The study of European integration has in the past been plagued by the so-called sui generis problem: ‘the EU is considered somehow beyond international relations, somehow a quasi-state or an inverted federation, or some other locution’ (Long 1997: 187). At the empirical level of analysis, few would deny that the EU does indeed display unique characteristics, be it in its scope, institutional design, decision-making procedures or supranational legal identity. Yet many students of international relations would probably instinctively echo Moravcsik’s claim that ‘although the EC is a unique institution, it does not require a sui generis theory’ (1993: 474).1 The danger perceived by students concerned with global trends in international relations is that theories developed specifically to explain one particular manifestation of a more general phenomenon become so embedded in the more unique characteristics of their object of study as to seriously limit their range of general applicability. It is in this space between the richness of empirical observation and the parsimony required by theoretical generalisation that the sui generis problem arises.
Long has suggested that ‘the sui generis problem . . . is at one level less acute with the CFSP’, given that ‘the CFSP is intergovernmental and is probably better characterized as a process rather than as an institution’ (1997: 188). Pijpers takes this argument one step further when arguing, with reference to realism and the study of CFSP’s antecedent, EPC, that ‘the traditional paradigm demonstrates that EPC is a less unique phenomenon than some integration theorists prefer to believe’ and that ‘considering the record of EPC so far, or its cooperation procedures, it is difficult to discover original aspects of the Twelve’s approach in world politics’ (1991: 31–2). Yet few analysts of European foreign policy cooperation, even those working within international relations theories, would probably go as far as Pijpers. Thus, Long concedes that, when analysing the CFSP, ‘the sui generis problem does not disappear altogether’, mainly because ‘the CFSP is not an ordinary multilateral institution or process’ (1997: 188).
In fact, one might even argue that the sui generis problem is exacerbated in the case of the CFSP: not only is the CFSP a unique form of international cooperation, it is also a unique form of European cooperation. While it may therefore fit uneasily into existing theories of international relations because it is too unique, it fits equally uneasily into traditional European integration theory because it is not as unique as the Community. In other words, the supranational institutions, majoritarian decision-making procedures and legal character on which traditional integration theory relies are not present in the CFSP, nor have national foreign polices been superseded by a single European foreign policy. As Regelsberger noted with reference to EPC, ‘supporters of the more traditional concept of integration, where competence in foreign policy was expected to transfer from the national to the supranational level may be little satisfied with the evolution of EPC’ (1988: 4). The same, of course, applies to analysts working within the framework of traditional integration theory.
It is this double sui generis problem that this chapter aims to address. To do so requires identifying the defining characteristics of CFSP which emerge from the numerous contributions of participants in and analysts of European foreign policy cooperation which have appeared over the years, and which any theoretical explanation of the phenomenon must, as a minimum, address. Naturally, this is a considerable task that would require more space than provided here. However, the aim is not to provide an empirically neutral and comprehensive account of CFSP, but rather to highlight issues of theoretical relevance, and it will therefore be done with the various theoretical alternatives in mind. Hence the discussion will be structured around the following key issues of relevance to the main theoretical debates: the key actors and the institutional framework for their cooperation; the nature of their interactions and their style of cooperation; and the possible impact of participation in CFSP on national interests and foreign policies. It will be argued that while the sui generis nature of CFSP presents an acute problem for international relations theory, it is less pronounced with regard to traditional integration theory. The conclusion is, thus, that traditional neo-functionalist integration theory, while in some respects problematic when applied to intergovernmental cooperation, nevertheless provides the most promising basis for further theorising about CFSP.
CFSP: a unique phenomenon
It appears natural when analysing European foreign policy cooperation to take as a starting point its intergovernmental nature. The key features of the original EPC framework – separate institutional framework, exclusion of autonomous supranational institutions, consensual decision-making, absence of legally binding commitment and enforcement mechanisms – are still present in the provisions of the CFSP, despite the introduction of a number of ‘federal detonators’ (Hill 1993b). And although the initially strictly observed institutional separation of EPC from the Community has been relaxed somewhat over the years, mainly through gradual codification of informal practice, the pillar structure of the TEU (retained in the Amsterdam and Nice Treaties) bears witness to the persisting desire of most member states to resist any formal supranationalisation of their foreign policy cooperation. Thus, ‘at the formal level of reasoning it seems difficult to question CFSP’s inherent intergovernmentalism’ (Jørgensen 1997a: 167).
Yet CFSP participants and analysts alike have long warned against simplistic application of such a general term to as complex and unusual a process as European foreign policy cooperation. As one practitioner has put it, CFSP ‘may be called “intergovernmental” [but] the connotation of ‘normal international practice’ is grossly misleading’ (Von der Gablentz 1979: 694). Thus, ‘even if EPC shares . . . conceptual elements with historically familiar forms of diplomatic cooperation, the intensity and quality of EPC activities, however, go beyond these accepted concepts in the way that makes this characterization appear no longer applicable in any satisfactory way’ (Wessels 1982: 14). CFSP may be inter-governmental at the level of formal institutional arrangements, but the practices of European foreign policy cooperation which have emerged over the years, and the impact which they have had, even at the institutional level, are not easily captured by this term as traditionally defined in opposition to supranationalism (Jørgensen 1997a; Øhrgaard 1997). Before embarking on the task of explaining CFSP, it is therefore necessary to try to understand the main features of this unusual, and perhaps unique, phenomenon.2
Actors and institutional framework
CFSP is often described as an instance of intergovernmental cooperation. As noted by Jørgensen, ‘intergovernmentalism typically implies that member states continue to control decision-making’ (1997a: 169), and a cursory glance at the formal institutional arrangements for European foreign policy cooperation appears to confirm this characterisation of CFSP. In the original EPC framework laid out in the 1970 Luxembourg Report, all decision-making powers were invested in the foreign ministers, who were also to act as the main public representatives of the European Union in international affairs through the rotating Presidency. In their work, the foreign ministers would enjoy the support of the Committee of Political Directors in the national foreign ministries, but mainly at the logistical and preparatory level of cooperation. No central institutions were created, the Community’s supranational institutions were given no formal powers in EPC, and foreign ministers would meet as the Conference of Foreign Ministers in the capital of the Presidency rather than in the Council in Brussels.
