4    Theories of European Integration
world in general and other instances of regional international cooperation in particular. It concludes by elaborating the structure of the rest of the book.
Theory
An observer enters a room and begins to take note of what he or she sees. Five things are noted: brown carpets on the floor, plain white walls, steps, inset spotlights in the ceiling and 61 people. A second observer of the same scene also notes five things. One individual is standing at the front of the room doing most, if not all, of the talking. This person is standing by a device projecting images and text onto a white screen on the wall. The other 60 are best described as being variously engaged with what the person at the front is saying. The room is tiered. The event lasts about 50 minutes. Our two observers have seen the same event, but they have chosen to observe different aspects of that event. The first observer has the sense that s/he has observed 61 people in a room with particular physical features, whereas the second has clearly seen a lecture.
Our first observer has no framework for ordering or making sense of what s/he sees. Unlike the second observer, s/he has no sense of which data might matter most; this person has no tools available to make sense of the event that is being witnessed, or at least to attribute meaning to that event. The second person fits what s/he sees into a predefined conceptual category ('lecture'). So, in many ways, the relevance of theory to inquiry in the social sciences is easily stated. Theories are necessary if we are to produce ordered observations of social phenomena. Theory, as Gerry Stoker puts it,
helps us to see the wood for the trees. Good theories select out certain factors as the most important or relevant if one is interested in providing an explanation of an event. Without such a sifting process no effective observation can take place. The observer would be buried under a pile of detail and be unable to weigh the influence of different factors in explaining an event. Theories are of value precisely because they structure all observations.
(Stoker, 1995: 16-17)
This fundamental point is often made to demonstrate that theory is important; that to 'do' social science properly, researchers need to conduct their inquiries under the auspices of a particular theoretical perspective. But it suggests something else: that it is impossible to make any statement about social phenomena in a theoretical vacuum. Like it or not, we are all
Introduction    5
informed by theoretical perspectives, even if we adopt an avowedly non-theoretical posture (Axford et al., 1997). Thus, what is written about European integration — by academics, students and journalists alike — is always grounded in a particular set of assumptions about the way in which the world operates. It is always instructive to tease these assumptions out of texts such as newspaper accounts of EU events or largely descriptive surveys of EU politics. To take a concrete example, it is often argued that a major turning point in the history of European integration came with the signing of the Single European Act (1986) (SEA), and that this was achieved thanks to a convergence of policies among the most powerful member-states around a broadly neoliberal, free market economic policy agenda. Yet this account is not an unproblematic and objective statement of truth. It relies upon a set of propositions that proclaim the centrality of state and intergovernmental interaction to the conduct of European integration. It also possesses a sense of what constitutes a significant event in European integration. The SEA and other 'history-making moments' are often treated as the key nodal points in the unravelling story of post-war European unity. Others suggest that these treaty revisions amount to nothing more than the formal consolidation of emergent practices and that the 'everyday politics' of the EU and the private actions of economic actors are equally, if not more, important to gaining a full understanding of the integration process. These are not just issues of empirical disagreement; they reflect differing assumptions about key actors, the environment within which action takes place and the relationship between structure and action. The fact that an issue such as the origins of the SEA is so hotly debated is indicative of the importance of teasing out the diversity of theoretical starting points (see for example Agnelli, 1989; Cameron, 1992; Moravcsik, 1991, 1993; Sandholtz and Zysman, 1989; Wincott, 1995b). Disagreement might also reflect alternative disciplinary starting points. To understand these deeper-lying questions is to acquire a more nuanced understanding of the debate and, therefore, of the core subject matter of 'EU studies'. Put another way, it is not just theories about the world that differ and generate disagreement. It is also important to recognize the disciplinary and historical context within which work arises. Knowledge is not neutral. We gather it according to agreed rules that change over time and which, in turn, influence the sorts of question we ask. In other words, knowledge has its own sociology and any attempt to recreate intellectual sequences needs to be aware of this.
Theorizing inteliectualizes perceptions. It is not just that theory helps us to identify that which is significant. Any event may involve multiple happenings that appear to be meaningful. To return to an educational theme, Andrew Sayer's example of an undergraduate seminar is instructive:
6    Theories of European Integration
It involves far more than a discussion of some issues by a group of people: there is usually an economic relationship (the tutor is earning a living); students are also there to get a degree; the educational institution gets reproduced through the enactment of such events; relations of status, gender, age and perhaps race are confirmed or challenged in the way people talk, interrupt and defer to one another; and the participants are usually also engaged in 'self-presentation', trying to win respect or at least not to look stupid in the eyes of others.
