156    Theories of European Integration
Conclusions
To be fair, these deep dilemmas apply to any attempt to theorize European integration. However, they matter particularly to intergovernmental accounts because these perspectives tend to reassert the importance of the 'national' versus the 'supranational'. Notwithstanding the densely fascinating plethora of issues raised by the governance literature, the survival of the nation-state as the primary authoritative unit in European politics remains the most compelling question. One common observation is that European integration or the growth of European-level governance capacity has arisen because states have come under pressure from both above and below (Wallace, 1996a). From above there are external pressures, nowadays described by the idea of 'globalization', that constrain the state's autonomy and limit its capacity to act (Armingeon, 1997; Kurzer, 1997 and Chapter 7 below). Theories that retain the idea of states (or governments or national executives) as the primary actors in the EU system need to grapple with this problem. If the state is guided by external forces, then how constrained is its agency? If domestic imperatives force the state into pro-integrative bargains, how effective is the state as a gatekeeper between the domestic policy and the world 'outside'? Or, intriguingly, does the state's pursuit of autonomy from troublesome domestic forces lead it into the arena of integration? If this is so, are states conscious of the trade-offs that they make at these points? Do they perceive themselves recapturing sovereignty, albeit collectively? Or is this the starting point for the historical institutionalist discussion of path dependencies and unintended consequences? The fact that these questions are all raised by state-centric and intergovernmental analyses of the EU renders these approaches valuable in their own right. Engagement with these issues will make intergovernmentalists more reflexive and self-aware. Whether they can remain intergovernmentalists depends upon their ability to keep supplying answers to the regular anomalies thrown up by the EU.
Chapter 7
Europe and the World: Contemporary International Theory and European Integration
The previous chapters have indicated that there is considerable richness in contemporary theoretical analyses of European integration. An implication, often amplified by students of policy analysis, is that the EU is no longer a subject simply for students of 'International Relations'. The argument here is quite compelling. After all, the focus of 'International Relations', by definition, concerns interactions among states or at least between members of national executives. In terms of integration, the only questions it is capable of asking are those about whether there should be more or less of it, or those about the interaction of national interests in an intergovernmental arena. International Relations is an academic discourse dealing with the polarities of nation-states or superstates and envisages the state of 'integration' as lying somewhere along a continuum between those two poles (Caporaso and Keeler, 1995). IR, runs the argument, is particularly ill equipped to deal with the complexity of the contemporary EU game. It lacks the tools to deal with the coexistence of multiple actors playing nested games and whose interests are not simply bound up with the final destination of the integration project. The EU is rather more than an international organization; it has a mature internal politics or 'normal politics' to borrow Caporaso and Keeler's phrase (Caporaso and Keeler, 1995: 56). The EU is not provisional; it is well established and heavily institutionalized. It follows that many political practices are embedded in its everyday games. Such things, it is argued, cannot be captured by IR paradigms.
This chapter begins with a more detailed scrutiny of this type of argument, thinks about the ways in which it has been countered and questions whether an 'International Relations' versus 'comparative politics' dichotomy is a helpful way of thinking about the analysis of integration and the EU more generally. It suggests that this is a largely false and highly constraining way of contemplating the disciplinary possibilities for studying the EU and integration processes. Moreover, it is argued that the 'IR—CP debate' rests on a narrow and largely anachronistic view of 'International Relations' in general and international theory in particular.
157
158    Theories of European Integration
The critique of Internationa! Relations as a 'home discipline' for EU studies
What is the problem with International Relations as a vehicle for the study of the European Union? The most usual argument is that IR is at best partial and at worst downright misleading in its depiction of the processes which sustain and develop the European Union. The usual view of how IR . develops its picture of the EU is put with some clarity by Paul Pierson:
Despite significant internal disputes, the dominant paradigm in IR scholarship regards European integration as the practice of ordinary diplomacy under conditions creating unusual opportunities for providing collective goods through highly institutionalized exchange. From this 'intergovernmentalisť perspective, the EC is essentially a forum for interstate bargaining. Member-states remain the only important actors at the European level. Societal actors exert influence only through the domestic political structures of member-states. Policy making is made through negotiation among member-states or through carefully circumscribed delegations of authority. Whether relying on negotiation or delegation ... Chiefs of Government are at the heart of the EC, and each member-state seeks to maximise its own advantage.
(Pierson, 1996: 124)
The point here is that IR is a discipline capable only of contemplating the interactions between states - and in the ways associated with 'conventional' forms of diplomacy at that. This means that even some of the 'intergovernmentalisť positions discussed here in Chapter 6 would not be appropriate for the formal IR canon. This leads some writers to play their trump card and to declare IR as largely irrelevant to EU studies. One variant of this 'International Relations is moribund' argument has been put most forcefully by Simon Hix (1994). Hix argues that the tendency in the academic community has been to study the Communities as instances of either interstate cooperation or emergent supranationalism. Such approaches, he argue, share the presumption that there is one significant line of political cleavage operating in the EU: advocacy of or hostility to further supranational integration. However, at the same time, the EU has gradually acquired many of the qualities normally associated with national political systems. In other words, alternative lines of political conflict concerned with (in particular) distributional questions ('who gets what, when, how' to recall some classic political science vocabulary) have become important. Like all political systems, the EU generates questions of interest articulation, representation and intermediation. Hix's point is
Europe and the World    159
that political science has an established set of tools for unravelling these issues and that these can be systematically imported and adapted for the study of the EU. This is not just a matter of disciplinary preference; the EU has become a genuine polity. IR may still work in areas where sovereign statehood is retained in the EU system. But 'where decisions are taken which involve cross-cutting party-political and national interests, decision and coalition theories from comparative politics are likely to have higher explanatory value ... this will probably be true for most areas of EC politics' (Hix, 1994: 23; see also Richardson, 1996b). It is this politics within the EC system (as opposed to politics among the member-states) that Hix identifies as important. The repertoire of 'comparative politics' approaches offers key insights into many aspects of the EU system such as the operation of interests, the relative influence of different actors and the importance of formal and informal institutions (see also Hix, 1999).
There are several presumptions here, but two stand out. Firstly, it is argued that the Euro-polity questions have supplanted the 'integration issue' as the most important for investigation in EU studies. Secondly, International Relations as a disciplinary starting point is incapable of asking the sorts of question necessary to unravel the complexity of EU politics because it is a disciplinary discourse of interstate interaction and little else. This chapter focuses on these claims. The first is, to some extent, a matter of argument and relies on the view that the EU has become rather more than an interstate forum. Even if one accepts the multi-actor complexity of the EU, it is still possible to assert the dominance of the member-state governments and the primacy of intergovernmental decisions. Whether this involves an adherence to 'International Relations' as a disciplinary home domain is quite another matter. As Chapter 6 tried to indicate, there are several intergovernmental approaches and many of these set their sights on the sorts of political games described by Hix as important. What is at stake for Hix is an issue of explanation - essentially 'CP' explains the EU better than 'IR'. Yet this claim also depends largely on the sorts of question that are regarded as important and this, in turn, is a derivative of particular disciplinary or theoretical starting points. So, to be pedantic, it could be argued that Hix's objection to IR (as he sees it) could be reduced to his own preference for the theoretical models of comparative political science and the questions thereby generated. Of course, there is a little more to the objection than this. In particular, the Hix position also threatens to expel IR-inspired integration theory from the EU studies repertoire because the latter allegedly seeks to develop a general theory of regional integration from very particular European experiences.
160    Theories of European Integration
In response, Andrew Hurrell and Anand Menon (1996) have argued that Hix underestimates the importance of international theory. They accuse him of making an untenable separation between the processes of 'integration' (where IR has some utility) and EU politics (the important domain where the tools of comparative analysis should hold sway). Moreover, they assert the continuing centrality of member-states to EU politics and, by extension, see a continuing role for theories which deal with the mechanics of interstate exchange. Finally, Hurrell and Menon argue that there are compelling reasons to use IR theories because of the necessity of locating the EU and its internal politics within the international system. The EU is not hermetically sealed from the international system and it is, therefore, proper to consider the impact of the international system by leaving space in EU studies for theoretical perspectives that account for its nature.
This response carries with it all kinds of assumptions of its own. One is that interstate exchange is largely the same as intergovernmental bargaining. Yet the assumption of much realist-inspired International Relations is that states are unitary actors, whereas much intergovernmentalist analysis presents a much less homogeneous image of the state. Liberal intergo-vernmentalists place emphasis upon the permeability of the state from below as domestic groups shape the bargaining preferences of national governments (Moravcsik, 1993, 1998), while scholars of the intergovernmental institutions — by implication — develop a much more pluralistic account of the state with their depiction of segmentation in Council cultures and severe coordination problems for different branches of national governments (Edwards, 1996; Wright, 1996). The second brand of scholarship does not deny that state actors are important, but throws open the question of how we are to best construe 'state actors'. A second assumption resident in the Hurrell-Menon critique is that the international environment is exogenous to the EU game. This may well be a route to parsimonious theorizing, but to make the assertion that EU actors are stimulated by external imperatives or pressures opens up its own particular can of worms. The immediate question is how external threat perceptions are realized and, to all intents and purposes, this is a theoretical question that raises substantial questions of the relationship between 'inside' and 'outside' which are far from unproblematic.
Hix's response (Hix, 1996) is accommodating to the extent that it acknowledges the usefulness of keeping the IR channels open for the analysis of the external determinants of actions within the Euro-polity. He also recognizes the difficulties of drawing boundaries between International Relations and Comparative Politics as disciplines. Yet, he adds a provocative rider to suggest that comparative political science may be
Europe and the World    161
better at answering some of the important and on-going normative questions in EU studies. When it comes to thinking about how we increase the welfare of the citizens of the EU, Hix suggests:
We may learn more from the ideas of Madison, Dahl, Easton, Rokkan, Olson, Lijphart, Schmitter, Rose and Majone than either from the likes of Morgenthau, Haas, Hoffman [sic], Waltz, Keohane and Moravcsik, or from the likes of Lodge, Wallace, Wessels and Nugent.
(Hix, 1996: 804)
There may be some objections to assigning particular authors to specific categories, but Hix is trying to take the argument further, by differentiating his notion of comparative politics from the general school of EU studies. His point is not elaborated, but it seems to mean that EU studies should not just be accumulating a cache of information about the EU. Rather, it should receive a further conceptual injection from conventional political science, the message being that the EU raises issues of concern to pluralists, systems theorists, students of public policy-making, analysts of party systems, and scholars of political cleavage formation.
Another slant on the question is offered by Bernhard Ebbinghaus (1998) who points out that the IR and CP approaches differ because of their tendency to generate different kinds of research question. However, it is also the case that, in their conventional guises at least, they share the imagery of the nation-state. In conventional (realist) IR, the world is composed of self-regarding unitary states, whereas comparative political science has tended to engage in the comparison of states and their political systems as discrete entities. Ebbinghaus is concerned that neither convention is able to capture the dynamics of a changing world with increasing transnational interpenetration of polities and societies. His recommendation is the mobilization of the multi-level governance metaphor (see Chapter 5) which has the virtue of recognizing both the multi-actor complexity of contemporary politics and the fact that political action occurs at and between several tiers of action. His argument echoes that of James Caporaso (1996; see Chapter 6) who intimates that much EU scholarship is hooked up with traditional Westphalian conceptions of the state.
