From conflict early warning to fostering resilience? Chasing convergence in EU foreign policy

ABSTRACT To prevent violent conflict and governance breakdown abroad, the European Union (EU) needs anticipatory analysis and preventive action at the EU and member state level that follow a coherent approach. The EU’s resilience agenda can complement early warning risk analysis for prevention, but it is unclear to what extent it has been operationalized. We compare the role that resilience plays in crisis early warning in diplomatic services at the EU and member state level in France and Germany. Drawing on the literature on Europeanization and diffusion, we seek to explain different levels of convergence regarding a resilience approach in early warning at the levels of strategy, analysis and action. We find that the diffusion item’s specificity, the number of sources and particular institutional contexts impact the resulting level of convergence. Member states see value in complementing risk analysis with a resilience perspective, but the EU has failed to provide a sufficiently clear source model. Our results contribute to the literature on EU foreign policy diffusion and coherence and show that the EU has not exhausted its potential to promote resilience as a tool for more coherent and effective conflict prevention.


Introduction
After failures to prevent mass violence in places like Rwanda and Srebrenica, the international community, including organizations like the UN and national governments, has engaged in extensive soul searching. 1 Despite much hope in improved early warning systems, the world is currently witnessing two of the most devastating humanitarian disasters in recent history in Syria and Yemen. With conflicts in Europe's neighbourhood and an apparent lack of great powers' political willingness to agree on peaceful solutions, conflict prevention is trending againand so is the topic of early warning (EW). 2 The European Union (EU) and several member states (MS) have developed early warning capacities in their diplomatic services in recent years. 3 Most EW tools focus on monitoring, forecasting and countering risks of conflicts and other crises. This special issue puts forward a more holistic framework for the prevention of violent conflict and governance breakdown that adds the possibility of not only countering risks directly, but fostering the resilience of a society to withstand risks and keep them from turning into threats. 4 The concept of resilience has trended in European foreign policy debates during the last years. 5 A proliferation of studies on the EU's so-called resilience turn has tried to define its resilience approach, including whether it implies a departure from maximalist normative ambitions of democracy promotion towards principled pragmatism, stability and securitya debate relevant to democratization. 6 The EU's Global Strategyits guiding document on foreign policyspells out the ambition to prevent conflicts and crises abroad and introduces resilience, defined as "the ability of states and societies to reform, thus withstanding and recovering from internal and external crises", as a key concept. 7 This transformative idea of resilience as adaptation is also reflected in this special issue's definition of resilience as the adaptive capacity of societies, communities, and individuals to deal with opportunities and risks in a peaceful manner. According to it, resilience and risks influence societies in opposite ways towards or away from potential violent conflict and governance breakdown. The higher the societal resilience, the more likely are societies to be able to fend off risks and adapt peacefully. 8 Applied to EW, a resilience perspective allows actors to not only monitor the multitude of possible future risks that are hard to predict and to counter but identify positive developments and entry points for fostering resilience. The EU has recognized this potential and attempts to implement a resilience approach in EW. 9 But whether and to what extent it has been able to do so is an open question, as many have argued that the resilience concept is too vague to guide policy action. 10 And it is unclear to which extent the EU's resilience agenda has translated into EU and member state practice. 11 Resilience building as a difficult task requires analysing resilience factors in a systematic way. And the extent to which the EU can act as an external resilience builder also depends on the level of coherence between EU institutions and MS' practice, that is, across the EU's multi-level governance system. While some argue that the EU Global Strategy and its resilience agenda have the potential to unite MS under a coherent framework for external action, coherence in foreign and security policy as the bastion of state sovereignty and intergovernmentalism remains particularly difficult. 12 To shed light on whether the EU has succeeded in operationalizing its resilience agenda and ensuring coherent engagement in the area of conflict and crisis EW and prevention, we analyse to which extent capacities for EW in foreign policy at the EU and member state level are tailored towards resilience building. More specifically, we build on the literature on Europeanization and diffusion processes to explore how far the use of resilience in EW as conceptualized by the EU has travelled from the supranational to the member state level in France and Germany and investigate the causes for the varying degree of convergence.