(Sayer, 1992: 3)
Events are multidimensional and theorists have to decide what they plan to explain from the array of multiple games embedded in any single situation. Theorists have to generate speculations or hypotheses about which of the games is to take precedence. But they also need to arrive at a view of the nature of the relationship between the different dimensions identified. Another example from the EU might help to reinforce the point. On the face of it, a meeting of the Council of Ministers is the primary forum for bargaining between governments in the EU system. The representatives of the national governments are there to elaborate and defend their national interests and to negotiate from the basis of these positions in a particular policy or issue area. But there is a lot else going on besides. Ministers have to attend to the problems of coordinating their position with those of their colleagues who sit in other Councils. Also, ministers have to engage in calculus about the appropriate way to present significant policy outputs to domestic constituencies from the point of view of their positions both as members of a government and as politicians in particular domestic contexts. Meetings of Councils are not without their institutional memories. Many Councils have evolved distinctive working practices and bargaining styles over time (Hayes-Renshaw and Wallace, 1996) and others such as the Council of Agricultural Ministers have a peculiar longevity of membership plus the common perception among members that agricultural ministries face similarly distinctive sorts of problems in relation to other government departments (Grant, 1997). Moreover, the member-state holding the presidency of the Council has issues of agenda management and brokerage to contend with as well as the conventional representation of national preferences. In short, the sociology - perhaps the anthropology - of the Council of Ministers is a feature of its operation. Theorists may find orthodox intergovernmental approaches appropriate, but it may also be the case that the rnultidimensionality of such events opens the space for alternative concepts and theories, perhaps involving refined notions of supranationality or drawing on policy analysis literature or even theories of social psychology. This leads to the increasingly popular conclusion that the EU and the processes of European integration
Introduction    7
are just too complex to be captured by a single theoretical prospectus. The debate generated by that observation is a central preoccupation of this book.
Theoretical debate could be construed as disputes about different ways of obtaining or producing knowledge. This is important because different theoretical perspectives produce and reproduce different types of knowledge. As Susan Strange put it in her discussion of the main perspectives used to study International Political Economy (IPE):
Each begins their analysis from a particular assumption that determines the kind of question that they ask, and therefore the answer they find. They are like ... toy trains on separate tracks, travelling from different starting points and ending at different (predetermined) destinations, and never crossing each other's path.
(Strange, 1994: 16)
The extent to which this is (a) a problem and (b) solvable is a matter for debate rather beyond the scope of this book (but no less interesting for that). The sense of dissatisfaction implicit in Strange's statement owes much to what she perceived as the rather stagnant nature of theories of IPE, rather than to the unsatisfactoriness of coexistent and largely self-contained paradigms per se. Indeed, there are strong and much-rehearsed arguments for the pursuit of academic work within a confined set of theoretical assumptions as the most efficient way to advance knowledge. Each distinct theoretical perspective has its own 'home domain' of description and explanation, which influences the level at which data is analysed (Alford and Friedland, 1985). Or, put another way, each theory begins with a 'basic image' of social reality (an ontology) upon which is built a theoretical superstructure including established ways of gathering knowledge (epistemology). This allows like-minded scholars with shared assumptions to advance knowledge significantly within the lingua franca provided by a particular theoretical discourse. While good theories are on the whole internally coherent, they run into conflict with one another over a range of issues such as which actors are significant, what is the 'dependent variable' (i.e. what is it that the theory is trying to explain?), which processes are important, and so on (Banks, 1985). The consequence of this depiction of theories as internally consistent paradigms may, of course, be viewed in a positive sense - as a way of advancing knowledge with efficiency and rigour. Moreover, any friction between perspectives can be understood as a zone of 'essentially contested concepts' (Gallie, 1956) where fundamentals, such as the nature of 'power', come under sustained philosophical scrutiny. In any case, there are always likely to be trans-paradigm mavericks in the social sciences who look to establish constructive dialogues and to accomplish theoretical syntheses. The key
8    Theories of European Integration
point is that there is clearly a relationship between theoretical assumptions and the way in which the processes and outcomes of something like European integration are depicted.