The 'IR versus CP' question to some extent revolves around the issue of what is the important stuff of EU studies. It is a matter of what scholars of the EU should be studying and how these phenomena are theorized. If the student of integration follows a writer like Majone and opts to treat the EU as an instance of a regulatory regime that can be compared fruitfully with other regulatory regimes (see Chapter 5), then it is unlikely that orthodox
162    Theories of European Integration
IR-derived theory will be of primary use. The student is much more likely to find theories of regulation the most useful route to explanation. However, the IR-CP question is rather more than a matter of choice. Most would argue that International Relations does not have the equipment to respond to the most compelling questions about the formal and semi-formal institutions of the EU and crucially cannot explain the emergence of policy outcomes, particularly of the day-to-day regulatory variety. Another point to bolster the CP case might be that the EU is unique, so in effect is nothing other than an instance of itself (and thereby incomparable with any other phenomenon in the contemporary international system). The only theoretically valid route from here would be to engage in longitudinal comparisons of different phases in the EU's history or perhaps to develop counterfactual methods (Anderson, 1995). Such thoughts open up the question of the time-frame of European integration. The overwhelming tendency has been to date the origins of the present EU to the period following the Second World War. However, others have traced European integration to processes of social and political mobilization dating back to the mid-nineteenth century (Klausen and Tilly, 1997). As with all comparative work, the longitudinal option must face the question: what to compare?
As we have also seen in Chapter 5 and with Hurrell and Menon's argument above, some analysts are prepared to give IR a role, albeit limited, in the generation of explanations about integration and EU/EC governance. For some, IR's role should be confined to the analysis of the systemic level - essentially the analysis of the external structural environment within which entities such as the EU evolve and operate (Peterson, 1995). Others see utility in thinking about integration as an international issue for particular stages in the EU policy process, notably those where interstate bargaining takes hold (say Treaty revisions). It is also evident that even in cases where intergovernmental bargaining is being analysed, there has been a serious attempt to wrestle analytical primacy away from traditional IR. This has been a particular project of institutionalists (Bulmer, 1994; Pierson, 1996; Pollack, 1996, 1997a).
Theorists, according to Rosenau and Durfee (1995) should be engaged in journeys up and down what they call the 'ladder of abstraction'. The question that should define all social enquiry and which must stand at the start of all theoretical work is: 'Of what is this an instance?' The fact that we are able to identify a particular set of events as, say, a military coup or as a revolution allows us to compare our set of events with other instances of coups d'etat or revolutions. This is also a crucial stage in theory building. In the case of revolutions, for example, writers such as Theda Skocpol (1979) have developed persuasive theories of revolution through exercises in comparative historical analysis. 'Of what is the EU an instance'
Europe and the World    163
is an interesting question with no shortage of potential answers. As suggested in the introduction to this book, the answer offered is likely to be crucial to the sort of theoretical analysis that emerges. Concept formation stands prior to investigation and further theory-building (Sartori, 1970). Advocates of the CP approach to the EU are offering the foundational proposition that the EU is an instance of a political system analogous to modern national polities (or at least that elements of the EU polity resemble significantly elements of the subject matter of 'conventional' political science). But that is not the only conceivable answer to this question. Indeed it is quite a stark assertion, given the tendency of some to treat the EU as something rather more complicated, provisional and atypical. For example, William Wallace and others have thought carefully about the extent to which the EU might be thought of as either a confederation or a regime (Wallace, 1983) and the erstwhile neofunction-alist Philippe Schmitter (1996b) has offered a considerable range of present and possible future characterizations of the EU. European integration first attracted the attention of scholars of International Relations because it appeared to be an early instance of post-national institution-building, or at least of quite intensive interstate cooperative activity. The fact that the European nation-state has not withered away in the way predicted by some functionalists and the fact that the EU is rather more complicated than a grand interstate bargain, does not in itself render the CP school's answer to the question triumphant. IR is quite capable of generating interesting research questions from newer bodies of theoretical work. It is possible then that there may be a multiplicity of answers to the question, 'Of what is the EU an instance?'
So what use is IR? It seems that to prove its utility international theory needs to be able to demonstrate its worth to EU studies in four ways:
1.    'IR' must transcend the 'would be polity'-nation-state continuum - or at least it must offer evidence that this polarity can be given new intellectual energy and can continue to pose meaningful questions about the EU
2.    'IR' must be able to pose interesting questions about the ways in which actors in institutionalized environments such as the EU derive their interests and thereby generate policy outcomes.
3.    'IR' must be able to pose alternative questions of the 'Of what is this an instance?' variety. The EU may be an instance of a polity and it may be an instance of interstate cooperation/bargaining/conflict, but these may not be the only alternatives.
4.    'IR' must show that it has the best tools to locate the EU within the dynamics of global change. This is an undeniable weakness of CP theories as they tend to treat exogenous input as largely given.
164    Theories of European Integration
International Relations: not just about states?
The remainder of this chapter offers an assessment of the extent to which IR theories might be able to do these things. Having said that, and as an important precursor to this discussion, it is also necessary contemplate the meaning of 'international theory'. After all, one accusation that might be levelled at advocates of the CP school is that they misunderstand the scope of International Relations as a discipline by caricaturing it as exclusively concerned with the interaction of states and/or governmental elites.
This is certainly one view of the substance of IR. However, as one recent introductory text notes, there are other views such as the idea that International Relations is about cross-border transactions or about the development of forms of world society - reflecting the fact that there is no real world essence that is 'international relations' (Brown, 1997). Any definition of what IR is 'the study of is likely to be partisan and exclusionary. It is true, as Brown notes, that many scholars within the discipline think of themselves as students of the interaction of states through diplomacy or violence within an overall context of structural anarchy (i.e. a situation without overarching authority on a world scale). Indeed, this is the starting point for much work, even that which selfconsciously seeks to reconstruct and reform the discipline. Martin Griffiths (1992) begins his recent re-elaboration of realist IR with the claim that the discipline conventionally revolves around two assumptions: that there is a fundamental distinction between domestic and international forms of governance and that explanations of the long-term patterns of state behaviour follow from this preliminary distinction. The claim here is that politics among states is quite different to politics within states. So, while writers in the idealist traditions of IR may be prepared to countenance the possibilities for collective security through interstate cooperation rather more than their realist counterparts, an IR purist would still argue that the domain of international politics cannot be viewed through the devices used by the political science of domestic polities.
But it is equally true that many others would dispute this basic image of IR, portraying themselves perhaps as observers of the sorts of global actors and post-national institutions that explicitly rival or threaten both the states system and the anarchic structural environment. Others prefer to think about the development of transnational social forces. Indeed, some scholars refuse to use the term 'international relations' on the grounds that the phrase presupposes an overly state-centric ontology (Rosenau, 1989; 1990). To use the term 'international relations' automatically privileges particular categories of actor in the mélée of political activity above the nation-state. So, to some extent, an ever-present question in IR is the issue of the basic units that constitute the international or global system and
Europe and the World    165
how these units relate to one another (Hoffmann, 1961). Such issues were at the core of the discipline's first great debate in the inter-war period where idealist conceptions of global society locked horns with an emergent hard-nosed proto-realism (Long and Wilson, 1995). What seems to be occurring in recent IR thinking is a re-opening of this particular 'Pandora's box' with the additional input of more critical theoretical sentiments (Cox and Sjolander, 1994; Smith, Booth and Zalewski, 1996). At the same time, the discipline of IR is being connected to some of the wider currents in social theory (Booth and Smith, 1995). Another way of posing the same issue is through so-called 'post-positivist' IR's depiction of the conventional terms of IR discourse as loaded in favour of Euro-centric or Westphalian assumptions about how the world operates. The paradox, of course, is that arguably the most telling challenges to these conventions is unravelling in the heartland of Westphalian Euro-centrism. All of this opens possibilities for the incorporation of non-state actors and non-state forms of authority into the IR analytical canon and at first sight the EU polity is swarming with such things.
It is certainly something of a straw man exercise to castigate IR and the theoretical work it generates as inappropriate tools for the analysis of European integration simply because the discipline is essentially about states and that the processes and practices of European integration and EU governance amount to rather more than interstate exchange. The primacy of particular actors is a matter for intellectual argument. Whether a focus on actors, their interests and preferences and their relations with other actors is enough is another matter entirely. Here the tendency to connect IR with broader currents in social theory has some importance. This is particularly true of those international theorists who have explored the agent—structure problem. This has been a central feature of constructivist International Relations and has direct applicability to the study of the EU and the questions that we ask about European integration (discussed later in this chapter).
It is also important to recognize that theoretical movements generated within the discipline of International Relations can have considerably wider applications. IR and its intellectual discourses are not separable from what goes on elsewhere in the social sciences. For instance, the resuscitation and re-invention of institutionalist work in political science, which (as Chapter 5 indicated) has been so influential in recent scholarship on the EU, was partially accomplished within the domain of comparative public policy. Yet it also owed something to developments in neoliberal institutionalist international relations, the growth of regime theory and the so-called neorealist/neoliberal synthesis (particularly Keohane, 1989; and see Chapter 6). Now, while much of the emphasis of institutionalist International Relations has been placed on the impact of institutions,
166    Theories of European Integration
norms and regimes upon state behaviour, the fact that particular actors tend to be given primacy does not obliterate the entire theoretical discourse. Similarly, constructivist IR has been especially concerned with rethinking state-centrism (see Wendt, 1992; 1994; 1995), but constructivist insights have also been applied to the rather more fluid and multi-actor Euro-polity (Jorgensen, 1997a, 1997b, 1997c; Rosamond, 1999).
The argument so far, therefore, is that International Relations deserves. more than summary dismissal. The rest of the chapter explores more fully what IR might be able to do for EU studies in terms of supplying useful theoretical concepts. The growth of regime theory and other thinking about cooperation in International Relations is the first port of call and this is followed by a discussion of the utility of constructivism as a theoretical perspective. This is followed by an exploration of how the EU's role in the global political economy can be given greater theoretical purchase and how the study of integration might usefully be attached to the current concerns with globalization and regionalization in world politics.
International cooperation, regimes, security communities and the EU
An on-going issue for scholars of international politics has been the question of why states cooperate and with what consequences? The problem for much IR has been to understand the rationale for cooperation. The neorealist answer is to suggest that the construction of international institutions may serve the mutual interests of the parties involved, or it might indicate the power of a hegemon to exercise its preferences. Either way, institutions in international politics tend to be regarded as both secondary to the real issues of power and inherently unstable. Institutions, like alliances, may dissolve once the interests of the involved parties or power balances alter. Liberals have traditionally seen the emergence of institutions, organizations and forms of cooperation as signals of a transformation from a state-bound world order towards a more peaceful settlement. More recently, and less idealistically perhaps, neoliberal institutionalists have tried to think about how to explain the undoubted existence of institutionalization of international politics in spite of (a) the persistence of a largely state-centred system (Keohane, 1989) or (b) the alleged decline of American hegemony in recent years (Keohane, 1984).