Our analysis contributes to the literature in three ways. First and foremost, it shows how resilience as a foreign policy concept diffuses between the EU and MS and what determines convergence. As a consequence, it helps understand the requirements for more coherence in EU foreign policy as a prerequisite for the effective fostering of resilience abroad. Finally, it provides insights into EW in EU and MS foreign policy and the potential added value of resilience in it.
In the following section, we first discuss the potential of the EU's resilience agenda for conflict EW. After that, we explain how we trace the diffusion of resilience from the EU to MS' anticipatory analysis in diplomatic servicesa potential process of horizontal Europeanization. We then present our results on the role of resilience in EW and the determinants of convergence. In the conclusion, we discuss the implications of our findings in the light of this special issue.

Early warning and the EU resilience agenda
Early warning (EW) refers to a systematic monitoring of risks, resulting in a warning about a potential undesirable event or development. This includes tools that inform impact mitigation of, for example, natural disasters in the short timeframe and tools that inform long-term structural prevention of conflicts and political crises. Few international organizations and governments have established fully fledged conflict EW systems that link warning to action. 13 One example is the European External Action Service's (EEAS) and the European Commission's EU Early Warning System (EU EWS), rolled out in 2012 as an "evidence-based risk management tool that identifies, assesses and helps prioritize situations at risk of violent conflict for non-EU countries, focusing on structural factors and with a time horizon of four years" and "identifies conflict prevention and peace building opportunities". 14 According to official EU documents, "early warning capabilities are indispensable" to "identify structural risks of conflict and prevent the emergence, re-emergence or escalation of violent conflict". 15 The EWS includes a global quantitative conflict risk analysis, political prioritization of country cases and an inter-agency process of qualitative case assessments, resulting in adjustments to programming across diplomacy, development, humanitarian, trade and other policy areas. 16 While EW processes exist in areas such as intelligence, development, defence and humanitarian disaster relief, few EU MS' diplomatic services have developed formal processes similar to the EU EWS. 17 These have not yet been systematically analysed in the literature. Existing studies focus on improving forecasts, on the gap between EW and early action due to lacking political willingness, which was also identified as a challenge of the EU EWS, and a potential for improved analysis and better communicated warnings. 18 To help overcome the gap between warning and action, the concept of resilience in EU foreign policy debates helps move the focus from risks to entry points for fostering dynamic adaptation. Indeed, the EU sees value in the resilience approach for EW and prevention. 19 This special issue's conceptual framework helps clarify the relationship between EW and resilience. Its central concept of societal resilience allows to clearly distinguish societies' ability to adapt to avoid conflict from autocratic regimes' resilience to democratic change that could be implied in state resilience, and which has traditionally been covered by the literature on democratization 20highlighting the relevance of resilience building in the context of democracy promotion. Following this framework, we assume resilience to affect societies in the opposite direction of risks, meaning that the baseline level of resilience can influence whether a tipping point for violent conflict or governance breakdown will be reached. 21 As a consequence, identifying and monitoring resilience factors in addition to risks facilitates finding both entry points to decrease the likelihood of a known risk turning into a threat and ways to increase the general level of resilience in the face of so-called unknown unknowns, moving from mere detection and warning to a holistic view on opportunities to increase resilience in specific areas at risk. 22 Take the example of advanced technologies in the hands of malign actors. Governments' ability to directly fend off the risk of conflict in a polarized society when forged content of politicians inciting hate spreads on social media. Beyond identifying risks, a resilience approach allows identifying entry points to, for example, support media literacy and investments in a pluralistic, free media environment which help raise the general level of resilience and counter escalation. 