Explaining European Intergration: the permanent structured cooperation

Explaining European integration: the

Permanent Structured Cooperation

Master Thesis

International Relations

Max van der Stelt

S4357507

Table of Contents

Table of Contents 2

1. Introduction

2. Theoretical framework

2.1. Overview of European integration theories 8

2.2. Neofunctionalism 11

2.2.1. The spillover 13

2.2.1.A. The functional spillover 13

2.2.1.B. The political spillover 14

2.2.1.C. The cultivated spillover 16

2.2.2. Critique on neofunctionalism 17

2.3. Liberal intergovernmentalism 18

2.3.1. National preference formation 19

2.3.2. Interstate bargaining 20

2.3.3. Institutional choice 22

2.3.4. The role of France and Germany 23

2.3.5. Critique on liberal intergovernmentalism 24

3. Methodology

3.1. Research design 25

3.2. Data selection 27

3.3. Operationalization of the variables 28 3.4. Strengths and weaknesses of the research design. 36

4. Empirical analysis

4.1. The neofunctionalist explanation 37

4.1.1. The functional spillover 37

4.1.2. The political spillover 41

4.1.3. The cultivated spillover 45

4.2. The liberal intergovernmentalist explanation 48

4.2.1. National preference formation 48

4.2.2. Interstate bargaining 50

4.2.3. Institutional choice 53

5. Conclusion

1. Introduction.

The adoption of the Treaty of Lisbon is one of the most significant events of the European Union (Gilbert, 2011). The treaty, which was signed in December 2007 and has been ratified by all EU member states, is a response to the negative outcome of two referenda on the Constitutional Treaty in 2005 (European Parliament, 2018). Even though it is not the official constitution of the European Union, it contains most of the meaningful elements. The Lisbon Treaty has significantly altered the working of the European Union on multiple terrains. One of these terrains is the Common Foreign and Security Policy (CFSP). This policy covers all aspects of foreign policy as well as the issues concerning security and defence. The Lisbon Treaty introduced a range of new CFSP actors, including the High Representative of the Union for Foreign Affairs and Security Policy, a position currently occupied by Federica Mogherini, and the permanent President of the European Council, who is currently known as Donald Tusk (European Parliament, 2018). Additionally, the treaty led to the creation of the European External Action Service (EEAS), the EU’s diplomatic service, and to the expansion of the Common Security and Defence Policy (CSDP), which are both situated inside the CFSP.

In this thesis, I will focus on the Common Security and Defence Policy. The CSDP sets the framework for both the political and military structures of the European Union and for military and civilian operations abroad (European Parliament, 2019). The Treaty of Lisbon introduced the possibility for the European Union to create the Permanent Structured Cooperation mechanism (PESCO) within the CSDP (European Parliament, 2016). In 2016, the EU Global Strategy started a process of closer cooperation in security and defence. The Member States acknowledged the importance of enhanced coordination, increased investment in defence and cooperation in developing defence capabilities (European External Action Service, 2018). These aims are sought to be achieved through the PESCO. This thesis will study the establishment of the Permanent Structured Cooperation.

Examining the establishment of the Permanent Structured Cooperation is interesting on an empirical as well as on a theoretical level. Empirically, because the policy measure has characteristics that are unique to the sector of security and defence: The PESCO has been designed in the form of ‘enhanced cooperation’. This entails that the EU Member States that are willing and have sufficient capacities can establish the Permanent Structured Cooperation (European Commission, 2019). So, not all Member States have to be on board. Furthermore, the Member States that have joined the

characteristic (European Defence Agency, n.d.). The Permanent Structured Cooperation is theoretically interesting, as it is a quintessential case of European integration. This leads to the following initial research question:

What explains the establishment of the Permanent Structured Cooperation in its current form?

In the following section, first the origin and functioning of the policy measure will be outlined. After that, a summary of the respective theories of European integration is given. And lastly, the concrete research question will be formulated.

The Permanent Structured Cooperation came into existence in 2017. However, as mentioned before, the adoption of the Treaty of Lisbon in 2009 already introduced the possibility to create the PESCO (European Parliament, 2016). In the meantime, the policy plan was shelved. In November 2017, ministers from 23 Member States signed the Joint Notification on Permanent Structured Cooperation, which was the first formal step in the establishment. The month after, Ireland and Portugal indicated their wish to join as well and the PESCO was formally adopted by the Council of Ministers (Mills, 2018).

At the moment, 25 Member States are participating in the PESCO. The only non-participant states are Denmark, Malta and the United Kingdom. Denmark has a defence op-out and does therefore not participate in the Union’s decisions and actions that have defence implications. Malta has a national commitment to neutrality and non-alignment. The country is afraid that any military alliance could infringe their commitment. The United Kingdom does not participate as it has decided to leave the European Union (European Parliament, 2018).

The Permanent Structured Cooperation comprises 20 commitments, which can be categorized into five key areas, that participating Member States subscribe to (European Defence Agency, n.d.). Firstly, cooperation on military investments. Secondly, bringing their defence apparatus in line with each other. Thirdly, making their forces more interoperable, flexible and deployable. Fourthly, cooperation on capability development, which will increase Europe’s strategic autonomy. And lastly, the development of major joint equipment programmes. Additionally, there is a commitment to defence expenditure criteria. The member states have the commitment to spend 20% of their defence budget on investment and 2% on research and technology (Council of the European Union, n.d.). Also, member states will take part in at least one project.

In March 2018, the European Council formally adopted the first set of 17 PESCO projects (European External Action Service, 2018). In November of the same year, a second set of another 17

projects was adopted. The 34 projects cover a wide range of areas concerning capability development and the operational dimension. Such as Cyber Teams, Military Disaster Relief and a joint EU Intelligence School (ibid.). As the PESCO and its projects have been established recently, it is impossible to measure and assess the compliance of the Member States involved. What is possible is to study the process leading up to the creation, and the creation itself, of the Permanent Structured Cooperation.

