Transgovernmental Networks and EU Security Governance

Article

Cooperation and Conflict 45(3) 312–330 The necessity of protection: © The Author(s) 2010 Reprints and permission: sagepub. Transgovernmental co.uk/journalsPermissions.nav DOI: 10.1177/0010836710378071 networks and EU security http://cac.sagepub.com governance

Simon Hollis

Abstract The remarkable increase in European security and defence integration in the past decade has presented a challenge to traditional integration theories. Although they remain relevant, these theories fail to take full account of the changing security architecture of Europe, which includes the rise of transgovernmental networks (TGNs). With a focus on EU civil protection, this article critically examines established definitions of TGNs and investigates how these networks influence the supranational and national levels of security cooperation. Findings point toward the emergence of an alternative form of European security governance that addresses the lack of authority in EU security policy.

Keywords civil protection, crisis management, European Union, security governance, transgovernmental networks, transgovernmentalism

EU cooperation on civil protection has experienced a rapid growth in preparing for and responding to natural and manmade disasters. In the last decade alone, common contin- gency plans and initiatives have been formed at the EU level in areas as diverse as floods, terrorist attacks, earthquakes and marine pollution (Boin, Ekengren and Rhinard, 2006; Kirchner and Sperling, 2007; Boin and Rhinard, 2008; Ekengren, 2008). Even though traditional integration theories failed to anticipate these and other developments in EU security cooperation, as Hanna Ojanen (2006) points out, they continue to provide value in understanding what drives cooperation in this sensitive policy field. However, this approach tends to overemphasize the explanatory power of traditional integration theories and underestimate the complexity of EU security governance. Cooperation is happening and integration is occurring, but this increasingly exists outside standard institutional frameworks and traditional state authorities. Security governance theory has attempted to depict this change by highlighting alternative and broader forms of security organization with particular emphasis on non-governmental forms of interaction that have ‘perforated’ state sovereignty (Krahmann, 2003; Webber et al., 2004; Kirchner,

Corresponding author: Simon Hollis, Hertie School of Governance, Quartier 110, Friedrichstraße 180, 10117 Berlin, Germany. [email: [email protected]] Hollis 313

2006; Kirchner and Sperling, 2007). Nevertheless, both of these perspectives on European security integration have failed to reflect on the Union’s increasing tendency to promote transgovernmental networks (TGNs) (Eilstrup-Sangiovanni, 2007; Eberlein and Newman, 2008; Lavenex and Wichmann, 2009, q.v. European Committee of Regions, 2009). This type of cooperation, which is characterized by fairly independent interaction between different governmental sub-units (Keohane and Nye, 1974; Sundelius, 1977; Raustiala, 2002; Slaughter, 2004), is an important and expanding form of governance. While the importance of networks has been well documented in other policy areas, such as social policy, telecommunications and the European Neighbourhood Programme (Sabel and Zeitlin, 2008; Walkenhorst, 2008; Lavenex and Wichmann, 2009), less is known about the influence and importance TGNs have on EU security policy, or how these networks relate to the sovereignty of nation-states.1 This article will address these deficiencies by critically examining how TGNs are defined and how these networks contribute to EU security governance. A set of formal and informal capacities to manage and mitigate an increasing number of major crises has emerged at the EU level in the last decade (Hollis, 2010), which has been tentatively described as the beginning of a ‘secure European community’ (Ekengren, 2007: 106). That is, a community that collectively deals with a wide range of transboundary threats that cross geographic and functional boundaries (Boin and Rhinard, 2008: 4), such as critical infrastructure break-downs, communicable diseases and terrorism. The supranational capacities designed to mitigate this potential include, for example, the Monitoring and Information Centre (MIC), a civil protection financial instrument and a European Centre for Disease Prevention and Control (ECDC) (European Council, 2007). Civil protection is a central aspect of this community which encompasses an all-hazard approach and pays little attention to the traditional divisions between internal and external security (Duke and Ojanen, 2006; Boin and Rhinard, 2008). This emerging policy sector will, therefore, be selected as the main unit of analysis in this article. Although a number of crisis-management capacities are established at the supranational level, supranational authority remains limited as member states continue to dem- onstrate considerable resistance to the sharing of responsibilities in the area of crisis management. In other words, cooperation in the event of a transboundary disaster is dependent on how member states chose to react (Kirchner and Sperling, 2007: 170–1). There is, consequently, a gap between EU capacity on the one hand and EU authority on the other (Börzel and Risse, 2008).2 This presents an unresolved puzzle for neo- functionalists: based on the classical writings of Ernst Haas, these scholars posit that the Union’s authority will incrementally increase through the spill-over of policy areas which are informed by divergent and expanding functional needs (Haas, 2004; q.v. March and Olsen, 1989; Sandholtz and Stone Sweet, 1997). This process of integration is, however, conspicuously absent in EU cooperation on civil protection, where there is often a ‘lag between rhetorical commitments . . . and the actual transfer of authority to Brussels’ (Boin and Rhinard, 2008: 19; Boin, Ekengren and Rhinard 2008: 27). Yet cooperation is increasing and integration is occurring in this field. To readily observe this form of cooperation analytical attention must focus less on structural outcomes and more on alternative forms of interaction such as TGNs. Indeed, the very lack of supranational authority that is understood to limit cooperation in the field of security and 314 Cooperation and Conflict 45(3) defence, can act as a catalyst by encouraging the growth of networks as a form of security governance. To be sure, this position does not negate neofunctionalism, but builds on it by broadening its field of perspective past the ‘mono-casual’ aegis of functional spill- over (Smith, 1996: part II). This article first sketches out how the EU cooperates in meeting transboundary threats, despite the limited authority it holds in this policy domain. This first section also provides a short synopsis of how this process is currently understood by scholars working on EU governance. Based on this literature and empirical observations, the second section provides a framework for understanding TGNs as a form of security governance within Europe. This framework is then used to examine a selected civil protection network, before providing an overview of other TGNs operating in the EU. Based on these findings, the influence TGNs have on security governance is elaborated upon, as well as the explanatory power of TGNs in European security research.

