On the emergence of ESDP and EU-NATO cooperation

Table of contents

Preface

Abbreviations

Table of contents

1 Introduction

2 Reflections on theory, philosophy of science, and methods
2.1 Puzzlement, causal mechanisms, and inductive process-tracing
2.2 Constructivism vs. rationalism?
2.3 ESDP as part of European integration

3 The new security agenda as a catalyst
3.1 Sovereignty and intervention revisited: changes in international law
3.1.1 Building consensus on intervention after the Cold war
3.1.2 Kosovo: outside the limits of consensus
3.1.3 The current state of consensus: responsibility to protect
3.2 The EU responds

4 Opening black boxes
4.1 The United Kingdom: governmental and policy change
4.1.1 British preference change on Europe
4.1.1.1 The Major government
4.1.1.2 The Blair government
4.1.2 The mechanism that led to St. Malo
4.1.2.1 New causal ideas as road maps
4.1.2.2 The impact of an epistemic community
4.1.2.3 The catalytic effect of the Kosovo crisis
4.1.3 A mechanism within the mechanism
4.2 The United States: burden-sharing and leadership
4.2.1 The Kosovo war
4.2.2 The incoming Bush administration

5 Toward Concordia
5.1 The path of partial Europeanisation
5.2 Berlin Plus
5.2.1 The negotiations
5.2.2 Problems, perspectives and prospects
5.2.2.1 Different purposes and perceptions
5.2.2.2 Implementation
5.2.2.3 Future

6 Conclusion

7 References

Preface

This thesis deals with a topic I have come across much earlier, and about which my astonishment never really ceased: the fact that there are military missions “out there” conducted under the flag of the European Union. In my view, the hoisting of the blue flag with the twelve golden stars in Macedonia in March 2003 changed the course of international politics – for the better or the worse – in introducing the European Union as an actor in military affairs, this ultimate stronghold of state sovereignty which deals with the final control over organised violence. From that moment on, the European project had an outward military face, which it will continue to display, at least in the European periphery, perhaps farther away – for the better or the worse, I insist, not daring to judge the consequences of this, neither for the European project itself nor for its neighbours or for the wider world as a whole.

The thesis can be read as a personal attempt to come to grips with this single moment in time and its implications, building on what I have learned during my studies of International Relations. It reflects my wondering about the EU as a military actor and about its cooperation with NATO, asking myself how one could best understand and explain this development, with the theories and methods of the discipline of International Relations at hand.

I first dealt with questions of European security in seminars of Dr. Ingo Peter’s at the Otto-Suhr-Institut für Politikwissenschaft of the Freie Universität Berlin, and I am very grateful for the insights I earned with him. I furthermore want to thank my supervisors, Prof. Bertrand Badie of Sciences Po Paris, and Prof. Thomas Risse of the Freie Universität Berlin, for valuable guidance and comments on earlier drafts. Then, I would like to thank all practitioners I have spoken to for their kindness in accepting my questions, and for their openness in answering them. This regards officials of different nationalities in Paris, Berlin, and Brussels, in service of their respective countries or of international bodies. As I have pledged to respect the confidentiality of our conversations, I in the text refer to these interviews mentioning only their place, not the name, the nationality, or the function of the respective person. I spoke with officials of different nationality in all three places, so that city names do not indicate any national views. All interviews were conducted between June and August 2006.

Most importantly, I want to seize the opportunity to most warmly thank my parents who invariably and invaluably supported me during all my studies – Danke Euch tausendfach!

Finally, I beg the reader’s pardon for all the harm I might have inflicted onto the English language, which is not my own, but which I adopted for practical reasons. I hope that mistakes and linguistic ineptitude do not obstruct the understanding of my central ideas and arguments, and I ask those who have a better command of this language than I do for leniency.

Frank Stadelmaier

Cité Internationale Universitaire de Paris, Maison Heinrich Heine

August 2006

Abbreviations

illustration not visible in this excerpt

1 Introduction

“[T]he [European] Union must have the capacity for autonomous action, backed up by credible military forces, the means to decide to use them, and a readiness to do so, in order to respond to international crises.”

This statement of the Franco-British St. Malo declaration of December 4 1998 remains the central turning-point in recent European history in security affairs.^[1] It elicited an astonishing development regarding the European Union’s willingness to act militarily, and also regarding its capacities to do so, as incomplete as these still may be. The development that followed could be seen as militarising the European Union, until then an exclusively civilian actor in international affairs, but also as lending it more visibility, more contours on the scene of world politics, notwithstanding its still considerable limitations in defence and security affairs. These limitations are geographical, concentrating first and foremost on Europe itself and its immediate neighbourhood, but also on some areas farther away in Africa, and they are functional, limited to what member states are willing and able to offer in terms of military means in a particular crisis. – And it is true that these European capacities are indeed limited, even if political will could be taken for granted, which of course is not the case, but depends on the circumstances of an arising crisis.

