Copying in EU security and defence policies: the case of EUNAVFOR MED Operation Sophia

ABSTRACT

Why did the European Union (EU) launch a naval operation in response to the migration crisis, despite the humanitarian character of the challenge at hand, doubts about the effectiveness of a military response, and the EU’s traditional focus on civilian means? Integrating institutionalist theory and the literature on crisis response, this article argues that EUNAVFOR MED Operation Sophia is an example of copying through contingent learning wherein the EU’s response to the migration crisis was shaped by naval missions Operation Mare Nostrum and Operation Atalanta. While the former set a precedent for a naval response to migration in the Mediterranean, the latter provided an off-the-shelf institutional blueprint for the design and implementation of Operation Sophia. In a crisis situation characterised by high uncertainty, and with little time to rethink policies or to create new structures, EU political actors used contingent learning to quickly evaluate potential policy responses and institutional reforms, leading them to the decision to copy past institutional designs and practices previously considered successful. This finding has relevance beyond the case of Operation Sophia, as it contributes to a better understanding of why a particular type of policy or action is chosen in times of crisis and urgency.

KEYWORDS:

• European Union
• EUNAVFOR MED Operation Sophia
• institutionalism
• copying
• contingent learning

1. Introduction

Why did the European Union (EU) launch a naval operation at the height of the migration crisis in the Central Mediterranean? EUNAVFOR MED operation Sophia was established in June 2015 under the EU’s Common Security and Defence Policy (CSDP). Sophia is one of the few widely agreed-upon policies to respond to the migration crisis (Riddervold 2018b, Riddervold et al. 2020). The choice of a military response to a civilian crisis, however, is surprising given the EU's operational tilt towards civilian missions and its limited experience in naval deployments (Smith 2017). Previous studies of Sophia have shown that the EU’s decision to act in response to the migration crisis was driven by the humanitarian crisis unfolding at its borders (Riddervold 2018a, 2018b), rising concerns about a smuggling-terrorism nexus (Boșilcă 2017), the security imperative of border protection (Moreno-Lax 2018), and Italian security and humanitarian policy goals (Nováky 2018).

Yet explaining the EU’s preference for a particular type of policy, i.e. a military operation, still remains unaccounted for. After all, the EU has a wide toolbox at its disposal, most of which is linked to various civilian instruments. Moreover, neither the persistent blockage in the negotiations on the reform of the Common European Asylum System (CEAS) and the inability of member states to agree on asylum policies, nor humanitarian concerns related to the loss of life at sea can explain the EU's choice of a naval operation. Instead, the EU could have focused on revising and rethinking current reform proposals, launching civilian search and rescue (SAR) missions, or finding ways to protect asylum seekers reaching European soil. From a security perspective, a military mission is hardly the most efficient solution to this type of challenge either. The launch of a mission appears even more striking in light of the widespread criticism attracted by militarised maritime interventions and restrictive border policies in the Mediterranean, which more often than not have aggravated both the security and the humanitarian situation (Moreno-Lax 2018). As late as 2015, even the EU Military Committee questioned whether a naval mission would be an appropriate response to the crisis (Johansen 2020, pp. 185–186). Adding to this, the CSDP is often referred to as a least likely area of EU integration, with member states often struggling to find common ground (Jørgensen et al. 2015). So why then, in spite of the civilian and humanitarian character of the challenge at hand, doubts about a naval mission’s effectiveness, and the EU’s traditional focus on civilian means, did the EU nonetheless decide to pursue a military response?

Integrating institutionalist theoretical lenses and the literature on crisis response, this article argues that the EU’s response is an example of “copying” (Verdun 2015), wherein the EU’s response to the migration crisis was shaped in significant ways by predecessor missions at both national and EU levels: most prominently, Operation Mare Nostrum and Operation Atalanta. From these two operations, we see conscious emulation of practices as well as the copying of pre-existing institutions. Further developing Verdun’s (2015) concept of copying, we suggest that in a crisis situation characterised by high uncertainty, and with little time to rethink policies or to create new structures ex novo, EU political actors use contingent learning (Kamkhaji and Radaelli 2017) to quickly evaluate potential policy responses and institutional reforms, leading them to the decision to copy past institutional designs and practices previously considered successful. This finding, we argue, has relevance beyond Sophia, as it contributes to a better understanding of why a particular type of policy or action is chosen in times of crisis and urgency – something that is often the case with the EU’s event-driven foreign and security policy. With a new naval mission, Irini, already agreed (Council 2020), a systematic study of the processes leading to the development of Operation Sophia also has broader relevance for our understanding of how the EU designs and conducts such missions.

