Spillover dynamics and inter-institutional interactions between CSDP and AFSJ: moving towards a more joined-up EU external migration policy?

ABSTRACT The EU’s response to the European ‘refugee crisis’ has involved parallel efforts to strengthen the EU’s capacity for external migration management in the domains of CSDP and AFSJ. To provide a swift response to the ‘refugee crisis’ in 2015, EU member states decided to utilise existing operational CSDP capabilities as a short-term foreign policy tool. Simultaneously, we have seen an impressive strengthening of the mandate and capabilities of ASFJ actors in external migration governance. Although institutional links between civilian CSDP and AFSJ have existed for over two decades, the parallel task expansion has created new demands for inter-institutional coordination between the two policy areas. Combining neofunctionalist theory with insights from research on inter-institutional overlap, this article argues that responses to neofunctionalist spillover pressures may trigger inter-institutional dynamics, which can involve both cooperation as well as conflict. Empirically, we find that these inter-institutional interactions have had important implications for integration in both policy areas, which cannot be fully understood by studying these policy domains in isolation.


Introduction
Since the European 'refugee crisis' in 2015, the European Union (EU) has taken several notable steps to strengthen its capacity for external migration governance (Niemann & Zaun 2023, this issue;Stutz 2023, this issue). In the domain of the EU's Common Security and Defence Policy (CSDP), the mandates of civilian CSDP missions were extended to include tasks related to migration and border management. EUCAP Sahel Niger and EUBAM Libya witnessed a gradual widening of their mandates, with the fight against irregular migration and the disruption of organised crime, including smuggling activities and human trafficking, being added to their task descriptions. This broadening of CSDP tasks was further institutionalised with the adoption of the Civilian CSDP Compact (CCC) in 2018 (Pirozzi and Musi 2019, 2).
The expansion of tasks and relevance of CSDP in the field of EU external migration policy since 2015 is somewhat puzzling, given that the EU had already in place other instruments and agencies mandated for external migration management within the Area of Freedom, Security and Justice (AFSJ), in particular Frontex. Moreover, against the backdrop of the 2015 refugee crisis the latter has been systematically strengthened by the 2016 Regulation on the European Border and Coast Guard and the updated version of the Regulation adopted in 2019 (Regulation (EU) 2016/1624 n.d.; Regulation (EU) 2019/1896 n.d.). Frontex has built-up its own capacity and resources in external migration management, including through the envisaged development of a new standing corps of 10,000 operational staff until 2027 (Tammikko 2019). Whilst loose institutional links between civilian CSDP and AFSJ have existed for over two decades (Parkes 2017), the increasing overlap of mandates and the simultaneous expansion of tasks and resources in the two domains has raised again the issue of how to foster coordination and avoid duplication and competition between their institutional structures and actors.
Combining neofunctionalist theory and research on inter-institutional overlap, this article develops a theoretical framework that seeks to capture inter-institutional dynamics in the evolution of European integration. Specifically, we argue that in situations where external challenges involve issues that cut across different EU policy domains, responses to neofunctionalist spillover pressures may trigger inter-institutional dynamics that can have a significant impact on the evolution of the implicated policy domains. Empirically, we highlight the relevance of institutional dynamicswhich can involve both cooperation as well as conflictbetween CSDP and AFSJ actors in the field of external migration governance. In particular, we argue that these inter-institutional interactions have had important implications for integration in both policy areas, which cannot be fully understood by studying these policy domains in isolation.
The article contributes to the research on EU external migration policy in three ways. First, it proposes a novel theoretical framework to explain dynamics in EU external migration policy by combining neofunctionalist theory with insights from the literature on institutional overlap. Through its particular focus on explaining policy-making dynamics and outcomes, and its contributions to theorising external and endogenous functional pressures across policy areas, neofunctionalism should be apt to account for changes at the interface of EU security and migration policies, in particular with regard to instances of incremental reforms such as the case at hand (Haas 2004, xvi). Second, the article also contributes to recent neofunctionalist theorising of CSDP (Bergmann 2019; Haroche 2020; Haakanson 2021) by integrating dynamics resulting from inter-institutional overlap. Third, the article contributes to this special issue's objective to shed light on the drivers of EU external migration policies post-2015 and its different underlying regulatory approaches. Concerning the latter, the EU's policy response at the migration-security interface corresponds to what the guest editors have described as a focus on remote control without taking into account third countries' and migrants incentives (see Niemann & Zaun 2023, this issue). Explaining how and why EU policy-makers have turned to instruments beyond the realm of migration policy to deal with the European 'refugee crisis' helps to better understand how the EU has sought to implement its 'joined-up approach' to EU external policies that the EU Global Strategy (European Union 2016, 49-50) has called for.
The article builds on secondary literature, policy reports, and official EU documents as well as eight semi-structured interviews with EU officials (EEAS, Commission, Frontex), member state representatives and independent experts (see Annex I). Methodologically, we draw on what has been described as 'minimalist process tracing' (Beach & Pedersen, 2019, 33-35) to explore evidence linking the individual steps of our theoretical argument to dynamics at the CSDP-AFSJ interface. We proceed as follows. First, we develop our theoretical argument on how neofunctionalist spillover pressures spur cross-policy, inter-institutional interactions and hypothesise the conditions under which they lead to increased cooperation or competition. Building on a short overview of the development of relations between CSDP and AFSJ until 2015, we then turn to the empirical analysis of integration dynamics and cross-policy institutional interactions between the two policy domains over the past six years.