Regime theory
In this intellectual context, the development of regime theory has been one of the major currents in International Relations scholarship since the early
Europe and the World    167
1980s. The idea began to have some purchase in the 1970s, notably in the context of John Ruggie's attempt to develop ideas about the importance of cognitive factors in international politics. These included symbols and referents shared by actors, particularly where exogenous shocks challenged the utility of particular relationships (see Ruggie, 1975). Later regimes came to be less thought of as exclusively 'ideational' phenomena and much more as mechanisms for delivering the effective governance of international or transnational relations. This came to be a theme in the literature on interdependence (Keohane and Nye, 1977; see also Chapter 4) and scholars such as Keohane developed the idea as they laid the groundwork for contemporary liberal institutionalism (Keohane, 1989; Suhr, 1997). Regime theory draws on several strands of scholarship, from IR and elsewhere (Zacher, 1991). Neorealist ideas about rational action can be used to show that states pursuing security can be persuaded to construct regimes to safeguard their interests. These have combined with imported ideas from microeconomics that aim to explain how frameworks of regulation can induce shifts in the behaviour of actors. The English school of IR, associated with writers such as Hedley Bull (1977), can be mobilized to explore the emergence of norm-governed behaviour. Even functiona-lism's emphasis on sector-specific organizations is an intellectual ancestor. Regimes are usually thought of as established and acknowledged practices, as ongoing patterns of behaviour. The definition most often cited is that of Stephen Krasner:
Regimes can be defined as sets of implicit or explicit principles, norms, rules and decision-making procedures around which actors' expectations converge in a given area of international relations. Principles are beliefs of fact, causation and rectitude. Norms are standards of behaviour defined in terms of rights and obligations. Rules are specific prescriptions or proscriptions for action. Decision-making procedures are prevailing practices for making and implementing collective choice.
(Krasner, 1983: 2)
This has almost become a consensus definition (though see Milner, 1993). The concept is usually applied to systems of rules that oversee more specific agreements (Keohane, 1993), where rules are defined very broadly. The issue at stake in the regime literature is how regimes make a difference (Levy et al., 1995). Where the involved actors are treated as utility -maximizers, regime theory tends to argue that the creation of a regime will alter the cost-benefit calculus of action. Put another way, regimes are an intervening variable between the structure of the international system and the behaviour of actors (Krasner, 1983). Others might go further and argue that the output of regimes is to reconfigure the international system. They may reinforce the power of a hegemon or they may emasculate its capacity
168    Theories of European Integration
to exercise authority. Indeed, for some, regimes are the basis for the modification of the anarchic international system. Under anarchy, states make short-term calculations based upon self-interest. In a system infested with regimes the calculus becomes longer-term and actors begin to expect reciprocity from others (Levy et al., 1995: 287—90).
Can the EU be studied as an instance of a regime? This has been the subject of some discussion. For the most part, analysts have expressed scepticism about the fit between the EU and the definition offered by regime theorists. Most argue that the EU occupies a plateau well above that of an international regime. William "Wallace (1993) notes that the peculiar legal personality of the Community, with its capacity to override national law, is an important difference, while Keohane and Hoffmann (1991) point out that the level of institutionalization, the spread of policy competence and the centralization of authority are all denials of the efficacy of regime analysis. An alternative interpretation has been offered more recently by Robert Breckinridge (1997). His argument is that comparing the EU with regimes is a false comparison because, according to the regime literature, international organizations are embedded within regimes. The function of international organizations, he argues, is to monitor, manage and modify regimes. So:
The regime aspects of the EU have not ceased to exist, replaced by the confederation, as Wallace would argue. On the contrary, the economic confederation is embedded in the regime as the international organisation was and as any future federation or unitary state will be.
(Breckinridge, 1997: 180)
Therefore, Breckinridge's argument is that the EU is embedded within a regime and that regime theory can help to unravel the nature of the rules and patterned behaviour that constitute the regime. With that in mind, he draws attention to the practices and norms of the 'Community method' as exemplified by the long-term reproduction of bargaining methodologies and forms of package dealing within the Community institutions, along with the persistence of basic unwritten rules of membership. This, he argues, allows scholars to engage in direct comparisons involving the EU and other instances of regional blocs in the global political economy. The EU may be functionally and organizationally different from NAFTA, Mercusor and the like, but all of these formations have an associated regime underwriting their activities and development.
The objections to this line of thinking might include the claim that perspectives such as constructivism (see below) are better able to capture the normative basis of the EU, because, unlike regime analysis, they are not wedded to rationalistic accounts of utility-maximizing actors. They might also latch onto one of the classic critiques of regime theory which argues
Europe and the World    169
that the study of regimes is a distraction from the more important analysis of the forms of power that lie in the background (Strange, 1983). It is certainly possible to discern changes in what might be called the EU's governing regime as alternative modes of decision-making rise and fall (Wallace, 1996b), but it might be just as (if not more) plausible to examine these shifts in terms of alterations to the balance of forces within the EC system and in the global environment.
Security communities revisited
The study of norms and rules in International Relations has also re-ignited interest in the Deutschian idea of security communities (see Chapter 2). This has been attributed to the shifting agenda of International Relations following the end of the Cold War. According to one account, IR has produced a number of distinct perspectives that seek to account for the absence of war (perhaps, in so doing, turning the traditional problema-tique of IR on its head) (Adler and Barnett, 1998a). This accounts for the growth of perspectives such as regime theory and liberal institutionalism. Deutsch had defined security communities as groupings of states where war was no longer a tenable means of dispute settlement. As a result, the concept might appear to have obvious purchase in the contemporary world. It might have added spice for students of the EU because the investigations of Deutsch and his colleagues had spawned a distinctive theory of integration in transactionalism. Might there be some utility in taking a second look? The answer from Adler and Barnett (1998a) is a decisive yes, but they qualify their argument by pointing to certain deficiencies in Deutsch's original formulation. They note that the transactionalist account was heavily reliant on behavioural reasoning (see Chapter 3). As a consequence, Deutsch became preoccupied with the achievement of security communities through intersocietal transactions. Furthermore, he was convinced that these transactions could be measured and quantified. So, his attention was focused on measurable indices of communication such as international phone calls and the cross-border traffic of tourists. The renewal of interest in security communities should, they contend, be attentive to the sociological origins of transactions and to the processes of social learning and communicative action that produce mutual identification among actors. This emphasis puts Adler and Barnett's deliberations firmly within the constructivist camp (see below). They develop a three-tier model of the development of security communities (Adler and Barnett, 1998b: 38). The first tier concerns the identification of conditions that precipitate the emergence of security communities. These include broad environmental factors such as demographic or technological change, shifts in the global economy or
170    Theories of European Integration
alterations in the pattern of external threats. But precipitating conditions also include broad epistemic shifts, what Adler and Barnett call 'developments of new interpretations of social reality'. Thus, ideational change as well as material change is considered an important factor. A similar story holds at tier two: the emergence of factors conducive to the development of mutual trust and collective identity. Here, there are structural and process variables including power (structure) and transactions and organizations (process). But Adler and Barnett also include knowledge as a component of the structure and social learning as one of the key processes. By knowledge, they are referring to 'cognitive structures ... [or] shared meanings and understandings ... part of what constitutes and constrains state action is the knowledge that represents categories of practical action and legitimate activity' (Adler and Barnett, 1998b: 40). It is these things that are transmitted via processes of social learning and socialization. The third tier, necessary conditions of dependable expectations of peaceful change, consists of mutual trust and collective identity among involved actors.
The transactionalist account originally developed between the 1950s and mid-1960s was distinctive because its notion of 'integration' was rather different from that of rival theories such as neofunctionalism. This also meant that its notion of European integration was less associated with the European Communities than other perspectives, in part because the accomplishment of a security community among a group of states was not dependent upon supranational institutional expression. Deutsch's theories directed researchers into quite particular empirical directions. Adler and Barnett's approach requires the development of indicators to measure communication and collective identification, and it also opens another route to comparative possibilities. The question remains though: what contribution does the study of the EU offer the literature on security communities? One answer would be to say that the use of the security community concept opens up renewed possibilities for thinking otherwise about 'European integration', so that the study of integration need not be coincident with the analysis of the EU.
To some extent, this is inescapable. Any reasonable definition of the geography of European security communities (or the European security community) would have to include non-EU members. But it would be difficult to factor the EU out of the equation. However, what emerges is the use of the EU as an independent variable used to explain the absence of war (or the persistence of a European security community), rather than the conventional theoretical twist of EU studies which is, to all intents and purposes, to explain the EU. The post-war European instance of a security community is examined by Ole Waiver (1998) who argues that the achievement of a security community was accomplished through a process
Europe and the World    171
of 'desecuritization'. This consists of the emergence of other issues of mutual concern to European states taking precedence and guiding their interactions. Needless to say, this development owed much to the emergence and growth of the European Communities. Paradoxically, however, the deepening of formal European integration has brought 'security' back onto the agenda, most obviously with the formalization of foreign policy cooperation and the aspiration to create a Common Foreign and Security Policy (CFSP). This could become a threat to the stability of the security community because it 're-securitizes' the agenda, which, in turn, may be the source of insecurity. Wsever's analysis relies on the idea of security as rooted in discursive action. The connection of an issue to security concerns brings with it meanings and expectations:
Security discourse is characterized by dramatizing an issue as having absolute priority. Something is presented as an existential threat: if we do not tackle his, everything else will be irrelevant ... By labeling this a security issue, the actor has claimed a right to handle it with extraordinary means, to break the normal rules of the political game.
(Wžever, 1998: 80)
It follows that the framing of an issue in non-security terms allows the continued pursuit of 'normal' procedures. Not only will actors have alternative notions of what 'security' is, they may also have absorbed different ideas about the 'Europe' that is now threatened. Wasver's use of the security community literature suggests that the EU can be used to explain both the rise and possible decline of security communities and also adds interesting new dimensions to the study of integration. Many of these are bound up with the constructivist turn in International Relations.