However, since December 1998 some decisive steps have been taken by member states to equip this odd institutional body which is the EU with a military arm for its Common Foreign and Security Policy (CFSP), whatever serious limitations the latter in itself may still have. All the limitations that are here alluded to, regarding the CFSP and the European Defence and Security Policy (ESDP) itself, do not hinder the fact that ESDP has been a very dynamic field in the development of the EU for the last eight years. Progress has indeed been more decisive and quicker than many would have thought after the Franco-British declaration of December 1998. Who would have thought in early 1999, the Kosovo crisis evolving, that the EU, helpless again in the face of a humanitarian and political crisis in the Balkans, needing the assistance of its powerful ally, the US, would by 2003 be in charge of a peace-keeping mission in Macedonia, by 2004 even of the Bosnia mission? And who would have thought that by 2006 it would have conducted two missions as far away as the Democratic Republic of the Congo (DRC), one of which being a straightforward peace-enforcement mission as early as 2003 (the mission Artemis in the Ituri region around the town of Bunia).

The importance of this astonishing turn in European history of security affairs, i.e. the emergence of the European Union as an actor on the scene of security affairs, is not to be underestimated. If European states are now beginning to take into their own hands the management of their neighbourhood’s security via their supranational integration project, including by military means, they may emancipate themselves in the years to come from the tutelage of the remaining superpower, after having been dependencies or protectorates of the two superpowers from 1945 onwards. We do not yet know the level to which this emancipation may rise, and what respective roles the states themselves and their supranational institution, the EU, will therein play in the end, but the inclusion of the EU in this process constitutes a distinctive feature, introducing a new kind of actor to the arena of military security affairs, an actor that is of another character than mere alliances. The course of this development will have important consequences on overall relations within the Western world, i.e. on the relations between Europe and the states that make it up on the one side, and the US on the other.

If the present development of ESDP goes on to transcend to some degree national European security policies, and if European states manage to co-ordinate and strengthen their capabilities in foreign and security policies, the EU may develop into an autonomous actor in world politics of its own right, leaving behind strictly state-centric traditions even in security and military affairs. This, i.e. having a well-established outward recognition, including in international security affairs, in turn could further the inner cohesion of the Union, its member states’ and citizens’ willingness to establish something of a real political Union at last. After the declined French and Dutch referenda on the Constitutional Treaty in 2005, ESDP, providing the EU with international actorness recognition in such a “hard” field of foreign policy as military affairs, may turn out to be a central instrument in creating a European Union that deserves the attribute of being political, and in establishing the Union’s political autonomy from other actors. In a more sceptic view, however, outward actorness on the whole range of international policies may replace rather than foster the inner-European integration project.

Anyhow, it is interesting that some of the gravest functional limitations of ESDP are dealt with in cooperation with NATO, an organisation that is very different to the EU in many ways, which itself has traditional claims in the management of European security affairs, and which represents the influence of the remaining superpower in Europe. NATO provides “assets and capabilities” to EU military missions the EU itself lacks, applying the so-called Berlin Plus agreements between the two organisations. This first and foremost means the use of SHAPE (Supreme Headquarters Allied Powers Europe) of NATO at Mons (Belgium) as a supreme military headquarters for EU missions. SHAPE basically provides a well-working permanent chain of command and control for multinational military operations, a basic tool for any multilateral operation, whether in the name of NATO or the EU. For EU missions, the Political and Security Committee of the European Council (PSC, or COPS for Comité politique et de sécurité) exerts supreme political control, and a EU “cell” has been established within SHAPE. Other NATO capabilities at the EU’s disposal might include military logistics like air transport and satellite communications (including, it is to be noted, US capabilities at NATO’s disposal). Berlin Plus so far has been applied on two occasions, for the EU mission Concordia in Macedonia in 2003, and for the EU mission in Bosnia since December 2004.

All questions relating to the institutional ties between ESDP and NATO are embedded in the context of the emerging project of a more or less independent European security framework, and inner-European and transatlantic tensions over what this means and how independent European security should be in the end. In this light, the ties with NATO as such seem to give a preliminary answer: European security in fact is not that independent, as it is coupled with NATO. On the other hand, however, Europeans interested in independence may judge the arrangement very helpful, for they profit from NATO’s expertise but keep the political control of their missions; moreover they may in this stage learn from NATO, to use it as a model for independent facilities to be established later on. In other words, a well-known question lingers in the background: does ESDP challenge NATO, or does it strengthen it? In exploring the question how ESDP and the institutional ties between ESDP and NATO came into being, I hope to shed some light on the underlying processes of the reconfiguration of transatlantic security relations too, although no definitive answers can be given yet.

In theoretical terms, the North Atlantic region represents a security community, built on shared values and institutionalised through NATO.^[2] Could it be that this security community slowly dissolves, or at least is complemented by a second, strictly European security community, not just in terms of shared values, but with its own security mechanisms, including external security? Are two security communities in the making, which are coupled with one another, one wider and looser transatlantic one, and one tighter European one? This thesis will not directly answer this, but it should give some hints. It would indeed be interesting to pose the question whether the foundations of shared values are at least partly eroding within the transatlantic community. However, this is not in the scope of this thesis; it rather treats the development of security institutions in the North Atlantic area as an indicator for the future of the North Atlantic security community. It is of significant relevance which security institutions complement NATO within the North Atlantic area, and under which frameworks of cooperation they do so.