23 Other contributions in this special issue provide insights into the opportunities of fostering societal resilience in specific cases in the EU's neighbourhood, including Bosnia, Iraq, Syria, Lebanon, Libya, Tunisia and the EU's Eastern Partnership Countries. 24 But whether EW capacities at the EU and MS level are tailored towards resilience building and what facilitates a coherent EU approach are open questions, which we aim to answer. Many authors have questioned the EU's preparedness to actually deliver on its promise to foster resilience. 25 Joseph and Juncos develop a framework of features that an EU resilience approach for peace should exhibit in practice and find that "the promise of resilience remains unfulfilled" in peacebuilding. 26 Similar features of resilience should also be noticeable in a resilience approach to conflict EW: (1) complexity or systemic thinking, (2) a long-term orientation (2) and a focus local capacities and human agency, as elements of a holistic approach that links resilient adaptation in one part of society to the ability to withstand risks. 27 Existing conflict EW approaches tend to focus on short-term forecasts of narrow concepts such as conflict fatalities instead of long-term futures thinking, which covers systemic effects and blind spots that forecasts based on past data cannot account for. 28 The few analyses of EU EW already point to shortcomings like the limited consideration of perceptions of local populations in contexts at risk of conflict and a need to linking MS' and EU capacities for coherence. 29 Existing research on resilience in the foreign policy of MS find that "a resilience paradigm is increasingly apparent in the foreign policy documents of Germany and France, particularly in their development and humanitarian agendas". 30 The temporal and spatial clustering of both EW systems and the adoption of the resilience concept by the EU and MS point towards interdependent processes. We thus aim to find out to which extent their EW capacities follow a resilience approach and what impacts the resulting level of convergence in the understanding and operationalization of resilience in EW between the EU and MShoping to shed light on how a more coherently operationalized resilience approach in EW could be achieved.

Combining Europeanization with a diffusion perspective
To better understand the processes that led to a resilience approach in EW and to investigate the varying degree of convergence between institutions, we build on Europeanization research and diffusion studies. Defined as "a process of incorporation of European Union norms, practices and procedures into the domestic level", 31 Europeanization is well-suited to analyse how resilience travelled from the EU to the member state level. It distinguishes between two basic mechanisms. Vertical mechanisms clearly demarcate the EU level of origin and domestic level of adoption.
Horizontal mechanisms "look at Europeanization as a process where there is no pressure to conform to EU policy models [and] involve a different form of adjustment to Europe based on the market or on patterns of socialization". 32 The latter mechanisms are suitable to capture how the EU influences the national level in the area of Common Foreign and Security Policy (CFSP), where the shadow of supranational hierarchy is absent, and MS rely on intergovernmental coordination. 33 A number of scholars have applied Europeanization to foreign policy. 34 While the mechanisms in the Europeanization literature are sufficient to investigate MS' motivation to include resilience in EW activities, we consult the diffusion literature to clarify what influenced our dependent variable, that is, the degree of convergence we observe. Like Europeanization, the diffusion approach does not assume that full or complete convergence is a necessary or likely outcome of the processes it analyses. 35 Defined more broadly as a consequence of interdependent decisionmaking, 36 diffusion research offers assumptions on convergence as one potential outcome, especially for cases like resilience in EW, in which there is no adaptational pressure. As demonstrated by Börzel and Risse, embedding Europeanization in diffusion research allows "to capture the more indirect ways in which the EU may affect domestic (…) change". 37

Assumptions
Taking into account the strengths and limitations of both approaches, we use the diffusion literature to derive assumptions that can explain varying levels of convergence between the resilience approach in EU and member state EW processes. In addition, we take Europeanization mechanisms to analyse MS' underlying motivations.
