Towards an integrated approach to EU foreign policy? Horizontal spillover across the humanitarian–development and the security–migration interfaces

ABSTRACT

Combining theoretical insights from works on policy integration and institutional overlap, this article develops a novel analytical framework to account for variations in policy integration in the EU’s external relations. We argue that the greater the level of institutional fit between different areas of the EU’s external relations, the greater the prospects for deepening policy integration through a ‘horizontal spillover’ mechanism. Specifically, we argue that network formation, policy entrepreneurship and political socialisation are key elements of this horizontal spillover mechanism leading towards greater policy integration. Empirically, we explore the relevance of our theoretical claims for the interface between the EU’s development policy and its humanitarian aid policy and between the CSDP and external migration policy in the AFSJ domain. Whilst both policy interfaces have been marked by growing functional pressures, we show that differences in institutional fit have resulted in varying degrees of progress in policy integration.

KEYWORDS:

• EU foreign policy
• policy integration
• migration
• development
• humanitarian aid

1. Introduction

Fostering the coordination and integration of different external policy domains has been a central ambition of the European Union (EU) since the Maastricht Treaty, as highlighted by the establishment of the ‘comprehensive approach’ in 2011 and the ‘integrated approach’ that was reinforced through the EU’s Global Strategy (European Union, 2016, pp. 28–32). Yet, in practice the EU has often struggled to implement more integrated foreign policy approaches, and the degree of horizontal integration has varied across different policy domains (Furness & Gänzle, 2017; Juncos, 2015; Shepherd, 2022). This raises the question of how we can explain varying degrees of policy integration despite the EU’s continued efforts to promote it.

Combining insights from research on policy integration and works on institutional overlap, we propose and unpack the causal mechanism of ‘horizontal spillover’ that links degrees of institutional fit to outcomes of policy integration. ‘Horizontal spillover’ concerns functional pressures facilitating steps towards greater policy integration between different EU policy domains. Here, relevant EU actors thus need to step out of the sectoral logic of vertical European integration and focus their efforts on steps towards policy integration across different EU policy domains. Specifically, we argue that network formation, policy entrepreneurship and political socialisation are key elements of the horizontal spillover mechanism leading towards greater policy integration. At the same time, we agree with neo-functionalist scholars that functional logics do not govern behaviour in a mechanical way, but are only as strong as they are perceived to be by (relevant/important) actors (Niemann & Ioannou, 2015, p. 198). Similarly, the initial causal condition for horizontal spillover processes to unfold is characterised by situations where relevant/important actors not only perceive functional pressures as plausible and compelling, but also consider the upgrading of policy integration as central for effective problem-solving. In particular, we argue that this will be the case when the institutional fit between the concerned EU policy domains is considered to be high.

We explore the relevance of our theoretical claims for two prominent policy interfaces of the EU’s external policy: (I) the interface between the EU’s development policy and its humanitarian aid policy, and (II) the interface of the Common Security and Defence Policy (CSDP) and external migration policy in the realm of the Area of Freedom, Security and Justice (AFSJ). Our case studies combine the congruence method with process tracing. The congruence method rests upon the assumption that a causal relationship between two variables can be established based on the consistency between a theory’s prediction about the values of the independent and dependent variables and the outcome of a case (George & Bennett, 2005, pp. 181–204). Researchers have to assess whether the values of the variables in each case are consistent with the theoretical expectations about them.

Given the novelty of our theoretical claims about the causal relationship between institutional fit and policy integration, probing its initial plausibility by applying the congruence method is a necessary first methodological step. As consistency alone is not sufficient to make causal inferences, however, we combine it with a ‘theory-building, process-tracing’ approach (Beach & Pedersen, 2019, pp. 16–18). The latter constitutes a common approach to probe the plausibility of novel theoretical arguments by searching for a theorised causal mechanism that may also be generalisable outside of the examined individual case(s). To make evidence-based claims about the plausibility of the novel causal mechanism of ‘horizontal spillover’, we unpack the mechanism in its component parts – i.e., network formation, policy entrepreneurship and socialisation – focussing particularly on the actors and activities transmitting causal forces from cause to outcome (Beach, 2016, p. 464).

Being subject to growing crises-related pressures, horizontal interdependencies – in terms of governance resources and synergies – have moved into the focus of both the humanitarian – development and the security – migration nexuses. At the same time, the two nexuses also entail important differences in terms of the institutional fit of the policy areas involved, hence alternating on the independent variable, providing the basis of our congruence analysis. With respect to our subsequent process-tracing analysis, the humanitarian – development nexus, which is marked by a high degree of institutional fit, can be considered a ‘typical case’ (Gerring, 2007, pp. 91–93) for the horizontal spillover mechanism to unfold, whereas the low degree of institutional fit at the security – migration nexus makes it more prone to sectoral integration dynamics. Whilst both case studies include an assessment of policy integration in earlier time periods, the time period on which the empirical investigation focusses spans from 2015 to 2022. We selected this specific time period to investigate whether the ‘refugee crisis’ in 2015 that produced strong functional pressures for vertical integration also resulted in horizontal spillover pressures (Niemann & Speyer, 2018).

Our empirical analysis builds on twelve semi-structured interviews with EU officials from Directorate-Generals for migration and home affairs (DG Home), for humanitarian aid (DG ECHO), for international partnerships (DG INTPA), from the European External Action Service (EEAS), member state representatives, as well as on a series of official EU documents and the relevant scholarly literature. The article proceeds as follows. We first introduce a novel analytical framework, proposing our theoretical explanation of policy integration. Our empirical analysis probes the validity of our conceptual arguments based on two case studies focussing on the interface between the EU’s humanitarian aid and development policy and the interface between the CSDP and the AFSJ in external migration governance. Our main findings are presented in the conclusion.

2. Analytical framework: The inter-institutional politics of policy integration

Policy integration concerns ‘the management of cross-cutting issues in policy-making that transcend the boundaries of established policy fields and which do not correspond to the institutional responsibilities of individual departments’; and it aspires to the ‘horizontal sectoral integration’ between ‘different departments and/or professions in public authorities’ (Meijers & Stead, 2004, p. 2). As such, the demand for policy integration is not specific to the EU, but instead concerns most contemporary, functionally differentiated governance systems that organise their governance activities into sectoral ministries and decentralised agencies (Jordan & Lenschow, 2010, p. 150).