EPC did, however, rapidly evolve into an unusually dense process of cooperation. The Luxembourg Report foresaw two annual meetings of foreign ministers and four annual meetings of the Political Committee (Part Two, Articles II.1(a) and III.1). In these early days of EPC, European foreign policy cooperation was still at an experimental stage, and ‘in November 1970, when the Foreign Ministers of the then six EC member states met for the first time to discuss matters of foreign policy, there were strong doubts that similar meetings would follow’ (Regelsberger 1988: 3). Yet the workload of EPC soon increased to a level which could not be adequately dealt with on such a relatively limited basis. Little more than a decade on from its inception, foreign ministers would meet formally more than once per month, and in times of international crises might find themselves consulting with each other on an almost daily basis.3
The density of cooperation was not, however, limited to the ministerial level of EPC and ‘on a practical, day-to-day basis, governments could not ignore the fact that delegation and communication below the European Council/Presidency level were required, especially in the absence of an EPC bureaucracy’ (M. E. Smith 1998a: 313). In 1973, the Copenhagen Report noted that the Political Committee had met nine times in the preceding year and, as a result of its increasing workload, had established a number of sub-committees, specific working parties and expert groups, and increased consultations between member state embassies and within the UN (Annex, Article 2). In addition, it sanctioned the establishment of a Group of Correspondents charged with monitoring the implementation of political cooperation and the running of the ‘correspondance européenne’ (COREU) telex network linking national foreign ministries (Part II, Article 3).4 A mere three years into its existence, EPC had thus developed into an ‘inter-diplomatic structure which does not simply limit cooperation to the highest level, but anchors it firmly in the diplomatic machinery of [member states]’ (Wessels 1980: 23). Furthermore, in recent years we have been witnessing ‘the beginnings of staff exchanges among foreign ministries and shared embassies’ (Hill and Wallace 1996: 6).
To be sure, none of these developments alter the fundamental fact that CFSP has remained essentially intergovernmental at the formal level of the institutional distribution of power. Indeed, no formal supranationalisation of European foreign policy cooperation has taken place. One can detect a process of what Allen has termed ‘Brusselisation’ – ‘a gradual transfer, in the name of consistency, of foreign policy-making authority away from the national capitals to Brussels’ (1998: 54) – but this has essentially been a matter of logistical rationalisation: meetings of working groups previously organised in the member state holding the Presidency were switched to the rooms of the Brussels-based Secretariat established with the Single European Act (SEA) in 1986 (Regelsberger 1997: 71–4).5 Similarly, while the Commission has seen its formal role upgraded from a right to be consulted to sharing the right of initiative and having co-responsibility for consistency and external representation, it has not acquired Treaty powers akin to those it enjoys within the Community, nor has it hitherto been very successful at expanding its limited competences under the CFSP (Allen 1998; Peterson 1998). The ‘federalist detonators’ contained in the CFSP provisions (Hill 1993b) have, as far as the institutional distribution of power is concerned, been left untouched by member states, mainly, it would appear, as a result of their ‘fear that any new decisions will set precedents for the CFSP that may bind them later or that will involve the Commission or [the European Parliament] to a greater degree than they desire’ (M.E. Smith 1998b: 154).6
From this observation, however, ‘it does not necessarily follow that the CFSP is a typical example of intergovernmental cooperation’ (Jørgensen 1997a: 168) in the traditional sense of tight governmental control with policy-making. Inter-governmentalism is a term used to describe cooperation between governments, but because of the intensity of cooperation within EPC, ‘governments did not monopolize the system to the extent assumed by intergovernmental approaches’ (M. E. Smith 1998a: 308). Without necessarily accepting that ‘transgovernmentalism was the key feature of EPC from the beginning’ (ibid.: 313),7 it is safe to say that transgovernmental interaction and interpenetration certainly have been a key feature of CFSP. This becomes abundantly clear when one examines the nature of interactions and style of decision-making within CFSP.