Constructivism and European integration
Constructivism has become difficult to ignore in contemporary International Relations scholarship, certainly since the early 1990s, although a number of writers have been exploring constructivist themes, without using the heading, for many years (see, for example, Ruggie, 1998). It is bound up with the move towards greater meta-theoretical reflection upon international politics and the desire to interrogate established categories and concepts. It represents the connection of international theory with long-standing sociological concerns with the social construction of reality (Berger and Luckmann, 1967). Moreover, constructivism connects IR to some important strands in social theorizing. With so much to draw on, it is hardly surprising that rather than there being a single constructivist approach, there are many constructivisms. The disputes are quite profound
172    Theories of European Integration
and tend to revolve around the epistemological implications of ontological starting points - in effect, can we use conventional rationalistic research procedures to investigate a world that we regard as socially-constructed? (Matláry, 1997). That said, most constructivists working within contemporary International Relations agree that the structures of world politics are social rather than material (Checkel, 1998). This means that structural properties such as anarchy are not fixed and external to the interaction of states. Rather, anarchy is a social construct ("Wendt, 1992), something that is inter-subjectively understood by states and which is reproduced through their interaction. So, state behaviour does not just derive from the anarchic international environment; it also helps to make it. So constructivists
all agree that the structures of international politics are outcomes of social interactions, that states are not static subjects, but dynamic agents, that state identities are not given, but (re)constituted through complex, historical overlapping (often contradictory) practices — and therefore variable, unstable, constantly changing; that the distinction between domestic politics and international relations are tenuous.
(Knutsen, 1997: 281-2)
This places constructivism in a particular position in the debates about agency and structure, normally labelled structurationist (Duvall, Wendt and Muppidi, 1996; Hay, 1995; Wendt, 1987). Structurationists write about the complex relationships between structures and agents, so that neither structural determinism nor intentionalism are viable theoretical starting points. Agents are bound by structures, but they are also capable through action of altering the structural environment in which they operate, albeit in ways that may be structurally contained. This duality of structure and agency, first explored by Giddens (1984), has been summed up as follows:
The structurationist points of departure are the rules, norms and patterns of behaviour that govern social interaction. These are structures, which are on the one hand, subject to change if and when the practice of actors changes, but on the other hand structure political life as actors re-produce them in their every day actions.
(Christiansen and Jorgensen, 1999: 5)
Social interaction is the mechanism for the reproduction of structures. This means that constructivists object to the rationalism that characterizes the mainstream perspectives in international relations. This is because constructivists treat the interests and identities of actors as endogenous to interaction. Rationalists, including neorealists and liberal institutional-ists derive their accounts of actors' interests from an analysis of their
Europe and the World    173
material position. Institutions, such as those formal and informal environments provided by the EU, are arenas (often created by states) in which those interests are bargained. For institutionalists, this usually means that institutions facilitate the procedures of bargaining by providing atmospheres óf transparency and trust. Constructivists, on the other hand, treat interests as socially constructed - as derivatives of processes of social interaction. They also maintain that identities are socially constructed, that actors' accounts of self and other and of their operational context are also the products of interaction (Checkel, 1998; Wendt, 1992, 1994; Wind, 1997).
This begins to show how constructivist approaches might make a difference in the study of European (Union) governance. As Janne Haaland Matláry points out with some vigour, such a research programme poses some fairly fundamental challenges to mainstream contemporary theories of integration such as intergovernmentalism:
It is state-centred and hence privileges the state conceptually and even reifies it; it is centred on instrumental interests that are a priori assumed to be the most important ones; and it exogenises interest formation. By privileging one type of actor - the state - and one type of interest - that of money and/or power, it imposes categories on the empirical material that select and limit what can be found.
(Matláry, 1997: 206)
Having said that, Matláry argues that there are some issue areas (she mentions fishing quota disputes between the EU and Norway) that do not require constructivist methods to produce an explanation. In some cases, it might be possible to generate plausible and verifiable accounts of events using rationalistic methods. The key, she implies, is to practise theoretical reflexivity. That is to say that investigators should always be theoretically aware and conscious of the assumptions that underlie their arguments. Of course, it is also true that rationalism and constructivism are ontologically opposed. Constructivists would shift the research agenda of EU studies into the analysis of the role of ideas, the impact of shared beliefs, the effects of dominant discourses and the processes of communicative action (Risse-Kappen, 1996). In conventional rationalistic accounts, such factors are treated as epiphenomena - matter explained by other, more crucial factors.
The linkages between constructivist investigation and the fluid image of multi-level governance offers an obvious way in which the insights of a branch of International Relations can connect with policy work that deals with symbols, norms, understandings and belief systems (Christiansen, 1997b; Radaelli, 1999; Richardson, 1996b: Risse-Kappen, 1996). Thomas Christiansen argues that the emerging reality of post-territorial, multi-level
174    Theories of European Integration
governance in Europe cannot be captured by conventional concepts. It is important, he argues, to depict the fluidity of the system that is structured through institutions, the economy and identity: and whose agents operate at and between several levels of action. Yet it is also imperative to understand how 'Europe' comes to be understood as a legitimate space for political action:
A constructivist epistemology ... must conceive of territorial units on all levels as social constructs ... view the political significance of [these] in the processes for which they provide containers, and such research must address the agency/structure problem, meaning that no level in the studied process must, ex ante, be assumed to be primary.
(Christiansen, 1997b: 54)
Another potential way of using constructivist work is through the development of more critically attuned notions of intergovernmentalism. In many ways, the central achievement of constructivist International Relations has been the substantial re-evaluation of state-centred concepts (Wendt, 1992, 1994). In the EU context, writers have begun to argue that processes of intergovernmental bargaining can be better captured through a non-rationalistic frame. Knud-Erik Jorgensen (1997c) has argued the depiction of the EU's Common Foreign and Security Policy (CFSP) as 'intergovernmental' may relate the formal institutional reality, but it cannot capture the emerging norms and rules of the game, in short the governance regime of CFSP. Christiansen and Jergensen (1999) look at one of the principal focal points of liberal intergovernmentalists: the processes of Treaty reform. They argue that conventional models are inherently actor-centred in that they read treaty reforms as the product of bargains based upon the negotiations of actor's (exogenously derived) interests. This has the twin failing of losing sight of the structural environment in which the bargains take place and aggregating and unifying actors into implausible collectivities such as 'the state'. The actors in the process of treaty negotiations and the preceding Intergovernmental Conferences (IGCs) are diverse 'civil servants, Commission officials, MEPs, national ministers and Prime Ministers, rather than personified states' (1999: 3). The structure includes the established formalities and routine practices of IGCs and the mosaic of path-dependent institutionalization, which defines and contains the preferential possibilities of the involved actors. They also introduce the structural component of IGC discourses, a set of signifying themes within which the whole process occurs. Such approaches allow for the understanding of integration as an international process, but within a framework that is able to capture the institutional complexity and maturity of the EU.
Europe and the World    175
The EU and international 'actorness'
'International Relations' is often assigned importance because of a need to understand the environment within which the restructuring of European governance is taking place. But it also has to be remembered that the EU has a presence on the global stage and that it can be construed as an actor in the global system in its own right. This 'external' presence is manifested in four broad ways: trade policy, development cooperation, foreign policy and interregional dialogue. Foreign policy has a treaty basis (the Common Foreign and Security Policy occupies title V of the Treaty on European Union), but is conducted through heavily intergovernmental decision rules that are quite distinct from the possibilities for multiple actor interaction found within 'pillar one' of the treaty (European Community). On occasions this yields joint positions among the member-states, but it remains difficult to make the case for construing the EU as a unitary foreign policy actor. Probably the most important manifestation of the EU's external presence is its role in trade issues. Because the EU is a customs union with a common commercial policy and a common external tariff, it has long been clear that the Commission should conduct trade relations with outside parties through agreements conducted under the rubric of the General Agreement on Tariffs and Trade (GATT) and, more recently, the World Trade Organization (WTO). The Commission also negotiates economic cooperation agreements that establish privileged relations with third countries and association agreements creating particular sets of reciprocal rights and obligations. In all of these circumstances the EU acts as a bloc and negotiates with single states, with blocs of states and within international agreements such as GATT. Relations with other regional groupings remain embryonic, but processes such as the biennial Asia—Europe Meetings (ASEM) certainly operate with the presumption that 'Europe' has some kind of distinctive presence (Camroux and Lechervy, 1996).
Viewing these elements of EU activity suggests that the governance turn and its associated public policy perspectives do not have the tools to achieve explanation or understanding in this domain. The most obvious issue is that the appearance of actor-like entities such as the EU challenges traditional state-centred images of the 'international system'. However, the literature on EU external relations has been conspicuously cautious about possible transformations of the international system. Instead, attention is focused upon the EU's capabilities as an international actor and, by extension, to the capacity of the EU to mimic the features of a nation-state within the international system (Allen and Smith, 1990; Hill, 1994; Sjostedt, 1977; Smith, 1996; Whitman, 1997). The usual conclusion of such deliberations is to recognize that the EU is not a state, or at least that its
176    Theories of European Integration
external attributes do not resemble those of a state as normally understood. As Martin Holland remarks,
the notion of an international actor is wedded, at least historically, to the concept of the nation, sovereignty and the broad tenets of realpolitik ... the mismatch between the language of international affairs and the institutional and procedural realities of the EC has created an oasis for theoretical dispute and occasional obfuscation.
(Holland, 1996b)
As Taylor (1982) notes, even while there are discernible external 'products' of the EU, they do not arise from a unified policy-making process (that which would be expected from a state), but via a form of loose intergovernmentalism.
Here, of course, there is synergy with much of the new governance literature's claim that the EU cannot be placed on a continuum running from loose intergovernmentalism to 'superstate'. The multi-level governance metaphor is an emphatic attempt to break with the rigid imagery associated with such a continuum (Marks, Scharpf et al., 1996; Marks, Hooghe and Blank, 1996; Richardson, 1996a). Indeed, the external relations issues are not really amenable to exploration through what is usually thought to be the traditional IR problematique on the EU (more or less integration?/nation-state or superstate?). But the question remains: if it is not a state, what is the EU in the global system? This is, of course, a fundamentally theoretical question.
Christopher Hill's discussion of foreign policy cooperation in the EU (Hill, 1994) attempts to conceptualize the EU's world role through a discussion of 'actorness' and 'presence'. 'Actorness' is about (a) the delimitation of one unit from others, (b) the autonomy of a unit to make its own laws and (c) the possession of various structural prerequisites for action at the international level (including a legal personality, a set of diplomatic agents and the capability to conduct negotiations with third parties). Presence is about 'the reality of a cohesive European impact upon international relations despite the messy way in which it is produced' (Hill, 1994: 107). Hill chooses to represent the Community as a subsystem of the international system that generates international relations 'collectively, individually, economically and politically'. This is rather distinct from an entity - like a normal state - that produces a foreign policy. Richard Whitman (1997) seeks to take the debate further by actually mapping the capabilities of the EU in a variety of manifestations that reveal the structural facts of the EU's actorness and presence. For Whitman, these various exhibits (which include the formation of rationales for external action, the existence of established procedures for developing relations with the outside world, the issuing of communiques of external relevance,
Europe and the World    177
the provision of forms of assistance to the outside world and the placement of EU representatives in other parts of the world) represent steps towards the fulfilment of an international identity for the EU. This theme is taken up by Dave Allen and Mike Smith's discussion of the EU's 'structured presence' in the international arena (Allen and Smith, 1990; Smith, 1996). They make the point that this presence has two elements. The first is that the EU exhibits distinctive forms of external behaviour. The second is that the EU is perceived to be important by other actors within the global system. So, actorness is not only about the objective existence of dimensions of external presence, but also about the subjective aspects embodied in the validation of a collective self by significant others.