The cooperation between NATO and EU, institutionalised through the Berlin Plus agreements, provides the EU with sophisticated military planning capabilities, and NATO with surveillance and technical control over missions the EU conducts – although political control rests with the PSC. Out of this framework of cooperation, which has proven to work in every-day practice, problems notwithstanding, grow considerable political tensions between the two organisations, and some of their member states. Most lines of argument oppose the United States and European states who are interested in more EU autonomy, notably France. Perceptions what Berlin Plus actually means differ: for example, the US originally defended the point of view that Berlin Plus grants NATO a right of “first refusal”, and only if NATO (i.e. above all the US itself) is not interested in conducting a particular mission, the EU may use NATO resources to mount a mission on its own. To Europeans, Berlin Plus simply places NATO resources at the EU’s disposal at the latter’s request, without such a condition of “first refusal”. The EU indeed rapidly started conducting missions without NATO support and supervision, notably in the Congo in 2003, and again in 2006.

However, the cooperation with NATO essentially enabled the EU to conduct its very first mission in Macedonia, beginning on 31 March 2003, the symbolic importance of which is not to be underestimated. How has it been possible that, first, the European Union mounts its own security policy, and second, that it does so with considerable help of NATO? Why did it happen this way, and why did it happen then, i.e. between 1999, with the first official EU statements of political will, and 2003, the year of the first missions? The problem I am trying to solve thus is: how do we explain the emergence of a European security policy and its ties with NATO? This issue is intrinsically relevant because it touches upon the very bases of European and transatlantic security relations; it represents the ongoing processes of reconfiguration of transatlantic security relations after the end of the Cold war, and it adds a completely new dimension to the EU itself.

I proceed from “macro” to “micro”. In a first step, underlying changes in the deep structures of the international system need to be discussed, which rendered the observed development possible in the first place, without directly triggering it. I notably refer to changes in the ideational structures of the international system occurring after the end of the Cold war concerning humanitarian interventions. In a second step, proceeding to the state level, some “black boxes” need to be opened, to explain specific puzzles. Especially the British approval of an EU security policy, given at St. Malo in December 1998, has to be explained in detail, for it was of paramount importance for ESDP’s lift-off directly after. Furthermore, the US consent not only to some vague European security policy, but to an EU-led ESDP backed up by NATO resources and capabilities needs to be addressed. Finally, the process that led to Berlin Plus and Concordia needs to be traced, giving some hints on underlying political problems and prospects, too. An interesting question is why the first EU mission finally happened at that point in time and not another – the timing indeed is quite surprising when we think of the fact that March 2003 was precisely the month when the Iraq war started, which caused considerable transatlantic and inner-European tensions.

But first, some general words on theory are due, to provide the reader with some broad lines of the understanding of international politics and the science thereof that are implicit to the following chapters. As I do not refer to one single theory, this is necessary to make my underlying approach clear.

2 Reflections on theory, philosophy of science, and methods

This thesis defends the point of view that no single theory of International Relations (IR) can explain the development that led to the emergence of ESDP and EU-NATO cooperation on its own. Rather, the complexity of international political life calls for consideration of different theories, understood as different foci on the topic. This admittedly may be an unorthodox way of approaching problems in IR, but if we take seriously the claim that different theories are nothing else than different explicit and elaborate ways of interpreting the world, it may well fit to apply different theories to one complex social phenomenon. This enables us to see “reality” from different angles, revealing different problems and offering a range of solutions to them. I in this respect follow Walter Carlsnaes’ basic arguments concerning the analysis of foreign policy (although there is a difference between explaining foreign policy as such and other parts of international relations), stressing the different logic of theory application for empirical purposes from meta-theoretical dissection; though built on different, sometimes irreconcilable theoretical presuppositions, applying different approaches helps us in appropriately treating “real-life” issues in empirical analyses.^[3]

2.1 Puzzlement, causal mechanisms, and inductive process-tracing

To give order and meaning to this multifaceted approach, the course of the study is guided by the idea of the “puzzle”.^[4] In considering the consecutive steps of the development, the thesis tries to identify and answer those questions that relate to the most surprising aspects of the development. This way, a complex causal mechanism is established that draws on different insights of IR theory, and that hopefully offers a convincing explanation on the questions posed above, in reconstructing the process that led to the first EU mission in 2003.^[5] As Alexander George and Andrew Bennett note, the causal mechanism approach indeed often calls for cross-cutting different IR theories.^[6] The implicit claim is that the very complexity of the “real” socio-political world calls for this cross-cutting of grand theories, none of which can explain all facets of “real” life on its own. The causal mechanism approach can be seen as linked to the scientific realism school that opposes itself to positivism, notably in calling for “real” accounts of the “real” world, as opposed to “as if”-assumptions. Positivist “as if”-assumptions may provide reasonably well established correlative relationships between macro-variables, but they can not explain the micro-processes of causality that work within these relationships; thus they “cannot distinguish a good predictive relationship from a good causal explanation”,^[7] i.e. they can not tell if there is real causality or just correlation, in which the two factors under consideration may be the effects of another third factor that causes both.