Coping with crises in Europe The energy debacles in Ukraine, the terrorist bombings in London and the forest fires in Portugal have highlighted Europe’s interconnectedness and vulnerability. Largely in reaction to these and several other crises, the EU has developed a crisis management capacity to prevent, prepare for and respond to natural and manmade disasters in a func- tionally interdependent Europe. The MIC, the Joint Situation Centre (SitCen) and the EU’s Judicial Cooperation Unit (EUROJUST) represent just some initiatives that have been instigated in the last decade aimed at tackling a wide range of common threats. However, it is not just past events that have shaped this policy area. Whether it is the result of an expanding democratic deficit, maintaining regulatory control or guarding national sovereignty, the effectiveness of EU integration in crisis management is restricted by the resistance of member states to concede political authority to the EU (Börzel, 2006; Etzioni, 2006; Kirchner and Sperling, 2007; Eberlein and Newman, 2008). However, this does not mean that cooperation has receded; rather, other forms of cooperation have been created to address the lack of authority of the EU. One approach has been the Open Method of Coordination (OMC). Through this non- regulatory process the Council produces guidelines and benchmarks from which member states create annual goals that are then subject to a peer review process (Borrás and Greve, 2004: 181; Eberlein and Newman, 2008: 28–9). The prominence of this approach, which was first devised for welfare cooperation, has received much acclaim by academics and practitioners (q.v. special issue in Journal of European Public Policy 11(2): 2004), and its versatility has commended it as a means of providing a more effective EU crisis management capacity (Ekengren, 2006). Another approach is through TGNs which have been suggested as an alternative to the OMC (Eberlein and Newman, 2008). This form of cooperation compensates for the lack of EU authority (vertical interaction) by promoting horizontal or heterarchical interaction (Eberlein and Newman, 2008: 30). Transgovernmentalism covers many aspects of the OMC, such as its emphasis on networks and learning; however, transgovernmentalism distinguishes itself from the OMC by paying specific attention to forums of cooperation beyond that of the EU ‘comitology’ committee system and intergovernmental cooperation (q.v. Christiansen, Hollis 315

2007). Transgovernmentalism also broadly defines its unit of analysis to include other areas of cooperation, such as informative networks. Although much has been written on TGNs, security networks within Europe have been rarely touched upon, which is argu- ably a growing and fundamental form of cooperation that will shape European security integration in the future. Understanding what networks are currently in place within the selected field of civil protection will provide a first glance at how significant these networks are and what potential they may have.

Network scholarship and transgovernmentalism Scholarly attention on transgovernmentalism was first hinted at by Karl Deutsch, James Rosenau and Karl Kaiser in the late 1960s.3 Their concerns were summed up by Kaiser, who stated: ‘because of the ensuing neglect of the processes of interpenetration, overlap, and societal interaction there is no theory connecting the two fields of national and international politics’ (1971: 792). This theoretical and conceptual deficiency was soon answered by the seminal work of Robert Keohane and Joseph Nye, who defined transgovernmentalism as ‘sets of direct interaction among sub-units of different governments that are not controlled or closely guided by the policies of the cabinets or the chief executives of those governments’ (Keohane and Nye, 1974: 43). As Anne-Marie Slaughter and David Zaring note, these tentative steps at ‘un-packing the state’ (2006) grew into current network scholarship, which also takes into account other nongovernmental, private and public networks. Thus, more recent studies have highlighted the growing support the EU gives to horizontal governance, which covers a wide range of policy areas from education and employment to health care and telecommunications (Sabel and Zeitlin, 2008; Walkenhorst, 2008). Counter terrorist and judicial networks have also been subject to investigation, revealing a number of horizontal networks within the EU, such as the Counter Terrorism Group and the Police Chief Task Force (Celaya, 2009: 28; q.v. Lugna, 2006; Edwards and Meyer, 2008; den Boer, Hillebrand and Nölke, 2008). Based on, or in the support of, Slaughter’s thesis on a disaggregated state, many of these networks are characterized as transgovernmental, albeit subject to variances in what constitutes a TGN. This body of literature has focused on a relatively ‘low’ area of political cooperation leaving those areas that affect the heart of state sovereignty, namely security and defence policy, aside.

Security governance The implication of security networks overlaps with another branch of literature that comes close to the network perspective and places attention on the ‘high’ area of political cooperation. Security governance provides an alternative method for observing the ‘regulation of international politics’ by analysing security cooperation outside traditional state authority (Kirchner and Sperling, 2007: 18). Both network scholarship and the security governance literature emphasize a heterarchical system of interaction, challenge the unitary state assumption, and agree on a broad concept of security (Webber et al., 2004; Kirchner, 2006; Slaughter and Zaring, 2006: 214; Kirchner and Sperling, 2007). Despite these similarities, the security governance literature tends to bracket 316 Cooperation and Conflict 45(3) or overlook governmental networks by drawing a distinction between governance and government, which often results in overemphasizing non-state security actors and ignoring TGNs. However, this does not mean that TGNs are to be excluded; scope for their inclusion certainly exists in the standard definition of security governance given by Webber et al.:

The coordinated management and regulation of issues by multiple and separate authorities, the interventions of both public and private actors (depending upon the issue), formal and informal arrangements, in turn structured by discourse and norms, and purposefully directed toward particular policy outcomes. (2004: 4)

Based on this broad definition, Emil Kirchner and James Sperling document transgovernmental activities in the area of judicial and police cooperation, revealing a fascinating array of security networks and institutions (2007: 138–51). Yet, the overall conclusions derived by Kirchner and Sperling reflect the main difference between theories of security governance and network scholarship. They argue that state sovereignty is perforated due to an expanding concept of security, the ‘obsolescence of traditional alliances’ and the dispersal of authority away from the state through the rise of other forms of security governance such as public-private security partnerships (2007: 16, 1–22). By contrast, network literature argues that the state has adapted to the new security environment by promoting interaction between governmental sub- units, whereby the state becomes disaggregated but retains its authority (Kirchner and Sperling, 2007: 16; Slaughter, 1997). Consequently, it can be hypothesized that sovereignty is not necessarily a barrier, as Kirchner and Sperling put it, but rather a catalyst for network and cooperative innovation in the face of increasing interdependence. Investigating the characteristics and the extent of governmental security networks in Europe will provide initial grounds to support or negate this hypothesis.