There is by now a considerable body of literature on policy integration within states and a number of different international organisations (for an overview, see Tosun & Lang, 2017). In the context of the EU’s foreign policy, however, only a few works have applied a specific focus on policy integration, mostly in relation to the external dimension of environmental and climate-related issues (e.g., Dupont, 2016; Kaplaner et al., 2023). In part, this situation may be explained by the compartmentalisation of scholarship, in which distinct research communities often study the integration and functioning of different EU policy areas in isolation from each other. At the same time, however, the demand for policy integration in the EU foreign policy domain can be understood to be particularly strong. The EU’s foreign policy system is marked by a high degree of institutional fragmentation and differentiated integration (Amadio Viceré & Sus, 2023; Rieker, 2021), whilst it simultaneously faces a multitude of complex external challenges that often require an integrated foreign policy response.

Whilst the Lisbon Treaty formally abolished the pillar structure introduced by the Maastricht Treaty, different EU policies continue to be governed by different procedures, mandates and legal frameworks. Against this backdrop, research has explored multiple policy interfaces, such as the ‘security – development nexus’, the ‘security – trade’ nexus, the ‘security – migration nexus’ as well as the ‘climate – energy – trade nexus’ (Furness & Gänzle, 2017; Karyotis, 2007; Leal-Arcas & Alvarez Armas, 2018). Often, horizontal aspects of EU foreign policy have been addressed through the conceptual lens of ‘coherence’ and ‘consistency’ (e.g., Juncos, 2015), which overlaps partly with the perspective of policy integration (Hogl & Nordbeck, 2012). What have received less attention, in turn, are the EU’s efforts to forge a horizontally more integrated approach to its external relations (EU, 2016, pp. 28–32) through policy integration, which has faced new urgency in light of recent crises.

2.1. Assessing horizontal policy integration: Communicative, organisational and procedural dimensions

To assess the degree of policy integration, we can rely on works that have distinguished three main institutional instruments through which policy integration in the EU unfolds – they include communicative, organisational and procedural instruments (Jordan & Lenschow, 2008; Persson, 2007). Communicative instruments range from comprehensive, long-term strategies, such as the EU’s Global Strategy, to policy documents on very specific aspects of EU external action, such as the ‘Mini-concept on possible civilian CSDP efforts to address security challenges linked to irregular migration’ (Council of the EU, 2020). Those communicative instruments provide strategic awareness, joint narratives, long-term visions and objectives, and guidance for operational implementation steps (Hogl & Nordbeck, 2012, p. 119; Jordan & Lenschow, 2008). Here, the degree of policy integration can be assessed by looking at how key actors within the respective policy domains frame their strategic priorities, the central challenges they face and appropriate policy responses in their strategic communication (Carrapico & Barrinha, 2017). Progress towards a higher degree of policy integration involves an increasing approximation and specification of core common strategic priorities, shared definitions and understandings of major challenges, and the identification of cross-sectoral policy responses to address these challenges in central communicative instruments. A low degree of policy integration at the communicative level, in turn, only allows for relatively few and/or rather general and vague common objectives, strategic priorities and problem definitions that span institutional boundaries, whilst important differences persist.

Organisational instruments involve formal and informal institutional mechanisms for policy coordination, through which actors exchange information, consult and develop policy initiatives, deliberate on policy choices and assess policy outcomes. In the case of a high degree of policy integration, this involves the creation of inclusive, formal organisational mechanisms. This may include the establishment of a central authority that is entrusted with joint analysis, supervision, coordination and implementation of tasks related to the policy integration process. The Commission’s DG NEAR – European Neighbourhood Policy and Enlargement Negotiations – would be an example of a centralised organisational structure that is tasked with coordinating common, overarching policy objectives of the European Neighbourhood Policy, such as security, economic prosperity and democracy, which cut across various functional areas of different DGs (Noutcheva, 2015). Still, even in the presence of centralised organisational structures, this is often paralleled by coordination mechanisms that bring actors from the involved policy areas together – such as interdepartmental working groups, tasks forces and liaison officers. By creating ‘new arenas and open[ing] previously compartmentalized decision-making networks’ (Hogl & Nordbeck, 2012, p. 119), the policy integration of organisational instruments may alter the decision-making context and provide for more coordinated policy responses. In a weaker institutionalised form, policy integration may involve the evolution of organisational mechanisms that are more ad hoc and/or informal, including the informal networks of relevant actors for consultation and the exchange of information. There will be fewer ‘hard’ cross-cutting organisational mechanisms, in turn, and they will remain weak and ineffective in terms of altering the decision-making context.

Procedural instruments, by contrast, involve changes to the procedures for administrative standard or the introduction of new ways of working together and taking joint decisions and initiatives across pre-existing boundaries (Hogl & Nordbeck, 2012, p. 119). Here, progress in policy integration is associated with the streamlining of internal procedures and administrative standards to facilitate information exchange and policymaking across different sectors and institutional domains (see Table 1).

Table 1. Dimensions and degrees of policy integration and their indicators.

