of European law, and contributed to the mobilization of numerous social—mostly women's—groups. It is clear from these and other chapters that where institutions provide mechanisms for their own updating—say, through litigation—they do not simply reflect societal pressures or lock in commitments, they also provide potential for generating ongoing societal change through feedback. To the extent that such feedback loops exist, international institutions are never the analytical endpoint of a narrative that simply runs from demand to mobilization, to official decision‐making and public policy, to (p.234) institutional lock‐in. Instead, the creation and day‐to‐day operation of these institutions provide another convenient point to enter a different debate about the effects of international institutions on domestic politics.

In any case, the new partly comes out of the old through social interactions. This does not happen in a simple way. Institutions are not mindlessly self‐reproducing, but are responsive to changes in the environment. Given all the economic arguments one could amass in favour of monetary integration, for example, the resulting form and content of EMU are substantially different from what a team of experts would construct if they were starting from scratch. As McNamara argues in this volume, the EMU and European Central Bank emerged out of a consensus that crystallized in the 1970s. An institutional theory that depended solely on the economic data would not get the story right. McNamara uses a more sociological approach that is attentive to the broader features of the environment as well as to institutional legacies.

2.6. The Demand for, and Supply of, Institutions

Most functional theories of institutions stress the demand side. This is true of classical functionalism of Mitrany as of international regime theory (Keohane 1984). The early functionalists stressed the demands arising from society and the way that institutions responded to those demands. Institutional economists and the regime theorists in international relations stress the role of institutions in terms of locking in bargains, signalling commitments, and monitoring the implementation of agreements. One theme of this book is that institutions have a supply side and an independent impact too. Every chapter in the book is quite explicit about how extant frameworks structure subsequent institutional innovation. How they do so, and the extent to which the demand and supply sides are linked, deserve some attention.

A first issue concerns the project's focus on how processes of institution‐building achieve their logics over time. The book contains a wide range of important stories whose endpoints were never known by the relevant actors at any ex ante moment. Instead, the authors trace sequences and analyze patterns whose logics are typically unfolding at the same time that actors' strategies, modes of supranational governance, and institutions are also being formed or unmade. The implied assertion is that we miss a great deal of importance if we restrict our attention to what game theorists commonly model as ‘institutional design’ games, wherein relatively prescient actors, responding to a stable distribution of preferences, bargain among themselves to construct optimal rules to govern their collective future. The fact that, at any given point in time, relevant or potential rules can be hidden or unknown or desired—by some group or individuals, but not others—provides one of the logics of institutionalization the book identifies.

As already noted, many of the chapters explore what the introductory chapter calls endogenous processes of institutionalization, in which actors seek to (p.235) build new institutions through, among other things, (re)interpreting existing ones, clarifying vague ones, and exploiting institutional voids. Héritier's contribution offers an overview of these techniques, as they have developed over time. She begins with a primary institutional order, that given by the Treaty of Rome, but notes that the restrictive nature of some decision rules, particularly unanimity voting, can freeze the status quo. While this voting rule provides assurance that the Pareto principle will not be violated, one can also ask about the forgone welfare of those who want change. From a strictly utilitarian standpoint, there is little difference between a change that worsens someone's utility and a stasis that prevents an equivalent gain from occurring. Yet, for many Council of Ministers' decisions through the life of the EC, one vote was enough to prevent change. From the interplay of demand—the preferences of the members—and the existing supply of institutions, Héritier describes the evolution of informal procedures for making new rules or changing existing ones. Today, running alongside the extant treaty‐based rules of legislative procedure, we find ad hoc negotiations, litigation and adjudication, the propagation of soft law, and other informal norms within comitology, among others. Other chapters, of Cichowski, Shapiro, and Smith, take up some of these same mechanisms of institutional change, in light of the development of discrete arenas of governance.

A second issue concerns the interaction between positive and negative integration. Of course, there exists an important literature on this question (for example, Fritz Scharpf 1999: Ch. 2; Weiler 1999: Ch. 2). In it, negative integration—the process through which barriers to cross‐border activity within Europe are removed—is taken to be a set of activities that are more or less completely distinguishable from positive integration—the process through which common, supranational public policies are made and enforced. Further, it is commonly asserted that negative integration proceeds more smoothly and less painfully than does positive integration, since it enables Member State governments to reap large and diffuse joint gains; positive integration, in contrast, regularly pits these same governments against one another, to the extent that deciding on one form of regulation or intervention as opposed to another will have distributive consequences for identifiable national constituencies and given restrictive decision rules (Moravcsik 1993; Scharpf 1999).

Scholars have not done a very good job at actually assessing these arguments in light of the actual evidence, and we are somewhat sceptical of the claims made. Building on insights developed by the early neo‐functionalists, various contributors to European Integration and Supranational Governance argued that negative and positive integration are often causally linked to one another through feedback loops, and that these linkages regularly generate dynamics that serve to expand the domain and scope of supranational governance (for example, Sandholtz 1998; Stone Sweet and Caporaso 1998a). We will not repeat these arguments here, since positive integration is one of the topics of this volume. Indeed, one of the tasks assigned to each of the contributors was to explain why we observe so much institutional innovation in the EU despite (p.236) powerful forces favouring inertia and stalemate—for example, restrictive decision rules, opposed Member State government interests, and the complexity of a veto‐laden legislative process. The volume, we submit, provides good reasons to think that the distinctions between positive and negative integration are less absolute than is commonly thought, and that positive integration has proceeded much further than many of our theories have predicted.

Notes:

(1) Until recently, a comparative social science of integration did not exist. The pioneering study is Mattli (1999).

(2) The basic economic model is designed to explain allocative decisions, especially allocation of economic resources: land, labour, and capital. The right‐hand side of the explanatory equation includes preferences of actors, their resources, and the available technology. Economists such as North (1990) have sought to incorporate institutions into the basic model, allowing the system of rules to interact with other explanatory factors in determining allocation. This move remains controversial among many economists because, by endogenizing institutions, North denies that stable outcomes are thereby necessarily efficient, but rather could be the effect of embedded, but suboptimal, institutional arrangements. Nevertheless, for most economists, institutions are omitted or—what amounts to the same thing—treated as parameters, or constants, in the basic model.

(3) Some of these points had already been made in light of certain conceptual problems that have afflicted international regime theory (see Kratochwil 1989: Chs 1–2; Stone 1994). Stone (1994) elaborated a continuum on which the rules constituting various regime forms could be situated, with the EC occupying one extreme. The continuum captures three dimensions: degree of normative precision, degree of formality, and degree of organizational capacity to monitor compliance and punish non‐compliance. In a special issue of International Organization appearing just as this book is going to press, a research project on the legalization of international politics adopts, as an analytical or heuristic device, a continuum that largely reproduces these same elements (see Goldstein et al. 2000).