If theory is inescapable, then it might appear that the traditional arguments for theory are not worth stating. Not so. The rehearsal of the case is useful since, while theory may be inevitable, not all analysts appear to be conscious of this important point. So, for the purposes of theoretical self-consciousness it is worth reminding ourselves what theory is for. However, here again we begin to run into ambiguity. There are clearly different types of theory, all of which have alternative purposes. Theory is sometimes thought to be about the generation of law-like statements. Others conceive of theory as the instrument with which investigators can test hypotheses or propositions about social phenomena. For some, theorizing is an activity with normative (value-laden) consequences; for others, it is a political act: the way in which we criticize the present with a view to maximizing the prospects for human freedom in the future. Finally, theory may also involve the contemplation of the process of theorizing itself (Burchill, 1996: 8). Most - if not all - of these purposes have found their way into the broad field of integration theory. To take an example, the neofunctionalist perspective, which is discussed at length in Chapter 3, was the creation of writers operating in the behaviouralist Zeitgeist of post-war American political science (Eulau, 1963). Their approach was largely consistent with the attempt to connect natural scientific methods to the study of social reality. The study of the European experience was thought likely to yield law-like generalizations about regional integration that might be applied to other instances elsewhere. It was thought to have failed, by its practitioners as much as anyone, because of its empirically dubious depiction of the events of European integration through the 1960s and early 1970s. Neofunctionalists have also come under sustained criticism for their supposed normative purposes. One pair accuse the neofunctionalists of acting as a kind of theoretical 'Trojan horse' for the aspirations of US foreign policy in the 1940s and 1950s (Milward and Sorenson, 1993). This view sees neofunctionalist integration theory as a vehicle for lending legitimacy to the project of creating a federal Europe via the integration of national economies. Less aggressively, neofunctionalism is often thought of as the formal intellectual depiction of the political strategies of a group of post-war European politicians such as Jean Monnet and Robert Schuman.
So, how do we judge theories? What might be the basis for preferring one over another? As this book progresses, it will become obvious that these questions have been central to the enterprise of integration theory. These issues are explored in greater depth in Chapter 8, but for now it is worth reiterating that the alleged failings of neofunctionalism, as docu-
lntroduction    9
mented by intergovernmentalists like Stanley Hoffmann (1966) and neofunctionalists themselves (notably Haas, 1975a), concerned its failure to correspond to the observed 'reality' of the integration experience as exemplified by the EC. Such reasoning may seem commonsensical, but it is worth remembering that this is only one way in which theories might be evaluated. Indeed, it could be argued that the selection of this particular evaluative criterion reflects deep assumptions about both the nature of the social world and the processes/purposes of theorizing. More 'constitutive' (Burchill, 1996) or 'critical' (Cox, 1981) approaches to theory would approach the issue in rather different ways (see Chapter 8). In his discussion of International Relations theory, Scott Burchill presents six criteria against which theories might be evaluated:
1.    a theory's understanding of an issue or process;
2.    a theory's explanatory power of the theory;
3.    the theory's success at predicting events;
4.    the theory's intellectual consistency and coherence;
5.    the scope of the theory;
6.    the   theory's   capacity  for  critical  self-reflection   and   intellectual engagement with contending theories.
(Burchill, 1996: 24, emphasis in original)
In other words, theories may stand or fall according to rather more than whether they can successfully describe a phenomenon or predict the consequences of that phenomenon. One particular concern of this book is to think about theories of European integration both in terms of the criteria they appear to set themselves and, in a critical sense, as manifestations of particular forms of knowledge production located in particular contexts. Therefore, the book also seeks to think about theories in the light of the context — in terms of the social sciences as well as the 'real world' of integration practice — in which they arose and in terms of the relationship between theories of European integration and the practice of integration and EU governance.