Neofunctionalism and inter-institutional politics
Neofunctionalism has made important contributions to our understanding of crisisinduced integration dynamics (Lefkofridi and Schmitter 2015;Schimmelfennig 2015), including with respect to the European refugee crisis. Research shows that the refugee crisis that peaked in 2015 acted as a catalyst for further integration in the field of EU migration policy, creating additional functional pressures to enhance the EU's capacity for external border management (Niemann and Speyer 2018). At the same time, the crisis contributed to integration dynamics in the EU's Common Security and Defence Policy (CSDP), not least by expanding the mandate and tasks of civilian CSDP missions in the domain of external border management (Bendiek and Bossong 2019;Müller and Slominski 2021).

Neofunctionalism and European integration: transcending a compartmentalised perspective
Previous works employing a neofunctionalist perspective have approached integration dynamics emerging from the refugee crisis in a somewhat 'compartmentalised' way, often focusing on a particular policy sector or issue area (Kriesi et al. 2021;Niemann and Speyer 2018). This 'compartmentalised' viewwhich also dominates studies of integration dynamics in the domains of CSDP and AFSJmight be owned to the neo-functionalist concept of 'functional spillover', which follows a sector-based logic. Functional spillover pressures relate to situations in which progress in an integrated policy sector cannot be successfully achieved without extending integration to new, related sectors (Haas, 1958, 297). They may also occur when the policy goals pursued in a little or non-integrated sector can be more effectively pursued with the support of legal and policy instruments located in more integrated, related sectors (Haroche 2020, 857). What has received little attention, however, is that functional spillover pressures may give rise to different institutional responses in the affected EU policy areas. As we argue below, these different institutional responses may range from cooperation to open conflict between functionally overlapping EU policy domains.
Functional overlap between different EU policy areas and sectors also has important implications for other spillover dynamics identified by neo-functionalists. 'Cultivated spillover' relates to situations where supranational EU actors benefiting from integration promote policy initiatives that expand their own tasks, whilst previous integration steps might also create sites for elite 'socialisation' and 'experiential learning' that have been described as 'political spillover' (Lindberg & Scheingold, 1970, 119). In functionally overlapping EU policy domains, different supranational actors may seek to promote greater integration in their respective policy fields and sectors, which can enhance both, the need for coordination and cooperation as well as the potential for rivalry and conflict. As Stetter (2004) has argued for EU foreign policy more generally, the original design of the Maastricht Treatywhich has separated EU foreign policies into loosely connected pillarshas over time given way to a cross-pillar politics setting. This cross-pillarization has been the result of efforts to address the 'functional indivisibility' of foreign policy through cooperative arrangements. However, growing coordination and cooperation between functionally overlapping policy domains and sectors may not easily be achieved. As existing research shows, EU politics also involve rivalry between different EU actors and institutions (Fiott 2015)which have their own policy-preferences, institutional selfinterests and organisational cultureas well as bureaucratic politics within them (Smith 2013;Simpson 2000).
To understand the institutional implications of spillover dynamicsbe they functional, cultivated, or politicalit is thus important to pay close attention to inter-institutional dynamics between functionally overlapping policy domains. These interinstitutional dynamics are particular relevant in situations of external events and crisis such as the European refugee crisis, which have been identified as a forth type of neofunctionalist integration pressure (Bergmann and Niemann 2018). Importantly, 'external spillover' may trigger further integrative steps by 'affect[ing] the overall demand conditions for integration' (Haakanson 2021, 3;cf. also Nicoli 2020). Yet again, if external shocks and crisis produce a growing functional overlap between EU policiesas it has been the case with the refugee crisiswe cannot understand the implications of external spillover without also considering interaction dynamics between these policies. Importantly, external spillover pressures also present a window of opportunity for supranational actors, and their sector-specific bureaucracies, to expand their own tasks, influence, capabilities and resources in a particular domain of the EU's external action (Müller 2016), which may not easily align with functional, problem-solving solutions.

Neofunctionalism and Inter-institutional Interactions: Cooperation or Conflict?
The proliferation of international, and particularly regional organisations, over the last few decades has led to significant rivalries and overlaps between organisations and the rise of international regime complexity (Alter and Raustiala 2018;Haftel and Hofmann 2019). This has inspired a growing scholarship on inter-organisational relations and the politics of institutional overlap (Biermann & Koops, 2017;Faude and Fuss 2020;Haftel and Lenz 2021;Hofmann 2011;.