Policy integration	Degree of policy integration and indicators
Communicative instruments	Low: – Persistence of distinct narratives and strategic visions – Only general/vague objectives, strategic priorities and problem definitions that span institutional boundaries – Emphasis on policy domain-specific rather than integrated policy responses/ implementation strategies High: – Increasingly integrated narrative and strategic vision – Ever more precise shared objectives, strategic objectives and problem definitions that span institutional boundaries – Greater emphasis on integrated policy responses/ implementation strategies
Organisational instruments	Low: – Mainly weakly institutionalised, informal networks and ad hoc mechanisms for horizontal coordination (e.g., for exchanges of information on initiatives, activities) – No, or only few, weak and ineffectual formal cross-cutting organisational mechanisms for coordinated policymaking High: – The existence of a ‘central authority’ entrusted with joint analysis, supervision, coordination and implementation of tasks related to the horizontal integration process between EU institutions/ member states – Broad range of formal cross-cutting organisational mechanisms – including interdepartmental working groups, tasks forces, liaison officers
Procedural instruments	Low: – Few or no efforts for the streamlining of administrative standards and procedures for joint policy assessment and joint policymaking High: – Streamlining of a broad range of administrative standards and procedures for joint policy assessment and joint policymaking

2.2. Explaining varying degrees of policy integration – functional pressures and horizontal spillover

To explain the degree of policy integration across different EU policies, we draw on insights from the literature on inter-institutional overlap (Hofmann, 2011; Hogl & Nordbeck, 2012; Koschut, 2018). The concept of institutional overlap maintains that overlapping international institutions do not evolve and operate in isolation from each other, but influence each other’s development and governance effectiveness (Dunoff, 2012). Drawing on this insight, we posit that the degree of institutional fit between the distinct areas of the EU’s international relations is a key condition for determining whether policy integration materialises (Drezner, 2000). In particular, we argue that a high degree of institutional fit can produce what we call ‘horizontal spillover’, which leads to policy integration between different policy domains (cross-sectoral integration). Conversely, a low degree of institutional fit will likely facilitate greater integration within a policy domain (sectoral integration), which is well-studied by classical integration theories, such as neofunctionalism and liberal intergovernmentalism. Here, policy integration across different EU policy domains will remain weak.

Institutional fit relates to the compatibility of core aspects of the organisational cultures prevailing in the institutions that underpin the respective EU policy domains as well as of their bureaucratic self-interests. Institutional culture involves ‘deeply embedded patterns of organizational behaviour and the shared values, assumptions, beliefs, or ideologies that members have about their organization or its work’ (Peterson & Spencer, 1990, p. 14). With respect to EU foreign policy institutions, organisational culture includes beliefs about an institution’s core political mission or vision, as well as key assumptions and substantive norms, beliefs and values held by its members related to the governance of a particular policy domain. Perceived incompatibilities between core aspects of institutional culture can pose severe hurdles for horizontal cooperation and boundary-spanning socialisation processes, as the relevant actors are deeply pre-socialised into sectoral, policy-specific communities with their own missions, values and beliefs. In turn, in situations where EU actors understand their respective missions – and thus their governance resources in terms of synergies and complementarity – we expect far-reaching cross-policy integration steps to be more likely than in situations of perceived incompatibilities of institutional cultures.

Besides a commitment to their core mission and other central aspects of their organisational culture, EU institutions and their sector-specific bureaucracies also have institutional self-interests (Smith, 2013). Accordingly, they assess the consequences of greater policy integration, and related organisational changes, in terms of its implications for the survival and growth of their own bureaucracy. Institutions marked by similar and/or compatible organisational cultures and self-interests are ‘expected to interact more easily and with better results’ (Knoben & Oerlemans, 2006, p. 76). Conversely, institutions dominated by actors that feel threatened by other governance actors from a different EU policy domain in vital interests and properties of their organisational culture (i.e., integrity of their political mission and core beliefs and values) are less likely to support far-reaching steps towards policy integration. In a nutshell, the degree of institutional fit influences whether institutional overlap between different policy domains results in policy integration.

To explain the causal relationship between institutional fit and policy integration, it is important to further elaborate the causal mechanism of horizontal spillover, unpacking it in its component parts that involve ‘network formation’, ‘policy entrepreneurship’ by supranational actors and ‘political socialisation’. In situations marked by a high degree of institutional fit, functional pressures imply that policymakers come to realise that problems of substantive interest can no longer be addressed effectively within a particular policy area but require a more integrated approach (Bergmann, 2019; Niemann & Ioannou, 2015). Functional pressures may result from significant external developments and crises, after which issues that were previously regarded as separate and parallel are increasingly perceived as functionally interconnected. They may also result from more incremental, long-terms trends, such as the parallel growth of different EU policies (Allwood, 2015, p. 10), or from the prioritisation of cross-cutting issues, such as gender policy or sustainable development, that the EU seeks to mainstream across different EU policy domains (Geyer & Lightfoot, 2010).

When functional pressures for policy integration emerge, a high degree of institutional fit facilitates a process in which actors from individual policy sectors reach out to actors from other policy domains to form new horizontal links or networks (Adelle et al., 2015). At this point of horizontal spillover, actors behaving as entrepreneurs for greater policy integration may be motivated by collective objectives, such as the need to build more effective and coherent EU policies, or they may seek to expand their own resources and powers. What is important here is to acknowledge that functional pressures or normative ambitions towards greater horizontal integration rely on agency to produce concrete results, with actors articulating concrete objectives and working towards their implementation (Peters, 2015, p. 78).

In a second causal step, top-level actors – who may seize those newly established horizontal links – initiate and craft political proposals that seek to establish and institutionalise formal organisational mechanisms for horizontal cooperation. In some instances, top-level actors – such as the President of the European Commission, the EU’s High Representative for Foreign Affairs and Security Policy or the European Council – may formulate political objectives towards greater policy integration (Amadio Viceré, 2016), which are then implemented at lower levels across different policy domains. Such top-level actors may rely on hierarchical control, their cross-cutting authority and steering from the top to create incentives towards greater policy integration.

Once certain mechanisms for cooperation, coordination and information exchange between different EU policy areas have been established, we may witness processes of cross-institutional socialisation as a third causal step leading towards greater policy integration. Cross-cutting organisational mechanisms for policy coordination – including intergovernmental working groups and task forces – often resemble policy networks that provide for professional exchange and social interaction among peers (Slaughter & Zaring, 2006). Socialisation refers to a process ‘through which agent properties and preferences change as a result of interaction’ (Checkel & Moravcsik, 2001), as they pick up new knowledge and ideas about policy (Adolph, 2013, p. 15) and become accustomed to common procedures, rules and norms. This suggests that policy integration can trigger a self-reinforcing dynamic, as the actors that become involved in formal cross-cutting organisational mechanisms become familiar with the problem-understandings, norms and values that prevail in other policy domains. It is important to highlight that socialisation does not necessarily mean a strong similarity, or even a progressive fusion of distinct organisational cultures, but it may also lead to a situation in which key actors in one policy area perceive core organisational values and practices in another EU policy domain as compatible, legitimate and effective to achieve common governance objectives (Weber et al., 1996, p. 1216).