The EU’s new security role in Europe: Towards a theory of understanding At face value, the concept of transgovernmental networks appears clear, especially when anchored to Keohane and Nye’s oft-quoted 1974 definition. However, clarity soon fades upon closer inspection: some authors have departed from the original definition by including executives, while others only provide a simple description that over- looks the complex nature of TGNs.4 Alternative names have also been used to explain the same phenomena; the distinction between policy networks, policy communities and horizontal networks (q.v. Atkinson and Coleman, 1992), is not always clear and, as such, only compounds confusion. In an effort to make sense of these competing definitions and concepts of TGNs, four general characteristic traits are below set out as a means of locating and understanding TGNs. These are: membership, structure, autonomy and function of transgovernmental networks. The relationship between these variables will also be referred to as they are considered to influence the end function of TGNs. It should be noted that this delineation is not a search for a perfect model. Rather, it provides appropriate parameters and a rough guideline for defining and locating Hollis 317 important variables of change to stimulate and give clarity to the way one thinks about transgovernmental networks.

Membership The membership of TGNs is defined as specialized governmental sub-units who interact or network with their counterparts in other governments. At the operational level, membership can include judges, fire fighters or the police; at the strategic level, membership can include national regulators, police chiefs or officials (q.v. Lavenex and Wichmann, 2009). Thus, membership includes operational and bureaucratic fields and excludes executives – that is, ministers or heads of state – and those who reside outside their country of residence, such as military attachés and members of the foreign office. However, this ideal type of network membership rarely materializes. Instead, and in a slight devia- tion from previous definitions of TGNs, actors from NGOs or academic institutions should not be immediately discounted. Providing their representation remains a minor- ity, actors other than national practitioners should be included as a sub-variable that can influence the function of a network.5 Other inconsistencies include the frequency of participation, and the re-occurring attendance of participants, which provide an indicator of the importance members place on a network, as well as the extent to which collegiality may arise. The country a participant represents may also influence the end function of a network, depending on the amount of autonomy that exists between members’ representative countries and the EU.

Structure The structure of government networks is based on horizontal and informal cooperation (Keohane and Nye, 1974). Informal in this context means bilateral or multilateral interaction through regulated meetings with limited binding rules or voluntary contact through electronic mail or intranets. Horizontal means a regular dispersal of authority to members of a network (Kaiser, 1971). These voluntary and informal attributes mean that TGNs often rely on high levels of trust, peer-to-peer ties and a ‘sense of collegiality’ for decisions to be made and cooperation to continue (Keohane and Nye, 1974: 45; Slaughter and Zaring, 2006: 215). Thus, instead of a centralized and intergovernmental system where cooperation is set within a formal hierarchy, members ‘make pledges of good faith that are self-enforcing’ under the notion of ‘positive comity’ which is understood as ‘enduring cooperation between government agencies’ (Slaughter, 1997: 190–2). Therefore, a standard rule for the structure of TGNs is that decisions must be non-binding and made by consensus. Problems of definitions arise when this model is applied to empirical data, revealing nuances in the structure. Although it might be expected that a centralized form of cooperation would also be formal, this is not always the case: while some networks are highly informal, others reflect a hybrid of decentralized and centralized structures. These differ- ences have resulted in a patchwork of definitions pertaining to the structure of networks. Slaughter, for example, includes the possibility of bilateral and multilateral agreements and a Memorandum of Understanding (MoU) in her definition of TGN structures, albeit 318 Cooperation and Conflict 45(3) with no executive or legislative control (1997: 190). Others, however, maintain a stricter definition where transgovernmental cooperation is based on direct personal contact (Eilstrup-Sangiovanni, 2007: 9). The range between formal and self-enforcing structures of cooperation is also illustrated by Burkard Eberlein and Abraham Newman, who investigate ‘incorporated’ TGNs within the EU, which is essentially ‘another type of policy network’ that is more formal in character (2008: 30–1). The varying degrees of formality in TGNs are significant as this may affect the choices available to participants: a high level of formality might, for example, dissuade some members from joining due the sensitive nature of national security.6 Keeping this in mind, a TGN must remain voluntary, have few binding rules and no contractual agreements, decide by consensus and maintain a reasonable distribution of power.

Autonomy The degree of autonomy or the freedom of self-governance in transgovernmental networks is characteristically loose. This means that executives have little control over transgovernmental interactions. According to Keohane and Nye, a network is not transgovernmental when the members are ‘controlled or closely guided by the policies of the cabinets or the chief executives of . . . [their respective] governments’ (1974: 43). However, this is subject to a ‘matter of degree’ (Keohane and Nye, 1974: 43), which can usher in confusion over where transgovernmentalism begins and where intergovernmentalism ends. The scope of self-governance is not only restricted by the state, but also by formal international organizations. Keohane and Nye note that international organizations influence TGNs through its secretariat, its ‘aura of legitimacy’ and its reserve of information (1974: 51). A good example of this can be found in the extensive comitology system of the EU. Here, networks, such as the Permanent Network of National Correspondents (PNNC) in civil protection, consist of government officials from the EU-27 including a permanent chairmanship from the Commission. While the membership may qualify the PNNC as transgovernmental, this network is not a TGN because the chairperson can set the agenda and thereby maintain a strong amount of control, limiting the degree of autonomy. Investigating the links networks have to international organizations, such as financial links, as well as the harmonisation between a network’s mandate and the international organizations, will offer appropriate indicators of the degree of autonomy. The autonomy is considered one of the most important variables as it will have a direct affect on the membership, structure and function of a TGN. If there is a high amount of autonomy, the importance of membership, as a variable, will be higher: participants will have more authority and greater flexibility in how the agendas are formed and how the structure of the network can change. On the other hand, if autonomy is low, more con- straints will be placed on the structure, agenda and function of the network. A certain hierarchy can therefore be seen to exist between the variables, where autonomy will influence the structure and membership of a network. However, this does not mean that autonomy is independent from the effect of other variables. The type of membership, for example, may also influence the autonomy of a TGN: if the network includes represen tation from insurance companies, NGOs or academics, their influence may restrict the Hollis 319 autonomy, especially if this small group of actors contribute to the financing of a network. The most important variable in terms of influence, however, is the function of a network.