Our assumptions on convergence build on a meta-review of causes for (non-)convergence conducted by Klingler-Vidra and Schleifer. 38 First, we expect more convergence the more specific a diffusion item is. The idea is that a high degree of specificity gives adopters more precise information about what to reproduce and how to do so, leaving less room for interpretation and leading to higher convergence. 39 Second, we expect more convergence the more similar the institutional context and mandate of MS' EW processes are compared to the EU model. As Klingler-Vidra and Schleifer argue, more similar formal institutions between receiving and sending actor increase the probability of convergence. 40 This allows focusing on micro-level administrative practices rather than the Europeanization approaches' focus on actor constellations and veto players to explaining varying domestic change. 41 Third, we expect less convergence the more different sources were considered. Diffusion processes with a single, central source are assumed to cause more convergence than multiple-source mechanisms, in which actors might combine elements from several sources. 42 The literature already shows that EU MS have adopted resilience based on different sources. 43 Finally, we assess actors' underlying motivations to include a resilience approach in EW: socialization or strategic calculation. The Europeanization literature usually focuses on the market or socialization patterns when explaining horizontal Europeanization. 44 Moumoutzis argues that strategic calculation can drive the incorporation of EU foreign policies by MS in areas without adaptational pressure. According to him, socialization takes place when national foreign policy makers interact and "become convinced of the appropriateness of the EU way of doing things and internalize EU behavioural rules". 45 Strategic calculation takes place whenever national policy makers choose EU foreign policy practices and procedures over alternative courses of action because they consciously consider alternative options and choose the option that allows achieving their goals most effectively. 46

Methodological approach
The limited number of existing conflict EW systems in EU MS' diplomatic services and little publicly available information limit our case selection (Interview #1). A research project with the EEAS, the French Ministry of Europe and Foreign Affairs (MEAE) and the German Federal Foreign Office (FFO) as partner institutions enabled us to obtain preliminary information about existing EW approaches. In addition to a document study of public sources, we conducted two full semi-structured interviews with representatives of each case (EU, France and Germany), as well as several background conversations. While the latter were not authorized for citation, we used them to triangulate our findings. 47 To increase the variation in our observations, we structure the empirical section according to three levels of EW: strategy, analysis, and action. The strategic level relates to the explicitly formulated understanding of a resilience approach. The level of anticipatory analysis relates to the actual operational procedures of EW, such as the EU EWS analytical components. Finally, we define the category of action narrowly as the political action potentially triggered by the EW analysis (not including other potential areas of resilience building in, for example, humanitarian and development policy). At all three levels, we take the EU approach as the source model for diffusion.
Going beyond a mere comparison of whether a resilience approach, according to the EU definition of resilience and the three features of (1) complexity and systemic approach, (2) long-term thinking and (3) a focus on local capacities, is present, we apply congruence methods as presented by Beach and Pedersen to test our assumptions the determinants of (non-)convergence. 48 Several authors argue that the unique nature of CFSP makes Europeanization less likely to occur, its influence weaker, or at least more difficult to detect. 49 To establish that non-binding EU stimuli caused domestic change, Moumoutzis suggests to process trace the observable implications of alternative explanations of foreign policy change 50an approach that is also frequently used in diffusion research. 51 While a fully-fledged process tracing analysis goes beyond the scope of this article, congruence methods allow assessing mechanistic evidence for our assumptions without making the study of the underlying causal mechanism explicit. 52 The weaker causal inferences that can be made on this basis allow us to draw "tentative conclusions about a causal process that potentially links a given cause and outcome". 53 4. Convergence of the resilience approach in early warning

Early warning in French and German foreign policy
As in the EU's approach to EW, presented in chapter 2, the French and German governments also recognize the value of EW for conflict prevention. The MEAE's 2018 strategy on "Prevention, Resilience and Sustainable Peace (2018-2022)" mentions the "SyAI" EW system that has been developed by the ministry's Crisis and Support Center, designed to "anticipate crises using a comprehensive approach". 54 In its 2017 "Guidelines on Preventing Crises, Resolving Conflicts, Building Peace", the German Federal Government has committed itself to further developing its EW capacities to an inter-ministerial working group for "joint situation assessments of potential crises". 55 The EW processes at the MEAE and FFO as the counterparts of the EEAS are the comparison cases for our analysis. 56 In the following section, we present the operationalization of resilience in all three cases, including the EU, and our evidence on the causes of (non-)convergence according to the assumptions described in the previous chapter at the three levels of strategy, analysis and action, before we evaluate the results.