In sum, we argue that institutional fit can advance a horizontal spillover mechanism that leads to greater policy integration. To test our conceptual argument for explaining policy integration, we first assess whether the degrees of institutional fit and policy integration in the two cases are consistent with our theoretical expectations (congruence method). Second, we explore whether the causal mechanism of horizontal spillover is present and can be traced empirically in the case studies (process-tracing approach).

3. Policy integration at the humanitarian – development interface

At the EU level, humanitarian aid and development policy have evolved as separate domains, with each having its individual policy objectives, legal bases, institutional responsibilities and funding instruments. Although development policy has a longer history as an external policy of Europe, it was only with the Treaty of Maastricht in 1993 that a formal mandate and legal basis for an EU development policy was established (Orbie, 2020, p. 425). Humanitarian aid, by contrast, was only legally incorporated into the realm of EU external action with the Treaty of Lisbon in 2009. Before the 1990s, there had been neither a distinct legal basis nor an institutional body responsible for humanitarian aid operations. In 1991, the European Commission decided to create a dedicated European Community Humanitarian Office (ECHO) in response to the significantly increasing spending on humanitarian action against the background of the end of the Cold War (Versluys, 2008, p. 209). In 1999, ECHO became fully incorporated into the European Commission as a General-Directorate. Despite the different institutional trajectories and policy objectives of EU humanitarian aid and development policy, we find in the period of investigation that policy integration at the humanitarian – development interface has moved towards a relatively strong level of policy integration, which can be related to a high degree of institutional fit that unfolded a horizontal spillover mechanism between the two domains.

3.1. A high degree of institutional fit: Shared understandings of complementarity

Despite the different institutional trajectories and policy objectives of EU humanitarian aid and development policy, there has been a relatively high degree of institutional fit between the two domains. In terms of the normative underpinnings of the two policy communities, there is a shared belief about the need for global solidarity and assistance for vulnerable individuals and groups in situations of crisis and disaster. Both the European Consensus on Development in 2006 and the European Consensus on Humanitarian Aid in 2008 strongly build on the notion of solidarity as a key driver of the EU’s engagement in development cooperation and humanitarian aid (European Union, 2006, paras. 13, 55; European Union, 2008, paras. 1, 72). As one interviewee stated, shared understandings and compatibilities in their organisational cultures are also due to the fact that when ECHO was created in 1992, many staff members joined from the Commission’s DG Development (DEVCO) (interviewee 6). In addition, another key contributing factor leading to a high degree of institutional fit is the mutual recognition of distinct norms governing the respective policy areas. EU humanitarian aid is bound to the humanitarian law principles of neutrality, impartiality and independence, i.e., the autonomy of humanitarian objectives from political, economic, military or other objectives. Development policy, by contrast, has always been embedded in the EU’s foreign policy system and its strategic objectives (Friesen et al., 2020, pp. 7–8). However, this difference is largely perceived by policymakers in both domains as legitimate and compatible, as it has informed the traditional division between short-term humanitarian relief, which was guided by neutrality and independence, and longer-term development support, which was more strategic and political (interviewees 6, 7, 8, 9). In other words, shared understandings and views of compatibilities between the more short-term-oriented humanitarian aid and long-term-focussed development cooperation have been an incremental part of the institutional culture in both policy domains (interviewees 6, 7, 8, 11). Hence, despite their specific missions and mandates, the fit of institutional cultures between the humanitarian aid and development ‘communities’ has traditionally been relatively high.

3.3. Policy integration at the humanitarian – development interface

The ambition to bring the two policy areas closer together and foster their complementarity first gained traction in the mid-1990s with the concept of linking relief, rehabilitation and development (in short: LRRD), which came up in policy debates at the global level. In 1996, the European Commission published a first Communication outlining the EU’s own approach to LRRD. The Communication emphasised the need for a smooth transition between short-term humanitarian aid efforts and medium – to long-term-oriented development interventions (European Commission, 1996, iii). By adopting the LRRD concept and by framing humanitarian assistance and development cooperation as elements of a continuum, the EU used the LRRD as a tool to create a joint narrative for linking its humanitarian aid and development policies (interviewees 6, 8). In a second Communication on LRRD in 2001, the Commission abandoned the sequential logic in linking humanitarian aid and development cooperation and highlighted that, in situations of complex crisis, different instruments might have to be used at the same time ‘as part of an integrated approach towards preventing crises and disasters’ (European Commission, 2001, p. 3).

In 2012, the European Commission proposed an EU approach to resilience that was also endorsed by the member states. Building on the EU’s experience in dealing with the food crises in the Sahel and the Horn of Africa, the strategy noted that strengthening the resilience of partner countries and societies ‘lies at the interface of humanitarian and development assistance’ (European Commission, 2012, p. 5). Hence, at the communicative level, we witnessed the emergence of an increasingly integrated narrative and joint strategic vision for the cooperation between EU humanitarian aid and development policies, starting from the 1996 Communication on LRRD towards the 2012 Communication on resilience. This has also led to the development of more precise objectives on how to move forward the integration of EU humanitarian assistance and development cooperation, driven particularly by the European Commission.

At the same time, however, policy integration remained limited at the organisational and procedural levels in this period before 2015. The traditional separation of institutional responsibilities for humanitarian aid with DG ECHO and for development assistance with DG DEVCO also translated into different funding instruments and different procedures for managing the funds. Their lack of flexibility and divergent management procedures, implementation modes and timeframes had made policy integration difficult (Van Elsuwege et al., 2016, p. 53; interviewee 6, 10, 12). Moreover, at the level of Council working parties, different working parties on development policy (CODEV) and humanitarian aid (COHAFA) were created, although the representatives of several member states in these working groups come from the same national ministries.