These considerations all have important theoretical implications and, while serious theoretical analysis is still very much in its infancy, Knud-Erik Jorgensen (1992) has gone some way to laying out clearly the theoretical possibilities. At one level, of course, these external dimensions of European integration defy conventional classification. In its global activities the EU would seem to be in the « = 1 category. Jorgensen's solution is to pay attention to the deeper questions of epistemology and ontology. He presents a typology of potential theoretical treatments of the external dimensions of the EU, which is illustrated in Table 7.1.
Alternative ontological positions are represented by the choice of agency versus structure, whereas the broad epistemological alternatives are presented by the interpretative-objectivist distinction. Different categories of theory can be located in each box. So, agency-objectivistic theories ask about the motivations of actors and, by posing such questions, tend to seek rationalizations of actor behaviour. This in turn feeds into a preoccupation with understanding the linkages between interests and action, where behaviour is understood in terms of exogenously defined interests. As a starting point, this runs into some difficulty because the identification of these interests presupposes - to a degree at least - unitary actors. It might be possible to assert that the EU is a customs union and therefore, that it possesses definite and definable interests in external trade matters. But the obvious  empirical objection  is that the positions held by the EU in
Table 7.1    Possible ontological and epistemological positions on the EU's external role
Interpretative                    Objectivistic
Agency Structure
Source: Jergensen, 1996: 6.
178    Theories of European Integration
international trade matters are always the consequence of delicate negotiations and that different member-states can have radically different positions on matters associated with trade. For example, certain member-states might be happy to open their markets to the import of cheap Japanese automobiles; whereas others with threatened indigenous car manufacturers may be more inclined to support measures to constrain such incoming trade. The argument boils down to a level of analysis issue. It is indeed possible to read the EU as an agent with articulated preferences, but the basis of those preferences may best be understood as the product of the interaction of interests below EU level and not in terms of the generation of specific 'EU interests'.
Structural objectivism operates at the systemic level of analysis (rather than focusing on the interests and behaviour of identifiable agents). The point here is to treat the international system as a 'deep structure' within which activity occurs and by which it is contextualized. So, the EU acts externally because of imperatives set by, amongst other things, global markets and shifts in the global security structure. Such a 'structure-primitive' account still requires the identification of the EU as an agent. But because agents are seen as being constituted by structural imperatives, this may have greater theoretical purchase. After all, the galvanization of market integration in Europe from the mid-1980s is often seen as the consequence of changes in the logic of global markets (Sandholtz and Zysman, 1989) and, by extension, the motivation for the development of external activities such as ASEM could also be read in these terms. (Specifically, ASEM might be seen as a way of responding to (a) the competitive challenges posed by East Asian economies and (b) a response to the development of various forms of Asia-Pacific regionalism). The other advantage of the structuralist position is that it does not necessarily require the privileging of particular actors. Structural imperatives can be seen as impacting upon the interests of states, firms and supranational bureaucracies. The key is how these imperatives interact with the interests of these actors and how actors work in concert or in opposition to one another thereafter.
One major objection to structural objectivism, however, is the criticism supplied by constructivists about the treatment of interests as exogenous givens rather than endogenous to the processes of interaction. The key research question in terms of the EU is how is its 'actorness' constituted? The problem with rigidly structuralist accounts is that they presuppose the existence of the actor. Interpretative approaches may be better at identifying how agents or structures come to be constituted. Yet, as Jorgensen argues, this is best accomplished through the development of a structura-tionist dissolution of the structure-agency dichotomy. Investigation from this premise would think about how the global structural environment
Europe and the World    179
contributes to the collectivization of an EU identity, but also about how actors within the EU define the global structural environment so as to create the rationale for a cohesive EU identity (Rosamond, 1999). It would also investigate the extent to which the development of EU cohesion in external affairs related to the perceptions of other (external) actors about the EU's actorness.
One of the central problems with thinking about the EU as an international actor is that the whole idea of actorness in the global political economy has been thrown into confusion in recent years by the extensive debates about globalization. It is to these that we now turn.
Globalization, regionalization and European integration
The phenomena of globalization and regionalization have attracted huge attention in recent years in the literatures of International Relations and International Political Economy (IPE). Both are posed as challenges to the traditionally state-centric Westphalian order. Globalization is said to reduce the capacity of the state to govern effectively in key policy areas, while the appearance of regional orders might suggest that new sites of authority are emerging. Neither term begs easy definition. Regionalism, or the processes of regionalization, has been used since the late 1980s to refer to the appearance and consolidation of various economic arrangements among groups of geographically proximate countries. These include the North American Free Trade Agreement (NAFTA), Mecusor in South America, Asia-Pacific Economic Cooperation (APEC), Southern African Development Cooperation (SADC), along with the relatively venerable Association of South East Asian Nations (ASEAN). The revitalization of the EC/EU following the single market initiative of the late 1980s usually brings European experiences into wider discussions of regionalism (Coleman and Underhill, 1998; Fawcett and Hurrell, 1997; Mansfield and Milner, 1997; Mattli, 1999). The appearance of regional blocs or trading areas is a phenomenon that demands the attention of international theorists (Hurrell, 1995).
The same is true of globalization, a yet more slippery and elusive concept. Globalization is usually used to describe the breakdown of discrete economic spaces (economies). Moreover, it is often said to herald the consequent loss of executive capacity by territorially bound national governments, or at least radical residualization of the state in economic governance (Cerny, 1995, 1996; Strange, 1996). Globalization, in this sense, is bound up with the liberalization of global finance and the rapid rise of instant transborder dealings in financial commodities. It is associated with the multi- and transnationalization of production activities and the growth
180    Theories of European Integration
of global trade. Globalization theorists portray a world where the conventions of territorial space have been rendered meaningless by the proliferation of transborder economic activity (Schölte, 1996). For some, globalization is a recent phenomenon occasioned by a mixture of technological innovation and progressive market liberalization, particularly since the early 1970s. For others, it is a rather longer-run phenomenon, bound up with earlier technological innovations and/or the unfolding logic of capitalism as a world economic system.
Needless to say, the debates about these phenomena - particularly globalization - are incredibly wide-ranging. Considerable doubt has been cast on whether the image of globalization is a valid depiction of the changing economic order. For some talk of globalization is wild hyperbole. The consequence is that political agency, and that of the state in particular, is rather less constrained than might be supposed (for a flavour of the debate see inter alia Boyer and Drache, 1996; Held et al., 1999; Hirst and Thompson, 1996; Kofman and Youngs, 1996; Strange, 1996). But this also raises questions of the relationship between agency and the processes of globalization. Is globalization sponsored by particular actors? Or is it best conceived as an external process that impacts upon actors' interests and thereby affects their preferences and the possibilities they enjoy for strategically motivated behaviour? If globalization is conceived as the world-wide spread of neoliberal ideas and economic practices, then the authorship of globalization might be associated with the possession of power. Once again, issues both of structure and agency and of the relationship between economics and politics are predominant.
It is impossible to do justice to the full range of debate here. The questions most pertinent for the purposes of this chapter seem to be the following:
1.    What is the relationship between globalization and regionalization?
2.    How does the EU relate to both of these phenomena?
The first question tends to beg the second. It is often supposed that the wave of regionalizing activity that began in the mid-1980s was occasioned by the challenge of globalization. So, to include recent European integration in the equation presupposes that the activities of the EU can be treated as an instance of regionalization. If regionalization is defined as the consolidation and formalization of economic integration among a group of geographically proximate economies, then the EU fits the pattern, although the stage of economic integration reached could be said to be more advanced than other counterparts. By the end of the 1990s, the EU was well on the road to the achievement of a common market and was embarking on the first stages of monetary integration, while the likes of
Europe and the World    181
NAFTA and APEC were still only aspirant free trade areas (Dent, 1997). Also, the EU exhibited altogether more institutionalization and patterns of supranational governance than other regionalisms. These may be significant divergences, but it is still possible to argue that globalization has been a significant stimulus to both integrative activity and the widespread questioning of established patterns of governance.
In international economics and much IPE, the norm is to think about regionalism in terms of its effects on trade. Regions are seen either as inward-looking fortresses that impede the progress to a global free-trading order or as accelerators of free trade. On the latter option, by promoting intra-regional liberalization, regional orders can be seen as stepping-stones to the globalization of the precepts of liberal trade. The key is whether regions remain open or closed to the outside and whether they can be shown to create new trade rather than simply diverting international transactions (Anderson and Blackhurst, 1993; Cable and Henderson, 1994). This raises the broader question of whether regionalization is merely a territorially-bound form of globalization - that, to all intents and purposes, regionalization and globalization are manifestations of the same process of economic integration and heightened interdependence. In this way regions might be read as globalizing entities. On the other hand, regions could be construed as loci of resistance to globalization, as protective blocs erected by economic actors confronted with the 'chill winds' of globalizing imperatives. Here, the very existence of regions, as well as the tendency in some instances to pool elements of authority, could be seen as the attempt of actors (most obviously states) to recapture executive capacity that has been lost domestically. This feeds into a further set of issues concerned with the origins of regionalizing tendencies. Are regions deliberate and controlled acts of institutional design by policy actors or do the political decisions that institutionalize regionalism represent a kind of catch-up by policy-makers? This debate taps into the distinction made by William Wallace (1990) between formal and informal integration and by others between de facto and de jure regionalization (Bressand and Nicolaides, 1990; Higgott, 1997) (see also Chapter 1). Placing emphasis upon formal integration (cooperative agreements, treaties, acts of collective legislation) suggests that regionalization is guided by the deliberate agency of identifiable actors (usually states). This then creates the space for non-state, private economic actors to engage in regionalizing activity. For example, the creation of the Organization for European Economic Co-operation (OEEC) and the European Payments Union (EPU) in the late 1940s are usually seen as the stimulus for a substantial growth in intra-European trade (Tsoukalis, 1997: 9—10). On the other hand, the growth of cross-border economic activity in terms of production,  investment and trade could be taken as the stimulus for
182    Theories of European Integration
political decisions to institutionalize or gain control over the regionalizing process. Approaches to this problem come in different theoretical guises, but essentially the issue boils down to which actors analysts choose to nominate and the capacities they are willing to ascribe to political agency. One argument often used is that globalization engenders outward policy orientations for states and that this often manifests itself through the emergence of regional accords. The impact of globalization may also manifest itself via the 'internationalization' of the state as particular ministries become closely connected to or find themselves at the behest of the forces of global finance. Such arguments found in the IPE literature often portray states as involved in three-level games (domestic politics, regional cooperation  and global engagement)  and generate research questions about the capacities for state autonomy in this context (Coleman and Underhill,  1998).  These sometimes generate the counter-intuitive proposition that globalization actually strengthens the autonomy of the state,  by reorienting its  imperative activity away from traditionally troublesome domestic constituencies (which in turn become marginalized). One way in which states pursue this autonomy is via the institutionalization of cooperative activity and the cooperative sponsorship of formal economic integration. The point here is that an IPE frame of reference offers a  coherent set of research questions, which link the internal dynamics of the EU, the domestic politics of member-states and the processes of global change. An exercise of this sort is carried out by Helen Milner (1998) who draws on rational choice and two-level game assumptions to compare the processes behind the Treaty on European Union and the North American Free Trade Agreement (NAFTA). She argues that national leaders are likely to seek international agreements in circumstances where their economies are interdependent and where domestic conditions allow. Thus, the processes of financial liberalization in key EC member-states created imperatives for the governments to seek international cooperation. Financial liberalization was the product of particular domestic imperatives, but its execution meant that governments lost effective control over key areas of economic governance. But the successful accomplishment of the international negotiations could only be achieved if national leaders had the support of key domestic constituencies - hence the multiple protocols and opt-out clauses in the TEU.