Alexander Wendt prominently drew attention to the notion of scientific realism within IR.^[8] At stake is the understanding of science, or more precisely: the understanding of the right relationship between theory and reality. In simplifying terms, positivists are content if their theoretical accounts match with observed phenomena, even if their “as if”-assumptions do not necessarily reflect reality. Scientific realists, on the other hand, want to explore how the world “really” works, what it is “really” like; theory must reflect reality as best as possible.^[9] Moreover, if we follow basic insights from critical theory, we come to see theory and reality as intimately intermingled, so that an “objective” positivist approach dealing with observed phenomena in a “neutral” way simply is not possible.^[10] In order to minimise the risk of distorted theory circling back into the same social and political reality that is under consideration, e.g. through its application by real-life actors, what might turn social science into a self-fulfilling prophecy,^[11] we indeed need to try to come as near to social science as a truthful portrayal of reality as possible. What “reality” is, of course is a crucial point here. If we acknowledge that there is a reality “out there”, i.e. that it exists independently of our perceptions, interpretations, and discourses, science precisely serves to address the question how this reality looks like, via empirical research. In this endeavour, individual researchers should try to give as convincing an account of “reality” as possible, and it is other researchers’ task, via scientific discussion, to judge the appropriateness of these accounts, which always remain subjective to a certain degree. Differing from some strands of “post-modernist” relativism, scientific realism builds upon the conviction that “social facts exist independently of the observer and can be the subject of defensible causal inference”,^[12] but it does not content itself with broad covering laws, like positivism; it wants to know how causality works in detail.

The causal mechanism approach responds to this call for details, as it concentrates on processes (“X leads to Y through steps A, B, C”), not on merely static correlations (“if X, then Y”).^[13] This necessitates a more detailed analysis of the subject under investigation, in principle trying to push the limits of research to the most detailed level of analysis as possible. In practice, however, degrees of detail may differ, “depending on the particular research question and research objectives under investigation”,^[14] and according to the information at hand. This thesis tries to establish causal mechanisms in an inductive manner without pretending to always go to the lowest and most detailed level possible, which often simply is not possible due to lack of information. The causal mechanism approach produces what John Ruggie, referring to Durkheim and especially Weber, calls a “narrative explanatory protocol”, showing “why things are historically so and not otherwise”.^[15] He calls for “the successive adjusting of a conjectured ordering scheme to the available facts, until the conjecture provides as full an account of the facts as possible”.^[16]

The method that best corresponds to the challenges of the “real” social and political world’s complexity is process-tracing. In George’s and Bennett’s words, “a ‘causal mechanism’ invokes an ontological causal process, and process-tracing is an operational procedure for attempting to identify and verify the observable within-case implications of causal mechanisms”,^[17] i.e. process-tracing consists of identifying causal mechanisms through close-up investigation of a given case. It is to be noted that George and Bennett speak of an ontological causal process, i.e. causal mechanisms are indeed assumed to exist in reality, and not to be confounded with theories and models which are situated exclusively on the analytical level. This thesis tries to inductively reconstruct such a causal mechanism, with the help of different theoretical insights. Different theoretical schools serve to offer reasonable interpretations of the chaotic empirical world. As David Dessler puts it, the task consists of “using existing theories and laws and acquiring a more precise characterization of […] the event itself”.^[18] As stated above, I in this enterprise let me guide by the idea of the puzzle, for it provides a good orientation of what is interesting and profitable to dwell upon, in order to establish a meaningful causal mechanism, and what is not.

2.2 Constructivism vs. rationalism?

Colin Wight writes that “the contemporary meta-theoretical framework the discipline [of IR] employs is: a bar to constructive dialogue; a hindrance to much-needed research into issues of vital concern; a confused misrepresentation of the issues; and most importantly, a construct of those working in the field”.^[19] He may be right to say so; some words are nevertheless due on the philosophical foundations this thesis rests on. Concerning the discipline’s rift between the meta-theories of rationalism and constructivism, one might have noted that the approach followed here is nearer to the constructivist programme. However, there are huge problems in defining the ultimate nature of the two approaches, and we might be better able to grasp the issues if we understand these two labels as just the current cipher for much older questions in meta-theory of social science, notably concerning ontology, i.e. how the social world works, and epistemology, i.e. what we can know about it, and how.^[20] These are important questions and I shall give some hints on this thesis’ understanding of them, to better situate the following empirical research.

The debate between rationalists and constructivists has now ravaged the discipline of International Relations for more than a decade, even two, depending on where you start to consider diverse publications on the matter a “debate”.^[21] Some have proposed that the two are complementary rather than inescapably opposed to each other.^[22] Anyway, no final word has been spoken yet, and possibly never will. Since questions of meta-theory are to a certain degree beyond empirical validation, belonging to the border zone between philosophy and science,^[23] judgements will necessarily remain subjective to a certain degree. Difficulties in identifying the real core of the debate notwithstanding, there remain some broad ideas on what divides the two sides. To begin with, rationalists are methodological individualists, whereas constructivists emphasise the interdependent effects of both structure and agency. Second, for constructivists, ideas and identities play a more important role, because they are constitutive for the individual’s preferences and behaviour, whereas for rationalists they are merely regulative.^[24] Third, constructivists tend to be interpretativists, i.e. they draw on Max Weber’s notion of Verstehen and Erklären,^[25] viewing social science and its methods as different from natural sciences, calling for a distinctive interpretative and to some extent emphatic understanding of social action, while rationalists hold that social action can be explained (Erklären only) according to general laws imposed “from the outside”, just as natural phenomena can, without drawing on Verstehen.^[26] Generally, constructivists attack rationalists on the ground that the latter are materialists who do not take into consideration ideational factors that exist independently of material interest.^[27] However, all this is not really agreed upon in the end, and rationalists might argue that constructivists draw a distorted image of rationalism, to better place their own views in the debate, and vice versa.^[28]