What makes PESCO different from already existing cooperation on defence within the EU, and therefore an interesting case to study, is that the commitments undertaken by the participating Member States are legally binding. What this essentially means is that the commitments undertaken by the member states have to be met (Biscop, 2018). After joining PESCO, it is not possible for a state to opt-out, whereas that is possible when it comes to defence policy in general. For instance, Denmark has opted out on the whole CSDP. Within the PESCO, the Council of Ministers will annually assess whether the commitments are fulfilled. Every year, member states are required to produce a National Implementation Plan, which clarifies what steps are taken towards the targets. This does not guarantee that that all targets are met by every state, but it does mean that a failing state will have to explain their shortcomings to fellow member states, the public and parliaments (ibid.). And, if a participating Member State fails to fulfil the criteria or is no longer able to meet its commitments, the Council of Ministers, acting by qualified majority voting, may suspend the state from further participating in the cooperation (Mills, 2018). Although the decision to participate was voluntary for each member state, joining PESCO and the commitments that come with it, leads to collective engagements that are hard to abandon without any harmful consequences to their national reputation.

The European Union is a complex institution. For it to be able to come to policies and regulations, it is necessary to converge and incorporate the interests of the Member States. This makes the process of further integrating within the European Union a complicated affair. As can be observed, the Permanent Structured Cooperation is a rather demanding policy initiative. The establishment of an integrative measure with such stringent guidelines and commitments is therefore remarkable and puzzling. The question can be raised why a state would choose to establish - and simultaneously voluntarily join – an agreement that can potentially undermine its sovereignty and autonomy, as its commitments are legally binding and as non-performance can be penalized. In line with this reasoning, it can be questioned why the commitments and structure of the Permanent

There are various theories that can provide an answer to this phenomenon of European integration. In this thesis, two influential theoretical frameworks that each provide a different explanation will be applied to the case of the Permanent Structured Cooperation. The two theories are neofunctionalism and liberal intergovernmentalism. This thesis will employ these theories as they are the two most suitable grand theories of European integration that can explain this case. The other grand theories that specifically try to explain the outcome of the integration process are federalism and intergovernmentalism (Bergmann & Niemann, 2013, p.1). Federalism is less suited to explain this case of European integration as it tends to be a normative rather than an analytical theory. The discussion among federalist scholars revolves around the explanation of why states should form a federation and not on why they might do so. The theory of intergovernmentalism could potentially be applied, but it is not too different from liberal intergovernmentalism, which is a further development of intergovernmentalist theory (Bergmann & Niemann, 2013). Therefore, this thesis will feature the theories of neofunctionalism and liberal intergovernmentalism. Now, the important aspects of, and differences between, these two theories will be outlined.

Neofunctionalism attaches a significant importance to the influence of supranational institutions and to the emergence of organized interests. The main concept that explains further European integration is the spillover. The spillover is used to describe the process of integration and to identify the propellants of this process. The theory advances three different forms of spillover: the functional, political and cultivated spillover. All of these spillovers, which will be discussed comprehensively in the following chapter, hold to the notion that the role of the political and social elites is the determinant factor in the process of integration, and can provide an explanation for the creation of the Permanent Structured Cooperation. Additionally, neofunctionalism considers both national and international, as well as both governmental and non-governmental, actors to be relevant.

The theory of liberal intergovernmentalism provides a vastly different answer to the query of why European integration takes place. It brings multiple aspects of international relations together in one coherent approach. The theory argues that European integration can best be understood as the result of rational choices made by national governments. These choices result from the effects of a three step process: (1) the formation of national preferences, (2) interstate bargaining and (3) institutional choice. Liberal intergovernmentalists maintain that nation states are the main actors in this process. While the national preferences are shaped by an interplay between the government and domestic societal actors, the second and third step only attribute significant relevance to the state (Moravcsik, 1998).

The theories of neofunctionalism and liberal intergovernmentalism significantly differ in their predictions on how European integration occurs and in which actors are relevant in the process. Where neofunctionalism believes that further integration results from the effects of a spillover, liberal intergovernmentalism argues that integration follows from advancing nationally shaped interests on the international level. The theories also hold discordant views on which actors are relevant in the process of integration. In neofunctionalist thought, states, European institutions and interest groups all play a substantial role. In contrast, the only actors that are of significant importance in the liberal intergovernmentalist framework are states and, to a lesser extent, domestic societal actors. As can be seen, the theories fundamentally differ in their approach towards European integration, and provide conflicting explanations to the initial research question of what explains the Permanent Structured Cooperation. This theoretical puzzle leads to the following concrete research question:

Which theory, neofunctionalism or liberal intergovernmentalism, better explains the establishment of the Permanent Structured Cooperation in its current form?

The theoretical relevance of this study is that it deals with – and attempts to provide an answer to – a problem of competing theoretical frameworks. As the theories maintain conflicting views and explanations, this thesis will reveal which one of them is more capable of explaining the process of European integration, and has therefore more explanatory value than the other theory. The societal relevance lies within the fact that the thesis will offer an explanation to a recent, far-reaching case of European integration. This will contribute to the understanding on what factors and which actors are decisive in the process of European integration. The insights gained from this thesis can further explain the workings and dynamics of the decision-making process that leads to the realization of European policy. This will not only be of interest to scholars, but also to the actors directly connected to the process of integration. They will be better informed on the elements they can tend to in their position, as well as on the degree of influence of their role and the role of the other actors involved.

The remainder of the thesis is structured as follows. In the second chapter, the theoretical frameworks are outlined, explaining the key concepts of - and deriving hypotheses from - the theories of neofunctionalism and liberal intergovernmentalism. The third chapter concerns the methodology of the study, which expands on how the analysis will be executed. Subsequently, in the fourth chapter, the empirical analysis is conducted. The fifth and final chapter concludes the thesis by summarizing and discussing the findings of the empirical analysis.

2. Theoretical framework.

2.1. Overview of European integration theories.

There are various theories that have sought, and still seek, to explain the process of European integration. This section will present an overview of the most recognizable and influential theories, before elaborating the two theories that will be applied in this thesis.

Federalism is one of the oldest theories of integration. Federalists argue that nation states are unable to provide their citizens with political and economic safety, and that therefore a federal system would make sense (Bergmann & Niemann, 2013). It envisions Europe as a project that brings separate, autonomous states together in order to create a union with a central authority, wherein the states hold on to a certain degree of power. The theory primarily highlights the end product of European integration – the formation of a federal system – without explicitly explaining what the pathway ought to look like. Federalist thought has many dimensions, ranging from calls for a world government on one end to near anarchism on the other end (Rosamond, 2000, p. 24). This has made federalism an elastic and controversial concept in the politics of European integration.