Function The function of transgovernmentalism is traditionally understood as either ‘transgovernmental policy coordination’, where policy can be regularized, or ‘transgovernmental coalition building’, where a coalition is built to influence a government’s decision based on the like-mindedness of the network’s members and domestic conflicts of interest (Keohane and Nye 1974: 44–50). Inspired by Slaughter (2004), Sandra Lavenex and Nicole Wichmann (2009) identify three types of TGNs, offering a more explicit account of function. These are informative, implementation and regulatory networks. The former does not produce any regulatory instruments, but is focused on information exchange and learning of best practices. The implementation networks, like policy coordination, cooperate to implement and enforce existing national or supranational laws and rules. The latter has an explicit mandate aimed at creating common rules and practices in a given policy area (2009: 86). The overriding principle in locating TGNs through these various functions is the degree of indirect control that a TGN has on the political agenda and legislative processes of a state or international organization. The type of function a TGN has is dependent upon the initial aims defined by the government or organization that designs the network. The origins of networks are, therefore, an important aspect to take into account. Of course, the function can change over time, which is dependent upon the degree of autonomy: it is assumed that a low amount of autonomy will mean that the initial goals and aims of a network will not change independently from the founder. The structure and membership is also influenced by the function of a network. The structure, for example, is usually decided upon at the first meeting where national or supranational influences will be most strongly felt.7 The membership will consequently fit in with the structure, autonomy and function of a network. The inter-relationship between these four variables is depicted in Figure 1 by the vertical multi-directional arrow. The Figure also contrasts transgovernmentalism and intergovernmentalism as two ideal types of governance, where a horizontal arrow reflects the variants between the two forms of governance.

Transgovernmental Intergovernmental

Membership Governmental Sub-units Executive Structure Decentralized Centralized Informal Formal Voluntary Compulsory Horizontal Vertical Degree of Non Ministerial Ministerial Autonomy

Function Indirect Direct

Figure 1. Characteristics of Transgovernmental Networks vis–a-vis Intergovernmentalism Source: Adaptation from Eilstrup-Sangiovanni (2007) 320 Cooperation and Conflict 45(3)

Civil protection in European government networks

Civil protection cooperation through TGNs appears to be an expanding form of governance. A growing awareness of the need to enhance collective cooperation – as reflected in an increase in member state participation in the EU mechanism for civil protection (Hollis, 2010) – coincides with a rise in the number of networks formed in the last decade.8 The number of TGNs in the area of crisis management is also expected to increase, as long as the assumption holds true that member states are continually forced to adapt to an interdependent environment in the face of transboundary threats. Apart from the increasing number of networks, diverse methods of coordinating interaction are also observable. The development of technology means that specialized electronic networks, such as the European Forest Fire Information System (EFFIS) and the Global Monitoring for Environment and Security (GMES) system, offer efficient and effective methods of cooperation at the operational level. These and other electronic networks, blogs and RSS-feeds provide a valuable complement to more traditional methods of cooperation, such as face-to-face network meetings and telephone calls, which should not be ruled out as an emerging form of transgovernmental cooperation. If there is a growing number of TGNs in civil protection, the question of how they contribute to security governance must be answered. In conjunction with the definitional framework, the following empirical description includes a specific case study on the EU Fire Safety Network, which is complemented with a broader description of other EU civil protection networks. As previously noted, the Permanent Network of National Correspondents (PNNC) represents national civil protection agencies of the EU-27 as well as Norway, Iceland, Liechtenstein and Croatia (House of Lords, 2009).9 This network presents a suitable point of departure by providing a reference point to locate other networks that their respective agencies or departments participate in.10 The methodologi- cal tools used to trace these links are through interviews with members of the PNNC, tracing reported network relationships on their respective websites and analysing minutes of meetings and other primary sources.

The EU fire safety network in the field of civil protection: A case study The proposal to form an autonomous network in the area of fire safety was based on a recommendation from an EU report on the ‘prevention of fire and other incidents’ (EU-FSN, 2004). As a result, the civil protection fire safety network emerged in 2004, together with the approval of the Commission’s Civil Protection Unit and the civil protection committee. Minutes of the network meetings are used as the main source material for this case study. A total of nine meetings are examined, starting in Brussels, 6–7 October 2004 and ending in London, 24–25 October 2008.11 The membership of the civil protection fire safety network includes national officials from crisis management agencies in EU member states or EEA countries. Since November 2008, the network has 22 members from the EU and two from EEA-countries. The network meets biannually for two days at a participant’s country. On average, the same official will attend between four and five network meetings before being replaced. This, combined with an average attendance rate of 63 per cent, places limits on the Hollis 321 possibility for a sense of collegiality to arise. A quarter of the network, however, has an attendance rate between 75 per cent and full participation and a very low turnover rate. This group of countries includes Sweden, Latvia, Estonia, Finland, Ireland and Hungary. In contrast, larger states, such as France, Germany and Italy, are fairly inactive. The voluntary and horizontal structure of the network is one possible explanation for the high activity among small and regionally defined countries. An even dispersal of authority in these networks means that smaller countries will have a larger influence than intergovernmental cooperation. The small group of countries also points towards the possibility that a sense of collegiality was previously formed around similar identity groups before the establishment of the network. Although it is horizontal and fairly informal, the network does have some formal rules of procedure, albeit restricted to bare essentials such as secretarial support or the composition of the network. Moreover, the formal rules of procedure stipulate that the rules can be changed at any meeting, that there are no voting procedures and no core group. A chairperson is also elected for a two- year period with the possibility for re-election. Cooperation tends to be formed based on ‘positive comity’ and high levels of trust as there are not contractual agreements. The existence and validity of the network are often, for example, tabled whereupon members will decide if collective cooperation is beneficial and should continue (EU-FSN, 2005a: 7, 2006a: 8, 2007a: 6, 2007a: 9). The EU fire safety network enjoys a high degree of autonomy from national executives and EU officials. As noted, the origins of the network began at the EU level. Commission officials were, therefore, present at the inaugural meeting where it was agreed that a Commission representative should participate. The Commission, however, never appointed a representative, which has provided for greater freedom of self-governance. Moreover, all operational and administrative costs come from the member states. This flexibility has provided the network with the ability to react quickly, which has been essential for fulfilling some of its aims. For example, participants were alerted that the Commission was going to present a standard on Reduced Ignition Propensity (RIP) cigarettes to the General Product Safety Committee of the European Parliament within a matter of days. Due to its loose structure and high autonomy, the network could act quickly by influencing the proceeding with a collective letter, which proved to be a cru- cial piece of evidence (EU-FSN, 2007b: 3). It should also be noted that cooperation between the network and other similar TGNs has increased. Although this may not have a direct affect on the autonomy, the agenda and aims of the network could be affected if contacts between networks increase. Consequently, the autonomy of TGNs are not only affected by nation states and international organizations, but also by the possible cross- fertilization of ideas between these networks. The function of the EU fire safety network is not easily confined to one category: it is informative, as it encourages information exchange and learning of best practices; it is implementive, as it discusses the best strategies of enforcing existing rules; and it is regu- lative, as it aims to promote new laws at the national and supranational level. In order to implement these functions the network uses a number of tools that coalesce around the overall function of networking. Members of the network agreed, for example, that RIP cigarettes should be endorsed in national and supranational legislature. In order to achieve this, the network wrote various letters to the Commission, the chairman attended 322 Cooperation and Conflict 45(3) meetings in the Commission and Parliament, and various contacts with relevant NGOs and national media were used. In an effort to muster greater collective support for the initiative, active networking between similar TGNs such as the International Association of Fire and Rescue Service (CTIF), Fire Prevention Commission and the Federation of the European Union Fire Officers Association (FEU) was also carried out. More gener- ally, members were encouraged to influence their PNNC colleagues to promote fire prevention as a topic on the agenda of the civil protection committee (EU-FSN 2004: 2).