Anyhow, my goal here was just to give a broad idea of the differences between the two schools, and I now concentrate on more concrete matters. Picking up some lines of debate that seem to be important in situating my own approach, I start with logics of behaviour. We have a choice of three in contemporary IR debates: first, the logic of consequences that posits that actors behave in a rationalist utility-maximising way, considering means-ends constellations on the basis of cost-benefit analysis. Second, the logic of appropriateness that posits that actors are rule-followers, i.e. behave according to social norms, some of which they internalise, so that they are not even aware of the underlying rules of their own behaviour (constitutive norms and rules). Third, the logic of communication that draws on Habermas and posits that actors adjust their standpoints and behaviour in processes of enduring deliberation, open for “the better argument”.^[29] Max Weber claimed that there are different logics of behaviour, and he was certainly right. The three logics work at the same time and are intermingled with each other. Which of the three dominates in a given situation is a matter of empirical study.

The same then necessarily is true for the debate over regulative versus constitutive norms. Sometimes social norms regulate consequential behaviour, by altering costs for example, sometimes they constitute norm-driven behaviour, by being socialised into actors’ mind sets.^[30] Other issues in the debate might simply be a matter of focus. Walter Carlsnaes perceives of foreign policy actions as the result of a whole chain of explanations.^[31] The final action is related to the intentions of the actors, which in turn are related to their mental disposition, which in turn is related to material and social structures (the actors’ environment if you will). In a step-by-step framework, Carlsnaes proposes to use approaches from different theoretical camps, “as long as they are used when and if analytically appropriate”.^[32] The bottom line is that a preliminary choice for one of the two meta-theories that structure the field might eclipse our understanding of what is going on in the “real” world, as I already argued above concerning the discipline’s substantive “grand theories” (“substantive” theories as opposed to meta-theory). While this understanding openly exposes itself to attacks on the coherence of its “scientific” foundations, it is best suited, in my view, to grasp and explain reality.^[33]

Thinking about epistemology, i.e. what we can know about the social world, and how, the choice seems to be between objectivism and interpretativism,^[34] the latter reproaching the former not to take into account that all our knowledge about the social world is reflexive, i.e. that it is not more than our interpretation of the actors’ interpretation of the world (Anthony Giddens’ “double hermeneutics”), far from representing something like “objective reality”. But Colin Wight may be right in saying “if the point is simply that the world is socially constructed then it would be difficult to find many social scientists […] who think otherwise”.^[35] In this sense, the choice on this issue seems to be more clear-cut, and I am very ready to align this thesis with the interpretativist notion of epistemology; but I also point to the fact that verbal accounts of empirical “reality” for the sake of comprehensibility can not constantly stress this, and need to simplify in a way that sometimes might seem to suggest an objectivist world view. If this might be the case in the following pages, the present caveat should always be remembered, so as to understand the account as an interpretation of “reality”, not an objective report.

If we wanted to go deeper into these issues, we should probably rely on genuine sociologists, such as Emile Durkheim, Max Weber, Norbert Elias, or Anthony Giddens, and on real philosophers of science, such as Karl Popper, Imre Lakatos, or Thomas Kuhn. All I try to do here is to take some well grounded pragmatic stances that allow to take into consideration different viewpoints on the empirical level. As for the rationalism-constructivism divide, constructivism is much more reconcilable with such a pragmatic open-ended approach: it allows for all three logics of behaviour, combines agency and structure effects, material and ideational factors, thus is built around a wide and complex ontology. The goal is to understand reality, and theory therefore has to be complex, and to some degree open-ended; artificial parsimony is not always useful. Anyhow, as Colin Wight shows, there is not much clarity in IR over the foundations and the nature of many of the most hotly disputed theoretical issues surrounding the rationalist-constructivist divide. The dichotomy in itself may be an unhelpful simplification, given the vast array of different approaches to meta-matters: “a ‘thousand theoretical flowers’ into two will not go, and hence the current framework bursts at the seams. Simply adding a new ‘middle ground’ category does not help”.^[36] In the end, we as political scientists should concentrate on issues that are more important to a wider public of decision-makers and public opinion than the production of theoretical swamps. This means we should concentrate on real-life issues, of which we students of international relations are not short of.