To substantiate our argument that copying is key to understanding Operation Sophia, the article proceeds as follows. We first provide a brief overview of the mission. Then, we introduce our analytical framework, discussing the concept of copying through contingent learning as a decision-making mechanism in times of crisis. The next section contains the analysis and is divided into three parts. The first part explores how the massive drownings near Lampedusa in April 2015 marked a critical juncture in the EU’s response to the migration crisis, facilitating the transfer of existing strategic and operational templates into new settings. The second part examines how the Italian operation Mare Nostrum served as an example of a naval response to a migration crisis. The third part then investigates the specific ways in which the EU counter-piracy practices and structures in the Indian Ocean have offered an institutional template for Operation Sophia in the Mediterranean, and have significantly shaped its format and focus. The concluding section summarises the main findings and discusses their empirical and theoretical implications. It must be underlined, however, that we do not assess Sophia’s impact and relation to the EU’s broader migration policies or the effectiveness or normative validity of Sophia. Other studies explore these aspects extensively. For instance, Cusumano (2019a) examines the organised hypocrisy caused by the tension between norms and interests in EU foreign policy and migration management; Aalberts and Gammeltoft-Hansen (2014) discuss the broader politics attached to SAR operations in the Mediterranean; Tazzioli (2016) and Tazzioli and Garelli (2020) critique the use of military operations in maritime migration governance; Little and Vaughan-Williams (2017) and Pallister-Wilkins (2015) analyse how humanitarianism supports deterrence-based discourses; Cusumano and Villa (2019) find that maritime rescue operations are not a pull factor for migration, while various NGOs have raised human rights concerns linked to the EU’s naval approach (Amnesty International 2017. Also see Riddervold 2018b).¹ Although all of these issues are key to understanding both the EU’s migration policies and the role of Sophia, our focus remains on explaining the EU’s choice of a military response and by this illustrating the relevance of copying in EU security and defence policies more broadly.

2. EUNAVFOR MED Operation Sophia

Operation Sophia was established on 22 June 2015 as the first EU maritime security operation in the Central Mediterranean. The core mandate was to “undertake systematic efforts to identify, capture, and dispose of vessels used or suspected of being used by smugglers and traffickers” (Council of the EU 2015). This type of disruption could be seen as a maritime implementation of processes of containment that also hampered migrant movement upon arrival in Europe (Tazzioli and Garelli 2018). While not explicitly included in its official mandate, the operation also conducted some limited number of SAR operations on the most trafficked maritime routes (Cusumano 2019b).

The operation was originally planned in four sequential phases. During the first phase, Sophia patrolled the high seas and gathered intelligence on smuggling activities and methods. The second phase distinguished between two different operational areas. During phase 2A, EU ships were allowed to board, search, and divert smugglers’ vessels on the high seas, subject to the conditions and limitations arising from applicable international law. Phase 2B would extend these tasks into Libyan territorial waters with an appropriate UNSC mandate, the consent of Libyan authorities, and a “legal finish”, i.e. handling suspected smugglers either by prosecuting them in another member state, or in Libyan courts. The third phase would expand the EU mandate to take “all necessary measures” against suspected vessels – including onshore in Libya. Finally, the fourth phase envisaged the withdrawal of forces and the completion of the operation. The lack of a legal basis however prevented the full implementation of the mandate, and Sophia only reached phase 2A of the mandate, in October 2015 (EEAS 2016).

In June 2016, the Council bolstered the operation by adding two supporting tasks: first, capacity building and training of the Libyan navy and coastguard (LNCG), and second, contributing to the implementation of a UN arms embargo off the coast of Libya (Council 2016). In July 2017, the mandate was further expanded to include surveillance activities and information gathering on the illegal trafficking of oil exports from Libya (Council 2017). Additionally, a new mechanism was set up to monitor the training of the Libyan forces, and information exchange on human trafficking with member states and law enforcement agencies (e.g. Frontex and Europol) was intensified (Council 2017).