The EU Global Strategy has been described as ‘the big turnover moment’ (interviewee 6), as it put on paper the ambition to use separate resources such as expertise and funding in a more complementary way. The Global Strategy’s ambition to move forward in terms of policy integration can clearly be seen with respect to addressing push factors for migration. It states the need for ‘overcoming the fragmentation of external policies relevant for migration’ and to ‘develop stronger links between humanitarian and development efforts through joint risk analysis, and multiannual programming and financing’ (European Union, 2016, p. 50). At the communicative level, this resulted in a further strengthening of the narrative and vision for a closer cooperation between EU humanitarian aid and development policies. The discourse moved from LRRD towards the ‘humanitarian – development nexus’ (European Commission, 2016), and has further developed towards the concept of the ‘humanitarian – development – peace nexus’ (hereafter: HDP) since 2017, aiming also at the further integration of conflict prevention and peacebuilding efforts into humanitarian aid and development assistance efforts. The 2017 European Consensus on Development emphasised ‘the nexus between sustainable development, humanitarian action, peace and security’ (European Union, 2017, para. 65) and committed EU institutions and member states to ‘implement humanitarian action and development cooperation in a more coherent and complementary way’ (European Union, 2017, para. 69).

The emergence of an integrated strategic vision under the umbrella of the HDP nexus at the communicative level has also led to increased integration at the organisational and procedural levels. At the headquarters level in Brussels, officials at both the political and technical levels have increased the frequency of meetings for discussing issues related to the HDP nexus approach. In particular, the EEAS, DG ECHO, DG INTPA and the Service for Foreign Policy Instruments (FPI) have established focal points for the HDP nexus that interact on a daily basis (interviewees 6, 9).

During the Czech Council presidency in the second half of 2022, the working parties on development cooperation and international partnerships and on humanitarian aid – CODEV-PI and COHAFA – held several joint meetings in which they discussed the lessons learnt on the HDP nexus approach from six pilot countries and how to enrol the approach to additional countries and additional thematic areas such as food security (interviewees 6, 7, 8, 9). This process also led to a Presidency Paper on Good Practices on the operationalisation of the HDP nexus, promoting a shared understanding of the nexus approach across EU institutions and EU member states (interviewees 6, 9, 11). At the country level, the lessons learnt process on the pilot countries demonstrated that the HDP nexus approach led to the creation of effective coordination structures, such as ‘Nexus Task Forces’, that further institutionalise policy integration and involve EU delegations, ECHO field offices and embassies of EU member states (interviewees 7, 8).

In addition to this organisational dimension, new procedures to facilitate the HDP nexus approach have been implemented in the context of the newly established Neighbourhood, Development and Cooperation Instrument (NDICI-Global Europe). One new procedure is the joint Conflict Analysis Screening (CAS) for 68 conflict-affected and fragile countries that the EU cooperates with. Incorporated into the NDICI-Global Europe regulation at the request of the European Parliament, the CAS requires the EEAS and all relevant Commission DGs to base the multi-annual programming for all conflict-affected and fragile countries on a joint conflict analysis. As one interviewee mentioned, a ‘conflict analysis that gives us a joint vision of where the problem lies in any case is a first step (…) to see things the same way, that’s a good innovation in the first step’ (interviewee 7). In sum, policy integration at the humanitarian – development interface has clearly moved towards a strong degree of policy integration in the past few years, including a deeper integration at the organisational and procedural levels (interviewees 6, 7, 8, 11).

3.4. Horizontal spillover mechanism: Realisation of functional pressures, policy entrepreneurship and initial socialisation effects

The dramatic increase in instability and crises, both at the international level as well as in the EU’s immediate neighbourhood, also has important implications for the humanitarian – development nexus. For the EU, the 2015 refugee crisis as well as its peacebuilding experience in a growing number of protracted conflict situations increasingly gave rise to the view that humanitarian and development resources were complementary tools to jointly address a complex mix of challenges (interviewees 9, 10, 11). Accordingly, there was a growing realisation of functional pressures for greater horizontal integration and the need to better realise synergies.

Whilst there still are certain formal constraints for horizontal coordination and network formation – including the absence of a formal coordination mechanism between CODEV-PI and COHAFA – actors have invested in network formation through more informal meetings, workshops and joint study trips. Under the Czech and Spanish Presidencies of the Council of the EU, for instance, all meetings on the HDP nexus were held in the framework of joint CODEV-PI and COHAFA meetings. At an informal workshop held in Madrid in 2023, for example, group discussions and a policy simulation allowed EU and member state diplomats to gain a better understanding of their own roles and those of their counterparts in the EU’s approach to the HDP nexus (interview 11).

Such lower-level networking activities were paralleled with the political agency that was being exercised by supranational actors and EU member states working towards a more integrated approach. A group of EU member states – including Sweden, the Czech Republic, Ireland, Luxembourg, Spain and more recently France – gave considerable attention to the HDP nexus (interviewees 10, 11). At the same time, actors operating on humanitarian and development cooperation in the field articulated the need for a more integrated approach, producing a sense of urgency in the Commission, especially DG ECHO, to improve horizontal cooperation (interviewees 9, 11, 12). To operationalise the HDP nexus, the 2017 Council Conclusions asked the Commission to implement this ‘nexus approach’ in six selected pilot countries – Chad, Iraq, Myanmar, Nigeria, Sudan and Uganda – with the goals of strengthening joint risk and conflict analysis, coordinating programmatic approaches better and using available financing instruments and modalities in a more flexible and coordinated way (interviewee 7; Council of the EU, 2017, para. 10).

The 2017 Council Conclusions on operationalising the HDP nexus also contributed to the development of a more integrated vision. In this regard, key member states acted as policy entrepreneurs for policy integration by asking the Commission to strengthen the interlinkages between the three policy areas. Yet, the Council Conclusions remained vague about what the ‘triple nexus’ narrative would actually mean in practice and what role its ‘peace’ component would have, particularly with regard to the implementation of the nexus approach in the six pilot countries. Therefore, the Secretary-Generals and Directorate-Generals of the EEAS, DG NEAR, DG INTPA, DG ECHO and FPI in 2020 wrote a joint letter that was sent to the senior management of these DGs and services, to heads of EU delegations and heads of ECHO field offices (interviewees 6, 7, 8, 9). As interviewees noted, the joint letter was possible because of the strong collective will across the newly appointed European Commission to move policy integration forward, also in light of the COVID-19 pandemic, which clearly illustrated the need for close coordination between EU humanitarian aid and development assistance (interviewees 6, 9). The letter also spelt out and operationalised the five key elements of the HDP nexus approach. They included (a) joint conflict analysis to develop an understanding of the context and causes of fragility, (b) joint planning to design collective outcomes, (c) coordinating in implementation, (d) agreeing on joint financing mechanisms or adequate funding tools and (e) joint monitoring and evaluation, with the objective of identifying lessons learnt and best practices (interviewees 6, 7, 8, 9).