2.3 ESDP as part of European integration

Let me finally come to a more “substantive” issue of theory: one may object that the EU is not really the actor with regard to European security policy, but merely its member states which have discovered the EU as a cover and instrument for their own traditional security policies, all the more so as the second pillar of the European Union (CFSP and ESDP) is intergovernmental in nature, as opposed to the “communautarised” first pillar. This begs the well-known question if international organisations are to be seen as instruments of the member states, arenas for discussion, or actors on their own.^[37] In most cases, elements of all three alternatives are present. In the special case of the EU, which in its sui generis character marked by the integration project is more than an “ordinary” international organisation, it would be hard to argue that it is nothing more than an instrument of member states. It transcends specifically national approaches, even transforms them in the processes of “Europeanisation”, the definition of which is a matter of debate, but which here might be best conceived of as a mutual process of member states collectively shaping EU policies and these policies in turn reshaping member states’ interests and identities – the latter constitutes the meaning of “Europeanisation” in a narrow sense and indeed is the most interesting point in the process.^[38] It can be assumed that these processes of Europeanisation also work in the intergovernmental field of common foreign and security policy, of which ESDP is a part. The EU and its policies can not durably be controlled by, say, a directorate of the largest states, notwithstanding their undeniable importance in European decision-making processes. Therefore, ESDP, as CFSP in general, represents the group of EU member states, which is more than the simple addition of its parts. And even if it is just this group that manifests itself as a new actor in security affairs, it differs from individual member states, and its symbolic and institutional glue are the European integration project and the EU. Speaking of the EU as an actor in security affairs thus at least means speaking of this group of member states (except Denmark which has opted out of ESDP), bearing in mind its own specific group dynamics, including the processes of Europeanisation that in due time might lead to new perceptions and policies not thought of by any single member state, and representing more than the smallest common denominator.

It can be argued that the processes of Europeanisation with the creation of ESDP have reached security policies and will lead to a slow redefinition of interests and identities, via elite socialisation and the “coordination reflex”. In the long term, this is expected to lead to growing convergence in national European preferences in security policies.^[39] This view might be attacked on the grounds that we often find cases where national interests, or preferences if you like, dictate member states’ policies, and where no Europeanisation of interests can be detected. But Europeanisation is an incremental process, not a switch that changes everything once turned. Jakob Øhrgaard offers a useful account of the political processes under the second pillar, proposing a “flexible” understanding of neo-functionalism, extending it to inter-governmental integration.^[40] In his words, “it is safe to say that transgovernmental interaction and interprenetation certainly have been a key feature of CFSP”^[41], due to working groups, exchange of personnel and working procedures, leading to the creation of a strong transgovernmental network. This creates socialisation and mutual expectations effects that lead actors out of classical negotiation situations, and into free deliberation processes where consensus is tried to be reached, rather than a smallest common denominator.^[42] There is not much reason why these processes should not be as valid for ESDP as for any other part of CFSP. Besides, Øhrgaard’s flexible neo-functionalism accepts and incorporates the fact that consensus is not always reached and that there are instances of national defection. The logic of “upgrading of common interests” is of special interest here: it consists of concentrating on issues where the actors think that a consensus might be reached, while they postpone contested issues until the processes of Europeanisation will have rendered consensus possible.^[43]

The processes of Europeanisation are slow, and probably they are slower in the intergovernmental second pillar than in the first, but this thesis is based on the assumption that indeed they exist. It thus assumes that ESDP in the long run will be more than an intergovernmental instrument of some European states, but an integral part of the European integration project which to a certain degree transcends national interests, identities and policies, and lends some form of actorness to the EU itself, even if in foreign and security policies it may remain coupled to the actorness of its member states as a group.^[44]

The thesis in some parts refers to concepts of national interest or preferences, whether of different European states or the US. Whenever it does, these should be read as socially constructed, generated by complex deliberation processes that take place within governmental agencies and society; different interest groups, the media and public opinion all are of importance to these processes. Ideational and identity factors are as much at the basis of these processes as are material factors. Moreover, the construction of different states’ interests is influenced by the internalisation of international norms and by the impact of transgovernmental and transnational networks that exceed state boundaries. Dealing with one of the last resorts of state dominance in international relations, external security in the Western world, the thesis thus none the less is not to be read as a realist approach. Furthermore, the very issue under consideration, the emergence of ESDP, exceeds classical concepts of international security, for it too exceeds state boundaries, coupling security policy to a supranational integration project.

Both classical realism represented by E.H. Carr, Hans J. Morgenthau and in a more differentiated manner by Raymond Aron, and especially structural realism represented by Kenneth Waltz, build upon a strict inside-outside distinction along the frontiers of nation states, which are considered the only “real” actors in international politics, especially in the field of security affairs, i.e. in questions of war and peace, which are of central interest to realists.^[45] The crucial point is that the development of ESDP constitutes a real-world breach of these (historically grown) theoretical assumptions. It transcends nation state frontiers and introduces a supranational actor to the very core of traditional state activity, the application of organised violence, after we have already seen the (re-)introduction of non-state actors like private military companies and terrorist networks into this field of war and peace.^[46] Analytically speaking, the state is joined in its core area of controlling organised violence not only by sub -and trans -national entities, but also by at least one supra -national actor. This is not to say, of course, that the state has become obsolete, but that today it is by no means the sole actor in “inter-national” politics, not even in security affairs – if it ever was. The emergence of ESDP thus is another aspect of the protracted “turbulence in world politics”,^[47] in that it calls into question the dominance of the sovereign state. At the same time it is an answer of a group of states to that problem, in that it reorganises this group’s sovereignty in security affairs, by “pooling” it.^[48] Nevertheless, “pooling” in this case also means transcending individual sovereignties, as it is part of the European integration project and its dynamics; again, to consider ESDP solely as an instrument of European states would mean neglecting the complexity of the European Union’s incremental processes. States will still exert control, but in the context of specific group dynamics, which include processes of socialisation and the redefinition of interests and identities.