Strategy: understanding of resilience as guidance
For assessing convergence between the EU and MS' resilience approach in EW at the strategy level, the EU's definition serves as the source model, originating in the Global Strategy and further refined as the "capacity of societies, communities and individuals to manage opportunities and risks in a peaceful and stable manner, and to build, maintain or restore livelihoods in the face of major pressures". 57 As explained in Section 2 above, we trace the approach according to the definition and the core features of complexity, a focus on bottom-up capacities (beyond the central state) and a long-term orientation.
Interview partners in Berlin and Paris highlighted that the 2017 EU joint communication greatly helped understand what the EU resilience approach means for foreign policy (Interviews #2, #1, #4). They acknowledged that the concept is inherently broad and cannot give more specific guidance than is included in the existing EU documents. One counterpart highlighted that resilience as a complementary concept to risks constituted "an eternal truth", which "could not be ignored" when writing guidance on conflict prevention (Interview #2). Even though a critical voice described resilience as a rather "empty" concept and "a container" that could be replaced by another term exercising a similar function (Interview #4), counterparts considered the overall specificity of the diffusion item at the level of strategy as medium to highalmost as specific as a definition of a broad guiding concept in strategy documents can get.
A comparison with the French and German documents shows great similarityor convergencebetween MS and the EU at the level of the resilience definition. The key idea of an ability to adapt peacefully and the core elements of complexity, a focus on bottom-up capacities and a long-term orientation also feature in the definitions of France and Germany. 58 Officials from all three institutions displayed similar understandings of resilience as the opposite of risk or fragility. Both national documents explicitly refer to the EU's resilience agenda, in the French case in relation to EU external action and in the German case with reference to the EUGS. 59 The German document explicitly acknowledges EU EW efforts and commits to working with the EU and other partners on EW. 60 The French conflict prevention strategy carries resilience in its title and specifies the concept throughout the document. In the German conflict prevention strategy, resilience does not occupy a comparably central role and is less significant at the decisionmaking levels at the German MFA (Interview #4). All three sources are outcomes of drafting processes in similar institutional contexts across the foreign policy, development and humanitarian realm. Some ambiguity remains with regard to the role of complexity or systemic thinking, local capacities and a long-term orientation. EU officials observe that the concept of resilience increasingly moves into the realm of development policy, while at the EEAS "nobody has been able to really pin it down" for foreign policy practice (Interview #1; similar in #5). No working unit at the EEAS is tasked with resilience (building) and references in working documents have decreased (Interview #1). EU development policy emphasizes local capacities and a long-term orientation, while the EU EWS' forecasting tool conceptualizes resilience more narrowly as a state's resistance to conflict (Interviews #1, #4).
Both national strategic documents refer to other sources beyond EU foreign policy guidance, including the EU consensus on development and the areas of disaster risk reduction, development and defence. The MEAE developed its resilience definition resilience based on a range of sources and each working unit dealing with the concept operationalizes it to its own needs (Interviews #3, #6). In both France and Germany, resilience has been used in the context of humanitarian and development policy, including disaster risk reduction and natural hazards, with reference to EU humanitarian and development policy documents such as the 2012 Supporting Horn of Africa Resilience (SHARE) drought recovery strategy, before it appeared in the EUGS, other EU foreign policy documents and the conflict prevention discourse (Interviews #2, #3, #4, #6). 61

Analysis: operationalizing resilience
The operationalization of resilience for EW analysis varies considerably. The EEAS has developed the most institutionalized conflict EW mechanism that may serve as a source for MS, but the specificity of resilience as a source model for analysis is low. On paper, EU recommendations for resilience in foreign policy include the integration of "appropriate indicators of resilience alongside the risk factors currently monitored", as a "monitoring of resilience indicators could help identify the tipping point at which pressures overwhelm coping mechanisms". 62 The EU EWS's quantitative risk assessment does not measure resilience, but there is a desire to include resilience indicators in the future (Interview #1). 63 The in-depth qualitative analysis of country cases does not systematically include resilience factors. Sources of resilience sometimes come up in unstructured assessments, especially towards the stages of identifying entry points for preventive engagement (Interview #1).