In relation to the EU, the concept of institutional overlap has traditionally been applied to institutional overlap with other international organisations such as NATO or OSCE in terms of mandate, membership and resources (Hofmann 2011;Koschut 2018). However, core insights from the literature on institutional overlap can also inform our understanding of intra-EU dynamics. Whilst on paper the EU is governed by a single institutional framework, in practice the powers and responsibilities of different EU foreign policy actors are based on distinct policy-making methods in different domains of EU foreign policy (Keukeleire & Delreux, 2014). This fragmentation of the EU policy framework creates a situation in which interactions across different actors and institutions become a central feature of EU external policy-making (Furness and Gänzle 2017). The ambition to forge a more 'comprehensive' or 'integrated' external EU policy approach through a more coherent and holistic use of different policies and instruments has thus also involved efforts to enhance institutional cooperation and coordination among the various actors involved (Council of the EU 2014, 2018a.
The EU's unified yet fragmented institutional framework provides for inter-institutional interaction, which can involve cooperation as well as conflict among actors and institutions that have their own foreign policy preferences and bureaucratic selfinterests. Understanding the range of possible modes of interaction as a continuum, we differentiate between cooperation and competition as its opposite sides. These different forms of interaction also influence the way institutions in these domains evolve and the extent to which functional synergies can be realised between them. Cooperation pertains to a situation where institutions respond to spillover pressures through cooperative solutions across different policy fields, trying to establish a certain division of labour that demarcates competences, divides tasks, realises synergies and avoids duplications. As such, cooperation benefits solutions that advance the EU's collective problem-solving capacity in response to functional pressures. In a weaker form, cooperation may also entail efforts to strengthen coordination among different policy areas, even if a certain overlap in tasks, mandates, and resources persists. Competition, in turn, involves situations where supranational actors and sector-specific bureaucracies from different EU policy domains seek to exploit spillover pressures to advance their own institutional self-interests and preferences, even at the cost of duplication of mandates, tasks and resources and a lack of coordination. For instance, rather than relying on existing resources and instruments in a related policy domain, competition involves efforts by EU actors to build-up their own resources and instruments. Hence, competition facilitates a zero-sum view of the relationship between functionally overlapping policy domains, where the strengthening of integration in one EU policy domain is perceived to possibly weaken the other.
Cooperation and competition are ideal types depicting the dominant mode of interaction between institutions from different EU policy domains during a given period. In practice, however, interaction may often not be as clear-cut and can involve both, elements of cooperation and competition. We posit that two main sets of factors will influence the mode of inter-institutional interaction in situations of functional overlap: (i) the compatibility of institutional self-interests and organisational logics and cultures of the main actors and institutions involved, and (ii) the complementarity of their governance capabilities and resources. Hence, if actors do not feel threatened by other governance actors from an overlapping domain in their vital institutional self-interests and/ or organisational preferences, and are convinced that joining forces will enhance their collective governance effectiveness, we would expect them to opt for cooperative solutions across policy fields. Conversely, in situations where actors from different EU policy domains perceive each other as a threat to their own interests and of little use to advance their collective governance effectiveness, we would expect them to engage in competitive behaviour. Certainly, one may also find situations in-between, where actors either perceive institutional overlap as a threat to their own interests and preferences but also see the added value for collective governance effectiveness, or do not feel threatened by institutional overlap but do not perceive it as enhancing collective governance effectiveness. In those situations, we expect actors to engage in a certain degree of coordination efforts, without fully exploiting the potential synergies and functional advantages of close cooperation.
Having developed a conceptual framework to account for inter-institutional interactions in relation to functional spillover dynamics, the subsequent section applies our theoretical considerations to CSDP and AFSJ. Whilst inter-institutional interactions between CSDP and AFSJ actors pre-date the 2015 European refugee crisis, our main focus is on post-crisis integration dynamics that are also subject of this Special Issue (Niemann & Zaun 2023, this issue).

The internal-external security nexus and CSDP-AFSJ cooperation pre-2015
To respond to international challenges related to migration, the EU can draw upon two distinct institutional 'apparatuses': CSDP military and civilian missions, and AFSJ agencies, including Frontex, Europol, and Eurojust (Parkes 2017, 49). The existence of these two separate apparatuses mirrors the institutional divide between internal and external security competences that the Treaty on European Union in 1992 established, creating the CFSP for external security matters and AFSJ (then Justice and Home Affairs -JHA) for internal security issues. In both pillars, gradual institutionalisation processes led to a proliferation of Brussels-based committees, working groups and coordination fora since the 1990s, and an expansion of tasks and mandates (Shepherd 2015, 162).
In the realm of CSDP, the so-called Petersberg Tasks included in Article 17 of the Nice Treaty on European Union defined the scope of EU crisis management activities. They comprise humanitarian or rescue operations, peacekeeping operations, and combat missions for crisis management, including peacemaking operations. With the adoption of the Lisbon Treaty, the task catalogue was further expanded to include joint disarmament operations, post-conflict stabilisation as well as the fight against terrorism and the support of third countries in combatting terrorism within their territories (TEU, Art. 43). However, the Lisbon Treaty did not explicitly include tasks related to border management as part of CSDP operations and missions' mandate. Within AFSJ, it was the 1999 Tampere European Council Conclusions that first called for a stronger external profile of AFSJ activities and 'clear priorities, policy objectives and measures for the Union's external action in Justice and Home Affairs' to be defined (European Council 1999, para 61). In the course of the 2000s, AFSJ agencies gradually expanded their activities beyond the EU's internal borders and 'went global', including Frontex capacitybuilding activities in the Western Balkans, Eastern Europe and Northern Africa (Trauner 2016, 1-2;Parkes 2017, 49-50).