Functionalism has also been a central element of the study of integration. David Mitrany, the key figure of functionalist theory, states that the starting point should not be a question about the ideal form of international society, but rather about what its essential functions should be (1933, p. 103). The theory does not have foundational propositions that are shared by all functionalists. Nevertheless, there is a broad-ranging agenda, with at its core the prioritization of human needs or public welfare (Rosamond, 2000, p. 33). According to functionalist thought, the creation of transnational institutions has two effects: a transfer in popular loyalty from the nation states towards these institutions and a reduction in the likelihood of international conflict. While functionalist theory does have concrete predictions on the process of European integration, it does not provide a complete theoretical framework that is suitable for the study of the creation of the Permanent Structured Cooperation.

Transactionalism is closely linked to functionalism. The theory has similar thoughts on integration: the functions of the institutions are more important than their form. The main concern of transactionalist theory involves achieving security within a region or among a group of states (Rosamond, 2000, p. 42). Similar to functionalism, successful integration revolves around reducing the likelihood of states using violent means to resolve their differences. The theory’s guiding

hypothesis is that the sense of community among states depends on their level of communication (ibid. p. 44). Again, not a comprehensive theoretical framework is presented.

The gaps that federalism, functionalism and transactionalism leave when it comes to integration dynamics are subsequently picked up by multiple theories, all with different notions on what drives European integration. The theory of neofunctionalism, which was introduced by Haas (1958) and Lindberg (1963), is one of them. Its framework is based on the idea that integration is the result of different spillovers. It is theorized to take place due to: (1) the interdependence of policy sectors, whereby cooperation in one sector leads to cooperation in another sector. (2) Socialization processes, that form the perception among elites that interests can better be addressed at the international level. (3) Supranational institutions becoming agents of integration (Bergmann & Niemann, 2013).

In contrast to this framework, the theory of intergovernmentalism minimizes the influence of supranational institutions. Instead, it is argued by Stanley Hoffmann, the main proponent of intergovernmentalist theory, that European integration is shaped by the interests of states and the outcomes of EU bargaining (Hoffmann, 1966). Moravcsik's liberal intergovernmentalism finds its origins in intergovernmentalist theory. The two frameworks mainly differ in where they believe national interests and preferences originate from. Intergovernmentalism holds that they depend on the state's perception of its relative position in the states system, while liberal intergovernmentalism maintains that they emerge from dynamic processes in the domestic politics of the state (ibid.). Additionally, intergovernmentalism puts emphasis on the distinction between high and low politics. Here, it is argued that European integration will not take place in areas of high politics because they concern key aspects of state sovereignty on which there is no solidarity between states, such as security and foreign affairs (Hoffmann, 1964, p.90). In areas of low politics, for instance the areas of trade or agriculture, integration is possible. However, as there has been integration in the areas of high politics, Hoffmann’s distinction is no longer supported by empirical evidence (O’Neill, 1995, p.65). Therefore, liberal intergovernmentalism rejects the idea that interests are hierarchical and revolve around high and low politics. Instead, the theory argues that national interests reflect the direct, issue-specific concerns of domestic actors (Moravcsik, 1998, p. 6).

Apart from the theories explained above, there are more approaches that cover European politics. However, these focus on understanding and analysing the European Union as a type of political system, rather than looking at the process of integration. The most influential theories of this sort include the governance approach, policy networks, new institutionalism and social

they do not focus specifically on the process of European integration.

Now, the existing literature that specifically focuses on the EU’s foreign and security policy will be summarized, in order to further embed the theoretical frameworks used in this thesis. Four main research traditions can be identified that study the formation of state preferences concerning the depth of cooperation in the area of foreign and security policy. The first approach, rational institutionalism, looks at the congruence between anticipated policy outcomes and the government’s interests (Koenig-Archibugi, 2004, p. 143). It maintains that the level of support for supranational integration depends on the degree of this congruence: the more congruence, the higher the level of support. The second approach is connected to realist theory and focuses on differences in resources rather than differences in interests. The approach argues that strong governments, whose resources allow them to conduct effective foreign policy by themselves, should not see as much need for integration as weaker countries with less resources (ibid. p. 144). The third and fourth approach have their origins in the theory of constructivism. The third approach concerns the role of collective identities. It holds that governments of countries with a strong identification with Europe will be more supportive of supranational integration in the area of foreign and security policy than governments of countries with less identification (ibid. p. 147). The fourth approach bears upon the constitutional culture present in the country. When there are multiple layers of governance present within a country, its government will find it easier to accept a supranational transfer of power in the Common Foreign and Security Policy (ibid. p. 148-149). In that case, the creation of an additional layer of governance is viewed as desirable if it leads to an increase in policy effectiveness.

There are both significant resemblances and differences between the four approaches discussed above and the two theories that will be analysed in this thesis. The first two approaches overlap with liberal intergovernmentalist thought in the way that policy preferences are formulated and on when integration is perceived as desirable. Where they differ is that Moravcsik provides an all-encompassing theory that involves the entire process of European integration. The third approach has a reasonable degree of affinity with the political spillover of neofunctionalist theory, but it has a different focus on which actors are subject to Europeanization. Here is where the analysis conducted in this thesis also adds value to existing research: two grand theories of European integration will be applied to a specific case of the Common Foreign and Security Policy. These grand theories employ a broader framework than the aforementioned approaches, and can provide a more complete understanding on what factors and actors are important in the process of integration.

In the remainder of this chapter, the theoretical frameworks of neofunctionalism and liberal intergovernmentalism will be outlined.

2.2. Neofunctionalism.

The theory of neofunctionalism was developed by Ernst Haas and Leon Lindberg in the 1950s and 1960s. The theory holds certain core assumptions on international relations and contains a model that attempts to explain the growing level of European integration. In the years that followed, the original works of the authors were revised both by themselves and by other writers, such as Philippe Schmitter and Joseph Nye (Niemann, 2006, p.13). In this section, the initial assumptions of neofunctionalist theory will first be presented. Subsequently, the revisions that have been put forward will be discussed, which will lead to a neofunctionalist framework that will be used in this thesis.

Neofunctionalist theory is based on a number of underlying assumptions (Niemann, 2006, p. 15). The assumptions can be divided into two sets. The first set of assumptions provide us with a background on what the concept of integration entails and on which actors are important in the decision-making process. The second set of assumptions expands on how neofunctionalists view the dynamics of European integration. Both sets of assumptions are explained to provide relevant context. After outlining the relevant underlying assumptions, the core concepts of neofunctionalism will be discussed.