Counterparts in France and Germany are barely aware whether the EU operationalizes resilience in EW. Nevertheless, they make use of the resilience concept in their anticipatory analyses. Even more distinctly than at the strategic level, they consider resilience to have practical value and highlight that analysing resilience as the opposite of risks allows finding entry points for action.
The French MEAE's crisis centre, which is responsible for humanitarian and consular action and crisis management including EW, has developed the Atlas index for natural disaster risks. The index involves hazard, exposure, vulnerability and resilience, defining the latter as the ability of a system, a community or a society exposed to risks to resist, absorb, receive and correct the effects of a threat, in a timely and effective manner, especially through the protection and restoration of its essential structures and core functions. 64 Resilience is most clearly operationalized and measured as the non-resilience of state, i.e. a state's inability to correct the effects of a natural disaster (Interview 3). 65 The focus on natural hazards is the result of a different context and mandate of the unit tasked with EW compared to the EU level. Developing the index, analysts considered a variety of sources, including the EU INFORM disaster risk index, which is closer to their mandate of assessing natural disaster risk than the EU EWS, focused on conflict risk (Interview #6). The MEAE's crisis centre has started developing a conflict forecasting tool, which is not in use yet. Analysts are working on ways to measure resilience, for example, through proxies like past likelihood of escalation from peaceful protest to violent unrest (Interview #3, #6).
At the German FFO, the conflict analysis team makes use of quantitative analysis by the ministry's own data analysis project and additional qualitative and classified sources that assess both drivers of crises and potential sources of resilience as opposing categories across socio-cultural, economic, political and other factors (Interview #4). 66 Resilience factors play a special role in the qualitative attempts at suggesting specific avenues for action, which are then considered by the inter-ministerial working group. The analysis team is located in the Directorate-General responsible for crisis prevention, stabilization, post-conflict peacebuilding and humanitarian assistance (DG S). In contrast to France, the FFO's crisis centre not located in the same DG, while working units for humanitarian assistance are also located in DG S. Those units fund efforts for anticipatory humanitarian action including forecast-based financing, but this is not formally linked to the anticipation of crises in the FFO's own crisis EW process.
The institutional context and mandate of the EW process in Germany are more similar to the EU level than in the case of France, as the focus has been on crisis EW in a broader sense that is supposed to inform action across government portfolios.

Action: fostering resilience in practice
The link between EW and action is formalized at the EU level, as the EU EWS's results inform adjustments in programming and action across a wide range of policy fields. But its specificity as a model that operationalizes resilience is low. In this area, the EU also struggles to put resilience building as a foreign policy into practice and many see the concept more fitting for development policy, which is concerned with long-term structural prevention of crises and conflicts (Interviews #1, #6).
At the national MFAs, resilience building as an instrument is not considered to provide added value. The strategic documents are vague regarding how specific foreign policy instruments can build resilience, although the documents establish a link, for example with training and equipment of state security sectors by Germany. 67 Only the French document mentions strengthening "sources of resilience". 68 The results of the Atlas Index have been made available across the MEAE, but there is no formal link between EW analysis and action, and "conflict early warning has not yet found its place" (Interview #3, similar #6).
In Germany, so-called crisis early detection papers resulting from the FFO's process provide suggestions for action based on resilience factors and available tools for action. They are taken into the ministerial bureaucracy for further decision-making. But resilience building is not a distinct field of action. Counterparts instead refer to instruments that are either more specifically conceptualized or more established in their institutions and discourseslike prevention, stabilization or security sector reform (Interviews #2, #4). 69 The concept of resilience thus gradually "disappears" on its way from the abstract, strategic level to concrete action (Interview #2). Table 1 gives an overview of the findings for each institution at the levels of strategy, analysis and action regarding the assumptions of (a) more convergence with the EU level when the source model of resilience was specific, (b) more convergence in the case of EW units with a similar institutional context and mandate, (c) more convergence in the case of single sources from which resilience emerges, and (d) the decision to include resilience as either a result of socialization considerations or strategic calculation.