In a nutshell, the AFSJ and CSDP can be understood as 'distinct policy areas acting under different legal regimes, governed by different stakeholders (military, diplomatic, or home affairs) and implemented by different EU entities' (Huisman 2021, 184). At the same time, awareness of the increasing need to strengthen ties between CSDP and AFSJ has been growing among policy-makers in Brussels and European capitals since the 2000s. In the EU's strategic discourseincluding in documents such as the 2003 European Security Strategy, which noted that 'internal and external aspects of security are indissolubly linked' (European Union 2003, 29), and the 2010 Internal Security Strategy (ISS)an understanding emerged of the interlinkages between internal and external security challenges, coupled with calls for enhanced linkages between the EU's different institutions responsible for addressing them (Shepherd 2015, 158-162).
However, it was not before 2010 that the discourse on the 'internal-external security nexus' was met by concrete action to strengthen institutional ties between CSDP and the AFSJ domain. This development was also related to the creation of the European External Action Service (EEAS) by the Treaty of Lisbon, which is specifically tasked to improve cooperation and coherence both between member states and EU institutions as well as across various EU external policies and actors (Furness and Gänzle 2017). In 2011, the EEAS' Crisis Management and Planning Directorate (CMPD) drafted a roadmap for strengthening ties between CSDP and AFSJ that the Political and Security Committee (PSC) and the Committee on Operational Cooperation on Internal Security (COSI) then jointly revised and adopted (Council of the European Union 2011). The roadmap identified 12 priority lines of action, aiming to establish new cooperation frameworks and linkages between the EEAS' CSDP bodies and AFSJ agencies, both at the strategic and operational level. Interestingly, in terms of thematic focus the 2011 roadmap referred to 'piracy, terrorism, organised crime, financial transactions' and did not include migration management, although it was already a relevant issue with regard to the geographical areas of interest mentioned by the roadmap, 'i.e. Western Balkans, Horn of Africa, Southern Neighbourhood, Sahel' (Council of the EU 2011, 4).
The implementation of the roadmap led to a working arrangement between the EEAS and Frontex established by an exchange of letters between the Deputy Secretary-General of the EEAS and the Executive Director of Frontex in 2014 (Council of the EU 2015a, Annex II). Inter alia, the arrangement put in place specific procedures for the cooperation of Frontex and CSDP missions on the ground, for exchanging of information between the two instruments, and for technical assistance and institutional support. Another result of the roadmap was the initiation of regular joint meetings between PSC and COSI as well as between DG Home, the EEAS, PSC and COSI (Shepherd 2015, 164). Taken together, the quest for strengthening the internal-external security nexus and for enhancing the coherence between CSDP and AFSJ has been a long-standing one. However, earlier efforts notwithstanding, the notion of CSDP as an instrument of EU external migration policy, and the resulting demand for much closer coordination and coherence with AFSJ actors, only gained real momentum in the context of the 2015 European 'refugee crisis'.

Institutional overlap between CSDP and AFSJ and the 2015 Migration Crisis
The substantive increase of refugees coming to Europe in 2014 and early 2015 laid bare the dysfunctionalities of the EU's asylum and migration policy to that date (Schimmelfennig 2015, 976-977). Apart from exposing the weaknesses of the Schengen system and the Dublin principle, it created strong pressures to upgrade the EU's external border management and to strengthen the competencies and capacities of Frontex (Niemann and Speyer 2018, 30). Against this backdrop, EU institutions and member states have made growing efforts to strengthen the EU's external migration policy (see Tittel-Mosser 2023;Nantermoz & Zaun 2023;Cardwell & Dickson 2023, all this issue). An extraordinary European Council convening in April 2015 agreed on a number of urgent response measures, which included 'strengthening our presence at sea', 'fighting traffickers in accordance with international law', 'preventing illegal migration flows', and 'reinforcing internal solidarity and responsibility (European Council 2015)'. The 2015 EUCO statement constituted the starting ground for further enhancing the EU's external migration governance capacities, including through the expansion of CSDP tasks and the strengthening of Frontex in the AJFS domain. However, as our analysis will show, functional pressures emanating from the 2015 refugee crisis were simultaneously addressed by different actors within both domains of CSDP and AJFSgiving rise to inter-institutional dynamics that shaped further integration steps in significant ways.