One of the assumptions holds that neofunctionalism was established as – and perceives itself as – a general theory of integration (Haas, 1961, 366ff). Here, it is important to describe what neofunctionalist authors understand as integration, and therefore the definitions are given. Haas and Lindberg have different notions of the concept of integration. Both will be discussed in order to present a complete picture of what integration entails. Haas defines the concept as follows:

"the process whereby political actors in several distinct national settings are persuaded to shift their loyalties, expectations and political activities towards a new centre, whose institutions possess or demand jurisdiction over the pre-existing national states. The end result of a process of political integration is a new political community, superimposed over the pre-existing ones." (Haas, 1958, p.

16).

It should be noted that Haas expects states to shift their loyalty towards the integrative body and that he identifies an end result where integration can be completed. Lindberg has a partly overlapping, but significantly different interpretation of integration. His definition is twofold:

independently of each other, seeking instead to make joint decisions or to delegate the decision-making process to new central organs. And the process whereby political actors in several distinct settings are persuaded to shift their expectations and political activities to a new centre." (Lindberg,

1963, p.6).

As can be seen, Lindberg only talks about a shift of expectations, not one of loyalty. Also, he does not envision an end date for the process of integration. This makes his approach more reserved than the approach of Haas. This assumption ties in to another assumption, which maintains that integration is understood as a process. This process evolves over time and can have its own dynamic. Thus, rather than looking at isolated events, neofunctionalists seek to analyse the broader picture (Niemann, 2006, p. 15). Also, neofunctionalists have a pluralist notion on international relations. They reject the realist belief that states are the only relevant actors on the global playground. There are multiple important actors that can engage and build coalitions across borders (Haas, 1964). This is essential for the neofunctionalist framework as a whole as well as for this thesis, as it attaches great value to the influence of international organizations, such as the European Union. States cease to be fully sovereign, as they voluntarily enter into arrangements with their neighbours in order to cooperate and resolve conflicts (McGowan, 2007).

Now, the second set of assumptions will be discussed. Neofunctionalism identifies a couple of dynamics that drive the integration process forward. Actors are rational and mainly self-interested (Haas, 1970, p. 627). However, interests can change during the integration process, as actors can learn from their experiences. The process is also driven by supranational institutions, which push for further integration by influencing the interests of governments (Niemann, 2006, p. 16-17). This will come back when the concept of the spillover is outlined.

Based on the previously outlined assumptions and dynamics of integration processes, neofunctionalist writers developed three main arguments: the spillover hypothesis, the elite-socialization hypothesis and the hypothesis on supranational interest groups (Niemann & Schmitter, 2009, p. 6-9). These hypotheses combined make up the theoretical framework of neofunctionalism. Each of them will be discussed thoroughly, and will then be connected to the case of the permanent structured cooperation.

2.2.1. The spillover.

The spillover is the main concept of neofunctionalism. It refers to a process in which cooperation in one political sector leads to cooperation in other sectors, in order to ensure that the goals that were set in the original sector are achieved (Haas, 1958, p. 291ff; Lindberg, 1963, p. 10). This means that political cooperation is expanded over multiple areas, which was not necessarily intended at the outset. Some sectors are so interdependent that they cannot be separated from the rest (Tranholm-Mikkelsen, 1991, p. 4). Integration of one area at the regional level is only practical when it is combined with the integration of other areas, as further integration is the only solution to the problems that arise. An example of this phenomenon is given by Jensen (2013, p. 63). He states that an original aim could be the free movement of workers across EU borders. What can occur is that a nurse that is certified in one country is prohibited from working in another country, due to differences in national educational systems. This then results in setting goals in the area of education policy in order to overcome this obstacle to the original aim of free movement of those workers.

The process of generating additional goals in new fields is the essence of the spillover concept. There are three different kinds of spillovers: the functional, political and cultivated spillover. Each spillover identifies different reasons for a spillover to occur and each will be discussed separately. Here, it is important to state that not all three spillovers have to be present for the theoretical framework of neofunctionalism to have sufficient explanatory value. If one of the spillovers is found to be present, this is enough of an indication for the theory to be relevant.

2.2.1.A. The functional spillover.

The functional, or technical, spillover entails that one step towards cooperation functionally leads to another (Haas, 1958, p. 297). Schmitter (1969) states that functional spillovers are triggered by frustration and/or dissatisfaction with the performance towards common goals set in a certain sector, which will result in the search for alternative means in order to reach those goals (p. 162). He adds that it is likely that this will involve collective action in other sectors than the one originally contemplated. This comes from the fact that some sectors are so interdependent that it is impossible to treat them in isolation. Attempts to integrate certain functional tasks will inevitably lead to problems which can only be solved by further integration (Tranholm-Mikkelsen, 1991, p. 4-5). Further integration then leads to an increased delegation of authority to regional institutions (Schmitter,

commitment to the original sector. The momentum that is created by a spillover is assumed to have a certain degree of automaticity, it will gradually and progressively continue and ultimately lead to merged national economies which are overseen by centralized institutions (Rosamond, 2005). The example introduced above on the free movement of workers is an illustration of the functional spillover.

What should be empirically observable in the creation of policy in order to be able to refer to it as a functional spillover? The answer lies in the motives of the policymakers. When the process of further integration continues via the adoption of policies in one sector, and these are motivated by developments in other sectors, where cooperation has been ongoing priorly, we can speak of a functional spillover. It can also be a functional spillover when there is an increased commitment to resolve the occurring problems within the relevant sector itself. These references ought to be made by the actors directly involved in the policymaking process: the national governments and the European institutions. Let us make this more concrete and connect it to the Permanent Structured Cooperation. The PESCO is part of the security and defence sector of the European Union. The functional spillover takes place if we observe that, during the creation and adoption of the PESCO, it was argued by the actors present in the decision-making process, that this was necessary due to dissatisfaction with, or failure in, other policy sectors or the security and defence sector itself. It also occurs if the PESCO was seen as a logical step after successful cooperation in other sectors. In short, the functional spillover can be observed by analysing the motivations and discourse of the actors involved.

For the functional spillover, the following hypothesis will be tested:

H1: If integration is the result of a functional spillover, then the Permanent Structured Cooperation is seen as a solution to existing problems in other sectors or the security and defence sector itself.

2.2.1.B. The political spillover.