Interactions and decision-making
One consequence of the growing density of the CFSP agenda was the emergence of a strong transgovernmental network of national diplomats sharing ‘professional expertise and professional pride’ (Hill and Wallace 1996: 11). Through its regular meetings, it provided ‘a constant European training on the job for an important branch of government in finding European solutions for their problems’ (Von der Gablentz 1979: 694), resulting in ‘an important group of actors, as well as maintaining national loyalties, orient[ing] itself towards the development of common European positions’ (Wessels 1982: 15). At the level of individual relationships, according to Nuttall, ‘the Political Directors [were] on first-name terms’ and among the European Correspondents, ‘the esprit de corps of the group [was] even stronger than that of the Political Directors and many of them [became] personal friends’ (1992: 16 and 23). As the intensity of the process increased, so more and more officials, whether based in national foreign ministries, at embassies in third countries or on secondment to foreign ministries of other member states, would come into contact with each other and gradually come to see each other as ‘no longer ‘foreign’ but as colleagues’ (Hill and Wallace 1996: 12). Thus, as CFSP developed, ‘the more ‘Europeanised’ became the diplomats, including even those who had never dealt directly with European affairs before’ (Regelsberger 1988: 36).
The socialisation (or Europeanisation) effect to which this gave rise was further helped by the initial absence of formally specified procedures, providing participants in the process with the freedom to establish the ‘rules of the game’ as they went along, thus creating a sense of ‘ownership’ of the process. As Michael E. Smith points out, ‘CFSP insiders consistently stress the value of habitual processes of socialisation, building trust, and the adoption of pragmatic working habits in a decentralised system with no real compliance mechanisms’ (1998b: 151). Perhaps the most significant of these working habits was an unusually high degree of communication and information exchange which gradually contributed to the emergence of a ‘communauté d’information’ (de Schoutheete de Tervarent 1986: 49). This was given its most tangible expression in the ‘coordination reflex’, a phenomenon which participants in EPC all claim rapidly became a significant factor in the definition of national positions and in the search for common positions (Von der Gablentz 1979; Hurd 1981; de Schoutheete de Tervarent 1986; Nuttall 1992; Tonra 2001) and which was formally recognised as early as the Copenhagen Report (Preamble). Naturally, the coordination reflex did not exclude the possibility that a member state might ultimately decide to pursue its own objectives unilaterally, regardless of the views of its partners; but it did increase the cost of such actions as they would inevitably come to be viewed as resulting from open and direct disregard for partners’ views and could not be explained away as misunderstandings resulting from lack of information (Hurd 1981: 389; Hill 1997: 9). As Wessels points out, despite the absence of formal enforcement mechanisms, ‘the habit of cooperation and the self-obligation (coordination reflex) imply sanctions which are not based on legal rules but on group expectations of mutuality’ (1980: 23).
The initial lack of clearly specified procedures, and the development of a strong transgovernmental network, had a significant impact at the level of decision-making. In line with intergovernmentalist principles, and in contrast to cooperation within the Community framework, the formal rule of the process was that decisions taken were to be based on consensus. The corollary of this rule was often assumed to be that decisions reached would inevitably reflect the ‘lowest common denominator’ of national positions, given that any member state could, in principle, at any time prevent the emergence of a consensus by reference to its particular national interest. This was not, however, the case. As Wessels has pointed out, ‘the regular reference to unanimity points to only one aspect of reality; the consensus principle does not mean a stiff inflexibility of national positions, but it leads to dynamic adaptations’ (Wessels 1980: 23). Two former participants put it more directly: ‘it would be erroneous [. . .] to conclude that a compromise reached after lengthy deliberations would only reflect the lowest common denominator of national briefs’ (Von der Gablentz 1979: 698); ‘a lowest common denominator would only result if the procedure followed were that all the national positions were put on the table and whatever coincided became the European position’, but ‘this [was] not the case’ (Nuttall 1992: 314).
Instead, EPC operated ‘by talking incessantly’ (Nuttall 1992: 12) with a view to achieving common positions. At least two factors facilitated this approach. First, although formal decisions were taken during meetings at ministerial level, these meetings were not conducted in a vacuum, but were preceded by extensive preparatory work at sub-ministerial level. Thus, ‘decisions [were] made after painstaking and frequently long drawn-out processes of consultation, negotiation and coordination which [were] characterized by a complex inter-bureaucratic network between the member-states’ (Wessels 1982: 13). The inter-diplomatic socialisation effect and coordination reflex described above further reinforced this decision-making style as ‘the preparatory stages of national policy-making [would] already [be] infused with shared information and consensus building’ (Hill and Wallace 1996: 12). The ‘special code of conduct among Community diplomats’ which emerged included, in addition to the commitment to inform, a commitment to consult with a flexible attitude, of which ‘to face partners with a fait accompli [was] considered to be a particularly grave contravention’ (Von der Gablentz 1979: 691). The result was that, at all levels, ‘EPC [did] not operate under the perpetual threat of veto’, but instead participants would make ‘genuine efforts to reach a positive outcome’ (Nuttall 1992: 12).
The second important factor in facilitating consensus-building within EPC was the role played by the Presidency. As in the Community, the Presidency was expected to look after the interests of the group as a whole and to ‘raise its horizons beyond the pursuit of immediate national interests’ (Wallace 1983b: 5). The Presidency offers the holder, particularly when a small member state, the opportunity to achieve ‘enhanced visibility and presence on the international scene’ (Lorenz 1996: 236) as a representative of the European Union as a whole, but this depends on the ability to ensure backing from one’s partners and thus avoid paralysis. In CFSP, the successful President would therefore, first and foremost, need to ‘acquire the confidence and understanding of its partners’, then ‘endeavour to know the precise limits of possible concessions of each partner’ and, eventually, ‘identify the areas of possible agreement and seize upon them quickly in order to achieve common action’ (de Schoutheete de Tervarent 1988: 79). The weak institutional framework and relatively undefined scope of CFSP made this both an easier and a more delicate task. The absence of a neutral mediator such as the Commission could, potentially, make agreement more difficult to achieve, but it also meant that there was ‘no Commission to act as a scapegoat for failure or alibi for inaction’ (Wallace 1983: 4). Similarly, the absence of a predetermined agenda could be a paralysing factor, but it also enabled member states to agree implicitly that some sensitive issues, for domestic or historical reasons, were the ‘domaines réservés’ of member states and thus not a subject for deliberation. On balance, however, both factors probably played a positive part in ensuring the commitment of member states to a process which, due to its flexibility, was unlikely to pose a direct threat to any perceived national interests.