Many studies cite the importance of global economic change as a crucial determinant of the advances in formal economic integration. What might be called 'globalization', competitive threat', 'informal integration' or 'global economic change' often appears in the literature as a form of external 'regulatory shock' that forces a policy response from within the EU. These changes 'outside' impact upon the interests of key policy actors and are said to have accelerated the momentum towards the liberalization
Europe and the World    183
of the European economy and galvanized the deeper institutionalization of governance functions at the European level. This breaks down into a number of more specific accounts. One is that global economic change — especially technological changes — affected the preferences of key sections of European business and opened a strategic opportunity for the development of alliances with the Commission, which in turn generated the momentum for the single market programme in the mid to late-1980s. This combined the de-fragmentation and liberalization of the European economy with the enhancement of supranational regulatory capacity (Sandholtz and Zysman, 1989). More state-centred accounts observe the impact of global shifts upon either the executive capacities of (member-state) national governments (Armingeon, 1997) or the interests of domestic constituencies that are central to the formulation of governmental preferences (Moravcsik, 1993a). Another common line of argument is that the Europeanization of economic governance arose as national models of economic policy-making manifestly failed to cope with the pressures of globalization, a perception aided and abetted by the growth of informal transnational networks within high technology sectors (Bressand, 1990). A way of combining some of these insights is to see European integration, and especially that since the mid-1980s, in terms of the dilemmas of nation-states emasculated on the one hand by the forces of globalization and overloaded by demands from the domestic arena on the other:
European integration can be seen as a distinct west European effort to contain the consequences of globalisation. Rather than be forced to choose between the national polity for developing policies and the relative anarchy of the globe, west Europeans invented a form of regional governance with polity-like features to extend the state and to broaden the boundary between themselves and the rest of the world.
(Wallace, 1996a: 16)
This proposition is attentive to the specificities of the European experience, but also recognizes the advantage of treating the EU as an instance of region-building in the context of globalization.
These studies tend to make a sharp separation between the EU and the outside. Globalization, therefore, becomes a set of dynamics generated outside the EU that impact upon the inside. This sharp analytical separation works best if integration is viewed as a response or a form of resistance to globalization. It works less well if European integration is treated as a facilitator of globalization. If the EU is a globalizing project, then analysts clearly need to think about how the linkage between the European and the global is accomplished (Rosamond, 1995). Work here remains in its infancy, but a number of possibilities are being canvassed. One method is to reject the image of the EU as a unitary entity and note
184    Theories of European Integration
how different components of the EU polity and different actors within the EU polity have distinct relationships with globalization. This would amount to looking for ways to combine the insights of the branches of globalization theory emphasizing diversification rather than homogeniza-tion, with the powerful metaphor of multi-level governance (Ebbinghaus, 1998: Rosamond, 1999). Constructivist insights might also be used to think about how knowledge about globalization is socially constructed within the EU polity and how this in turn might promote particular policy possibilities while downgrading others. This could then be attached to a wider set of theoretical propositions about how intersubjective understandings of external context influence policy choice (Rosamond, 1999).
One significant recent attempt to give European regionalism a comparative and historical gloss is provided, from a rather different vantage point, by Walter Mattli (1999). Defining integration as 'the process of internalising externalities that cross borders within a group of countries' (1999: 199), Mattli argues that the demand for regional rules arises amidst a variety of circumstances. These include the potential for gains arising out of increased cross-border transaction which in turn grows as technology improves. That said, greater cross-border economic activity heightens uncertainty for market actors and imposes - in the parlance of economics - 'externalities' upon them. Therefore, like Stone Sweet and Sandholtz (1997), Mattli's argument is that the impetus for integration is rooted amongst market actors. Where these demands for integration are lacking, then regional integration schemes are unlikely to prosper. Mattli's argument also insists that authoritative actors need to have some rationale for integration to proceed. Additionally, the presence of a regional hegemon to lead, coordinate and broker the activity of other states is important (Mattli, 1999: 50-7). Mattli's contribution brings together several strands of literature in political economy, though the real added value of his contribution lies in his thoughts about the external effects of regional integration (Mattli, 1999: 59-64). Actors excluded from the initial delineation of the region are likely to lose market access and, therefore, to undertake compensatory action. This has two historical variants. The first is to seek to join the integration scheme, an option. The second is to set up a rival regional organization (which itself must satisfy the demand and supply conditions).
Mattli's discussion shows how far rationalistic perspectives can take us in the search for a general theory of regionalism, but it inevitably runs into the sort of objections raised by critics of rationalism such as constructivists (see above). That said, he does indicate the interesting questions that can be followed using a 'political economy' frame of reference. Whereas the governance literature is largely preoccupied with the political activities of multiple actors pursuing their interests in the context of the institutiona-
Europe and the World    185
lized environment of the EU, the political economy literature begins with a focus on private market behaviour and draws conclusions about political outcomes. The governance literature seems to be good at providing a kind of 'steady state' theory of European governance. Political economists cannot provide the equivalent of a 'big bang' theory to act as a rival, but there is a sense that they can do a better job at explaining the initiation, collapse and expansion of regional integration schemes.
Conclusions
International theory is not easily dismissed. This is not because traditional IR problematiques continue to be relevant to the study of European integration. Rather, recent reflections in the International Relations literature have thrown up new ways of thinking about old concepts and have provided substantial challenges to images of the world built around images of the Westphalian nation-state. This has not just been a matter of exploring new ontologies of world politics, but it has also included the critical examination of established concepts. With the study of European integration in mind, these developments have several implications. Firstly, the imagery of dispersed authority and multi-actor complexity present in much IR seems to connect well with some of the developing themes in EU studies. Secondly, the growth of critical investigations of established concepts has meant that even relatively staid depictions of the EU such as intergovernmentalism can be given new life. Thirdly, and probably most importantly, there are clear possibilities for making productive theoretical connections between elements of the 'governance turn' outlined in Chapter 5 and some of the main themes in contemporary international theory. It follows that the use of 'International Relations' need not necessarily imply rigid a priori depictions of what the world is like which correspond poorly to the complexity of EU governance. Rather, receptiveness to IR suggests receptiveness to themes and methods that open exciting possibilities for the study of the EU.
Chapter 8
Integration Theory and Social Science
Integration theory is an elusive concept. What to include under this heading is a matter for some debate and is likely to revolve around how we define 'integration'. This book has sought to show that integration theory has been tightly bound up both with the evolution of what is now called the European Union and with the efforts of successive generations of scholars to grapple with this entity. It also shows that a narrow focus on 'EU studies' and the conceptual work thereby generated is not enough and that the theorization of European integration is only fully understood with reference to wider currents in the social sciences. The theoretical analysis of European integration and shifting patterns of European governance has happened amidst several mood-swings in the academy. The earliest work discussed in this book (Chapter 2) arose in two contextual environments. On the one hand, the inter-war period saw only marginal disciplinary formalization in the political sciences. The 'theories' of writers such as Coudenhove-Kalergi (1926) were not theories in the contemporary understanding of that term, but normative visions which contained several propositions about the dynamics of the (European) international system. They are not 'functionally-equivalent' to latter day theories. On the other hand, early functionalism found expression in the early stirrings of formal International Relations and its first great debate about the essence of the international system (Long and Wilson, 1995).
The 'classic' integration theorists, Deutsch and Haas most notably (Chapters 2 and 3), developed their work in the context of key movements in the American social scientific Zeitgeist. Deutschian transactionalism is an expression of the behavioural movement that came to be dominant from the 1950s (Eulau, 1963). In a fascinating autobiographical account, Deutsch (1989) explains the various intellectual intersections that led him to think as he did about international integration. His formative encounters with behaviouralism occurred as a series of behaviouralist scholars arrived at MIT in the early 1950s, and his interest in the use of large-scale computer modelling was ignited by a year's study leave at the Center for Advanced Study in the Behavioral Sciences at Palo Alto. He also recalls the happy occurrence of Talcott Parsons walking into his office at Palo Alto:
186
Integration Theory and Social Science    187
[Parsons] began to explain to me his view of the basic functions of every social system — pattern maintenance, adaptation, goal attainment, and integration. I found this idea fascinating. Soon the small blackboard in my office was covered with their graphic representations, and we kept discussing them for several days.
(Deutsch, 1989: 19)
These reflections raise a number of issues about the significance of individuals to theoretical development and the reliability of autobiographical reconstructions of personal intellectual journeys. Yet, this example shows clearly how a theorist of integration developed ideas within broader academic movements. The input of Parsonian ideas gave direction to the processes Deutsch and his colleagues wrote about and the emphasis on the quantitative study of behaviour supplied means to conceptualize and measure transactions. More broadly still, these influences also imported a particular brand of positivism that suggested possibilities for mapping the social world according to the criteria of natural science (Hollis and Smith, 1991: 28-29; Adler and Barnett, 1998a). Indeed, International Relations' second great debate was that between realists and behaviouralists. This sometimes ferocious encounter saw behaviouralists attacking realists for importing a priori assumptions into their analysis rather than relying upon that which was observable. The fact that both schools saw themselves as occupying radically distinct intellectual terrain should be a lesson to those who might seek to package together IR as a discipline united by particular guiding themes.
Similarly, the neofunctionalism of Haas and others drew heavily of the imagery and logic of pluralist political science (Harrison, 1974: 237). As the school developed, so theoretical structures were formalized and re-interrogated (see especially Lindberg and Scheingold, 1971). This found expression in the increasing resort to quantification and the use of mathematical notation, all trends in political science in general and US political science in particular. The point to make here is that the neofunctionalists were not simply trying to do a better job at explaining regional integration. They were also following rules and norms about how theoretical enquiry should be conducted; rules that were standard working practices well beyond the small community of neofunctionalists. 'Routine models of conduct' (Christiansen and Jorgensen, 1999: 6) are not just subjects to be studied. They are also part and parcel of the process of studying.