The meaning of 'integration'
So, the process of theorizing is, to a very large extent, a mechanism for the generation and organization of disagreement. Put more positively, being theoretically conscious sharpens the sense in which analysts are aware of their own assumptions about the way in which the world works. For students of European integration, this is a particularly poignant lesson. As
10    Theories of European Integration
suggested at the beginning of this chapter, the unfolding events in Europe after 1945 offered a generation of social scientists an alluring set of events to describe, categorize, explain and predict. During the 1950s, the quite extraordinary processes of international cooperation and the subsequent institutionalization manifested initially in the European Coal and Steel Community (ECSC) and later in the European Economic Community (EEC) and Euratom were genuinely novel. Here, at the very least, was an instance of quite intensive international cooperation among a group of states. Moreover, this was perceived as a radical experiment, especially from the vantage point of 1950s social science (Caporaso, 1998). The heavy institutionalization associated with the early communities suggested something yet more profound (the Treaty of Paris of 1951 which established the ECSC set in place an institutional pattern, involving clear elements of supranationality, which has endured until today). Thus, an immediate theoretical controversy was to develop around the question of whether the communities constituted a new 'post-national' political system in which the authority of national governments was destined to recede. To accept this proposition led to two sorts of conclusion. Firstly, Western Europe was undergoing a quite profound period of transformation in which the established patterns of political authority were being radically reordered. Secondly, it was possible to see an unfolding logic to this transformation, where a new sort of state form above the nation-state would be the outcome. To deny the proposition would be associated with the assertion that the nation-state possessed historical durability. States controlled the integration process and any outcome would be fundamentally intergovernmental or at least would reflect the preferences of the most powerful states in the game.
These events provided an important stimulus for theoretical work in their own right. For example neofunctionalism (see Chapter 3) can be read at one level as a theory provoked entirely by the integrative activity among the original six member-states. The study of European integration became a major site for debates in the academic discipline of International Relations which, by the 1950s, had spawned the sub-field of International Organization. Practitioners here were concerned with collective international and transnational institutions and the emergence of significant non-state actors in the world polity. Integration theory emerged in this context, tapping into pre- existing concerns, and generated new debates. In some ways, integration theory became a pioneering site for the development of non-state-centred forms of International Relations scholarship, and many sub-fields of present-day IR and International Political Economy are rooted in the endeavours of functionalists, neofunctionalists and transac-tionalists. Also, as academic work on the Communities grew throughout the 1960s and 1970s, so rival integration theories became the 'pet theories'
Introduction    11
of an emerging sub-discipline - EC (and later EU) studies. Greater attention to the EC, along with developments in the Communities themselves, began to open up the questions of (a) whether the EC had acquired systemic properties and, therefore, (b) whether theories derived from IR any longer offered the best frameworks for analysis.
So, the attractions for social scientists were and, no doubt, still are obvious. But what was it that they were trying to explain? Writing in 1971 and reflecting on no less than a decade and a half of intensive theoretical work on integration in Europe, Ernst Haas argued that '[a] giant step on the road toward an integrated theory of regional integration ... would be taken if we could clarify the matter of what we propose to explain and/or predict' (Haas, 1971: 26). This was a succinct statement of the so-called 'dependent variable problem' in integration theory - what is it that theorists are trying to explain when they contemplate the processes of institution-building and integration that have characterized the post-war European Communities? As Haas put it:
the task of selecting and justifying variables and explaining their hypothesized interdependence cannot be accomplished without an agreement as to possible conditions to which the process is expected to lead. In short, we need a dependent variable.
(Haas, 1971: 18)
At one level this is a matter of definition. Is integration an economic or a political phenomenon? If it is an economic phenomenon, what levels of interdependence need to be achieved among a group of national economies for them to be described as 'integrated'? Is the achievement of a free trade area the appropriate condition? Or is the end point of economic integration a customs union, or a common market, or full economic and monetary union? Does economic integration imply political integration? Or, at least, what levels of common institutionalization are associated with an integrated economic space? Do all customs unions/common markets/ monetary unions have similar levels of institutionalization? Does economic integration generate the momentum for political integration? Or, to turn the issue on its head, does political integration create the space for economic integration to flourish? Turning to political questions, does integration amount to the dissolution of national authority within a given geographical region? If this is so, does integration consist of the replacement of traditional structures of governance with new types of institution and new forms of authority? Or is integration accomplished when a group of geographically-adjacent states reach an accommodation -perhaps in terms of a federal union or a system of common security, or in terms of a widespread sharing of core values among elites and masses
12    Theories of European Integration
across nations? In short, what does it mean to say that Europe is integrated or is in the process of integrating? Posed this way, the issue also becomes a matter of whether we should understand integration as a process or as an outcome.