National interests and CFSP
In principle, the institutional design of CFSP, with its core unanimity requirement and lack of legal obligations,8 protects its members’ national interests. While the original commitment to work ‘for a harmonisation of views, concertation of attitudes and joint action when it appears feasible and desirable’ (Copenhagen Report, Part Two, Article I(b))9 has been upgraded to an obligation on member states to ‘ensure that their national policies conform to the common positions’ (TEU, Article J.2(2)) and a provision that ‘joint actions shall commit the Member States in the positions they adopt and in the conduct of their activity’ (TEU, Article J.3(4)), nothing in the CFSP provisions can in practice prevent a member state from unilaterally pursuing its national interests, even if in contradiction of agreed common positions. The existence of such national interests has often been seen as the most serious obstacle to the emergence of a truly common European foreign and security policy. Before EPC had even been established, Hoffmann (1966) warned that a ‘logic of diversity’ in the sphere of foreign policy would not only prevent the integration process from spilling over into this traditional area of ‘high politics’ but would also, ultimately, put a brake on economic integration. While this prediction has yet to come true, in the wake of CFSP’s perceived failures in a number of areas, most notably the Gulf crisis and the disintegration of Yugoslavia and in its marginalisation during the US-led airstrikes against Afghanistan, this concept continues to attract the attention of analysts of European foreign policy cooperation (Hill 1997; Zielonka 1998).10
There can be little doubt that national interests have continued to coexist within CFSP instead of being replaced by a distinctly ‘European’ interest. As Ifestos has warned, ‘an observer should avoid searching for the existence of an autonomous ‘European interest’ and instead examine where, when and how the national interests of the member states converge, or at least are not in conflict’ (1987: 106). On the other hand, if CFSP decisions rarely reflect the lowest common denominator of national positions, and if the high degree of socialisation has had an impact on the actors’ perceptions of their interests, then it should be possible to detect even subtle changes in member states’ definitions of these interests. Thus Michael E. Smith identifies a number of areas in which this appears to have taken place, both on regional issues such as the Middle East, Central America and South Africa, and on substantive issues such as nuclear non-proliferation and security issues more generally (1998c: 30–8).11 While the substantial policy impact of CFSP has undoubtedly been stronger in the small member states, at times compelling them to create foreign policy in areas where none previously existed,12 even the larger member states have at times, more or less enthusiastically, adapted their national foreign policies to common CFSP positions.13 And once these common positions have been more or less formally codified in the ‘acquis politique’ member states have, to a large degree, found themselves bound by them in their subsequent actions, thus adding to the ‘de facto binding quality’ (Wessels 1982: 15) of their commitment in CFSP, increasing the expectations of third countries and creating ‘a certain knock-on effect from one issue to another’ (Hill 1983: 199).14
Perhaps the most tangible evidence of this knock-on effect is the gradual widening of the scope of European foreign policy cooperation to include, first, ‘the political aspects of security’ (London Report, Preamble), then ‘the political and economic aspects of security’ (SEA, Title III, Article 30.6(a)) and, eventually, ‘the progressive framing of a common defence policy . . . which might lead to a common defence, should the European Council so decide’ (TEU, Article J.7(1)). The European Union – even with its Rapid Reaction Force – is arguably still some way from having an operational security policy by comparison with, say, NATO, in whose shadow it remains. However, from an internal and historical viewpoint the very acceptance of security and defence policy as a core policy competence marks a significant development. As noted by Kirchner with reference to the SEA, ‘considering that the subject was taboo for such a long time, the explicit reference to security cooperation in the EC [was] a big step forward’ (1989: 8). Admittedly, the dynamics behind this development were essentially external, with developments in the international system in the 1980s initially putting security considerations firmly back on the agenda and the end of the Cold War redefining European security interests. But pressures also came from within, especially in terms of a British determination to re-assert leadership within the European Union and seeing defence and security as an area of comparative advantage for so doing (White, 2001: 118).