After being extended three times successively, the operation was renewed for two additional six-month extensions but was drastically downsized through the suspension of its naval patrols (Council 2019) before being phased out on 31 March 2020. In February 2020, the EU Foreign Affairs Council agreed to replace Sophia with Irini, a new operation in the Mediterranean (Council 2020).

3. Crisis response and copying

Studies of crisis response often start from rationalist assumptions, expecting actors to choose the most efficient means to deal with a particular challenge or situation (for a discussion see Ansell et al. 2017). This, we argue, was not the case with Operation Sophia. Instead, drawing on institutionalist concepts and literature on crisis response, we argue that Sophia can best be understood as an example of copying. The climax of the humanitarian crisis in the Mediterranean functioned as a critical juncture, opening a window of opportunity for EU actors to suggest the development of a new policy. However, path-dependent policy choices have shaped the EU’s response, as its particular actions “were built on previous institutions or were inspired by structures that had been created before” (Verdun 2015, p. 231). In other words, rather than developing new policies and instruments from scratch in a time of crisis, the EU “copied” an existing successful policy tool to apply to a new situation. In the case of the migration crisis, the EU used the normative precedent set by the Italian Operation Mare Nostrum to pursue a naval response, while adopting the specific institutional structures of its own successful naval mission: Operation Atalanta, an anti-piracy mission. In this sense, the EU’s response to the migration crisis is reflective of Verdun (2015)’s account of the EU’s institutional response to the financial crisis.

Building on Mahoney and Thelen’s (2010) four ideal types of institutional change (“displacement”, “layering”, “drift”, and “conversion”), Verdun argues that various EU institutions created in response to the financial crisis are the output of a “copying” process, whereby new institutions are directly borrowed from existing institutions (2015, p. 232). We add to this argument by specifying the micro-mechanism by which actors decide to resort to such copying, drawing on the concept of contingent learning. We also contribute to the existing literature by showing that copying can be understood not only as the creation of new institutions by borrowing from earlier institutions but also as the establishment of new actions or policy tools based on existing ones. Crisis, where rapid response is needed and actors have limited opportunities to assess alternatives, makes copying an especially viable option in the event-driven security and defence policy areas.

More precisely, theoretically, our concept of copying rests on the combination of two key explanatory concepts of stability and change: path dependence and critical junctures. While critical junctures allow us to understand change – i.e. a decision on a new policy, institution or action – path dependence linked to copying through contingent learning helps explain why new decisions often emulate existing institutions and policies.

Institutional perspectives have traditionally focused on how institutions persist over time and on how behavioural patterns within such institutions are reproduced due to particular norms and structures (DiMaggio and Powell 1983). A key assumption is that institutions are “‘sticky’ or inertial, as expressed in the idea that institutions are path-dependent and reinforced through feedback effects” (Ansell 2020, forthcoming December 2020). Previous institutional choices and experiences are, alongside institutionalised socialisation and learning processes, key to understanding how and what new institutional decisions are made in a given situation (March and Olsen 1998, Olsen 2009). As a result, earlier decisions influence and constrain the range of options considered by the decision-making actors. Several studies have suggested that processes of path dependence and socialisation are important for understanding institutional choices in the EU’s foreign and security polices (see for example, Mayer 2008, Juncos and Pomorska 2020, forthcoming December 2020).

Institutions, however, are not immutable. Policy change and new institutional arrangements can emerge from “critical junctures”, defined as “relatively short periods of time during which there is a substantially heightened probability that agents’ choices will affect the outcome of interest” (Capoccia and Kelemen 2007, p. 348, emphasis in original). Critical junctures act as windows of opportunity that actors may use to suggest institutional or policy changes that might not be possible or likely under normal circumstances (Keeler 1993, Cortell and Peterson 1999). While external crises do not necessarily act as a critical juncture and do not always predict that policy or institutional reform will take place (Capoccia 2015), they can if politics dictates that leaders meet the moment with action. Oftentimes those most prepared to meet the moment are those with policy solutions readily prepared (Roos and Zaun 2016).