Both Leon Lindberg's elaboration of the definition of regional integration (1963: 4—5) and Haas's look backwards at the early 'pretheorizing' of integration identified this particular ambiguity (1971: 6-7; see also Pent-land, 1973 for an extended discussion). Indeed, all of the possibilities canvassed in the previous paragraph were represented in the integration theory of the 1950s and 1960s. One problem was that integration theorists, while focusing on a common set of events, evidently had different conceptions of process and outcome in mind. Karl Deutsch's work (for example, Deutsch et al., 1957; see also Chapter 2) clearly understood integration as the creation of security communities (or zones of peace) among states in a region. This did not require the transcendence of formal statehood. Alternatively, many writers define integration precisely in terms of the radical reordering of both the conventional international order and of the existing authoritative structures of governance. Haas defined integration as 'the voluntary creation of larger political units involving the self conscious eschewal of force in relations between participating institutions' (1971: 4) and elsewhere as
the process whereby political actors in several distinct national settings are persuaded to shift their loyalties, expectations and political activities toward a new center, whose institutions possess or demand jurisdiction over pre-existing national states. The end result of a process of political integration is a new political community, superimposed over the preexisting ones
(Haas, 1968: 16)
Others, notably the authors of previous texts on integration theory, put it more starkly. Michael Hodges offered integration as 'the formation of new political systems out of hitherto separate political systems' (1972: 13). Reginald Harrison, like Haas, pointed to the importance of central institutions: '[t]he integration process may be defined as the attainment within an area of the bonds of political community, of central institutions with binding decision-making powers and methods of control determining the allocation of values at the regional level and also of adequate consensus-formation mechanisms' (Harrison, 1974: 14). The difficulties of definition were memorably summed up by Donald Puehala (1972) who compared the quest for a definition of integration to blind men being confronted with the task of defining an elephant. This recalls the argument, introduced earlier in this chapter, about different starting points leading to different destinations. Different theoretical conventions
Introduction    13
have spawned differing methodologies in pursuit of independent variables (those factors that do the explaining). Thus, the transactionalist school relied heavily on the accumulation of aggregate survey data, whereas the neofunctionalist method often amounted to the theoretically-focused case study. Consequently, the understanding of different theoretical approaches to integration is vital to a developed understanding of 'integration' itself. As Haas puts it: 'it is they [the approaches] rather than the nature of things which lead students to postulate the relationships between variables; it is they, not the nature of things, which lead us to the specification of what is an independent and a dependent variable' (1971: 19).
Where definitions have been advanced in more recent literature, they have tended to be rather more broad-ranging. To take an example, William Wallace defines integration as 'the creation and maintenance of intense and diversified patterns of interaction among previously autonomous units' (Wallace, 1990: 9). It should be clear that the classical phase of what is normally understood as integration theory was concerned with political integration. This is not to say that economic change was excluded from the analysis. Nonetheless, Wallace's discussion alerts the student of the EU to the relationship between economic and political integration. In particular, he makes a distinction between 'formal' and 'informal' integration. The former consists of outcomes (institutions, policies, legislative change) that have occurred as a consequence of deliberate political sanction. The latter are processes that have effective consequences without formal, authoritative intervention (see Wallace, 1990: 8—12). This connects to Richard Higgott's distinction between de facto structural regionaliza-tion on the one hand and de jure institutional economic cooperation on the other (Higgott, 1997). In both cases, the issue at stake is one of political economy: the relationship between political and economic processes in shaping change. Several permutations are possible here. For example, it could be argued that changes in the informal economic domain such as heightened capital mobility, increasing volumes of cross-border trade, alterations in the production process and shifting corporate strategies decisively structure and constrain the agenda of authoritative political actors. Faced with no alternative, governments seek closer cooperation through the construction of political institutions designed to 'capture' and control these economic processes. Here regional political integration is a consequence of regional economic integration. Of course, an alternative way of thinking would invert this argument to suggest that informal changes are, at the very least, facilitated by the deliberate sanction of government authority. Here economic integration can only happen because states produce policies that enable the flourishing of informal transborder economic activity. This debate also latches onto the question of whether the factors that initiate integration can be used to explain its
14    Theories of European Integration
maintenance. Do different periods of integration require different theoretical perspectives?