Somewhat paradoxically, the key to understanding these adaptations appears to lie in the very fact that CFSP has been perceived by its members to serve a number of useful functions for the pursuit of national foreign policies (Hill 1983a, b; 1996b). CFSP has at various times served as ‘amplifier’, ‘cover’/‘alibi’ or ‘legitimiser’ for member states in the pursuit of their national foreign policies.15 Yet while this instrumental, or functional, value of CFSP to member states would initially appear to confirm the intergovernmentalist interpretation of European foreign policy cooperation, a commitment such as this, based as it is on self-interest, need not make CFSP peripheral, let alone subordinate, to national interests. Indeed, the very fact that CFSP has proved so useful for its member states has transformed it into an interest in itself. In other words, because CFSP has been a key, and for some member states the key,16 instrument in the pursuit of national foreign policy, maintaining this system has come to be widely seen as a prerequisite for a successful national foreign policy. As Wallace pointed out almost two decades ago, ‘in little more than ten years, [EPC] has become an accepted and indispensable part of the national foreign policies of every member state’ and ‘its value to all member governments . . . is now clearly established’ (1983b: 14–15).17
This does not, of course, imply that foreign policies have become totally subordinated to CFSP, nor should it conceal the fact that the CFSP faces an ongoing crisis of commitment.18 Indeed, just a few years ago a trend was detected towards what Hill termed the ‘renationalisation’ of foreign policy, even in traditionally committed member states such as Germany and Italy (Hill 1997: 2). Since then the CFSP’s institutional structures and policy capacity have grown significantly. The installation of the High Representative, the development of military policy-making structures and the construction of the Rapid Reaction Force all point towards something more than just the ‘European rescue of national foreign policy’ (Allen 1996: 288) but something less than a quasi-federal EU foreign, security and defence policy (White 2001).
CFSP: a sui generis phenomenon?
It is clear from the above that CFSP presents a serious challenge to mainstream international relations theory. This challenge is two-fold. First, traditionally dominant strands of international relations theory, such as (neo)-realism or neo-liberal institutionalism, appear ill-equipped to account for some of the defining characteristics of CFSP. The traditional realist paradigm, with its emphasis on differing national interests and an overriding concern with sovereignty, may provide at least a partial explanation of why member states have resisted the supranationalisation of their foreign policy cooperation. Yet it cannot account for the density and quality of cooperation within CFSP, nor the significant impact this appears to have had on member states’ perceptions of their national interests and their growing commitment to the enterprise as evidenced in the Amsterdam Treaty and subsequent agreements on the creation of a European Security and Defence Policy (ESDP). This problem becomes even more acute in the structural version of realism (or neo-realism), which has traditionally rejected any notion of an institutional impact on national interests and behaviour on the grounds that the anarchic structure of the international system compels states to focus on their individual security at the expense even of potential mutual benefits from institutionalised cooperation.19 By contrast, neo-liberal institutionalists, while recognising the constraints on cooperation created by the anarchic structure of the international system, argue that international institutions can nevertheless help states realise some of the benefits of cooperation by reducing mutual suspicions and increasing behavioural predictability through the promotion of negotiations on the basis of shared information and commonly agreed rules within a stable framework.20 However, while international institutions may in this sense constrain and clarify the strategic options open to essentially self-interested sovereign states, neo-liberal institutionalists do not assume that they will seriously affect the fundamental interests and preferences of these states. In other words, they are essentially instrumental in states’ pursuit of their exogenously defined national interests.
More recent strands of international relations theory, such as multilateralism and social constructivism, attempt to address this problem of institutional impact on national interests. Multilateralism focuses on the generalised principles of conduct which are embedded in international institutions and which can, at times, compel states to sacrifice individual short-term interests for the greater mutual benefits involved in the very participation in multilateral cooperation.21 In other words, institutionalised cooperation becomes an interest in itself, coexisting alongside and competing with other national interests. Social constructivists take this argument one step further by arguing that interest formation must be seen as endogenous to institutionalised cooperation; that is, in part resulting from the process of cooperation itself.22 According to constructivists, states do not merely use international institutions as arenas or instruments in the pursuit of their interests, but are influenced by them to the extent of internalising the norms of behaviour that are embedded in them. These norms, in turn, affect the way in which states perceive their social identity, and thus their interests. However, whereas these theoretical frameworks may initially appear well suited to capturing the defining characteristics of the CFSP experience – and have indeed been shown to be so (Long 1997; Jørgensen 1997a; Jakobsen 1995; M. E. Smith 1998c) – they illuminate the second problem faced by international relations theory in explaining CFSP: while drawing attention to a key aspect of institutionalised cooperation, they often suffer from ‘empirical ad hocism’ in the sense that they have difficulties in generalising about ‘when, how, and why [social construction] occurs, clearly specifying the actors and mechanisms bringing about change [and] the scope conditions under which they operate’ (Checkel 1998: 325).23 These approaches therefore hardly help move the study of CFSP beyond the rich empirical descriptions and generalisations which already abound and towards a coherent theoretical explanation of the phenomenon.
As far as international relations theory is concerned, the sui generis problem in CFSP thus remains largely unresolved. Should the multilateralist or constructivist approaches develop into something resembling a coherent theoretical framework, then they would surely offer a promising way forward. In the meantime, however, it would appear worthwhile examining whether traditional integration theory, with its well-developed assumptions and causal mechanisms, might not offer a better starting point for any attempt at explaining the phenomenon of CFSP. In other words, although CFSP remains a sui generis problem for both international relations theory and integration theory, the problem may be less acute for the latter, given the proximity of CFSP to the Community and the identity of the individuals involved in the two institutions. The purpose of the following section is to examine this proposition.