Other EU civil protection networks In line with the definitional framework outlined in this study, the similarities and differ- ences of other civil protection TGNs are examined. A comparative analysis on membership, structure, degree of autonomy and function of networks helps to further elucidate the type and purpose of TGNs, and aids in understanding their influence in EU security governance. The analysed TGNs include: the EU Network on Information to the Public; the informal working group on forest fire prevention experts; the FEU; the Network on the Protection of European Capital Cities (Procapites); the SafeSeaNet (SSN), the Counter Terrorism Group; the Police Working Group on Terrorism; and the ATLAS network.12 The origins of these networks can help to explain their various functions, which are categorized in two separate categories. First, TGNs are established through an organic process, whereby a network will emerge out of an initial EU-funded project, exchanges (Interview UK, 2009) or through EU agencies, such as EUROPOL, EUROJUST, CEPOL, the ECDC and FORETEX. Examples of these networks include the EU Network on Information to the Public, the Civil Protection Fire Safety Network, the Network of National Training Coordinators and the SSN. Second, TGNs are also created in reaction to events or through dissatisfaction with the legislative status quo.13 The Police Working Group on Terrorism, for example, was created in 1979 out of a reaction to IRA bombings and the Baader Meinhoff terrorist exploits, and the ATLAS group was formed in reaction to the 9/11 terrorist attacks in New York (Benediek, 2006; Block, 2007). The FEU also falls in this category, as the causal motivation for creating this network was due to national fire fighting departments’ frustration that little legislative control or advice was provided for at the operational level (Nuesslar, 2008: 1). The first group of networks tend to have higher autonomy, consist of national crisis management representatives, have a fairly loose and decentralized structure, and gener- ally have an informative function. An explanation for the prominence for this type of function is due to the EU’s perceived value of networks. At the inauguration of the civil protection fire safety network, for example, civil protection representatives from the Commission noted that the network was important due to its role as a valuable ‘reference point’ that various Directorates General can contact for information and advice (EU-FSN, 2004a: 1). The loose structure and informal nature of the networks mean that TGNs are liable to quickly dissolve if no unanimous position of collective aim can be achieved. Indeed, this may have been the case for Procapites and the informal working group on forest fire prevention experts, which did not continue past the Commission’s funding expiry date. This flexibility also means that a TGN can more easily change its function Hollis 323 to adapt to internal or external changes, such as the configuration of membership or external events. An informative network, for example, might easily change to an implementive role as the case requires. A good example of this is the EU fire safety network. The second group tends to have a more formal structure and a low amount of autonomy and includes other actors such as the police. This is, perhaps, understandable con- sidering the highly political and sensitive subject of terrorism and other major disasters. The counter-terrorism networks represent an ‘operative’ function, which is essentially another type of function not emphasized by Keohane and Nye or Lavenex and Wichmann. The purpose of this type of network is to enhance collective operational capacities in a particular field of expertise.

Findings Based on the review of TGNs, it should be clear that the features that define a network are dynamic, and not static as a number of scholars presume. This means that ideal types of governing interaction – whether it is intergovernmental or transgovernmental – rarely exist in their pure forms. Rather, it is suggested that cooperation thrives between these two reference points, as Figure 1 illustrates. This reflects a mixed system of governance, where the membership, the structure, the degree of autonomy and the function of TGNs act as a guide for classification and as independent variables to identify change. This study also challenges traditional interpretations of transgovernmental networks in terms of how autonomy is limited by external influences, usually thought of in terms of the freedom of governance between the nation-state and TGNs. Emphasis is placed on the multifaceted influences not only from the state, but also from international organizations, the EU and other transgovernmental networks.14 The promotion of TGNs offers the EU a soft tool to promote cooperation in an area where the nation-state is unwilling to estab- lish any form of cooperation that could jeopardize its sovereignty. This, however, is not a day-to-day goal of EU practitioners, who understand TGNs as a valuable source of information and link to member states. This relationship is depicted in Figure 2, where the traditional view is represented as C1 and the influence of international bodies is represented as C2. The triangles reflect a nation-state (Kaiser, 1971), while the filled-in section of the triangles represent officials who function below the executive level and participate in transgovernmental networks. The multi-directional arrows represent the flow of transactions, communication and influence.

Figure 2. The Level of Interaction in Transgovernmental Networks 324 Cooperation and Conflict 45(3)