These issues are covered further in Chapters 7 and 8. The parts of this book that analyse the established schools of integration theory focus for the most part on political integration broadly defined, largely in response to and in the terms set by the literatures under discussion. However, it is worth noting that a large literature on economic integration has grown up in the field of International Economics. While the connections between the political and the economic are central to this book, a detailed analysis of the body of work on economic integration is rather beyond the scope of this volume (but see for example Balassa, 1962; El-Agraa, 1997; El-Agraa and Jones, 1981; Robson, 1998).
Theory and European integration
As we have seen, anyone seeking to engage academically with a subject must be conscious not only of their theoretical predispositions, but also of the nature of their subject matter. Combining these two aspects of theoretical awareness, it can be argued that the type of theoretical approach adopted will be related to the subject matter. In this sense 'subject matter' is not simply the events or phenomena to be interpreted, but the sorts of generalizations which intellectual inquiry aspires to make. Here James Rosenau's guiding question - 'of what is this an instance?' -becomes a vital prerequisite for any work with theoretical aspirations (Rosenau and Durfee, 1995). To answer the question requires an exercise in abstraction, and when faced with any phenomenon the question can induce multiple responses. The argument here is that (a) the study of European integration has to be theoretically-informed and (b) we need to be theoretically reflexive. To be blunt, we need to know what we want to get out of studying European integration, not in terms of what we want/ expect our answers to be, but in terms of where we seek to locate our investigations. The study of the EU/European integration seems to have at least four such locations.
The first of these approaches would be to understand the European Union as an international organization. The literature on international organizations (IOs) is substantial and ever developing (Kratochwil, 1995), but IOs are traditionally thought of as intergovernmental bodies designed in the explicit context of converging state preferences or common interests. For traditional liberal theorists of international relations, IOs constitute one of the principal means through which interstate harmony and, therefore, lasting peace can be secured. Quite a lot of the theoretical work reviewed in this book draws on this tradition, but the EU is evidently
Introduction    15
rather more than a straightforward instance of an intergovernmental organization. Whether it is dominated by state preferences is a moot point (see Chapter 6), but the EU is peculiarly institutionalized and the configuration of forces thereby created rather militates against the discussion of European integration in terms of established IO debates.
The second treats European integration - to coin the contemporary vocabulary — as an instance of 'regionalism' in the global political economy. The ultimate aim of such work is to offer reflections upon and possibly generalizations about the tendency of groups of territorially-adjacent states to cluster together into blocs. Inquiry of this sort can be motivated by a number of guiding questions. Most obviously, is it possible to make meaningful comparisons between the EU and other regional groupings such as the North American Free Trade Agreement (NAFTA), Asia Pacific Economic Cooperation (APEC) or Mercusor in South America? Do instances of regionalism arise in similar sorts of circumstances regardless of time or place? Do global economic and political pressures force or enable the creation of such organizations? How do variations in levels of institutionalization in regional blocs affect the interests and preferences of actors? Does the emergence of regional forms have implications for the construction of new identities and the deconstruction of established identities (at both elite and mass level)? Does regionalism accelerate or retard free trade and multilateral exchanges between states? Do regional agreements and institutions form a uniform threat to the nation-state and the international system of states? The pursuit of questions like these explains why many specialists in International Relations and International Political Economy regard the EU as worth studying.
The third broad approach aims to treat the EU as useful location for the study of policy-making dynamics. Here the EU is an instance of a complex policy system in which perspectives on policy-making developed largely in the context of national polities can be put to the test and perhaps developed. So, attention is turned to the interaction of interested actors and the processes of agenda setting, policy formulation, legislation, interest intermediation and policy implementation. The analysis of these processes raises questions about the location of power and the relationship between formal and informal policy processes. From this vantage point, the development of the EU affords an exciting opportunity to consider policy networks and the role of institutions in conditions where (old) national and (new) supranational politics overlap.
The final approach is less inclined to treat the EU and European integration as an instance of anything other than itself. Such an approach would regard the EU as a sui generis phenomenon. That is to say, there is only one EU and, therefore, European integration cannot be a theoretical
16    Theories of European Integration
testing site for the elaboration of broader generalizations. The inclination rather would be to treat the EU as an historically-rooted phenomenon, arising in utterly specific conditions and therefore without meaningful historical precedent or contemporary parallel.