CFSP: integration through intergovernmentalism?24
Neo-functionalist integration theory has traditionally been rejected as a useful framework for conceptualising the CFSP experience. At the heart of this rejection lies an often implicit supranational–intergovernmental dichotomy that posits a trade-off between intergovernmental cooperation and supranational integration (Jørgensen 1997a; Øhrgaard 1997). To be sure, if one accepts Haas’ formal definition of integration as ‘the process whereby political actors in several distinct national settings are persuaded to shift their loyalties, expectations and political activities toward a new centre, whose institutions possess or demand jurisdiction over the pre-existing national states’ (Haas 1958: 16), then clearly it is difficult to conceive of CFSP as a process of integration. In CFSP, no central institutions have emerged which possess jurisdiction over member states’ foreign policies, nor has CFSP led to the emergence of a political community defined as ‘a condition in which specific groups and individuals show more loyalty to their central political institutions than to any other political authority’ (Haas 1958: 5). Despite the commitment of its participants to the process, in CFSP ‘a pronounced shift of loyalties to a new centre of decision-making with jurisdiction over the national level cannot be identified’ (Wessels 1982: 13). As put rather starkly by Ifestos, ‘seen in terms of traditional integration theory, [EPC] is malintegrative, if it reinforces inter-governmental cooperation rather than supranational integration’ (1987: 208).
Yet one should be careful not to dismiss neo-functionalism merely on the grounds that CFSP has failed to develop the institutional characteristics associated with Haas’ definition of political community. Neo-functionalists, including Haas, were always more ambivalent about the nature of integration than the formal definition quoted above suggests. Even in early neo-functionalist writings there was an ever-present tension between integration as defined by a particular institutional outcome and integration conceived of as a process of cooperation and decision-making. Thus Haas argued that ‘conceived not as a condition but as a process, the conceptualisation relies on the perception of interests and values by the actors participating in the process’ and that ‘integration takes place when these perceptions fall into a certain pattern and fails to take place when they do not’ (1958: 11).25 Similarly, Lindberg suggested that ‘collective decision-making procedures involving a significant amount of political integration can be achieved without moving toward a ‘political community’ as defined by Haas’ and thus advocated ‘a more cautious conception of political integration, one limited to the development of devices and processes for arriving at collective decisions by means other than autonomous action by national governments’ (1963: 5).
As this section will seek to demonstrate, it is possible to distil from neofunctionalism a framework for conceptualising the form of cooperation which has developed within CFSP. The main challenge is to separate the processes that characterise integration from their institutional framework and ultimate outcome. This requires, first, an acceptance of the ambivalence which characterises neofunctionalist writings on this point and, second, that this ambivalence be exploited to relax some of the more deterministic assumptions of neo-functionalism concerning the actors, their motives and the institutional factors which condition their behaviour.
Actors and institutional framework
In terms of the actors involved and the institutional framework for their interactions, CFSP appears to pose at least two problems for neo-functionalism as traditionally conceived: first, ‘transnational coalitions which break the national monopolies for taking the final decisions [were] not existing in [EPC]’ (Wessels 1980: 22); and, second, ‘no parallel has been created within EPC to the role of the Commission as ‘motor’, ‘agent’, ‘guardian of the treaty’ and ‘executive body’ which can rely upon a bureaucracy with continuity, experience and specialized knowledge’ (Wessels 1982: 14). These characteristics would initially appear to correspond more closely with the intergovernmentalist focus on national governments setting the agenda and controlling decision-making than with the neo-functionalist emphasis on social and economic transnational actors as demandeurs of integration and supranational institutions serving as a focus for these demands and possessing autonomous powers to deliver the desired integration.
This version of neo-functionalism derives from an understanding of integration as an essentially economic process but, although widespread, this interpretation is too simplistic.26 Indeed, neofunctionalists always recognised the central importance of national governments in the process of integration. Thus, Lindberg explicitly acknowledged that ‘only the positive action of legitimate national authorities (the nation-states) can be the ultimate basis of integration’ (1963: 291) and that ‘political and economic integration cannot be expected to succeed in the absence of a will to proceed on the part of the Member States’ (1963: 11). Similarly, Haas noted that ‘the new central institutions depend on the good faith of the old power centres for the realisation of their aims, [in part] because of the real powers retained by national governments’ (1958: 58). As a result, in his specification of the relevant political actors involved in the process of integration, Haas explicitly referred to ‘policy-makers in government’ and ‘higher civil servants’ (1958: 17). The fact that CFSP remains within an intergovernmental institutional framework, where the main actors are governments and national civil servants, does not, therefore, constitute a valid reason for rejecting, a priori, the potential relevance of neo-functionalism to explaining the phenomenon.