The importance of networked governance

The networks examined in this article are ‘structured by discourse and norms’ and are ‘purposefully directed toward particular policy [and operational] outcomes’ (Webber et al., 2004: 4). They, therefore, represent an important sub-type of security governance. But how important are TGNs in relation to security governance and European integration? The influence these networks have on learning and creating best practices is gener- ally high, the influence they have on policy coordination is medium and the influence they have on creating new rules and practices is low. Put differently, all of the networks have an informative function, some have an influence on the legislative processes at the national and supranational level and very few, if any, have a direct mandate on creating new rules of procedure in a policy domain. If these networks have a predominately informative function, what influence does this have for EU crisis management? The importance of collegiality, learning and the exchange of knowledge can form a pretext for the shifting of loyalties (Haas, 2004), creating trust (Deutsch, 1957) and shap- ing identities (Risse, 2006). A sense of community or a ‘we-feeling’ may then develop (Deutsch, 1957), which encourages the transfer of authority to a ‘new centre’ (Haas, 2004; Risse, 2006: 77). However, it remains doubtful if networks can encourage this process. If a sense of collegiality and shifting loyalties develops in a network, the participants would need to meet on a regular basis. At best, the networks examined in this study meet biannually and usually on an annual basis. Furthermore, the turnover rate of participants occurs approximately every four years, which limits the scope for any sense of collegiality to form. This deficiency can be partly addressed through the use of specialized websites and newsletters; however, the relatively short life-span of these networks will usually mean that any sense of a common identity would have formed before and not after the formation of a network. An exception to this would be the operational function of some of the networks, where learning is essentially administered through ‘doing’ rather than through ‘thinking’ about cooperation (q.v. Pouliot, 2008). This would presumably increase the effectiveness of joint operations in the event of a major crisis and also entails more fre- quent cooperation. Another exception is that the membership of networks can include the same actors (Interview UK, 2009). Even structured committee meetings that have different executive functions in the Council and the Commission often include the same person- nel. Thus, vertical modes of governance and horizontal networks cannot be viewed in complete isolation, but rather as a larger entity where intricate and sophisticated forms of ‘collegiality’ can be reinforced and trust built. If this is true, it would certainly provide greater opportunities for coalition building among and between various networks. The importance of TGNs in promoting European integration in the area of security is highly dependent upon the degree of autonomy: if, for example, a TGN is informal, has a loose structure and a high amount of autonomy, it may contribute to integration by encouraging further cooperation and enhancing visibility in a given policy area. The high degree of autonomy, combined with the possibility of networking with other networks, also means that coalitions can be formed, where a TGN can become a ‘third’ source of power in addition to the state and the EU. If this is the case, disintegration may follow. On the other hand, it is also possible that the ‘operative’ networks will act as a force for European integration. For example, the proposal within the ATLAS group to have a Hollis 325 multinational response team (Block, 2007) could mean that German police may be sent to France to help in the event of a terrorist attack. If this proposal would ever eventuate, which could go against some member states’ interests, it could make it easier for the EU to institutionalize this type of cooperation in the future.

The power of networks The friction produced by the governing principles of national sovereignty and European solidarity have dominated traditional and current accounts of European integration. This relationship is exemplified in the area of security, where a gap between EU authority and capacity clearly exists. However, the need for cooperation – based on the functional and interdependent needs of EU member states – continues unabated in a Europe where transboundary crises are a real and present threat. Cooperation has, therefore, emerged through the rise of transgovernmental networks, which have emerged not only from the nation states, but also from the EU. The rise of an alternative form of governance does not necessarily mean a departure from neo-functional scholarship, but it does challenge its limited views on cooperation by looking outside established forms of cooperation. As this article has demonstrated, a number of TGNs are performing a role that influences the normative and legislative environment at national and supranational levels of governance. In this sense, networks can be seen as bridges connecting these two levels of governance at the bureaucratic and operational levels of cooperation. While this may be just one part of Puchala’s elephant, it remains an important aspect of European security cooperation as it touches a number of important aspects that are often overlooked by traditional integration theories and it provides a suitable platform for future research (Puchala, 1971). As a central issue, the changing position of power vis-a-vis the Westphalian state deserves greater analytical attention. Transgovernmental security governance opens an alternative view of cooperation that, while not distinct from state authority, bypasses established intergovernmental links. But is Slaughter’s thesis of a disaggregated state correct? Can the state really maintain its power through network adaptation in the face of an increasingly interconnected world, or are Kirchner and Sperling right when they say that power has dispersed beyond the state? The results of this study support the view that the power of the state is disaggregating and has the potential for proliferation. However, the end result will be determined by the degree of autonomy of a network and its ability to form coalitions. The influence the networks will then have can be determined by the membership, structure and function of a network. The case study and review of other networks show some signs of coalition building – such as between other networks – but this was not outside the interests of the EU or member states, although the possibility remains open in the future. An area of research that is currently underdeveloped is the importance of the notion of trust on governance. Most networks have non-binding rules on the behav- iour and practice of members. Although some are more structured that others, the idea of non-contractual cooperation is radically different to the neo-functional tradi- tion and more in line with reflective-based theories. The implication of trust as a driver of cooperation in this sense may provide an avenue of fruitful and pioneering 326 Cooperation and Conflict 45(3) research that may push the conceptual frontiers of contemporary research in European integration theory and security studies.

Conclusion The paradox of regional cooperation involving the unresolved conflict between national sovereignty and European solidarity is a constant ‘thorn in the side’ of the EU. The sensitive policy area of security only highlights this paradox, as the need to cooperate collectively to meet major crises has become increasingly poignant in the last decade. While this apparent disability will not be miraculously healed, it does provide a suitable context to investigate transgovernmental networks as an alternative form of security governance that bypasses the limited focus of intergovernmental and neo-functional theories. Whether it is a terrorist attack on a Russian pipeline, an oil spill in the Baltic Sea or forest fires in Southern Europe, transboundary threats can easily escalate, cross political bor- ders and affect millions of lives. Although the full extent of TGNs remains elusive and demands further research, the importance and necessity of coping with large-scale emer- gencies in Europe, through creating common practices and sharing knowledge, should be highly valued as building blocks for a more secure and networked Europe.

Notes I would like to thank Magnus Ekengren and the two anonymous reviewers for their insightful comments.

1. One exception to this is Eilstrup-Sangiovanni (2007). Krahmann (2003) also focuses on the multilevel networks of European foreign policy albeit with little attention on TGNs. 2. This is despite some shedding of national authority in the area of internal security, such as the introduction of QMV in the third pillar of Justice and Home Affairs (Risse, 2006). 3. For example, see Deutsch’s concept of the ‘linkage system’ (1966: 12–3). Also see Kaiser’s concept of multibureaucratic decision making (1971: 797–8). 4. For example, Smith (1996) provides no explicit definition of transgovernmealism, even though this is a central point in his analysis. He also includes members of the foreign office as transgovernmental actors, which is traditionally not included in TGNs (q.v. Keohane and Nye, 1974: 43–4). 5. For more on the difference between transnational and transgovernmental networks see Nye and Keohane (2004: 172). 6. The ATLAS network on counter terrorism is a good example of this, where the proposal for a legal framework was rejected by members who prefer an informal and more opaque form of cooperation (Block, 2007). 7. This was the case for the EU Fire Safety Network, which had direct input from the Commission in the first meeting. Corresponding meetings, however, were free from Commission representation (q.v. EU-FSN, 2004). 8. A majority of networks analysed in this article have been established only in the last decade. 9. In order to assist in the legislative process in the Commission and Council, a number of committees have been employed. The primary committee connected to the Directorate General for Hollis 327