Summary and evaluation of findings
The focus of our research on EW units in diplomatic services means that our evidence is strongest for the level of analysis, while it is limited regarding the levels of strategy and action, in which other parts of diplomatic services not interviewed for this article also play a role. But the great observed variation between the three levels allows for a more nuanced picture on added values and shortcomings of the EU's resilience approach in EW.
At the strategic level, we find evidence for the diffusion of a clearly defined EU understanding of resilience, which is taken up by MS, resulting in convergence towards very similar definitions of resilience. Interview partners acknowledge that strategic concepts are necessarily vague, and the EU has provided sufficient guidance at this level, but we find that the EU's resilience approach in foreign policy as the source model fails to mention the key features of complexity and a systems approach, long-term thinking and a focus on local, bottom-up capacities. Although MS also refer to other sources than the EU and our research did not allow tracing the initial diffusion of resilience into strategic discourses, we find evidence that the specificity of the diffusion item and similarity of the context in which strategic guidance documents on conflict prevention and resilience are developed favour convergence. Regarding the link between resilience and EW, the strategy documents remain vague in all cases including the EU.
We find evidence for both socialization and strategic calculation as mechanism at the level of strategic guidance in Germany. Here, resilience is both considered rather empty with no major role for overall strategy, but the strategy document still references the EU as important partner and resilience was seen as a truth that could not be ignored. While we consider the level of convergence greater in the case of France, where resilience is more central to the prevention strategy, the evidence is not sufficient to draw conclusions on the motivation. Perhaps our most surprising finding emerges at the level of EW analysis, where the EU does not provide a specific model and we thus cannot observe a direct link for diffusion. Instead, MS see an instrumental value in resilience as analytical category and operationalize it according to their needs. The availability and consideration of multiple sources may increase the uptake of resilience, but this happens in a very context-specific manner, for example resulting in an operationalization that only takes into account state resilience in the French disaster risk analysis. We consider convergence towards the EU understanding of the term higher in the German than the French case. Although not necessarily better, the German qualitative assessment of resilience is more similar to the unstructured assessments by the EU than the formalized equation in France. This seems related to the more similar institutional context and mandate for EW in Germany and the EU as opposed to France, with the crisis centre's focus on natural disasters and humanitarian needs, and absent link between EW and action. According to the diffusion literature, strategic calculation and learning tend to result in less convergence, as recipients tailor approaches to their needs and might take inspiration from multiple sources, which this example confirms. 70 At the level of action, the EU foreign policy approach to resilience also lacks specificity. In MS, we find evidence for strategic calculation resulting in the decision not to use resilience-building as an instrument. Diplomatic services instead dispose of instruments that are more clearly delineated and tested, like stabilization or mediation. Resilience (building) gradually disappears from medium to high convergence at the strategic level over independent operationalization at the analytical level to it having no role at the level of action, at which counterparts in all three ministries consider it lacking specificity as an instrument for foreign policy or more appropriate in the context of development policy.

Discussion: from early warning and (in)action to resilience building?
We have argued that the EU's ability to act as an external resilience builder to prevent conflicts abroad depends on a coherent implementation of the resilience agenda in EW at the EU and MS level. While this is not sufficient for effective resilience building, anticipatory analysis as a first analytical step increasingly informs foreign engagement. A resilience perspective in early warning can complement risk assessments and help find entry points for preventive action. We analysed to what extent a resilience is operationalized in conflict EW at the EU, as well as the French and German diplomatic services, andbuilding on Europeanization and diffusion approacheswhat influences the degree of (non-)convergence between the EU as the source and MS.
We find that the specificity of the diffusion item, the institutional context and mandate, as well as the number of sources matter for the degree of convergence. Convergence is strongest where specificity is highest and contexts are similar. A lack of specificity, a different mandate and other sources beside the EU considered in a process of strategic calculation and adaptation led each member state to adopt resilience in a very distinct way for EW analysis and to neglect resilience building an instrument for foreign policy action.