The political spillover occurs through a more deliberative political process. Through deliberation, national political elites or interest groups advocate for further supranational cooperation, as it is needed in order to solve specific problems. These actors are considered to be of substantial importance and they create further pressure for integration, which influences policy makers (Haas, 1958, p. 312-313; Lindberg, 1963, ch. 4).

The second and third core concepts of neofunctionalism, elite socialization and the formation of supranational interest groups, are connected to the political spillover and will be discussed

together. Elite socialization concerns the growth of support for supranational policy by participants of the decision-making process, such as officials and politicians (Tranholm-Mikkelsen, 1991, p. 5). It is argued that people who are involved in the international policy process will, over time, develop European loyalties and preferences. When the participants engage in the intensive decision-making process, they can develop a special orientation to that process and to their interactions with their peers. This is especially the case if the process is rewarding and when they value the system and their role within it. The European integration process leads to the establishment of so-called elite groups, who are loyal to the supranational institutions and hold European norms and ideas. These groups consist of officials and politicians who participate in EU decision-making. These elites, who have become ‘Europeanized’, then try to influence national elites towards engaging into further supranational cooperation (ibid.).

The formation of supranational interest groups, in addition to elite socialization, expect that organized interest groups also become Europe-oriented (Niemann & Schmitter, 2009, p. 8-9; Lindberg, 1963, p. 71). Corporations and business groups formulate and organize their interests with the intention to influence future EU policy. An example of this is the establishment of the Confederation of European Business, which is a common European group consisting of industrial and employers’ organizations that operate on the national level. Neofunctionalists also believe that the interest groups, such as NGOs, will aim to pressure the government and force them to speed up the integration process.

The concepts of elite socialization and supranational interest groups overlap a great deal. Essentially, they envision a similar process: there is a shift of focus from the subnational and national level to the supranational level. Where they differ is the actors that they ascribe this process to. Elite socialization concerns the officials and politicians that are directly responsible for the decision-making process. And supranational interest groups involve the surrounding actors that are indirectly involved, but still play an influential role. The process of socialization that takes place in both concepts can be viewed as an extended elaboration of the political spillover, as they predict similar processes and courses of action. Here, it should be noted that the process of socialization put forward by the concepts is important, but that they do not constitute a strong, independent hypothesis. Therefore, their proposed expectations will be connected to – and dealt with by – the political spillover.

The political spillover happens through the interaction between supranational, national and subnational actors. Both the supranational and the subnational actors focus more on European than

institutions by member state governments (MacRae, 2015, p.6). The socialization processes, with in particular the increased desire to look for European solutions, helps to generate a shift of expectations and loyalties towards the new centre, here the European Union, on the part of national, both governmental and non-governmental, elites (Niemann, 2006, p. 19). In order for the political spillover to occur, there should be calls for, and pressure towards, further European integration by either national and/or subnational actors. For PESCO, this means that national elites or subnational interest groups should have attempted to influence the supranational institutions to effectuate its creation.

The political spillover leads to the following hypothesis:

H2: If integration is the result of a political spillover, then national elites and/or interest groups must have advocated for the creation of the Permanent Structured Cooperation on the basis of it being a European solution to problems rather than a national solution.

2.2.1.C. The cultivated spillover.

The cultivated spillover concerns situations wherein supranational actors, with in particular the European Commission, use their mediating role between the Member States to push the process of political integration forward (Haas, 1961, p. 377; Tranholm-Mikkelsen, 1991, p. 6). The mediating institution directs the negotiations that revolve around the swapping of concessions in related fields (Niemann, 2006, p. 19). The outcome of classical diplomatic negotiations rarely moves beyond the position of the least cooperative partner (Tranholm-Mikkelsen, 1991, p. 6). However, if there is a mediator present, negotiations do move further than the minimum common denominator. When there is an autonomous and institutionalized mediator present, the common interests are upgraded and this improves the result.

Especially the Commission is in a privileged position of centrality and authority, which enables it to direct both the relations between states as well as the relations of groups within pluralist states. Neofunctionalists expect the Commission to gradually expand its mandate as the scope and level of integration increases, effectively stimulating the Commission's influence on the process even further (Haas, 1961, p. 377; Tranholm-Mikkelsen, 1991, p. 6). When it comes to PESCO, the cultivated spillover would foresee an influential role played by the European Commission as a policy entrepreneur.

H3: If integration is the result of a cultivated spillover, then supranational European actors, especially the Commission, have played an influential role as a policy entrepreneur for the creation of the Permanent Structured Cooperation.

2.2.2. Critique on neofunctionalism.

The theory of neofunctionalism has received criticism resulting from both empirical developments as well as from approaches with contrasting theoretical views.

There have been two notable occurrences within the European Community that have posed a challenge to the neofunctionalist conception of the spillover in combination with the degree of automaticity that comes with it. To recap, neofunctionalism assumes that the integration process tends towards continual expansion rather than contraction (Tranholm-Mikkelsen, 1992, p. 7). Both of the notable occurrences are a response to the integration process of the European Economic Community, which was developing at a rapid speed: the creation of the customs union and the Common Agricultural Policy progressed swiftly and the Commission played an important role (ibid.). The first occurrence is the French veto of British membership to the European Community in 1963. The second occurrence concerns the boycott of the Community’s institutions by French president De Gaulle in 1965, which is also known as the Empty Chair Crisis. Both of these developments involve a stagnation of European integration and have strained the assumptions of neofunctionalism (Eilstrup-Sangiovanni, 2006, p. 97; Tranholm-Mikkelsen, 1992, p. 7).

The most prominent theoretical critique – and most relevant critique for this thesis – has come from the camp of intergovernmentalism. Their criticism is threefold and is rooted in realist thought. Firstly, the importance neofunctionalism attaches to non-state actors is dismissed (Eilstrup-Sangiovanni, 2006, p. 97). Intergovernmentalists argue that states remain the ultimate authorities on integration and are able to resist integration when they want to. Secondly, they reject the belief that the decision-making process features a complex network of national, transnational and supranational actors that shapes these actors’ interests (ibid. p. 98). Instead, it is argued that governments adopt a certain position domestically before negotiating internationally. Thirdly, intergovernmentalists discard the notion of the spillover (ibid.). Integration is presumed to result from the calculated behaviour of states.