The rise of intergovernmental critiques from the mid-1960s (Chapter 4) was certainly - though perhaps not exclusively - connected to the dominant position established by realism within International Relations. In fact, both the early intergovernmentalism of Hoffmann (1966) and the
188    Theories of European Integration
later liberal intergovernmentalism of Moravcsik (1991; 1993a; 1998) are quite distinct from traditional state-centred accounts of international politics, not least because of their emphasis upon the domestic roots of national preferences and the operation of two-level games. The fact that intergovernmental theories of integration do not equate to realism/neore-alism is given substance by the existence of a separate small body of distinctively neorealist scholarship on the EU (see Chapter 6). Hoffmann, often thought of in EU studies circles as the apologist for realism par excellence, is perhaps better categorized as a dissident realist, an impression confirmed by reading some of his autobiographical reflections (Hoffmann, 1989: 1993). Moravcsik sees himself working squarely within the liberal tradition (1993b) and more specifically within the neoliberal institutionalism most associated with Robert Keohane (1989). It is also worth noting that intergovernmentalist contributions to the debates about European integration have tended to appear following neofunctionalist or supranational institutionalist flourishes. It may be an obvious point, but theoretical accounts do not develop in isolation as hermetically sealed bodies of knowledge. Each theoretical 'self has its theoretical 'other(s)' and for intergovernmentalist work on European integration, the other has most definitely been neofunctionalism. The initial reception of neofunctionalist work (see Chapter 4) suggested that its theoretical significance was not immediately apparent. Neofunctionalism was given substance as the primary theoretical account of integration by the efforts of neofunctionalist scholars, but its visibility and salience as 'integration theory' has had much to do with its intergovernmental critics. A recent scholar of the history of International Relations (Schmidt, 1998) makes a similar point in a broader context. The history of IR, it is argued, is reconstructed periodically to legitimize particular positions in contemporary debates. This means that central concepts and ideas, such as 'anarchy', need to be understood less as facts out there in the real world and more as the product of intradisciplinary conversations. This does not mean that intergovern-mentalists and neofunctionalists have been engaging in inward-looking conceptual jousts without concern for the unfolding developments of the European Communities. What it does suggest, is that we need to be conscious about the way in which theoretical trajectories are influenced by the interaction between alternative perspectives.
Intergovernmentalist work on the EC/EU has also been sustained by developments beyond the study of integration. Some of the rationalistic state-centred work on the interplay between domestic politics and international bargains (for example, Milner, 1997, 1998) owes much to the development of International Political Economy, where scholars have tended to question the firmness of the boundary between the 'domestic' and the 'international' (Underhill,  1994). The emergence of IPE-style
Integration Theory and Social Science    189
approaches to integration has clearly opened up the possibilities for comparative discussion and, in many ways, challenges the old realist-liberal dichotomy in international studies (Mattli, 1999). Yet, at the same time, it leaves open the 'rationalist—reflectivist' divide that has become such an issue (Keohane, 1988; Jorgensen, 1997a). Indeed, as Chapter 7 showed, the substantial turbulence in international theory in recent years has opened new avenues of thinking and forced significant re-evaluation of old cherished concepts.
The 'governance turn' in EU studies discussed in Chapter 5 draws on a wide array of theoretical developments in public policy analysis and comparative political science. In part, this has consisted of an insistence that traditional questions of politics can be applied to the EU because the EU is a polity (Hix, 1994). But it also reflects innovation and development in the policy sciences where there is renewed interest in the impact of institutions, the role of ideas and the place of symbols, norms and rules in political life (Richardson, 1996a). These approaches, which question exclusively interest-driven accounts of politics and policy change, connect well with movements in IR such as constructivism and open possibilities — as yet not fully explored - for the transcendence of the disciplinary divide between International Relations and Comparative Politics (Ebbinghaus, 1998; Jorgensen, 1997a; Risse-Kappen, 1996; Rosamond, 1999).
All of this suggests — obviously perhaps - that theories of integration need to be contextualized and that 'contextualization' means rather more than the idea that theories should be understood in relation to developments in the 'real world' of integration. They are only properly understood if we comprehend the broader social scientific concerns that gave rise to them and the social scientific environments in which they operate(d). In short, there are important 'sociology of knowledge' questions to consider when writing an intellectual history. It is not just the success of a theoretical enterprise in relation to the object of its studies that matters, but the ability of that enterprise to play the requisite academic 'games' successfully.
Evaluating integration theory
The few paragraphs above have suggested that there are complex issues involved in the evaluation of theories and that a proper appreciation of the significance and success of a theory needs to pay attention to the intellectual context in which it arises and operates. Theories can be evaluated in a number of ways. Pentland (1973: 19-20) argues that there are three sites of evaluation for any theory or, put another way, three avenues down which any theory may travel. One is the internal logic of
190    Theories of European Integration
theories. This relates to how well they develop their concepts, how rigorous they are in their quest to explain and how well they follow good practice in theory building. As we have seen in Chapters 3 and 4, neofunctionalists put a lot of effort into specifying and refining the concept of spillover in an effort to give scholars the capacity to measure that phenomenon and thereby activate the supposed explanatory power of their theory. The development of Andrew Moravcsik's work is also instructive in this regard. He has long maintained the primacy of intergovernmental bargains in European integration and the importance of domestic politics in the member states (Moravcsik, 1991), but his work has matured and become nuanced over time by thinking about the mechanics of national preference formation (Moravcsik, 1993a) and institutional choice (Moravcsik, 1998). It reflects a self-conscious attempt to follow a model of theoretical good practice. Too much work on integration, he argues, is un-theoretical and, therefore, flawed. It selects evidence to fit a pre-ordained conclusion, accepts at face value the ex post justifications of politicians and other involved actors and uses secondary sources that have drawn on similarly dubious methods to reach their conclusions:
One can find abundant support for any plausible conjecture about the causes of European integration. Only by deriving competing hypotheses from general theories, multiplying observations, and paying attention to the quality of primary sources can we transcend such indeterminacy and bias.
(Moravcsik, 1998: 11)
Moravcsik goes on to think about the limitations of neofunctionalism. He mentions its empirical deficiencies (see also below), but notes how the tortuous, yet necessary, re-evaluations of the late 1960s and 1970s stripped away so much of the theory's original power, that it ended up lacking the coherence to make predictions that could be tested. Neofunctionalists themselves acknowledged this (Haas, 1975a; 1976). But their 'failure' was not an isolated instance. Moravcsik suggests that the 'obsolescence' of neofunctionalism was symptomatic of a wider malaise afflicting most attempts to capture politics in terms of an overarching theory.
Such lessons are the staple of graduate classes in political science methods (King, Keohane and Verba, 1994) and they draw on Karl Popper's ideas about good theoretical practice (Popper, 1969; see also Hollis and Smith, 1991: 52-7). Popper's most famous argument is that theories should be constructed so as to render them capable of falsification. Empirical confirmation of a theory is not enough in itself - a point illustrated by the arguments made by Moravcsik about much EU studies work. One conclusion might be to suggest that what really matters is not how well the theory fits the reality (see also below), but the extent to which the
Integration Theory and Social Science    191
scholarship is reflective about its own assumptions and how rigorously the process of theorizing is conducted.
The second point of evaluation is the theory's intellectual context. This enables the evaluator to explore the family history of the theory and to check its resemblance and its debts to its living relatives. In some cases, the observer of theories will be able to identify theoretical cousins with whom productive contact might be made. This sort of contextualization - the kind of exercise carried out briefly above - has various uses. Historians of political thought often follow the dictums of Quentin Skinner (1978) who is most associated with the position that ideas and concepts are inseparable from their historical and intellectual contexts. Theories, ideas and concepts do not necessarily have timeless, trans-historical qualities. Their authors were burdened with preoccupations of the time and made their arguments by deploying particular forms of intellectual rhetoric in the context of the historically rooted intellectual games in which they were engaged. So, to read Hobbes as a defender of arbitrary dictatorship or Machiavelli as an apologist for amoral realpolitik might make the mistake of importing our contemporary concerns into the analysis without appreciating the contexts in which Hobbes and Machiavelli operated (Cox, 1981). In a rather different vein, writers influenced by Michel Foucault have argued that we need to be more attentive to the relationship between knowledge and power and how, in particular, dominant intellectual discourses are related to prevailing structures of power (see Devetak, 1996). In terms of integration theory, the general lesson would be well learned.
The fact that, for the most part, the theories under discussion have all arisen in the second half of the twentieth century removes the need for painstaking historical reconstruction. But any attempt to evaluate a theory has to be attentive to the social scientific and intellectual concerns of the time as well as to the prevailing political discourses of the period. So, from the comfort of a millennial armchair, armed with the sophisticated theoretical technology of the new governance literature, it is not enough to say that functionalists, neofunctionalists or transactionalists were wrong or naive. For example, we have to recognize that David Mitrany (Chapter 2) wrote in a particular way for an audience that was not as formally 'academic' (at least when he wrote A Working Peace System). True, his work was located within a broad intradisciplinary discussion, but he also wrote to intervene in practical policy debates at a time when thinking otherwise about structures of governance was widely practised. Andrew Moravcsik can be read as a careful adherent of widely accepted (if not always widely practised) theoretical norms. His is not a theoretical exercise in trying to change the world. His point is, as it were, to explain it. To say this is not to make a pejorative judgement, but to recognize that intellectual interventions are rule-bound and that intellectual work is a
192    Theories of European Integration
social exercise, influenced by dominant patterns of discourse and accepted modes of academic behaviour. At the same time, his state-centrism has the ring of familiarity about it. Politicians, especially members of national executives, view the integration process as an exercise in bargaining and diplomacy and these perceptions are reinforced through media coverage which is not inclined to explore the 'everyday' political economy of the EU (Richardson, 1996b; "Wincott, 1995b). State-centric explanations have a 'common sense' feel to them, but that should not substitute for theoretical or empirical verification.
In other words, the search for knowledge, indeed the establishment of what counts as valid knowledge, is socially located and socially constructed. Interestingly, Haas has reflected on these questions with characteristic insight:
Progress has occurred in international politics, but I also want to argue that progress has occurred because our conceptions of what constitute political problems, and of solutions to these problems have been increasingly informed by the form of reasoning we label 'scientific'.