Csdp missions and AJFS-Agencies: expanding mandates and tasks in external migration governance
In the CSDP realm, the EU could build on established and tested operational capabilities for immediately dealing with challenges related to external migration, which had previously not been part of its mandate. The EUCO statement mentioned 'current CSDP operations' as a tool on which increased efforts to support countries such as Egypt, Tunisia, Mali, Niger or Sudan in monitoring and controlling their land borders could be built upon (European Council 2015, 1). While the EUCO statement only vaguely circumscribed the potential role of CSDP operations in the EU's response to the refugee crisis, the Agenda for Migrationa Commission Communication presented on 13 May 2015had a much stronger wording on the role of CSDP in migration management. It specified that migration would become a specific component of ongoing CSDP missions already deployed in countries like Niger and Mali (European Commission, 2015, 5). Two days after the launch of the Agenda for Migration, the EU's Foreign Affairs Council expanded the mandate of EUCAP Sahel Niger to the prevention of irregular immigration and the combat of 'associated crimes' (Council of the European Union 2015b, 1). It was soon followed by similar mandate amendments of EUBAM Libya in 2016 and 2018 (ECRE 2019, 2). These incremental steps were subsequently institutionalised through the 2018 Civilian CSDP Compact (CCC) (Council of the EU 2018b). The CCC presented an ambitious and far-reaching reform agenda, laying out a number of strategic guidelines, commitments and proposals for action that should lead to a more capable, effective and joined-up civilian CSDP (Böttcher and Wolf 2019). As for the scope of civilian CSDP missions, the CCC further codified the widening of missions' mandatessuch as in the case of EUCAP Sahel Niger 2015 or EUBAM Libya 2016/ 2018. It stated that CSDP civilian missions should also 'contribute to the EU's wider response to tackle security challenges, including those linked to irregular migration' (Council of the EU 2018b, 4).
EU member states pushing for the broadening of CSDP missions' mandate were a central driver behind the closer alignment of migration policy objectives and civilian CSDP (interviews 3, 7; Parkes 2016, 2). Additionally, the importance of demonstrating the EU's ability to respond to the crisis in a timely manner significantly contributed to member states' decision to draw on existing instruments such as civilian CSDP missions. As neofunctionalists have argued, decisions in political arenas such as the Council are often taken under close deadlines and motivated by a sense of urgency (Haas 1970, 627). At the same time, supranational entrepreneurship also contributed significantly to the integrative steps (Bergmann and Müller 2021). Since the role and competencies of the Commission and the European Parliament in CSDP are limited, supranational entrepreneurship was primarily exerted by the High Representative of the Union for Foreign and Security Policy, who is also Vice-President of the Commission, as well as by the EEAS. Importantly, then High Representative Federica Mogherini and the EEAS played an instrumental role in moving the CCC initiative forward and provided its strategic direction (interviews 1, 2, 5). The Concept Paper on Strengthening Civilian CSDP that she had presented in May 2018 served as the main strategic basis for the discussions on the Concept among member states, with the result that many proposals of the Concept Paper also feature in the final version of the Compact (European Union 2018b, 4-5; see also European Parliament, 2019, 7).
Unlike it has been the case with CSDP, the AJFS actors could not draw on established operations and expertise in acting in third countries outside the EU's external borders. Following the 2015 refugee crisis, the expansion of the mandate and capabilities of Frontex has been a constant 'focus for reform' by EU institutions and member states (Parkes 2017, 49). Concerning the reforms and mandate expansion of Frontex, the Commission played a key role in driving the reform efforts forward, responding swiftly to functional pressures by presenting ambitious reform proposals (interview 4; Tammikko 2019, 5). Moreover, the Commission in concertation with the Council Presidency acted as a promotional broker between the Council and the EP in trilogue negotiations (Niemann and Speyer 2018, 32-33). In December 2015, the European Commission presented its proposal for a regulation establishing the European Border and Coast Guard (EBCG), which was adopted by the Council and the European Parliament as Regulation 2016/1624 in September 2016. The Regulation transformed the 'European Agency for the Management of Operation Cooperation at the External Borders of the Member States and the European Union' into the 'European Border and Cost Guard (EBCG)' agency, while keeping 'Frontex' as its short name. Most importantly, the Regulation significantly extended Frontex' mandate (Lavenex, Križic, and Veuthey 2021, 437-438). Among other things, it enabled Frontex to conduct regular vulnerability assessments of member states' capabilities in border management and to suggest binding measures to remedy member states' deficiencies. Concerning its international role, the Regulation extended Frontex' mandate to conduct joint operations with and within third states, granting it an important role in the management of returns of illegally staying third country nationals. In terms of resources, the decision to create a standing force of 1,500 border management experts to be deployable within three days meant a significant boost of Frontex' operational capabilities.
However, two years after the reform, many measures had still not been implemented, such as the creation of a 1,500 European reserve of border guards (Bossong 2019, 2). Hence, the Commission proposed another reform to boost Frontex' capabilities, which was adopted in November 2019 (Regulation (EU) 2019/1896 n.d.). The reform aims to further strengthen Frontex' ability to address border management-related challenges such as irregular migration, cross-border organised crime, and terrorism, both in the EU and abroad (Tammikko 2019, 3). Whilst Frontex had already been present in third countries in an advisory and technical support role, the amended EBCG regulation also allows for carrying out border management operations upon the invitation of third countries that are not neighbouring EU member states (Bossong 2019, 7). Moreover, the 2019 Regulation established the objective of creating a standing corps of 10,000 staff, with their own equipment, to be progressively build up until 2027. This significant upgrade of Frontex' operational resources went considerably beyond previous objectives of the 2016 Regulation. At the same time, it raised the question of how Frontex' new operational capabilities would relate to already existing operational capabilities in the CSDP domain, whose mandate had been expanded to migration related tasks.