Interactions and decision-making
It is in the interactions and the style of decision-making that neofunctionalists located the main sub-processes driving the overall process of integration forward. Indeed, as Lindberg argued, ‘if political integration . . . is going on, then we would expect to find change in the behaviour of the participants (1963: 9). One sub-process to which neofunctionalists attached great importance in achieving such change was ‘élite activation’ (Lindberg 1963: 9). One important way of achieving such activation was through the growing socialisation resulting from working together on transnational problem-solving. According to Haas, ‘as the beliefs and aspirations of groups undergo change due to the necessity of working in a transnational institutional framework, mergers in values and doctrine are expected to come about, uniting groups across former frontiers’ (1958: 14). Lindberg made a similar point: ‘there is strong evidence that this sort of interaction [between high policy-makers and civil servants] contributes to a ‘Community-mindedness’, by broadening perspectives, developing personal friendships, and fostering a camaraderie of expertise, all of which come from being involved in a joint problem-solving operation’ (1963: 286).27
Neofunctionalists foresaw that this growing socialisation among elites with influence on the decision-making process would gradually lead to changes in the style of decision-making. Increasingly, conflict resolution would come to replace simple bargaining. At least three forms of conflict resolution could be distinguished (Haas 1961: 367–9; Lindberg 1963: 11–12). The first, ‘lowest common denominator bargaining’, resembled the style traditionally associated with inter-governmental diplomacy, with the overall outcome determined by the least cooperative state. The two other forms of conflict resolution both involved some compromise in national positions. Conflict resolution based on ‘splitting the difference’ involved trading off concessions across issues with a view to finding a mutually satisfactory overall outcome, while ‘upgrading of common interests’ involved leaving the most controversial issues aside and concentrating on those issues on which agreement could most easily be achieved, in the hope that this would later facilitate agreement on the initially controversial issues. Of these three styles of conflict resolution, the two latter were perceived to have the greatest integrative impact, with conflict resolution based on the ‘upgrading of common interests’ representing ‘the true contribution to the art of political integration’ (Haas 1961: 369).
It should be noted, in this context, that neofunctionalists did ascribe a key role to central institutions in facilitating this type of conflict resolution. According to Lindberg, such institutions were necessary ‘in order to represent the common interests which have brought the Member States together, and in order to accommodate such conflicts of interest as will inevitably arise’ (1963: 8).28 However, Lindberg explicitly warned that this should not be mistaken for an argument in favour of supranationalism. What he was referring to were ‘Community’ institutions, and he was adamant that ‘the Council of Ministers clearly considers itself a Community institution and not an intergovernmental body’ (1963: 285). It could therefore not be assumed that national governments, acting within the framework of the Council, would always resort to lowest common denominator bargaining in the pursuit of their individual national interests. As Haas warned, ‘the supporters of Council powers may be . . . seriously mistaken in thinking that an intergovernmental structure automatically guarantees the prevalence of diplomatic decision-making techniques and thereby controls integration’ (1958: 487). Indeed, in a specific reference to the three possible styles of decision-making, Lindberg claimed that ‘conflict resolution in the Council usually follows an upgrading-of-common-interests pattern’ (1963: 285).
National interests and common policies
Implicit in the two processes described above – socialisation and upgrading of common interests – is the idea that the participants might eventually come to conceive of their interests differently, as a direct result of their participation in the enterprise. However, neofunctionalists did not expect this to happen in a purely random and self-generating fashion. Instead, they argued that these processes would exist in a mutually reinforcing relationship with a third process, that of ‘spillover’. This process, the one most often associated with neo-functionalism, operated in the following way: ‘earlier decisions . . . spill over into new functional contexts, involve more and more people, call for more and more inter-bureaucratic contact and consultation, thereby creating their own logic in favour of later decisions, meeting, in a pro-community direction, the new problems which grow out of the earlier compromises’ (Haas 1961: 369). Thus, participants in the process would become increasingly caught up in the web of their previous decisions, making further decisions in an ever growing number of related areas necessary lest the initial decision become ineffectual. The spillover process could manifest itself either in the expansion of cooperation into another, related area – ‘spillover in scope’ – or in a strengthening of the commitment to cooperation in the initial area – ‘spillover in level’ (Schmitter 1969: 162).
One common misinterpretation of neo-functionalism is that the spillover process would eventually gather such momentum as to become automatic, causing member governments to lose control with a process increasingly driven by trans- and supranational actors. While it is true that neofunctionalists did expect increased cooperation to limit somewhat the abilities of member governments to pursue their individual interests in total disregard of the interests of their partners, spillover nevertheless ‘assume[d] the continued commitment of the Member States to the undertaking’ (Lindberg 1963: 11). In other words, national governments would only allow spillover to occur if it was perceived as necessary to the realisation of their interests. Yet the decision to allow spillover would not necessarily be a conscious, strategic one, but would to a large degree be determined by pressures resulting from the unforeseen consequences of previous decisions, including the socialisation which cooperation to achieve these decisions would have brought about. Thus, while Haas did recognise that ‘the primary task of the Council, in its own view, is to safeguard the interests of member states’, he also claimed that in the Council, ‘national interests are always compromised; they are never maintained in the face of the ‘atmosphere of cooperation’ which prevails’ (1958: 489–90). Thus, national interests would be subtly redefined as a result of participation in the enterprise of cooperation.
According to Haas, the prospects for these three processes – socialisation, upgrading of common interests and spillover – would be further facilitated if a sense of ‘engagement’ prevailed among the participants: ‘if the parties to a conference enjoy a specific and well-articulated sense of participation, if they identify themselves completely with the procedures and codes within which their decisions are made, they consider themselves completely ‘engaged’ by the results even when they do not fully concur in them’ (1958: 522). In other words, a sense of ‘engagement’ might cause participants to be willing to sacrifice individual, short-term interests in the expectation that mutual commitment to the process of cooperation itself, and the atmosphere of compromise which it had fostered, would facilitate the achievement of greater and more important interests in the long term. This did not imply that participants had generally abandoned their individual interests, or were no longer acting in a self-interested way. Indeed, Haas explicitly recognised this pervading fact when he argued that ‘rather than relying on a scheme of integration which posits ‘altruistic’ or ‘idealistic’ motives as the conditioners of conduct, it seems more reasonable . . . to focus on the interests and values defended by the major groups involved in the process’ (1958: 13). But, as Lindberg noted, while ‘there is always the possibility of a calamity . . . it would seem almost impossible for a nation to withdraw entirely from integration’ (1963: 291).