Hence, the Council – and, particularly, the Commission and the EEAS – were key policy entrepreneurs promoting horizontal networking and crafting initiatives to foster policy integration. The European Commission has assumed an important role in overseeing the implementation of the HDP nexus. In November 2022, DG INTPA, overseen by an inter-service group, issued a comprehensive report on challenges and opportunities related to the implementation of the HDP nexus (Particip & ECDPM, 2022).

Whilst the degree of institutional fit had already been relatively high in the past and contributed to socialisation effects, since 2015 humanitarian and development resources have been increasingly considered as complementary tools to jointly address a complex mix of challenges related to the proliferation of violent conflict, fragility and migration pressures. Socialisation dynamics have been aided by the horizontal network formation activities described above. Providing for routine interactions through joint activities such as workshops fostered joint problem-understandings and awareness of common challenges, including existing institutional hurdles for more integrated policymaking (interviewees 10, 11, 12). Concerning the cooperation of EU-level actors, in particular, daily routine interactions between officials working on the HDP nexus across the different DGs and the EEAS have contributed to some sort of ‘esprit de corps’ among them (interviewees 6, 7 9). As one interviewee put it,

I even exchange with my counterparts [in different institutions] more often than with my colleagues here in the house sometimes (…).we are in constant touch with each other. We do everything together. I mentioned the workshops. We also did recently a joint study together. So the cooperation is very good. (Interviewee 6)

Although interviewees also mention staff rotation as a challenge to preserving socialisation effects over a longer time period (interviewees 7, 8), the evidence points to the emergence of a shared realisation of the added value of an integrated perspective on the cross-policy interface that ‘enabl[es] the different services to reach a common understanding of complex environments and opportunities for more strategic and programmed engagement drawing on the comparative strengths and mandates of the different services involved’ (Particip & ECDPM, 2022, p. 25; similarly, interviewees 7, 8, 9).

In sum, our findings suggest that policy integration at the humanitarian – development interface towards the HDP nexus became possible because of a high level of institutional fit, which led to functional pressures for policy integration being viewed as plausible and compelling. This, in turn, led to horizontal networking and new initiatives for policy integration, with initial socialisation effects resulting from new, informal organisational formats. Although policy integration at this interface is an ongoing process, our empirical findings clearly show how complementary organisational cultures can enable policy integration.

4. Horizontal policy integration at the security – migration interface

The Common Security and Defence Policy and the Area of Freedom, Security and Justice were established 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, p. 184). The parallel structures for external migration management reflect the institutional divide between internal and external security competences that was established by the Treaty on European Union (TEU) in 1992, creating the Common Foreign and Security Policy (CFSP) for external security matters and AFSJ (then Justice and Home Affairs) for internal security issues. Since their establishment, both policy domains have witnessed a progressive deepening of sectoral integration. AFSJ experienced a process of supranationalisation, a broadening of its mandate, as well as new instruments and operational capacities, including the foundation of Frontex in 2004, which serves as the EU’s agency for promoting, developing and coordinating external border management. Conversely, the CSDP (then the European Security and Defence Policy) was launched as part of the CFSP in 1999 to strengthen the EU’s capacity to intervene abroad, address crisis situations and bolster the capacity of weak states in areas such as security sector reform and the rule of law. Since 2003, the EU has conducted a broad range of civilian and military CSDP operations in third countries, acquiring valuable expertise in conflict prevention, crisis management and governance reform (Parkes, 2020, pp. 98–100).

4.1. A low degree of institutional fit between the security and migration policy domains

The fact that the AFSJ and the CSDP are conceived as distinct policy areas is also reflected in a low degree of institutional fit. The organisational cultures prevailing in their respective institutions involve important differences. Regarding their core mission, AFSJ actors have a long history of approaching migration challenges with a strong focus on the implications for internal security – this is at the core of their mandate, as specified in the treaties (interviewees 4, 5; see also Treaty on European Union, 2009, Art. 3). Their main focus is thus on how EU external migration governance serves the EU’s own political interests, primarily in terms of internal security interests. Here, irregular migration is often perceived as a potential threat to the Schengen framework and associated norms. And it has been noted that the focus of AFSJ actors – with respect to external migration governance – tends to be on ‘border controls, return and readmission and combating smuggling (…) instead of first ensuring full compliance with fundamental human rights standards and principles’ in the host countries (Carrera et al., 2015, p. 2). Conversely, CSDP actors traditionally place a strong emphasis on serving the interests of the host states, with the majority of CSDP missions being triggered by requests for assistance from the receiving country or international organisations. Promoting peace and security where needed through the EU’s external engagement is considered part of the CSDP’s core mission. Although the EU’s own political interests certainly play a role in civilian and military CSDP missions, there is also a strong commitment to contributing towards the security and governance capacities of third countries (European Union, 2016, p. 4; interviewees 1, 2, 3, 4, 5). This commitment to partner countries also translates into a strong emphasis on the external promotion of substantive EU norms such as the rule of law, good governance, democracy and human rights, all of which are central norms underpinning the civilian component of CSDP. Moreover, it involves a strong focus on being open to the host country’s perspective and local ownership in the implementation of CSDP missions, which is considered critical to achieve effective and sustainable results (interviewees 4, 5). These substantive differences between the organisational cultures of AFSJ and CSDP institutions were initially facilitated because the two policy domains evolved largely in isolation, and strategic priorities and core governance functions did not overlap in major ways. Accordingly, resource complementarity between the two policy domains has been deemed low (interviews 4, 5).