In the context of the EU, crises have often functioned as critical junctures enabling political actors to put forth new institutional and policy responses (Ansell et al. 2017, Riddervold et al. 2020). At the same time, due to path dependence, these suggestions are often affected by previous knowledge. The initial stage of crisis response, where action is required immediately, is an example of Kamkhaji and Radaelli’s (2017) concept of “contingent learning”. This process differs from traditional policy learning and isomorphism, including the policy learning often associated with critical analysis of the aftermath of crisis or disaster,² in that the requirement of immediate action means that many traditional forms of policy learning are unavailable to decision-makers. Crisis response is even more difficult when contending with transboundary crises, which further increase already heightened uncertainty (Ansell et al. 2010). As a transboundary crisis, there is often no obvious preset institutional “home” to manage the crisis response (Boin et al. 2013). The uncertainty of this situation thus makes copying more likely. Decision-makers often learn from existing models and historical analogies, though those decisions are filtered through individual cognitive biases (Levy 1994). These cognitive biases affect the perceptions of options available to a decision-maker in times of crisis (Kamkhaji and Radaelli 2017, p. 723). In line with this, and as we will see, in the case of the maritime response to the migration crisis, the possible solutions took the form of models already in use by the EU and one of its most directly affected member states, Italy, which sought leadership on the migration issue.

4. Analysis³

4.1. The Lampedusa shipwrecks as a critical juncture

Irregular migration in the Central Mediterranean had been a long-standing point of contention among EU member states (Klepp 2010, Aalberts and Gammeltoft-Hansen 2014). As migration increased following the Libyan civil war, rhetoric portraying migration as a crisis increased, beginning the process of mobilisation that would ultimately enable the EU to act (Jeandesboz and Pallister-Wilkins 2016). This framing had two tangible effects that impacted the policies ultimately pursued under Operation Sophia. First, migrants were framed as in need of rescue, leading to an emphasis on SAR missions. Second, migrants were depicted as a security risk (Guilfoyle 2016). Ultimately, these frames helped to mobilise support for a deterrence strategy, which could tackle both frames simultaneously (Dimitriadi 2014, Little and Vaughan-Williams 2017).

The death of more than 800 people in two consecutive shipwrecks near Lampedusa on 18 April 2015 functioned as a critical juncture in the EU’s decision to launch Operation Sophia. While the phenomenon of mass migration was not new and such occurrences were frequent, this episode was the deadliest single incident in the Mediterranean in decades. The shipwrecks opened up a brief period when new policy options that were previously unavailable now became possible. A number of member states, including Spain, Greece, and France, already supported closer integration in the field of maritime security. Yet other countries such as the UK and Germany had been more hesitant about a naval operation under the EU command. This changed, however, in the aftermath of the Lampedusa events. Prime Minister Cameron pledged to step up SAR efforts and called for additional measures to apprehend human traffickers (BBC 2015), while Chancellor Merkel provided reassurances that the government would “do everything” to prevent further loss of life at sea (Reuters 2015). As the then commander of Sophia summed up: “this very serious incident provided an impetus for the EU to take common action […]” (Senato della Repubblica 2016, p. 4). Tardy (2015, p. 1) similarly notes that the operation was set up “after a series of mass drownings”. Echoing these views, Blockmans (2016, p. 3) explains the surge of interest in migration issues in relation to “a rise in fatalities at sea”. Our interviews⁴ also suggest that the April shipwrecks, along with the intense media attention and public outcry that followed, acted as strong catalysts for an EU response.

During this window of opportunity, the EEAS, together with the High Representative of the EU/Vice-President of the Commission (HR/VP) staff in the Commission and supported by some member states, suggested the launch of a naval mission (Johansen 2020). And only two days after the shipwrecks, the Foreign and Interior Ministers convened in a special joint meeting adopted a set of immediate actions to respond to the migration crisis. These included, amongst others, the systematic capture and destruction of the vessels used by smugglers or traffickers, following the example set by Operation Atalanta in the Gulf of Aden (European Commission 2015). The proposal was endorsed by an extraordinary European Council meeting, where the HR/VP was directed to immediately prepare plans for a “possible CSDP operation” to disrupt migrant smuggling networks (European Council 2015). The decision-making process leading to the operation unfolded at an unprecedented pace. Normally, EU decision-making is slow, and even if the CSDP is largely event-driven, the decision to launch a mission often takes many months. Instead, the practical launch of an operation which would have normally required “between four and six months” of preparations, was agreed “within a week” by 28 member states with “initially very different positions on the subject” (Senato della Repubblica 2016, p. 5). The operation was established on 18 May and officially launched just over a month later, on 22 June. In the word of one of our interviewees “from that moment, a real marathon started so that the operation’s naval task force […] would reach full operational capacity on 27 July”.⁵

To make these rapid developments possible, however, the normal planning process was “fast tracked” by deliberately leaving out certain steps to enable the EU to deploy forces on short notice (Nováky 2018, p. 205). Moreover, Italy – which had been advocating for an EU counter-smuggling naval operation since 2013 – and HR/VP Mogherini pressed hard for an accelerated procedure (Nováky 2018).