2.3. Liberal intergovernmentalism.

Since the early 1990s, liberal intergovernmentalism has become one of the most important theories of European integration (Cini, 2013, p. 79). The theory draws on insights from traditional 'intergovernmentalism' and is partly a response to the already existing theory of neofunctionalism (Moravcsik & Schimmelfennig, 2009, p. 67). It aims to identify the motivations of social actors, states and leaders and to come to predictions of their behaviour. Liberal intergovernmentalism is, alike to neofunctionalism, a general theory that seeks to explain the broad evolution of regional integration. This means that liberal intergovernmentalism combines multiple theories and factors into a single coherent approach. Three theories, layered in a multi-stage model, are needed to explain integration: the construction of preferences, the bargaining for substantive agreements and the creation of institutions (ibid. p. 68-69).

Before discussing the theoretical components in-depth, the two basic assumptions liberal intergovernmentalism rests on will be outlined. First, in contrast to neofunctionalist beliefs, states are treated as the crucial actors when studying international institutions. The states achieve their goals by negotiating and bargaining with other states, rather than through a centralized authority that makes and enforces the political decisions for them (ibid.). This does not mean that liberal intergovernmentalists adopt realist assumptions, where state preferences and identities are uniform and interstate institutions are of insignificant value. It means that, within institutions such as the European Union, member states possess the decision-making power. So, where neofunctionalism attaches a great deal of value to non-state actors and a certain degree of autonomy to EU institutions, liberal intergovernmentalism argues that the non-state actors have a more limited role, and that states remain central in international politics (Moravcsik, 1998, p. 7).

The second basic assumption is that states are rational. They calculate the utility of alternative strategies and follow the one that benefits them most. Any outcome, whether it is the agreement to cooperate or the establishment of an international institution, can be explained as a collective outcome of interdependent rational state choices and intergovernmental negotiations (Moravcsik & Schimmelfennig, 2009). Andrew Moravcsik, who can be seen as one of the founders of liberal intergovernmentalism, states that rational state behaviour can be modelled as a process that takes place in two successive stages (1993, p. 481). In the first stage, governments define a set of interests. Then, in the second stage, governments bargain among themselves in an effort to realize those interests. He argues that these stages shape the demand and supply functions for international cooperation. On one hand, the demand side is shaped through a domestic preference formation

process, wherein the potential benefits of cooperation are identified by national governments. On the other hand, the supply side is shaped by a process of interstate interactions, which define the possible responses from the European political system to pressures from the governments.

The interaction of demand and supply is what shapes the foreign policy behaviour of states (ibid.). This conception of rationality leads Moravcsik to construct two types of sequential theories: a theory of national preference formation and a theory of interstate strategic interaction. Later, the institutional choice theory is added. The theories that were introduced before should be seen as three separate stages in a sequence, and cooperation outcomes are only explained at the end of the sequence.

2.3.1. National preference formation.

The formation of national preferences is the first stage. It entails that the goals of national governments vary in response to shifting pressure from domestic social groups, whose preferences are aggregated through political institutions (Moravcsik & Schimmelfennig, 2009, p. 69). The priorities and policies are determined by the national government’s politicians, who find themselves surrounded and constrained in their identities and purposes by the civil society (Moravcsik, 1993, p. 483). The preferences of states are issue-specific: they vary across states and over time according to the expected costs and benefits of potential policies. Moravcsik argues that geo-political interests and economic interests have had the most impact on European integration. From the geo-political perspective, it is in the interest of the states to preserve national sovereignty and control over their territory. Alternatively, from the economic point of view, national groups seek to realize profits from international trade and maintain the interdependence of actors in the financial markets (Moravcsik, 1998, p.20).

This process of formulating national preferences is described by Moravcsik as an interplay between governments and domestic societal actors. One of the governments’ main goals is to ensure the support of both voters and interest groups in order to remain in office. However, the varying interests of societal groups leave the government with two possible situations and corresponding strategies. First, if the interests of societal groups are split, governments do not feel pressured and have the freedom to take decisions in the international arena which they see fit. Second, if the societal groups are united they can strongly pressure and constrain the national governments (Moravcsik, 1993).

So, domestic societal groups, such as businesses or NGOs, articulate preferences that are then aggregated by the government. It is important to stress that this process takes place within the state. From all of the input that the government receives, it decides on a ‘middle point’. Where after the government becomes the actor responsible for realizing the collected policy desires at the international level. Different from neofunctionalist theory, there is no substantial outside (i.e. international) influence present that influences the formation of preferences.

This leads us to the national preference formation hypothesis:

H4: If liberal intergovernmentalism is correct, government preferences are instigated by domestic societal actors, not by international actors.

2.3.2. Interstate bargaining.

The second stage looks at the behaviour of governments and their willingness to delegate power to the international institutions. Moravcsik proposes a bargaining theory of international cooperation. Herein, states attempt to collectively overcome suboptimal outcomes and cooperate in a mutually beneficial way. The states also have to decide how the mutual gains will be distributed. In this process, collective and individual interests can conflict and hard bargaining can alter the willingness of states to cooperate. The theory argues that the decision to delegate power to international institutions, while the national governments maintain in strict control, is the most effective way to guarantee cooperation and commitment of member states, and to ensure that states will preserve their autonomy. The member states are willing to accept the involvement of international institutions as long as it strengthens rather than weakens their control over domestic affairs (Moravcsik, 1993, p. 507). So, international institutions do have an important role as they provide states with beneficial outcomes and minimize the chances of a state to defect.

Now, bargaining theory argues that the degree of cooperation – and the terms and conditions of this cooperation – depend on the relative bargaining power of the actors. Bargaining power can be drawn from several factors. In the European Union, asymmetrical interdependence plays a crucial role. Asymmetrical interdependence is a situation wherein there is an uneven distribution of benefits that result from a specific agreement, compared to benefits that would have been gained by alternative options. The alternative option involves what the state can realize unilaterally, without the agreement. The actors that have the least amount of ‘extra’ benefits to lose will attach less value to the agreement, enabling them to demand concessions as they can threaten with non-cooperation (Moravcsik & Schimmelfennig, 2009, p. 71). Additionally, the actors that have better information

about other actors' preferences and the workings of the institutions are able to manipulate the outcome to their advantage.

Although liberal intergovernmentalism does recognize the role of informational asymmetries in bargaining, it sees it differently than neofunctionalism does. Neofunctionalism holds on to the assumption that there are significant costs involved in international negotiation and that political entrepreneurs, which are often non-state actors, have access to information or contacts that can be used to influence national governments (Niemann & Schmitter, 2009, p. 22-23; Moravcsik & Schimmelfennig, 2009, p. 71) Liberal intergovernmentalism argues that these third parties are usually not required in order to reach an efficient interstate agreement, because they only rarely possess information or expertise that is unavailable to states. Thus, informational asymmetries are important, and can be viewed as bargaining power, while third parties are not (Moravcsik & Schimmelfennig, 2009, p. 71).