Having now concluded preliminary discussions, I will in the next chapters try to establish along the lines of puzzles a causal mechanism that explains the emergence of ESDP, the concluding puzzle being its specific links with NATO. I do nothing more than proposing an interpretation of “reality” that needs continuous adjustment or revision, for there is little objective truth in the (social) sciences; the result will be a reconstruction of the diverse dynamics that helped to bring about the outcome of a first EU military mission in cooperation with NATO, Concordia of March 2003. This account must to a certain degree remain subjective, as all interpretations are.

[...]

---

^[1] The declaration is reprinted in Rutten 2001.

^[2] Cf. Deutsch 1957, Adler/Barnett 1999.

^[3] Cf. Carlsnaes 2002, 341-343.

^[4] See for example Rosenau 1980.

^[5] On the utility of causal mechanisms, in addition to George/Bennett 2004, see for example Elster 1998.

^[6] George/Bennett 2004, 127-128: “much of the discourse in the study of international relations is structured among ‘schools of thought’ – neorealism, neoliberalism, and constructivism – that some scholars have consciously modeled after Thomas Kuhn’s ‘paradigms’ or Imre Lakatos’ ‘research programs’. This parsing of our field into contending ‘isms’ does not fit very well with the emphasis we and many other scholars have placed on causal explanation via causal mechanisms, which often cut across these schools of thought”. They go further in criticising the disciplines obsession with grand theories in asserting: “to the extent that it [the focus on large schools of thought] has diverted attention from empirical puzzle-solving and problem-driven research, the field has suffered”, 128.

^[7] George/Bennett 2004, 139. George and Bennett refer to Milton Friedman as a prominent defender of the positivist approach within the social sciences (economics being part of these).

^[8] See Wendt 1999. He already dealt with scientific realism before, see Wendt 1991.

^[9] Wight 2002, 43 (fn. 16), gives a short overview of scientific realism. He, as others, notably refers to the works of Roy Bhaskar, e.g. Bhaskar 1979. Wight 2002, 41-42 (fn. 8), also summarises positivism.

^[10] See Herrmann 2002, 127, for a brief elaboration on this.

^[11] Rational choice is a good example with ample societal and political consequences. Realism in IR is another one.

^[12] George/Bennett 2004, 131.

^[13] Ibid., 141.

^[14] Ibid., 142.

^[15] Ruggie 1998b, 861, emphasis in original. In my view, showing why things are so and not otherwise already is a painstakingly difficult task. We should not too much strive to find general covering laws, for, it seems, there are not that many of them. If we want to find ultimately capable grand theories, we rather should be trying to apply chaos theory to the social sciences.

^[16] Ruggie 1995, 98, cited in Adler 1997, 351-352 (fn. 37). Ruggie here refers to Charles Peirce, who calls this “abduction“.

^[17] George/Bennett 2004, 138.

^[18] Dessler 2003, 395, cited in George/Bennett 2004, 148, emphasis apparently in original.

^[19] Wight 2002, 24.

^[20] It is true that the debate first and foremost concerned the American IR community, and there are good reasons for this, namely the dominance of structural realism wedded to positivism, see Mayer 2003, 90.

^[21] Kratochwil/Ruggie 1986 normally is considered to be one of the first constructivist attacks on rationalism in IR. At the beginning of the nineties, something like a scientific debate shaped, including such influential publications as Wendt 1992, and, for the more rationalist side, Goldstein/Keohane 1993, to give just two examples. The constructivist momentum was to be intensified throughout the nineties; some examples are Dunne 1995, Finnemore 1996, Katzenstein 1996, Kratochwil/Lapid 1996, Weldes 1996, Adler 1997, Finnemore/Sikkink 1998, Ruggie 1998a, Risse/Ropp/Sikkink 1999, and Wendt 1999.

^[22] For James Fearon and Alexander Wendt, rationalism and constructivism simply concentrate on different features of understanding: “Rationalists and constructivists approach international life from different analytical standpoints, which has led them to ask characteristically different questions and develop characteristically different answers.” (Fearon/Wendt 2002, 68). For them, “the challenge now should be to combine insights, cross boundaries and, if possible, synthesize specific arguments in hope of gaining more compelling answers and a better picture of reality” (ibid.).

^[23] Indeed, what is the difference between science and philosophy? If there is a philosophy of science, is there also a science of philosophy? How do we define science? How do we define philosophy?

^[24] I do not have the space here to explain these terms in depth (and this wouldn’t be my ambition anyway). In short: Rationalists tend to assume that preferences are exogenously given, i.e. independent of personal experiences etc., and therefore they are basically the same for all human beings in whatever circumstances: individual power, security and wealth. Societal norms and other ideational factors just regulate individual behaviour within a cost-benefit logic (for breaking societal norms can cause considerable costs, for example). Constructivists, on the other hand, argue that norms are internalised through socialisation, and therefore constitutive for the individual’s behaviour. Internalised norms become an intrinsic individual motivation. And as norms may differ in space and time (from society to society), individual preferences may change, too, because they depend on underlying internalised societal norms. In other words, preferences in constructivist thought are endogenized, and subject to change over time, above all due to changes in societal norm structures that provide the basis for the individual’s socialisation, and therefore for his or her mental disposition. These changes in societal norm structures, in turn, are based on individual behaviour (agency), for it is individual action that constantly creates and recreates (perpetuates) the patterns of the societal norm structure – again we see that for constructivists agency and structure are interdependent and inseparably interwoven.