Each of these broad approaches is open to work from many theoretical perspectives and one of the purposes of this book is to investigate these perspectives in greater depth. (For instance one theme which emerges in this volume is the current debate about the relative merits of 'International Relations' versus 'Comparative Politics/Policy Studies' approaches to the EU. In particular, Chapters 5 and 7 discuss this matter which could be seen as a debate between the second and third of the approaches sketched above.) Having said that, it might be argued that work in the fourth category inclines towards crude empiricism with its tendency to chronicle the intricacies of the EU system. It is, therefore, less able to offer insight into broader theoretical issues (Rosamond, 1995) or indeed into more normative questions about the 'real' problems facing Europe and its citizens (Hix, 1996).
One response to this criticism - from the fourth position perhaps -would be to argue that it is folly to attempt draw generalizations from the study of the EU because it is such a unique organization that emerged out of a unique set of historical circumstances. It possesses an institutional and legal architecture quite unlike both national political systems and other international organizations. Such sentiments latch onto the fundamental problem of what integration theory might achieve and especially whether it could do anything more than make systematic generalizations about European integration.
For reasons which will be discussed in later chapters, this has been a particular problem for approaches drawing their ammunition from International Relations scholarship. The founders of integration theory imagined that generalizations would emerge from their intensive case study of the European Communities. The results could then generate hypotheses for the study of regional integration in a more general sense. Such aspirations are made explicit in the work of the early theorists of European integration, such as Karl Deutsch (1957), Ernst Haas (1961; 1968), Phillippe Schmitter (1971) and Joseph Nye (1971), and were more or less integral to the foundation of the Journal of Common Market Studies in 1962. For those with an interest in treating the EU as an instance of regionalism the same applies. The comparability of, say, the Association of South East Asian Nations (ASEAN) and NAFTA with the EU is thrown into doubt if it can be established that the EU is fundamentally different in a number of respects.
This begs the question of whether the EU represents an n of 1, as James Caporaso puts it (Caporaso, 1997). For some the answer is clearly 'yes'.
Introduction    17
for example, William Wallace has drawn attention to the specificities of the historical experience in Western Europe:
The experience of deep integration within Western Europe does not ... provide a model for others to follow. Its historical development was rooted in a stage of economic development and a security framework that have now both disappeared. The institutional structures that west European governments agreed to under those past circumstances have managed to respond to the very different challenges posed by economic and industrial transformation in the 1970s and 1980s. Political, economic, and security motivations have been entangled in the evolution of West European regional integration from the 1940s to the 1990s.
(Wallace, 1994: 9)
Indeed, it is reasonably clear that the initiators of some of the newer regional blocs - for example in the Asia-Pacific - have explicitly identified the EU as an example of 'bad practice' which should not be emulated (Köhler, 1995). So, perhaps practitioners of regionalism elsewhere in the world are ill-advised to seek to emulate the EU. But what about theoreticians? If the EU is unique, and is nothing other than an instance of itself, then we have a dilemma. Findings cannot be generalized to other cases because of this uniqueness and as a consequence general theories of integration are not attainable. Moreover, it has been argued that the uniqueness of the EU is also a barrier to theorizing the EU in general terms. As one analyst has put it, 'We do not have a general theory of American or German politics so why should there be a general theory of the EU?' (Hix, 1996: 804).