Furthermore, bargaining theory identifies three assumptions under which negotiations take place and refer to the respective factors that might characterize the negotiation outcomes (Moravcsik & Schimmelfennig, 2009, p. 69). The first assumption holds that governments are likely to reach an agreement if cooperation is perceived as the best option by each state. This means that if a government has an attractive unilateral alternative, then it will not take part in the international agreement when that agreement requires significant compromises, as they would be better off without international policies on the matter. States whose unilateral alternative is not as good, and thus rely on international cooperation, will be moved into the situation of having to make compromises. This is related to the concept of asymmetrical interdependence.

The second assumption entails the opportunity to engage in alternative coalitions, which can occur through strategic calculation by the states, and can result in threatening certain member states with exclusion. A state can threaten to cooperate with non-EU countries as well as threaten to cooperate among subgroups within the Union. This assumption is relevant for this thesis, as there are possibilities to form alternative coalitions. Although it is not possible to threaten to work among subgroups within the European Union regarding the PESCO, as every member state is eligible to join the cooperation, it is possible to threaten with non-EU cooperation. This is because the European Union is not the only organisation that focuses on security and defence. For instance, the North Atlantic Treaty Organization (NATO) deals with collective defence and overlaps with the Permanent Structured Cooperation on various areas (Zandee, 2018).

The third assumption, which concerns issue linkages, is also relevant. It maintains that issue linkages may occur when the states have diverging incentives on different issues (Moravcsik, 1998, p. 65). Issues can be linked when there are several policy options that are each an improvement of the status-quo. The potential gains on each issue become important and lead to push governments into swapping out agreements (ibid. p. 66).

Now, the assumptions on interstate bargaining are brought together. The dominant factor in Moravcsik’s view on the bargaining process is the bargaining power. This allows the state to influence the outcome, and is derived from asymmetrical interdependence and informational asymmetries.

This leads to the interstate bargaining hypothesis:

H5: Further integration is the result of interstate bargaining. The more bargaining power a state has, the more the state can customize the outcome.

2.3.3. Institutional choice.

The third stage explains the creation and the design of international institutions when the agreement has been made by states to coordinate their policies. International institutions are seen as instruments that can help states cope with unintended, unforeseen and unwanted consequences (Moravcsik & Schimmelfennig, 2009, p. 72). The institutions provide the necessary conditions for durable international cooperation in several ways. (1) They help reduce the transaction costs for further negotiations, as there is the possibility to build upon previous agreements. (2) Information is provided to governments that keeps them informed on the other’s preferences and strategies, effectively reducing uncertainties and possibly suspicions. (3) The institutions set up and monitor the rules of international cooperation. Hereby, they facilitate a check on the states’ behaviour and have the possibility to penalize non-compliance. So, international institutions set norms and procedures for interstate bargaining, which does not require a significant delegation of power by national governments. The theory predicts that governments delegate power to the institutions to the extent that is necessary for them to execute their role, nothing more (ibid. p. 72).

The European Union, different from national governments, has little economic capacity, no coercive capacity and a small administration (ibid. p. 73). This leads to the situation that most EU rules are implemented, administered and enforced by member states. Then, the question rises how it can ensure interstate cooperation and compliance over time. Liberal intergovernmentalism points to the pooling and delegation of sovereignty of member states, which they are willing to do in order to

make use of the aforementioned benefits provided by the supranational institutions (Moravcsik, 1998). Sovereignty is pooled when governments agree to decide upon certain matters by voting procedures other than unanimity, and sovereignty is delegated when supranational actors have the possibility to take autonomous decisions (ibid. p. 67).

Moravcsik especially stresses the desire of states for credible commitments. Governments can be worried about the future decisions of other states and therefore want assurances of their commitment to the agreement. Rather than relying on solely establishing rules in advance, they are willing to pool and delegate a part of their sovereignty by removing the possibility of a government to unilaterally decide on future decisions (ibid. p. 73). There are more reasons for why this process takes place. The alternative to delegation and pooling is the making of package deals. This way, it is more difficult to structure trade-offs and to deal with individual issues, as they are bundled (ibid. p. 74). Furthermore, the pooling and delegation can be used to pre-commit governments to decisions before the costs and benefits become clear enough, which could generate opposition.

Transferring sovereignty to an international institution strengthens the states’ commitments in several ways: (1) it can make it more visible when a state is not cooperating, (2) it can create a reputation for the member states that can be damaged as a result of non-compliance, (3) and re-establishing unilateral control requires the respective state to renegotiate the agreement, which is costly and risky (ibid. p. 74). So, liberal intergovernmentalism states that the choices by the governments to give up part of their sovereignty to supranational institutions is motivated by their desire for credible commitments. For this thesis, the desire to pool and delegate sovereignty is expected to translate into the creation of the Permanent Structured Cooperation.

This can be tested with the following institutional choice hypothesis:

H6: If states are unsure of the commitment of other states to an agreement, then they will favour the pooling and delegation of sovereignty.

2.3.4. The role of France and Germany.

In his works, Moravcsik argues that European reforms have rested on interstate bargains between Britain, France and Germany (1991, p. 20-21; 1998). The belief is held that the preferences and positions of these three states are the most important, and are decisive when it comes to policy

Permanent Structured Cooperation. In the analysis of this thesis, there will be a specific focus on France and Germany.

2.3.5. Critique on liberal intergovernmentalism.

The theory of liberal intergovernmentalism has been at the receiving end of criticism coming from various sources. First, the critique on the theory in general will be outlined, which will be followed by the critique that is specified to the theory’s three stages of integration.

Liberal intergovernmentalist theory has been criticized for remaining limited to treaty-amending grand bargains, which only constitute a small portion of the policy-making process (Moravcsik & Schimmelfennig, 2009, p. 80). So, the lion’s share of decisions do not fall under the scope of the theory. This would require the incorporation of a more detailed analysis of the effects of institutional rules on the dynamic between states and the European institutions (ibid.). Another criticism argues that the theory’s perception of integration, namely that it is a process promoted and controlled by states, is misleading. The critique, which is advanced by neofunctionalists, takes issue with the explicit focus on intergovernmental decisions (ibid. p. 81). Other factors influence the process of integration as well: unintended consequences can arise that are difficult to address and supranational actors can have a significant impact and put constraints on national governments.