(Haas, 1991: 189)
This view recognizes the ultimate contingency and social-rootedness of human knowledge. But it does not take this observation to a relativistic conclusion (where judgements about anything become impossible). Rather, for Haas, the social construction of knowledge becomes the means through which progress is achieved. Not all would agree with the explicit celebration of 'science', the logic of which might be to exclude heterodox or critical thinking, but Haas's position connects with a view of knowledgeable progress associated with Thomas Kuhn (1970). Kuhn's argument was that science can be periodized into phases where particular paradigms dominate research. Paradigms define the guiding assumptions of research, what counts as valid knowledge and how work should proceed. Work at the margins or beyond these working assumptions is not regarded as scientific or (therefore) valid. "Work progresses on the basis of these assumptions until the results produced challenge the paradigm to the point of unsustainability. There then follows a period of scientific revolution as a new paradigm displaces the old. There have been periodic attempts to apply the idea of paradigms to social science and even to integration theory. Michael O'Neill (1996) writes about the shifts from a supranational to an intergovernmental paradigm with a further shift towards a 'syncretic' paradigm as the insights of multiple perspectives become synthesized. But the real strength of Kuhn's account lies in his depiction of the way in which intellectual activity becomes institutionalized into sets of acceptable practices. In any case, neither 'integration
Integration Theory and Social Science    193
theory' nor 'EU studies' are disciplines in the sense of physics or, for that matter, political science and International Relations. It should be patently clear by now that all theoretical work on integration or EU governance draws on external referents. If we are looking for paradigms, then they are likely to be embedded in those broader disciplinary environments (Hollis and Smith, 1991).
The third and final point of evaluation concerns the ability of a theory to connect to reality. At first sight this seems to be the most straightforward criterion for judging the merits of respective theoretical accounts. It is true that theories develop and sometimes collapse in accordance with how well their propositions match what goes on in 'the real world'. Neofunctionalism provides a very obvious example from the work discussed in these pages. By the late 1960s, neofunctionalists ran into a series of empirical difficulties. The first was the residual obstinacy of the nation-state, manifested most starkly by the increasing impression made upon EC affairs by President de Gaulle. At the same time, the logic of functional spillover and the teleological account of the development of integration were also much less discernible. Ultimately, neofunctionalists were presented with the more plausible alternatives: (a) that national and nationalistic forces matter, (b) that the pattern of activity exhibited within the Communities might owe more to traditional international relations than they had first allowed and (c) that the Community displayed a politics that organized itself around multiple loci and not just questions of 'integration'. The upshot was that state-centred International Relations, theories of international interdependence and perspectives derived from policy analysis all seemed to be better candidates to offer explanatory power. It is very important to recognize that this was a game that neofunctionalists were prepared to play (Haas, 1975; 1976) because the evaluative standards they set for themselves were largely ones of empirical correspondence. The partial revival of neofunctionalism in the context of the Single European Act and the acceleration in formal integration in the 1980s showed the process in reverse. As the empirical winter thawed, so neofunctionalism re-emerged (Mutimer, 1989; Tranholm-Mikkelsen, 1991).
But the 'truth as correspondence' issue is not as simple as it might seem. In Modern Political Analysis Robert Dahl remarks that '[w]hether [an empirical] proposition is true or false depends on the degree to which the proposition and the real world correspond' (cited in Neufeld, 1995: 34). This relies on the nositivist supposition that the 'object' (the real world) can be separated from the 'subject' (the investigator). If accepted, this means that objective knowledge about the world is possible. This is not an issue to which those inclined towards positivistic and rationalistic investigation are indifferent. The comments of Moravcsik (cited above) indicate
194    Theories of European Integration
that there are ways through the minefield and that objectivism is possible provided that investigators follow certain rules. However, Mark Neufeld's study of International Relations takes the issue rather deeper into the realms of discourse analysis, hermeneutics and the philosophy of science:
If the paradigm (language game/tradition/discourse) tells us not only how to interpret evidence, but also determines what will count as valid evidence in the first place, the tenet of 'truth as correspondence' to the facts can no longer be sustained.
(Neufeld, 1995: 42)
The uncomfortable conclusion is that 'objective' facts are difficult to know because theories help to define the world that they describe. They have different answers to the question, 'Of what is this an instance?' and different criteria for selecting independent variables. Neufeld (1995) recommends theoretical reflexivity as the most useful way to encounter this problem. This involves being conscious about the status of our concepts and the rootedness of our theories. But it also requires reflection about our strategies and the politico-normative context of what scholarship entails. Social sciences and theories have different value bases. Robert Cox's famous dictum merits another citation:
Theory is always for someone and for some purpose. All theories have a perspective. Perspectives derive from a position in time and space, specifically social and political time and space. The world is seen from a standpoint definable in terms of nation or social class, of dominance or subordination, of rising or declining power, or a sense of immobility or of present crisis, of past experience, and of hopes and expectations for the future. Of course sophisticated theory is never just the expression of a perspective. The more sophisticated a theory is, the more it reflects upon and transcends its own perspective; but the initial perspective is always contained within a theory and is relevant to its explication. There is, accordingly, no such thing as theory in itself, divorced from a standpoint in time and space. When any theory so represents itself, it is the more important to examine it as ideology, and to lay bare its concealed perspective.
(Cox, 1981: 128)
Going back to integration theory, there is always the danger that theoretical work and the derivative empirical investigation follows the logic of particular values that are rooted in the object of enquiry. So intergovernmentalism might at times become the 'authorized version' of what  national   executive   actors   say   and   think   that   they   are  doing
Integration Theory and Social Science    195
(Chapter 6, also Rosamond, 1996). What's more, it can be an exercise in circular reasoning. As Christiansen and Jorgensen note: ' "Proving" that member states are in control of "intergovernmental bargaining" by starting with the input from member states is a tautology which ultimately obscures much of what is analytically relevant' (1999: 5).
Of course, we have also seen the close links between neofunctionalist theory and the strategies embedded in the Schuman Plan, the ECSC and the whole community method (Chapter 3). So, neofunctionalism could also be called an authorized version. Indeed, this accusation has been levelled by Mil ward and Sorensen (1993) who add spice to their quite venomous critique by making linkages between the largely American thrust of early integration theory and US foreign policy priorities of the time. For them, Haas and the other neofunctionalists were too eager to assemble grand theory at the expense of detailed attention to the historical record. Attentiveness to the minutiae of post-war reconstruction and the origins of the European Communities has been a long-standing preoccupation of Mil ward (see also Milward, 1984, 1992). Perhaps his major work on the period contains the most telling indictment of 'integration theorists' from this perspective. For Milward, the likes of Deutsch, Haas and Lindberg
simplified history unacceptably ... they all did so in the same way, by greatly exaggerating the incapacity of the state. From the beginnings of detailed historical research into the origins of the European Community, it became clear that nation-states had played the dominant role in its formation and retained firm control of their new creation
(Milward, 1992: 12)
A similar way of thinking informs another recent contributor to the archive-based history of European integration. Keith Middlemas (1995) is similarly dismissive of the partiality of theories of integration: 'I have tried not to confine myself to any one interpretation, whether federal, functional or intergovernmental, and to proceed empirically, taking account of all the significant players' (Middlemas, 1995: xiv, emphasis added). To argue from this position requires the establishment of an opposition between 'history' and 'theory', where the former is objective, empirical and inclusive and the latter is value-laden, conceptual and partial. Now both Milward and Middlemas are careful historians and are attentive to the quality of their sources and the contexts in which utterances are made. But, they rely heavily on the premise of 'truth as correspondence' to separate the subject and object. What count as valid facts? Who are 'all of the significance players'? To attend to these questions requires some kind of a priori judgement. Even historians are theoreticians - whether they know it or not.
196    Theories of European Integration
Concluding comments
The general arguments of this book have been stated often enough, but a final rehearsal might be worthwhile. Firstly, the study of European integration, the transformation of European governance and the politics of the EU polity has been a very fertile site for theoretical development. This is partly because European integration is an intrinsically interesting process and partly because creative scholars have flocked to study it. Secondly, in many ways, the identification of sub-fields called 'integration theory' or 'EU studies' is less than helpful. Obviously scholars specialize, but to map disciplines too much in this way artificially cordons them off from the wider social scientific contexts within which they arise and develop. A full appreciation of theories of European integration, therefore, has to be attentive to 'sociology of knowledge' issues. This will deepen understanding of the contexts of integration theory and help us to understand debates between different schools in their own terms. It will also help to give us a sense of theoretical family trees. It is tempting to portray integration theory's main cleavage as a long-standing confrontation between intergovernmentalism and neofunctionalism. This has obvious merits, but could lead to an overly static picture both of the positions themselves and the nature of the debate between them. Understanding these positions as they evolve in a broader social scientific context helps us to avoid basic simplifications such as the equation of intergovernmentalism with (neo)realism and helps us to trace theoretical lineages (for example, the relationship between neofunctionalism and regime theory, or domestic politics approaches and neoliberal institutionalism). Thirdly, attention to these sociology of knowledge issues connects to deeper reflection on the processes of theorizing. The observation that theory is inescapable is not meant to drive all students of the EU into metatheoretical frenzy. Rather, solid empirical work should be rooted in an understanding of the investigator's assumptions and the theoretical suppositions upon which they draw. We can choose how far to take this requirement. It might be a matter of careful theory building to ensure that hypotheses are properly generated and that conclusions are not drawn from inherently biased reasoning. Alternatively, it might lead to a break with rationalistic assumptions, not only about ontology (the nature of the world that we are investigating), but also about epistemology (the process through which we acquire knowledge about that world). Fourthly, the question of whether EU studies is an 'International Relations' or a 'Comparative Politics' question is a non-problem, relying on a false dichotomy between these two disciplinary domains. Of course, scholars from different vantage points in the disciplinary universe will see the EU differently and conceptualize it in different ways. But they may also use
Integration Theory and Social Science    197
similar methods to shed light on different things. For example, there is nothing to prevent constructivist analyses of the internal sociology of the European Commission, the processes of intergovernmental bargaining and the construction of the EU's identity as an international actor.
There is no doubt that 'integration theory' (if it can be called that any more) is in a good state of health. It was not always so, and the recent phase of theoretical reflexivity and innovation owes much to the spillover into EU studies of creative thinking across the political sciences. Grand theories of European integration have certainly had their day. In one sense they were never meant to exist! The neofunctionalists always intended to generate general theories of regional integration from the European case study and virtually all other 'theories' discussed here have slotted their analyses of the EU into broader theoretical contexts. Even the multi-level governance literature which, at face value, looks to be an attempt to depict a sui generis phenomenon does not fall into this trap. This is partly because MLG is more metaphor than theory, which allows alternative theoretical accounts to colonize it. Also, MLG analysis fits well with the increasing number of scholars depicting governance as fluid and authority as dispersed, in terms of both domestic politics and transnational relations. Therefore, theoretical endeavours on European integration are likely to develop most fully as sub-sets of other concerns. These include theories of regulation, epistemic communities, institutional choice, policy networks, path dependency, the role of ideas in policy-making, regionalism and regionalization, two-level games, transnational relations and constructivism - to name but a few. European integration may well be a totally unique enterprise without either historical precedent or contemporary parallel, but it is a ready source for comparative study in some of the most energizing and lively social science currently going on.
Theories of integration are also important because they grapple with one of the most remarkable experiments of the twentieth century. To contemplate the sources of apparently radical dislocations in patterns of governance in Europe and to seek to shed light upon the processes of institutional creativity and economic enmeshment that have been occurring is a vitally important enterprise - not just an academic game. Long may it continue. And long may it be theoretically astute.