Institutional overlap and inter-institutional interaction: cooperation and institutional self-Interest
The 2015 'refugee crisis' clearly served as a catalyst for a parallel expansion of the mandate and capacity of CSDP and AFSJ in external migration governance. Actors in both policy domains have worked towards expanding their own role and capacity, resulting in a growing need to 'sort out this messy overlap' (Parkes 2020, 102). Given the interinstitutional interaction of CSDP and AFSJ institutions and actors since 2015, we find evidence for accelerated efforts to further coordinate and improve cooperation with each other. However, we also observe persisting reluctance to fully implement the agreed coordination plans and to move towards close, synergetic cooperation.
Several efforts have been undertaken since 2015 to further strengthen ties and coordination between CSDP and AFSJ actors. Building on the 2011 Roadmap, the EEAS in July 2016 issued a Food for Thought Paper entitled 'From Strengthening Ties between CSDP/ AFSJ actors towards more security in Europe' (EEAS 2016). It was later turned into an EEAS non-paper on enhancing Cooperation between CSDP Missions/Operations and AFSJ Agencies and formed the basis of a 'Mini-concept on possible civilian CSDP efforts to address security challenges linked to irregular migration' that the Council adopted in 2020 (Council of the EU, 2020). The latter document's name demonstrates that the member states were still in the process of carving out civilian CSDP's contribution to migration management efforts. Nevertheless, at the administrative level several working arrangements and letter exchanges between the EEAS and Frontex, as well as between Frontex and civilian CSDP missions such as EUBAM Libya, have further formalised and developed cross-policy, inter-institutional cooperation (EEAS, 2021, 20).
In terms of concrete measures to strengthen cooperation, one example is that Frontex staff members can now also be deployed to CSDP missions in the context of the 'visiting expert concept' such as in the case of EUBAM Libya, where they are deployed 'to support the mission on a needs-driven and rotational basis' (Huisman 2021, 182). In a similar fashion, a Frontex liaison officer in Niamey is closely cooperating with the EUCAP Sahel Niger mission deployed there (interview 4; Parkes 2017, 53). A second example concerning cooperation on the ground is the joint participation of civilian CSDP missions and Frontex in the implementation of border management assistance projects with third countries. In the case of EUBAM Libya, the mission together with Frontex and the Italian Ministry of the Interior launched a joint pilot project to support the Libyan General Administration for Coastal Security in increasing its capacity to conduct search and rescue operations (see also Cusumano & Riddervold 2023, this issue). This includes 'training sessions on human rights obligations as border officers and conduct SAR operations in-line with a rights-based and do-no-harm approach' (EEAS, 2021, 22). A third example relates to the area of CSDP training, where the Joint Action Plan implementing the CCC seeks to advance cooperation and coordination between CSDP and other actors. In particular, the EU Civilian Training Group, a configuration of CIVCOM composed of training experts, is charged with promoting an effective use of training courses provided by different actors, including the European Security and Defence College, national training institutes, and several AFSJ agencies, including Frontex (Council of the EU, 2019; interview 4).
However, despite these efforts to further coordinate CSDP and AFSJ structures and practices, member states and EU institutions still need to craft a more structured approach towards inter-institutional cooperation and agree on a clear division of labour (Huisman 2021, 184). The latter is also supported by strong functional rationales, such as making effective use of complementary capabilities, avoiding the duplication of structures and resources as well as tackling persisting institutional constraints to a joinedup approach to external migration governance. Yet, important challenges in the cooperation between them persist, indicating that inter-institutional relations have somewhat stayed 'halfway' on their move from competitive towards more cooperative behaviour. At the strategic level, the complicated relationship between the Foreign Affairs Council and the AFSJ Council, and the PSC and COSI as their preparatory bodies respectively, provides some clear evidence of this. Despite the agreement to hold joint meetings of the two Councils and of the two committees since the adoption of the roadmap in 2012, the process of agreeing on a joint agenda of cooperation has been challenging and the outcomes of these meetings have been very limited (interviews 4, 5, 6; Johansen 2017).
At the operational level, cooperation between civilian CSDP missions and AFSJ agencies suffers from shortcomings related to inter-institutional information sharing, enhanced consultations from the planning stage towards improved cooperation in the field, and the joint, efficient use of resources and logistics (Böttcher and Wolf 2019). The fact that the December 2020 Council Conclusions on the Civilian CSDP Compact call for 'foster[ing] closer mutually reinforcing cooperation and synergies between civilian CSDP, AFSJ actors (including relevant ministries, agencies and Council working parties), as well as with Commission services' underscores the functional need for more effective arrangements for institutional cooperation (Council, 2020, 5).