Criticism has also been raised that focuses specifically on liberal intergovernmentalism’s stages of integration. The first stage, where national preference are formed, has been questioned by constructivist and postfunctionalist scholars. Constructivists claim that EU norms and rules shape national preferences. Postfunctionalists argue that domestic preferences follow broad, identity-based movements rather than functional, issue-specific interests (Kleine & Pollack, 2018, p. 1494). The second stage, which involves interstate bargaining, is criticized by constructivists on two points: the decision-making process of the European Union is more deliberative than Moravcsik’s theory allows for, and national governments share influence with the European institutions (ibid.). Historical institutionalists maintain that the model of institutional choice, which is present in the third stage, ignores the presence of internal feedback that can create pressure for further institutional choices (ibid).

3. Methodology.

In the previous chapters, the outcome to be explained as well as the possible causes have been outlined. The outcome this thesis aims to explain is the creation of the Permanent Structured Cooperation and how the policy was realized into its current shape. The possible explanations are those provides by the theoretical frameworks of neofunctionalism and liberal intergovernmentalism. I will investigate the causal relationship between the independent variables that are deducted from the theories and the outcome of the decision-making process.

Both of the theories propose rather complex causal frameworks on the process of European integration, which is why a qualitative, small-N research design would be most effective in analysing which theory provides more explanatory value. Therefore, this thesis will feature an in-depth case study of one of the most recent integrative policies within the domain of European security and defence: the Permanent Structured Cooperation. A case study should be understood as the intensive study of a single case for the purpose of understanding a larger class of cases (Gerring, 2001, p. 6).

Within the case study, I will be making use of the method of process tracing. The purpose of this research method is to go further than solely identifying correlations between the independent variables and the outcome (Beach & Pedersen, 2013, p. 1). It offers the possibility to trace causal mechanisms. A causal mechanism is a complex system that produces an outcome based on a number of independent variables. So, we can go beyond correlations by attempting to uncover the workings of the mechanism that links X and Y.

Process tracing is different from other forms of small-n case study methods, as it aims to make within-case interpretations on the presence of a certain causal mechanism rather than between-case interpretations. Now, how can we make causal inferences about the studied mechanism, as the tracing of the process only provides us with mechanistic, within-case evidence (Beach, 2017, p. 9)? Beach states that the right use of Bayesian logic will help in attaching value to the collected evidence. Bayesian logic concerns the idea of using newly gathered evidence to update our confidence in causal theories (ibid.). The confidence we had in the theories prior to the research combined with the weight of the new empirical material determines to what extent the confidence should be altered.

process tracing. This is a theory-first method that seeks to test whether a hypothesized causal mechanism exists in the analysed case. By providing evidence of the presence of the mechanism between X and Y, strong claims of causation can be made. There are multiple approaches to the way the study should be conducted, as various authors have different views on the method of process tracing. Beach argues that the researcher should follow a strict, step-wise approach. The first step is to theorize the causal mechanism, based on existing literature, logical reasoning and contextual conditions necessary for the mechanism to work (ibid.). The next step involves the operationalization of the mechanism, by outlining what exactly should be observed empirically (ibid.). Within the stage of operationalization, Beach states that there must be a strong focus on the certainty and uniqueness of the explanations, which is an extensive, demanding process. After that step, Beach believes the empirical analysis can be conducted. Beach suggests engaging in further case studies if supporting evidence is found, or to turn to theory-building if evidence is absent. Follow-up analyses will not be conducted, as they do not fall within the scope of this thesis.

A substantially less stringent approach to process tracing is provided by George and Bennett (2005). The authors believe that there is more than one method possible, and the choice depends on data availability, the theory involved and the research objective. Their proposed 'analytical explanation' is the most relevant for the study of this thesis and comes closest to the approach of Beach. Here, a narrative is written which includes a causal explanation with its roots in explicit theoretical forms (ibid. p. 158). The explanation may be selective on purpose, so the focus can lie on what are thought to be the most important parts of the explanation. This thesis will feature a mixture of both. The step-wise approach will be followed, but in a less specific way. So, it will not be as strict as Beach would like, but still maintains the general approach. And, the analytical explanation will be taken into account, without deliberately leaving certain parts of the theories out of the analysis.

3.2. Data selection.

In this thesis, the process of the creation of the Permanent Structured Cooperation will be examined. The analysis will look at the motivation and argumentation of the relevant actors and at the outcome of the policy making process. The actors involved are the European institutions, national governments, and interest groups. When it comes to the national governments, the analysis will follow Moravcsik's view on what states to study. As explained in his framework, Moravcsik argues that the preferences and positions of the three largest states are the most important, and are decisive when it comes to policy making. Therefore, when data on the level of states is necessary, there will be a specific focus on Germany and France. Not on the United Kingdom, as they are not involved in the PESCO and will, if everything goes according to plan, leave the European Union soon.

Two types of sources will be used to conduct the analysis: primary and secondary sources. The primary sources provide a first-hand account of an event (UNSW, 2019), in this case European integration. Primary sources include the documents directly published by the actors involved in the process. The secondary sources are descriptions or explanations of the primary sources. These include analyses, interpretations and evaluations by third parties.

As for the primary sources, the empirical analysis will examine publications by both the European institutions as well as by the member states. These include statements, speeches, press releases and official, legal documents. For the secondary sources, various articles written on the Permanent Structured Cooperation are studied. An example of an important source of information on the interests and motivations of states is the research conducted by the Armament Industry European Research Group (ARES). ARES is coordinated by the French think tank named the Institute for International and Strategic Affairs (IRIS), and brings together policy experts that work across European universities and think tanks (IRIS, n.d.). The research group focuses on security and defence policies and mostly writes about the Permanent Structured Cooperation. They have written policy papers on the perspectives of various EU member states on the PESCO, which can be useful for the deciphering of the states' motivations. As IRIS, and thus also ARES, are independent from any higher-up control and work together with experts connected to universities and other think tanks, the research can be considered relatively unbiased. Another source of secondary information are outlets by the media. These can contain relevant information and can be used to verify other sources of information. The database of LexisNexis has been used to search for news articles relating to the Permanent Structured Cooperation and the countries of Germany and France.