The results have important implications for the EU's resilience agenda. While other authors have found that resilience may lack specificity, our findings show that resilience fills an analytical gap and allows to bridge EW and early action by finding entry points for strengthening existing resilience. 71 MS see this added value and are willing to consider the EU approach to EW. This is a rare opportunity for the EU to shape member state practice in foreign policyan area characterized by high independence. But the EU has not provided sufficient operationalization of the resilience approach in early warning analysis and resulting action that MS could follow. The EU EWS is a model for linking risk analysis with action across a range of external portfolios, but methods to analyse resilience and tools to foster it are underdeveloped. The distinct features of the EU's resilience approachcomplexity, bottom-up local capacities and a long-term perspectivehave yet to find their place in it, as has the idea of working with sources of resilience, which other contributions in this special issue show to be valuable for fostering resilience against conflict. Changing this requires a clearer idea of what resilience-building can add to existing analytical and policy instrumentswhether as a distinct instrument of prevention with its own budget line and projects, or as an approach that should be mainstreamed, like conflict sensitivity.
Finally, an anticipatory approach to prevent conflicts that matches the conceptual underpinnings of resiliencecomplexity, unpredictability, unknown unknowns and a long-term strategic orientationneeds to cover the full spectrum of diffuse, global and regional risks, apply a variety of methods including structural risk assessment and foresight, and needs to systematically include the monitoring of resilience and its sources in order to identify tipping points and find entryways for action. 72 This has not been realized at the EU or member state level yet and requires better dataincluding survey data on people's perception of trust and governance quality.
Further research should focus on the underexplored value of resilience for a more comprehensive early warning and prevention approach, including case studies of analysis and action in countries at risks. The small snapshot we were able to cover shows that a systematic review of anticipatory analytical capacities across the fields of humanitarian action, defence, the strategic levels of policy planning and especially development policyan area in which the idea of building long-term resilience seems more naturally at homecould help identify gaps for a more strategic, long-term approach that is needed to achieve coherence and build resilience against conflicts and crises. Notes 49. Major, "Europeanization and Security Policy," 183-5; Moumoutzis, "Europeanization of Foreign Policy," 608. 50. Moumoutzis, "Europeanization of Foreign Policy"; for the application in other policy areas also see Moumoutzis and Zartaloudis, "Europeanization Mechanisms and Process Tracing". 51. Gilardi,"Transnational Diffusion," 459. 52. Beach and Pedersen,Causal Case Study Methods,273. 54. MEAE, "Prevention Resilience and Peace," 31. 55. Federal Foreign Office (FFO), "Preventing Crises," 111. 56. All three diplomatic services dispose of additional anticipatory analysis capacities, e.g. in the form of foresight and horizon scanning in policy planning units, which operate under a broader mandate. The French MEAE is also responsible for development policy, but the early warning system considered here had no link to this policy area. 57. EU, "Strategic Approach to Resilience," 3. 58. The German FFO defines resilience as "the ability of people and institutionswhether individuals, households, local communities or statesto cope with acute shocks or chronic stress caused by volatile situations, crises, violent conflicts and extreme natural events, and to adapt and recover quickly without compromising their medium and long-term outlook on life"; FFO, "Preventing Crises," 70. The French prevention strategy states that "Resilience, which concerns both states and societies, is defined as the ability to absorb and recover from shocks. Such capacity for resilience varies greatly, ranging from simply absorbing shocks to adapting and even going as far as transforming the system so that it can withstand any impact." And "in a broad sense since it encompasses not only the states' capacities to absorb but also to adapt and transform with a view to making sources of resilience sustainable and more robust", MEAE, "Prevention Resilience and Peace," 21-2. 59. MEAE, "Prevention, Resilience and Peace," 22; FFO, "Preventing Crises," 127. 60. FFO, "Preventing Crises". 61. See note 30 above for further reading. 62. EU, "Strategic approach to resilience," 19.