Our evidence further suggests that realising the full potential of cross-policy cooperation in response to functional pressures of the 2015 refugee crisis can raise concerns related to institutional self-interests that are deeply rooted in the EU's fragmented institutional set-up and diverging organisational logics and institutional cultures. Whilst AFSJ actors welcome the fact that CSDP resources will now also be directed to support external migration objectives, they also understood the 2015 refugee crisis as an opportunity to expand their own mandate and operational capacities in external migration governance (interview 4, 6). Relating to the conditions of effective cross-policy cooperation, there does not seem to be a strong concern among AFSJ actors that the task expansion of civilian CSDP may constitute a threat to their institutional interests, as the strengthening of Frontex' capacities is already underway and is a strong priority of many member states. At the same time, their motivation for further increasing the complementarity of CSDP and AFSJ governance capacities and resources has its limits, as the build-up of Frontex' standing corps will mean that AFSJ operational capacities will by far exceed the capacities of civilian CSDP missions (interviews 4, 6, 7). Hence, Frontex and DG Home rather argue that the build-up of staff capacities in the context of the Frontex' standing corps actually supports member states' quest for increasing the pool of civilian crisis experts, which will also benefit their recruitment capacities for civilian CSDP missions in the future (interview 7).
Similarly, CSDP actors see merit in expanding the scope of their mandate to migration related tasks and enhanced cooperation with Frontex. In fact, the political impetus for improving CSDP-AFSJ cooperation seems to stem primarily from foreign ministries, member state representatives in preparatory bodies such as the PSC and CIVCOM as well as from the EEAS (interviews 5, 6). Simultaneously, the strive for enhanced cooperation is motivated by CSDP actors' concerns that a strengthening of AFSJ's operational capabilities will impact on the future role of CSDP, specifically its civilian dimension where functional overlap is particularly substantive. As one interviewee stated, member state representatives within the CSDP domain perceive the strengthening of Frontex 'as a risk because what Frontex can do according to their new [2019] regulation is very close to what a CSDP mission can do' (interview 7). These concerns, in turn, also contribute to the persistence of institutional self-interests on the CSDP side.
Moreover, member state representatives in the two policy domains conceive the task of external migration governance quite differently. Ministers of the Interior and COSI representatives tend to view external migration governance through an internal security lens, understanding it predominantly as a way to preserve and enhance the EU's internal security. Their main focus is thus on how EU external migration governance serves the EU's own political interests, primarily in terms of security interests. Here, irregular migration is primarily perceived as a potential threat to the Schengen framework and associated norms. In contrast, foreign ministers and PSC representatives tend to perceive EU external migration governance as a contribution to international security and stability, viewing it as an aspect of foreign policy rather than an externalisation of EU domestic policies. Although the EU's own political interests certainly also play a role, the main focus is on how foreign policy measures such as CSDP missions serve the interests of the host states of those missions (interviews 4, 5, 6, 7, 8). These diverging institutional interests and logics have also led to a situation in which COSI has facilitated closer cooperation between the AFSJ agencies such as Frontex with the Commission to gain better access to EU budgetary resources 'in ways that CSDP deployments cannot' (Parkes 2020, 101-102).
In addition, as the EBCG standing corps and the civilian CSDP missions draw their personnel from the same staff poolnational civil servants and police officersforeign and interior ministers compete over the same resources (interviews 2, 3, 4, 5). Civilian CSDP has always suffered from shortages of resources and capabilities, primarily due to difficulties in recruitment and deployment of qualified personnel (Juncos 2020, 83). Although the implementation of the civilian CSDP Compact is an attempt to remedy this challenge, EU member states were not willing to agree on binding targets for individual contributions to civilian CSDP missions (Smit, 2019, 3), as they did in the AFSJ domain. Simultaneously, recruitment of personnel is less a challenge for Frontex operations. Financially, it is more attractive for interior and justice ministries to allocate staff to Frontex operations than to civilian CSDP missions (interviews 7, 8). It is also easier for ministries of the interior to recruit police officers and judicial advisors for Frontex missions, who usually work under their own authority (interview 8). The parallel build-up of the EBCG standing corps pushed forward by member states' ministers of the interior and the European Commission has made it even more difficult to reverse the trend of declining staff contributions to civilian CSDP missions (interview 5, 8). Finally, PSC and COSI are characterised by different institutional cultures, which further complicates their joint interactions (interview 4). While COSI representatives are Ministry of Interior officials based in capitals that fly in to Brussels for their committee meetings, member states' PSC representatives are based in Brussels, meet on a weekly basis and have established a certain 'esprit de corps' among them (Johansen 2017;Michalski and Danielson 2020).
At the operational level, the fragmentation of financial resources relevant for EU external migration governance across different policy domains and funding lines further complicates cooperation between civilian CSDP missions and Frontex in the field. Moreover, exchanging intelligence and information between civilian CSDP missions and Frontex faces substantive legal and technical obstacles. Whilst AFSJ agencies such as Frontex have a general need for actionable information and situational awareness on developing events in the field, CSDP military operations and civilian missions face legal and technical hurdles to share such data (Huisman 2021, 183). At the same time, CSDP operations and missions face limited access to knowledge and information produced by non-CFSP/CSDP instruments such as Frontex (Arnaud et al. 2017, 59-60). The underlying reason for these challenges of information exchange is the lack of a legal basis for the exchange of operational data. The 2019 Regulation on the European Border and Cost Guard only provides the basis for its cooperation with CSDP missions and operations on 'the promotion of European integrated border management standards' and 'situational awareness and risk analysis' (Regulation (EU) 2019/1896 n.d., Article 68). This means that the agency and civilian CSDP missions can only exchange aggregated data in the context of situational awareness and risk analysis, but not operational data such as information on individual persons (interview 4). According to one interviewee, the legal restrictions on how Frontex and CSDP missions can cooperate which were pushed through by the European Parliament in the negotiations on the 2019 EBCG regulationprovides a strong barrier for enhanced practical cooperation at the operational level (interview 7). Hence, the abovementioned limits have prevented CSDP and AFSJ actors from utilising each other's governance resources in a complimentary way, including funding and relevant operational information.