CFSP: integration through intergovernmentalism?
In the more flexible version proposed above, neo-functionalist integration theory clearly possesses significant explanatory power when applied to CFSP. The social-isation of participants, including government ministers, with its engrenage effect, the style of decision-making based on the upgrading of common interests, and the limited but gradual spillover, in both level and scope, are all evident in CFSP. More importantly, all these processes have been underpinned by a strong sense of engagement which has, to a large extent, made up for the absence of any legal obligations and enforcement mechanisms. Even the concepts used, independently of each other, by CFSP participants and analysts, on the one hand, and neo-functionalists, on the other, are strikingly similar: Europeanisation/socialisation, esprit de corps/elite activation, domaines réservés/upgrading of common interests, acquis politique/engagement, knock-on effects/spillover. Neo-functionalism, despite being associated with the supranational experience of the Community and thus often assumed to be of little relevance to explaining an instance of intergovernmental integration such as CFSP, would thus, initially, appear better suited to overcoming ‘its’ sui generis problem than most strands of international relations theory.
Yet even the most flexible reading of neo-functionalism cannot obscure the one difficulty which remains when applying integration theory to CFSP: in the absence of a political community in which supranational institutions possess the powers necessary to compel reluctant member states to honour their informal commitments, any process of integration will always remain vulnerable to set-backs, whether in the form of individual cases of veto, parallel unilateralism or defection, or, at a more general level, in the form of a trend towards the renationalisation of foreign policy.29 As suggested elsewhere (Øhrgaard 1997), one way to overcome this problem is to view the integration process as a three-stage process of socialisation, cooperation and formalisation, where the initial trust and commitment generated through socialisation permit a style of decision-making based on the upgrading of common interests and, eventually, the formalisation of the spillover in both scope and level to which this gives rise. CFSP would appear to have evolved in this way, with the Copenhagen Report, London Report, SEA, TEU, Treaty of Amsterdam and Treaty of Nice successively codifying informally developed practices and gradual policy expansion.30 Naturally, this conceptualisation of intergovernmental integration leaves the question of the ultimate institutional destination of European foreign policy cooperation unanswered. In this, however, it only reflects one of the most fundamental historical characteristics of CFSP.
More than a decade ago, two leading specialists deplored the failure of an ‘academic community unable either to relate EPC into any meaningful system theory, integration theory or international relations theory let alone create a new EPC general theory’ (Weiler and Wessels 1988: 229). They noted that ‘although there has been plenty of academic discussion and extremely fruitful analysis of many facets of EPC, the term theory as such has, one gets the impression, often been studiously avoided’ (ibid.: 232), resulting in ‘too many case studies, ad hoc ‘lessons’ from limited experiences and organizational description, but . . . too little theoretical mediation’ (ibid.: 230). While acknowledging that ‘the illusive [sic!] general theory continues to illude [sic!]’ (ibid.: 229), they nevertheless purported to ‘‘theorise’ unashamedly since we believe that this very sentiment (of the irrelevance of theory) has denuded discussion of EPC in the last decade’ (ibid.: 230). Theorising, they argued, would help ‘provide a coherent account and explanation’ (ibid.: 234) of the phenomenon of EPC.
One decade on from this self-critique, the literature on CFSP reveals that the academic community has made but modest progress towards providing such a coherent account and explanation of the CFSP experience. Arguably, the last decade has witnessed more theorising on CFSP than the preceding two decades.31 Yet in analyses of CFSP, theorising remains subordinate to case studies and organisational description,32 with the result that ‘the problems in grasping fully the phenomenon of EPC, and in analysing the factors that have led to success or failure’ (Pijpers et al. 1988: 259) remain unsolved. This continued failure may, as argued by Hill and Wallace, be put down to ‘the elusive character of the phenomenon’ (1996: 1) itself or, as suggested by Long, to ‘the paucity of things to theorize about’ (1997: 184), but these characteristics are not exclusive to CFSP, and the explanation for the failure to theorise is therefore likely to lie elsewhere.
As argued throughout this chapter, the problem lies in the sui generis problem which CFSP continues to pose for traditional theories, whether of international relations or European integration. As a result, the study of European foreign policy cooperation remains at a pre-theoretical stage, where individual concepts and partial explanations continue to appear to hold out the best promise for explaining CFSP. As some CFSP scholars have warned, ‘attempts to apply a theory to the totality of CFSP seem to cast the aims too high’ (Weiler and Wessels 1988: 237).33 Others have gone even further and rejected such attempts as ‘inappropriate and foolhardy’ (Holland 1991: 5), and for these authors, ‘the failure to create a [CFSP] general theory . . . is perhaps in fact to be welcomed’ (Bulmer 1991: 89–90). However, as suggested by Weiler and Wessels more than a decade ago, the sheer amount of empirical case studies and pre-theoretical analysis available increasingly requires, and indeed deserves, some form of ordering into a coherent explanation. This chapter has suggested one way of seeking theoretical parsimony without sacrificing the most defining empirical knowledge which has been generated about CFSP over the years. As such, it is meant merely as a contribution to an ongoing debate, not as a final answer to what remains, at this stage, a considerable challenge.