the Environment (DG ENV) in the European Commission is the Civil Protection Committee (CPC), which is made up of the members of the PNNC. The Council is also assisted by the Committee for Civil Protection (PROCIV), which will often consist of the same members from the CPC, albeit executing different functions. Apart from these committees, the general directors of the national emergency management authorities of the member states will meet biannually (q.v. European Commission 2009). 10. This is, of course, flexible in its approach as there are networks mentioned that do not have a direct connection to the PNNC but have been found through links emanating from the PNNC activities. 11. This material may be retrieved online at the Swedish Civil Contingencies Agency’s website. 12. For more on counter terrorism networks see Edwards and Meyer (2008); den Boer, Hillebrand and Nölke (2008); Block (2007); special issue of Journal of Common Market Studies 46(1) (2008). 13. It should be noted that the SSN also qualifies in this categorization. It was established in 2002 in reaction to the sinking of the oil tanker Erika in the Bay of Biscay in 1999 (EMSA, 2009). 14. The EU’s explicit interest in forming networks marks an interesting shift in how TGNs are utilized. During the 1980s and early 1990s, TGNs played a large role in the Europeanization of the European Political Community (Smith, 1996). These networks were purposively created by member states to contend against the European Community’s authority.

References Atkinson, Michael M. and Coleman, William D. (1992) ‘Policy Networks, Policy Communities and the Problems of Governance’, Governance 5(2): 154–80. Barrett, Scott (2007) Why Cooperate? The Incentive to Supply Global Public Goods. UK: Oxford University Press. Benediek, Annegret (2006) ‘EU Strategy on Counter-terrorism: Steps Towards a Coherent Network policy’, SWP Research Paper RP 12. Block, Lado (2007) ‘Europe’s Emerging Counter-terrorism Elite: The ATLAS Network’, Terrorism Monitor 5(3). Available at: www.jamestown.org; accessed on 30 September 2009. den Boer, Monica, Hillebrand, Claudia and Nölke, Andreas (2008) ‘Legitimacy under Pressure: The European Web of Counter-Terrorism Networks’, Journal of Common Market Studies 46(1): 101–24. Boin, Arjen and Rhinard, Mark (2008) ‘Managing Transboundary Crises: What Role for the European Union?’, International Studies Review 10: 1–26. Boin, Arjen, Ekengren Magnus and Rhinard, Mark, eds (2006) ‘Special Issue: Protecting the Union: The Emergence of a New Policy Space’, Journal of European Integration 28(5): 405–529. Boin, Arjen, Ekengren Magnus and Rhinard, Mark (2006) ‘Protecting the Union: Analysing an Emerging Policy Space’, Journal of European Integration 28(5): 405–23. Boin, Arjen, Ekengren Magnus and Rhinard, Mark, eds (2008) Security in Transition: Towards a New Paradigm for the European Union. ACTA-Series B 41, Stockholm: National Defence College. Borrás, Susana and Greve, Bent (2004) ‘Preface’, Journal of European Public Policy 22(2): 181–4. Börzel, Tanya A., ed. (2006) The Disparity of European Integration: Revisiting Neofunctionalism in Honour of Ernst B. Haas. Oxon, Canada; New York, USA: Routledge. 328 Cooperation and Conflict 45(3)

Börzel, Tanya A. and Risse, Thomas (2008) ‘Revisiting the Nature of the Beast – Politicization, European Identity, and Postfunctionalism: A Comment on Hooghe and Marks’, British Journal of Political Science, published online by Cambridge University Press: http://www. unc.edu. Celaya, Fernando (2009) ‘The Terrorist Threat is Being Materially and Normatively Shaped by National and Global Institutions of Law and Order: Spain and Beyond’, Athena Intelligence Journal 4(1): 7–34. Christiansen, Thomas, ed. (2007) The Role of Committees in the Policy-Process of the European Union: Legislation, Implementation and Deliberation. Edward Elgar. Deutsch, Karl (1966) Nationalism and Social Communication: An Inquiry into the Foundations of Nationality. Cambridge, Massachusetts; London, England: The MIT Press. Deutsch, Karl W., et al. (1957) Political Community and the North Atlantic Area. Princeton, NJ: Princeton University Press. Duke, Simon and Ojanen, Hanna (2006) ‘Bridging Internal and External Security: Lessons from the European Security and Defence Policy’, European Integration 28(5): 477–94. Eberlein, Burkard and Newman, Abraham L. (2008) ‘Escaping the International Governance Dilemma? Incorporated Transgovernmental Networks in the European Union’, Governance: An International Journal of Policy, Administration, and Institutions 21(1): 24–52. Edwards, Geoffrey and Meyer, Christoph. O. (2008) ‘Introduction: Charting a Contested Transformation’, Journal of Common Market Studies 46(1): 1–25. Eilstrup-Sangiovanni, Mette (2007) ‘Government Networks in International Security’, Working Paper, Robert Schuman Centre for Advanced Studies, EUI, Transatlantic Programme Series, RSCAS 2007(24): 1–53. Ekengren, Magnus (2006) ‘New Security Challenges and the Need for New Forms of EU Cooperation: The Solidarity Declaration against Terrorism and the Open Method of Coordination’, European Security 15(1): 89–111. Ekengren, Magnus (2007) ‘The Internal–External Security Challenge for the EU’, Studia Diplomatica LX(1): 81–107. Ekengren, Magnus (2008) ‘EU Civil Protection: An Ascending Sector’, in A. Boin, M. Ekengren and M. Rhinard (eds) Security in Transition: Towards a New Paradigm for the European Union, pp. 47–63. ACTA-Series B 41, Stockholm: National Defence College. Etzioni, Amatai (2006) ‘The Community Deficit’, Journal of Common Market Studies 45(1): 23–42. Haas, Ernst B. (1966) ‘International Political Communities’, in International Political Communities: An Anthology. New York: Anchor Books Doubleday and Company. Haas, Ernst B. (1975) The Obsolescence of Regional Integration Theory. Research Series, no. 25. University of California, Berkeley: Institute of International Studies. Haas, Ernst B. (2004) The Uniting of Europe: Political, Social, and Economic Forces 1950–1957 (3rd ed.). Notre Dame, IN: University of Notre Dame Press. Hodson, Dermot and Maher, Imelda (2001) ‘The Open Method as New Model of Governance’, Journal of Common Market Studies 39(4): 719–46. Hollis, Simon (2010) National Participation in EU Civil Protection. Report, ACTA-Series B 42. Stockholm: National Defence College. Kaiser, Karl (1971) ‘Transnational Politics: Toward a Theory of Multinational Politics’, International Organization 25(4): 790–818. Hollis 329