4.2. Policy integration at the security – migration interface remained limited over time

Consistent with our argument, we find that policy integration at the security – migration interface remained limited over time, corresponding with a low degree of institutional fit. Policy integration between the two policy domains in the period until 2015 had a low degree of institutional fit. In terms of communicative instruments, the ambition to enhance policy integration was addressed in a roadmap on ‘Strengthening Ties between CSDP and AFSJ Actors’ in December 2011 (Council of the EU, 2011). However, the roadmap mainly formulated some general and vague objectives aimed at overcoming institutional boundaries (Council of the EU, 2011, p. 4), rather than creating a strategic vision and shared narrative for doing so. Aside from a formal commitment to the general principle of cooperation in the EU’s discourse, there was little follow-up at the organisational and procedural levels, and first attempts to structure cooperation across the AFSJ-CSDP interface were described as being ‘mainly paper tigers without any effects on policy practice’ (Wolff et al., 2009, p. 13).

In the context of an increasingly instable and conflict-prone regional environment in which the EU faced augmenting migration pressures that culminated in the so-called 2015 refugee crisis (Seeberg, 2011), external migration governance increasingly appeared as a challenge that cut across the institutional boundaries between the AFSJ and CSDP domains. Still, despite a growing realisation of functional overlap, subsequent steps towards greater policy integration at the level of communicative, organisational and procedural instruments have remained rather modest. In the EU’s communications, the EU’s external response to the 2015 refugee crisis was largely formulated separately in AFSJ and CSDP documents, in which the institutes remained committed to their individual strategic visions and objectives in further developing their role in external migration governance. The Agenda for Migration – a Commission Communication presented on 13 May 2015 – outlined the EU’s immediate response to the crisis and also briefly referred to the role of the CSDP in migration management. It stated that migration would become a specific component of ongoing CSDP missions that had already deployed in countries such as Niger and Mali (European Commission, 2015, p. 5), yet there was no strategic CSDP document at that time actually specifying this as a CSDP task (European Council on Refugees and Exiles, 2019, p. 2). It was only in 2018 that the broadening of the mandates of civilian CSDP missions was codified in the Civilian CSDP Compact, stating that they should also ‘contribute to the EU’s wider response to tackle security challenges, including those linked to irregular migration’ (Council of the EU, 2018, p. 4).

Moreover, in 2016 the EEAS circulated a non-paper on enhancing cooperation between CSDP missions/operations and AFSJ agencies. The non-paper provided the thrust of a ‘[m]ini-concept on possible civilian CSDP efforts to address security challenges linked to irregular migration’ that the Foreign Affairs Council adopted in 2020 (Council of the EU, 2020). Yet, despite acknowledging cross-pillar cooperation as a general ambition, the main focus among CSDP and AFSJ actors remained on defining their distinct mandates, roles and visions largely on their own terms. Whilst it was clear that the migration crisis thus increasingly made external migration policy a focus of CSDP and AFSJ communications, no integrated strategic vision – including shared strategic awareness and priorities – for policy integration has been established.

Similarly, only limited steps were taken to work towards more policy integration of procedural and organisational instruments for external migration management. An exchange of letters between the Deputy Secretary-General of the EEAS and the Executive Director of Frontex in January 2015 established a working arrangement between the EEAS and Frontex (Council of the EU, 2015, Annex II). Among other things, the arrangement paved the way for specific procedures for flexible and limited cooperation between Frontex and CSDP missions on the ground, for the exchange of information between the two instruments as well as for technical assistance and institutional support (see Mavrakou, 2021, pp. 8–9). Other procedural measures concerned cooperation on the ground, including the joint participation of civilian CSDP missions and Frontex in the implementation of border-management assistance projects with third countries as well as the deployment of Frontex staff for CSDP missions as ‘visiting experts’ (Huisman, 2021, p. 182). Despite these improvements, cooperation between civilian CSDP missions and Frontex continued to suffer from important shortcomings, including the legal and procedural challenges of inter-institutional information-sharing, a lack of more systematic consultations during the planning stage for improving cooperation in the field, and a lack of capacity to realise synergies through more coordinated uses of resources and logistics (Mavrakou, 2021).

Concerning organisational instruments, progress towards greater policy integration has also remained relatively modest (Shepherd, 2022, p. 98, 132). Despite the agreement to hold joint meetings of the Foreign Affairs Council and the Home Affairs Council and of two committees (Political and Security Committee (PSC) and Standing Committee on Internal Security (COSI)) since the adoption of the roadmap in 2012, the process of agreeing on a joint agenda of cooperation has been challenging, and the benefits of these meetings have been very limited (interviewees 1, 2, 3). In sum, the evidence demonstrates that whilst policy integration at the security – migration nexus has moved forward a bit since 2015, progress has been limited and implies that there has been a lack of substantive efforts towards a stronger degree of policy integration.

4.3. The absence of substantive horizontal spillover

Marked by a limited degree of institutional fit, the enhanced problem-solving pressure resulting from the 2015 refugee crisis was not followed by a dynamic of horizontal spillover. Rather, the EU’s response was predominantly characterised by a sectoral logic within the CSDP and ASFJ domains, whilst the different elements of the horizontal spillover mechanism were largely absent. This is interesting, as in light of the 2015 refugee crisis, there was a growing realisation about functional pressures and interdependencies among EU officials, as it became apparent that both policy domains disposed of valuable resources and expertise (interviewees 4, 5). However, prevailing differences in their respective institutional cultures resulted in a situation in which actors from both domains only displayed limited political willingness and ambition to engage in steps towards greater policy integration.