^[25] Note the “and”! It is not the case that interpretativists would be content with emphatic Verstehen.

^[26] See e.g. Carlsnaes 2002, 335. Interestingly, not everybody in the natural sciences would subscribe to pure Erklären either. Another consequence of Verstehen is that constructivists go beyond causation as a strict cause-effect relationship, in introducing a notion of reasons (as opposed to causes) that enable specific causes to have specific effects in the first place. The aspiration for a united Europe is an example for such a reason that enabled more material causes (bipolarity, economic benefit) to work for a specific effect (the creation and development of the EC); see Ruggie 1998b, 869.

^[27] But Fearon/Wendt 2002 show that, quite on the contrary, modern game theory indeed recognises ideational factors in explaining human behaviour. In the end, those might be right who identify the structure/agency issue as the one underlying difference between the two schools of meta-theoretical thought, from which other cleavages result, rationalists concentrating on agency, constructivists underlining the interdependent effects of both. Indeed the structure/agency issue was present from the very beginning of the debate, see Wendt 1987.

^[28] The statement of Fearon (rationalist) and Wendt (constructivist) that “most of our difficulties arose not from clear substantive disagreements but rather from matters of presentation and efforts to gain clarity about just how ‘the other side’s’ argument works on specific points” reveals that indeed the real nature of the cleavage is difficult to identify, Fearon/Wendt 2002, 68.

^[29] Risse points out that there is no need that actors consciously and voluntarily submit their positions to such a deliberative situation in the first place, see Risse 2003, 114.

^[30] Cf. for example Checkel 1997.

^[31] See Carlsnaes 2002.

^[32] Carlsnaes 2002, 343.

^[33] This view in some ways is close to the critique on scientific method as a modern set of myths, put forward by Feyerabend 1975. This set of myths in his account justifies the scientific community’s authority in modern society without necessarily enhancing our understanding of the world. I do not fully concur with this relativistic stance, but draw from it the important insight that in order to understand the world, it might be necessary to ignore some scientific no-go-areas, like combining in a pragmatic endeavour different approaches that are theoretically “incompatible”. In the end, and I said this before, these approaches are nothing else than different ways to view the world, every one of which might provide valuable insights.

^[34] See e.g. Carlsnaes 2002, 334-336.

^[35] Wight 2002, 38.

^[36] Wight 2002, 40.

^[37] Cf. for example Rittberger/Zangl 1994: 25-26.

^[38] For an overview see for example Wong 2005. See also Tonra 2001.

^[39] Interestingly, the processes of socialisation and the „coordination reflex“ seem to be present in NATO, too, given national military officials’ attachment to it, and their will to coordinate in its integrated structures, which leads to the phenomenon that often European defence ministries are more attached to NATO than European foreign ministries or other governmental agencies; interviews Brussels, Paris. In this view, European military officials are expected to develop double international loyalties (in addition to their national one) as ESDP develops, both to NATO and the EU.

^[40] See Øhrgaard 2004; he especially refers to Haas 1958, Haas 1961, and Lindberg 1963.

^[41] Øhrgaard 2004, 30, emphasis in original.

^[42] Øhrgaard 2004, 31. He refers to the testimony of two actors who have published on common European foreign policy: “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, and: “a lowest common denominator would only result if the procedure followed were that all national positions were put on the table and whatever coincided became the European position […] this [was] not the case”, Nuttall 1992, 314. A more recent testimony, from 1996, is that of Pauline Neville-Jones, then political director of the British Foreign and Commonwealth Office: “The foreign policy process has become Europeanised, in the sense that in every international issue, there is an exchange of information and an attempt to arrive at a common understanding and a common approach”, cited in Aggestam 2004, 81. It is true that EU enlargement might have had a negative impact on these processes, for the number of participants has grown, and new members are not yet fully socialised to the EU style of deliberation, but even if this is the case, it seems more reasonable to see this as a temporary slowing down than as a veritable set-back. At least, this seems to be the experience with prior enlargements.

^[43] Øhrgaard 2004, 38-40.

^[44] Neo-functionalists like Haas and Lindberg do not deny the importance of member states’ willingness to pursue the integration project, especially where there are no supranational institutions to discipline them, as is the case with CFSP/ESDP. So, the pace of the Europeanisation of security policies will obviously also depend on member states’ willingness to let it happen. Personally, I here too think that both neo-functionalism and intergovernmentalism provide valuable insights of the “real” world.

^[45] See Carr 1939, Morgenthau 1948, Aron 1962, Waltz 1979. The catchy inside-outside theme of course is taken from Walker 1993.

^[46] Cf. for example Kaldor 1999, Singer 2001.

^[47] Rosenau 1990.

^[48] On the notion of pooled sovereignty, see Hoffmann/Keohane 1991.