The question is penetrating and raises all kinds of issues about disciplinary segmentation as well as the particular matter of how to study and theorize the EU. It is not just that the EU arose and evolved in historically specific circumstances. It is also the case that 'EU studies' has become a narrow specialism - largely a sub-domain of Political Science and International Relations. Simon Hix's objection is grounded in the view that bridges need to be (re) built between theories of comparative politics and EU studies and that the appropriate way to view the EU is as a polity rather than as an International Relations phenomenon. As Gary Marks notes, work drawing on this position has become commonplace in EU studies and generates comparative possibilities by 'slicing polities into subsystemic parts' such as interest groups, policy networks and so on (Marks, 1997: 2). This kind of work is presented by Mark Pollack (1997b) as following the guidelines for 'good' comparative social science (see King, Keohane and Verba, 1994). If the object of the exercise is to explain 'integration',  then the EU is indeed the only available case.  If other
18    Theories of European Integration
dependent variables are selected in conjunction with appropriate deductive theories then the n=\ problem vanishes. This, of course, favours the depiction of European integration in terms of the third approach outlined above. Having said that, Gary Marks (1997) suggests that comparison can go well beyond internal analyses of the EU system. Other organizations, he argues, exercise limited degrees of supranational authority and the fact that the EU is peculiarly supranational would not prevent the construction of a continuum. Also, as Alberta Sbragia (1992) has argued, the polity-like qualities of the EU beg comparison with other federal and quasi-federal systems. Germany and the United States can both be understood as 'federal' systems, but there are significant differences in the type of federalism that prevails. Clearly, two phenomena do not have to be identical in order to be compared effectively. Finally, Marks suggests that the radical processes of institutional change represented by the EU might be usefully compared to previous reorderings of authority such as the break-up of the Carthaginian Empire.
The pursuit of a general theory of European integration/the EU may indeed be misplaced. But, as Hix would no doubt acknowledge, the potential for EU studies to be a fertile site of social scientific theorizing is immense. Both US and German politics have been important venues for the development of particular branches of theory-driven political science, around ideas about pluralism, federalism, interest group liberalism, community power structures and the regulatory state (to name but a few). 'Integration theory' - defined here very broadly indeed - matters not just because of what it can tell us about the development of the EU or processes of regionalization, but also because of what it can tell us about the use of a fertile empirical location for the conceptual and theoretical development of the political sciences - again broadly defined.
The plan of the book
This book aims to do a number of things. The first is to provide an up-to-date cartography of theoretical work on European integration. This involves revisiting the classical 'pretheories' of integration, charting their development and their engagements with one another, and it involves a critical analysis of contemporary theoretical work. A second aim, perhaps more modestly realized, is to offer a 'sociology of knowledge' approach to this intellectual history. Rather too many accounts of integration theory fail to contextualize their material. Writers often succumb to the temptation of dismissing the theoretical work of several decades ago as 'wrong' or misguided. The point here, as stated above, is to set debates within their historical and intellectual contexts. Theorists obviously have
Introduction    19
eVe on the 'real world' as they see it. But - perhaps more importantly -
heórists  operate  within   particular   academic  contexts   and  particular
dons of what is (or isn't) good social science. A third and connected aim
•    to reflect upon the disciplinary questions about the social science of Furopean integration. If the book has an argument to advance here, then it
•     tnat 'international theory' has been too readily written off by contemporary writers seeking to offer theoretical treatments of the EU. To detach 'EU studies' from International Relations (again broadly defined) is to misunderstand and misread some very important developments in IR and IPE. It does not automatically necessitate a futile search for a general theory of European integration, but can open up important avenues of inquiry to complement the more public policy-oriented work on EU governance that has opened up in recent years.
The book is organized as follows. Chapter 2 explores the genesis of contemporary theoretical analyses of European integration. It examines the premises and claims of three early 'schools' of thought — federalism, functionalism and transactionalism — which found themselves attached in various ways to the unravelling experience of states in post-war Western Europe and to the growing formalization of the social sciences. Chapter 3 provides an extensive examination of the neofunctionalist perspective, perhaps the quintessential theory of integration, while the following chapter explores the way in which neofunctionalism and other self-conscious theories of integration were criticized by adherents and unsympathetic critics alike. Chapter 5 begins a survey of contemporary theoretical approaches to integration. The emphasis here is on two things: the attempt to resuscitate old 'paradigms' in the face of an apparently more receptive empirical 'reality' and the appearance of alternative ways of conceptualizing the EC/EU as a 'political system' in preference to an 'international experiment'. Chapter 6 offers a critical analysis of the predominantly state-centric character of much contemporary theoretical work on European unity, while Chapter 7 — in slightly more speculative mode - considers the broad canvas of contemporary international theory and suggests some productive linkages that might be made with the study of European integration. Chapter 8 takes time to consider some of the deeper questions about the 'sociology of knowledge' of integration theory its relationship to social scientific trends, the interplay between the theory and practice of integration and the knotty problem of the evaluation of theoretical work.