In sum, the analysis has shown that neofunctionalist spillover pressures for further integration have impacted on cross-policy, inter-institutional dynamics in the evolution of the EU's external migration policy. The integration pressures on CSDP and AFSJ have individually resulted in further integrative steps, and created an institutional overlap with regard to CSDP and AFSJ actors' competencies in external migration governance. As a result of this overlap, the evidence points to accelerated efforts to improve cross-policy cooperation since 2015. At the same time, this has played out primarily in terms of further institutionalising coordination formats, but it has not led to close, synergetic cooperation that serves to realise mutual benefits. Our results thus indicate that despite further integrative steps within the individual policy domains, the cross-policy integration of CSDP and AFSJ is still very much work in progress, with relatively limited results so far. A limited compatibility of institutional interests and policy preferences coupled with a limited degree of complementarity of governance resources such as funding and information have prevented CSDP and AFSJ actors to better exploit the potential for enhanced cooperation. Moreover, the findings suggest that the different institutional and legal set-ups in the CSDP and AFSJ domain constitute a substantive barrier for close cooperation at the operational level. In other words, the legacy of the 'pillar structure' of the European Unioncreated by the Treaty of Maastricht and abolished by the Lisbon Treatystill impacts on inter-institutional cooperation across EU external policy domains in addressing functionally indivisible, policy challenges (interview 4, 6).

Conclusions
Analysing patterns of inter-institutional interaction between CSDP and AFSJ since 2015, the article has gathered important evidence for efforts to address mutual interdependence and overlap in mandates and tasks through upgrading cooperation and coordination. Importantly, a number of new cooperation formats and coordination mechanisms were established. At the operational level, these include the deployment of Frontex visiting experts in civilian CSDP missions such as EUBAM Libya or the cooperation between Frontex liaison officers and civilian missions such as in the case of EUCAP Sahel Niger. Moreover, CSDP bodies and Frontex now cooperate more closely on the training of civilian experts and on capacity-building missions in third countries. At the strategic level, the adoption of the Civilian CSDP Compact has led to renewed efforts to increase coordination between CSDP and AFSJ bodies such as PSC and COSI or CIVCOM and the COSI support group.
However, actors in both domains have also sought to seise the opportunity created by the 2015 refugee crisis to expand their own role and tasks in the governance of external migration, with the integration process becoming subject to their respective institutional interests, organisational logics and culture. These parallel efforts of CSDP and AFSJ actors to advance their own roles, tasks and functions have also involved elements of rivalry and competition and have complicated functional solutions to establish a clear division of labour that demarcates competences, divides tasks, realises synergies and avoids duplications.
In theoretical terms, this article shows that in situations of institutional overlap, neofunctionalist spillover dynamics can give rise to inter-institutional interactions that shape integration outcomes in significant ways. Whilst cooperative inter-institutional interactions promote functionally efficient integration outcomes, competitive behaviour is likely to lead to duplications and functionally inefficient overlaps. Although we agree with previous neofunctionalist accounts that point to functionalist spillover dynamics and supranational entrepreneurship in accounting for the post-2015 integration patterns in CSDP and AFSJ (Bergmann 2019;Haakanson 2021;Niemann and Speyer 2018), this article also highlights the importance of transcending the dominant 'silo' perspective that focuses on explaining integration outcomes in individual EU policy domains. On the one hand, this isolated view neglects the development of horizontal mechanisms for cooperation as an important element of integration in situation of institutional overlap. This horizontal perspective on integration is particularly relevant when dealing with the EU's international relations, which involve several functionally related issues such as security, development, or trade. On the other hand, it fails to appreciate the fact that neofunctionalist spillover dynamics may also become subject to institutional self-interests and different organisational logics and cultures that prevail in the EU's fragmented institutional framework. As our analysis has shown, parallel efforts by CSDP and AFSJ actors to advance their mandates and tasks do not easily align with a functionalist, problem-oriented logic of integration that is frequently evoked by traditional neofunctionalist accounts. Moreover, our results indicate that efforts to integrate self-standing policy domains such as CSDP and AFSJ constitute a long-term task that requires to look beyond single decision-making events. Insights from research on institutional overlap and bureaucratic politics can thus contribute to a better understanding of how neofunctionalist integration pressures are negotiated between different actors and institutions across policy domains. In the case of CSDP and AFSJ, this process is still ongoing and current reformslike the built-up of Frontex operational capabilitieswill unfold their full effect on the respective EU policy domains only in the years to come.