With respect to horizontal networking, AFSJ actors primarily saw the migration crisis as an opportunity to strengthen their own mandate and capacity for external migration governance, rather than to systematically enhance links with the CSDP domain (interviewees 2, 3; Mavrakou, 2021, p. 12). Within the AFSJ domain, efforts focussed on further strengthening the EU’s framework in the areas of external border control, return, cross-border crime and asylum. Among other things, this included the strengthening of the mandate and resources of Frontex in external border management, including by establishing a European Border and Coast Guard standing corps (European Union, 2019). Among CSDP actors, in turn, external migration management was often seen as a way to make civilian aspects of the CSDP more relevant. Simultaneously, there was a clear ambition to do so without becoming dominated, contaminated or compromised by AFSJ governance objectives and actors. In particular, member states’ ambassadors in the PSC voiced concerns that strengthening the operational capacities of Frontex may come at the expense of the contributions of CSDP missions towards external migration management (interviewees 4, 5). This resulted in a situation in which actors from both domains sought to guard their own policy objectives and institutional self-interests, whilst only agreeing on modest steps towards greater policy integration that were, among other things, promoted by actors such as the EEAS.

Not only was there a limited ambition among the actors from the AFSJ and CSDP domains to function as policy entrepreneurs, functional pressures were also not accompanied by decisive political agency. There were few initiatives by EU actors, such as the EEAS non-paper on enhancing cooperation between CSDP missions/operations and AFSJ agencies, and they often lacked ambition (interviewee 5).

We also do not find evidence for substantive socialisation effects, which is the third element of the horizontal spillover mechanism. At the time of the 2015 refugee crisis, certain cross-cutting organisational frameworks for policy coordination between the AFSJ and CSDP domains had already been put in place. Notably, this included joint meetings of the Foreign Affairs Council and the Home Affairs Council and of two committees (PSC-COSI). However, despite such institutionalised forms of interaction, there is little evidence that would indicate a substantive transformative effect of elite socialisation towards a more unified institutional culture. Our interviews suggest that officials from both domains remained predominantly socialised in their particular institutional context, rather than developing shared problem-understandings, a common purpose or a normative commitment (interviewees 1, 3, 4, 5). The traditional perspectives on external migration governance identified above – an internal security lens shared by AFSJ actors that emphasises the EU’s own political interests, versus a foreign policy lens shared by CSDP actors that puts a focus on host states’ interests and needs – continue to represent core differences in terms of organisational culture (interviewees 4, 5). Actors in both domains thus largely maintained a distinct outlook that placed clear limits on their ambitions for policy integration, which resulted in a limited degree of policy entrepreneurship for policy integration (Trauner, 2016, pp. 1–2).

In sum, our findings suggest that the individual elements of the horizontal spillover mechanism were largely absent in the case of the security – migration interface, which is marked by a low degree of institutional fit and a weak level of policy integration.

5. Comparative discussion of findings and conclusions

Comparing the humanitarian – development and the security – migration nexuses, we find that policy integration has differed with respect to the two interfaces since the 2015 ‘refugee crisis’. At the humanitarian – development interface, policy integration has moved from a situation in which it was limited to the communicative level towards a comprehensive form that also involves a strong degree of integration at the organisational and procedural levels. At the security – migration interface, in turn, policy integration has remained rather modest, with progress being largely limited to the development of some informal, procedural coordination practices and ad hoc organisational arrangements. Here, the main integration dynamics mainly followed a sectoral logic, marked by efforts to strengthen the individual policy domains rather than the policy integration between them. In line with our central argument, we thus find that different degrees of institutional fit correspond to different degrees of policy integration.

Our first case study on the humanitarian – development nexus was marked by a high degree of institutional fit, based on an understanding of complementarity and division of labour between the policy areas – a factor that has been inherent in the organisational cultures of both domains. As our subsequent process-tracing analysis has shown, this shared understanding has greatly facilitated interactions and network formation between EU officials working in the two policy domains, political agency that was driven by actors from top-level EU institutions with the Council and different Commission DGs acting as the main policy entrepreneurs, as well as socialisation effects. This shows that the main components of the horizontal spillover mechanism have enabled substantive progress towards a deeper level of policy integration across the humanitarian – development interface.

Overall, the two case studies point to the importance of a high degree of institutional fit in order for horizontal rather than sectoral spillover dynamics to unfold. At the same time, the novel conceptual mechanism of horizontal spillover also addresses an important shortcoming in the theoretical literature on European integration, which has paid insufficient attention to the theoretical explanations of policy integration between different sectors. Whilst spillover mechanisms are well-known and at the core of neo-functionalist accounts of European integration, they have thus far primarily focussed on accounting for how functional pressures emerging from integration in one policy area result in steps towards greater supranational integration in terms of the expansion of the tasks and authority of supranational institutions in another policy area of the EU.

A promising way forward would be to analyse additional EU external policy interfaces, such as the climate – security or the trade – development nexus, to broaden the comparative perspective on policy integration at the EU level. Another pathway for further research could be to compare policy integration in foreign policy at the EU member state level, as the ambition for an ‘integrated’ or ‘nexus’ approach to external conflicts and crises is also shared by several member states such as Denmark, Finland, Germany, Italy and Sweden.

List of interviewees

Interviewee 1, EU official, July 2021

Interviewee 2, Independent expert, July 2021

Interviewee 3, EU member state official, July 2021

Interviewee 4, EU official, July 2021

Interviewee 5, EU official, July 2021

Interviewee 6, EU official, January 2023

Interviewee 7, EU official, January 2023

Interviewee 8, EU official, January 2023

Interviewee 9, EU official, January 2023

Interviewee 10, EU member state official, November 2023

Interviewee 11, EU member state official, November 2023

Interviewee 12, EU member state official, November 2023

Acknowledgements

The authors would like to thank the journal editors and the reviewers for very helpful comments on earlier versions of the article. The authors would also like to thank the interview partners for sharing their views and insights with them.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Julian Bergmann acknolwedges funding from the Bundesministerium für wirtschaftliche Zusammenarbeit und Entwicklung (BMZ)/ German Ministry for Economic Cooperation and Development. The views and opinions expressed in this article are solely those of the authors.

Notes on contributors

Julian Bergmann

Julian Bergmann is a senior researcher at the German Institute of Development and Sustainability (IDOS) and a senior lecturer (“Privatdozent”) at the University of Mainz, Germany.

Patrick Müller

Patrick Müller is Associate Professor at the University of Deusto, Bilbao, Spain, and a Guest Professor of European Studies at the Vienna School of International Studies, Vienna, Austria.