Keohane, Robert O. and Nye, Joseph S. (1974) ‘Transgovernmental Relations and International Organizations’, World Politics 27(1): 39–63. Kirchner, Emil (2006) ‘The Challenge of European Union Security Governance’, Journal of Common Market Studies 44(5): 947–68. Kirchner, Emil and Sperling, James (2007) EU Security Governance. Manchester and New York: Manchester University Press. Krahmann, Elke (2003) ‘Conceptualizing Security Governance’, Cooperation and Conflict 38(5): 5–26. Lavenex, Sandra and Wichmann, Nicole (2009) ‘The External Governance of EU Internal Security’, European Integration 31(1): 13–102. Lugna, Lauri (2006) ‘Institutional Framework of the European Union Counter-Terrorism Policy Setting’, Baltic Security and Defence Review 8: 101–27. March, James G. and Olsen Johan P. (1989) Rediscovering Institutions: The Organizational Basis of Politics. New York: The Free Press. Nye, Joseph S. and Keohane, Robert O. (2004) ‘Transgovernmental Relations and International Organizations’, in J. S. Nye (ed.) Power in the Global Information Age: From Realism to Globalization, pp. 171–91. London; New York: Routledge. Ojanen, Hanna (2006) ‘The EU and Nato: Two Competing Models for a Common Defence Policy’, Journal of Common Market Studies 44(1): 57–76. Pouliot, Vincent (2008) ‘The Logic of Practicality: A Theory of Practice of Security Communities’, International Organization 62: 257–88. Puchala, Donald (1971) ‘Of Blind Men, Elephants and International Integration’ Journal of Common Market Studies 10(3): 267–84. Raustiala, Kal (2002) ‘The Architecture of International Cooperation: Transgovernmental Networks and the Future of International Law’, Virginia Journal of International Law 43(1): 1–92. Risse, Thomas (2006) ‘Neofunctionalism, European Identity, and the Puzzles of European Integration’, in Börzel, Tanya A. (ed.) The Disparity of European Integration: Revisiting Neofunctionalism in Honour of Ernst B. Haas, pp. 75–94. Oxon, Canada; New York, USA: Routledge. Sabel, Charles F. and Zeitlin, Jonathan (2008) ‘Learning from Difference: The New Architecture of Experimentalist Governance in the EU’, European Law Journal 14(3): 271–327. Sandholtz, Wayne and Stone Sweet, Alec, eds (1997) European Integration and Supranational Governance. Oxford University Press. Slaughter, Anne-Marie (1997) ‘The Real New World Order’, Foreign Affairs 76(5): 183–97. Slaughter, Anne-Marie (2004) A New World Order. NJ: Princeton University Press. Slaughter, Anne-Marie and Zaring, David (2006) ‘Networking Goes International: An Update’, Annual Revue of Law and Social Science 2(2): 11–29. Smith, Michael E. (1996) ‘The Europeanization of European Political Cooperation’, Institute on Global Conflict and Cooperation. Availabe at: http://www.ciaonet.org/wps/ smm01/index.html, accessed on 29 April 2009. Sundelius, Bengt A. (1977) ‘Trans-governmental Interactions in the Nordic Region’, Cooperation and Conflict 12: 63–85. Walkenhorst, Heiko (2008) ‘Explaining Change in EU Education Policy’, Journal of European Public Policy 15(4): 567–87. Webber, M., et al. (2004) ‘The Governance of European Security’, Review of International Studies 30: 3–26. 330 Cooperation and Conflict 45(3)

Interviews and Documents

EMSA (2009) ‘European Maritime Safety Agency Extranet’. Available at: https://extranet.emsa. europa.eu/index.php?option=com_docman&Itemid=120; accessed on 8 May 2009. EU-FSN (2004) ‘EU Fire Safety Network: 1st Meeting’, Brussels, 6–7 October. Available at: www.raddningsverket.se; accessed on 30 September 2009. EU-FSN (2005a) ‘EU Fire Safety Network: 2nd Meeting’, Frankfurt, 30–1 May. Available at: www.raddningsverket.se; accessed on 30 September 2009. EU-FSN (2006a) ‘EU Fire Safety Network: 4th Meeting’, Amsterdam, 22–3 May. Available at: www.raddningsverket.se; accessed on 30 September 2009. EU-FSN (2007a) ‘EU Fire Safety Network: 6th Meeting’, Prague, May 31–June 1. Available at: www.raddningsverket.se; accessed on 30 September 2009. EU-FSN (2007b) ‘EU Fire Safety Network: 7th Meeting’, Athens, November 27–8. Available at: www.raddningsverket.se; accessed on 30 September 2009. European Commission (2009) ‘22nd meeting of the Directors-General for Civil Protection of the European Union, of the European Economic Area and of Croatia, Turkey and the former Yugoslav Republic of Macedonia’, European Commission Working document DG 22/0/1. Prague, 22–4 April 2009. Draft annotated agenda. European Committee of the Regions (2009) ‘The Committee of the Regions white paper on multilevel governance’, CONST-IV-020: CdR 89/2009: Brussels; Belgium. Rapporteurs: Luc Van den Brande & Michel Delebarre. European Council (2007) Council Decision of 5 March 2007 establishing a Civil Protection Financial Instrument in Council decision 2007/162/EC, Euratom. House of Lords (2009) ‘Civil Protection and Crisis Management in the European Union: Report with Evidence’, European Union Committee 6th Report of Session 2008–9. London, House of Lords. Interview UK (2009) British PNNC official, 18/03/2009. Nuesslar (2008) ‘FEU Developments’, paper presented at the 51st Annual Conference of Chief Fire Officers Association, Galway, Ireland. Available at: www.galway.ie/en/ Services/ FireService/CFOAConference2008; accessed on 8 May 2009.

Author Biography Simon Hollis is a PhD candidate at the Berlin Graduate School for Transnational Studies (BTS), a joint endeavour of the Freie Universität Berlin (FUB), the Hertie School of Governance (HSoG), and the Social Science Research Center Berlin (WZB). He is also an affiliated researcher in the programme for European Security (EUROSEC) at the National Swedish Defence College. His research interests include regional security studies, crisis management, European integration and political theory. Address: Hertie School of Governance, Quartier 110, Friedrichstraße 180, 10117 Berlin, Germany. [email: [email protected]]