Interviews⁶ indicate that with little time available to plan the operation, EU policy-makers often fall back on familiar solutions and previous expertise. As argued by one naval officer at the Operational Headquarters (OHQ) of Sophia, “the experience gained by the EU in establishing Atalanta was useful during the initial stages of Sophia, when the latter operation was started at short notice”.⁷ In establishing a mission in this situation, the EU turned to its own and its member states’ experiences. And the two most salient familiar solutions, each of which Sophia drew upon, were the Italian naval Operation Mare Nostrum and the EU-led Operation Atalanta. Mare Nostrum had already operated successfully in the same waters, while Atalanta, the EU's anti-piracy naval operation, had deterred and disrupted Somali-based piracy, and had protected shipping off the Horn of Africa and in the Western Indian Ocean.

4.2. From Mare Nostrum to Operation Sophia

Existing political debate in Italy had already linked migration and security issues long before the emergence of the critical juncture at the EU level around the migration crisis (Angelescu 2008, Cusumano 2019a, 2019b, Panebianco 2016). The Italian government quickly implemented Operation Mare Nostrum in October 2013 after a separate large loss of life near Lampedusa. Mare Nostrum was primarily a naval operation, as opposed to being run by the Italian Coast Guard, setting a precedent followed by Sophia. Naval vessels patrolled well beyond Italian territorial waters with an explicit SAR mission, ultimately rescuing over 150,000 migrants (Carrera and den Hertog 2015). While the Italian government had maintained seaborne rescue missions since 2004, Mare Nostrum represented a large increase in the budget and capacity of Italian rescue operations and proved unsustainable for the Italian government (Patalano 2015). Costs eventually exceeded €9 million monthly, well beyond the initial budget, and were borne almost entirely by Italy.

Given the quickly increasing costs, Italy sought an EU-level response and terminated Mare Nostrum at the end of October 2014. The EU initially responded with Operation Triton, operated by the European Union’s border control agency, Frontex. This mission was pushed for by the Italian government, which also developed its own successor naval-led military mission, Mare Sicuro, at the national level, but with reduced costs and scope (Dibenedetto 2019). Triton, similarly, was narrower in scope than Mare Nostrum, even with the involvement of other EU countries. Much like Mare Nostrum, Triton was criticised as a potential pull factor for migrants, although unlike Mare Nostrum it lacked an active SAR component and focused exclusively on border surveillance. The British Foreign Office, for instance, had announced that the UK would not participate in Triton fearing an “unintended pull factor”, whereas the German Interior Minister had felt dubious about the Italian Operation Mare Nostrum, seen as a “bridge to Europe” (DW 2014). Triton also had a reduced geographic scope compared to Mare Nostrum and a smaller budget (though this budget was increased following the April 2015 critical juncture). While Triton is a more direct successor mission on paper, albeit with much smaller scope, copying from Mare Nostrum clearly influenced the decision to launch Operation Sophia.

The Italian government sought a CSDP solution as a replacement for Mare Nostrum as early as 2013, though it was unsuccessful at persuading other member states (Nováky 2018, p. 201). Ultimately, as the politics of the Central Mediterranean migration crisis changed following the April 2015 critical juncture, this Italian goal was realised. The Italian experience of a naval mission to respond to the migration crisis made this an easy solution for the EU to choose in a crisis situation, especially given the prominence and well-publicised successes of Mare Nostrum. As argued by Johansen (2020, p. 172), “the Lampedusa accident created a new momentum for the Italian proposal for a CSDP naval operation”. Following the accident, a 10-point action plan was quickly drafted by the HR/VP together with the Commission, where “the Italian naval operation re-emerged” (Johansen 2020). And this time, due to the Lampedusa events, the Ministers of Foreign Affairs felt pressured to respond immediately, rapidly accepting the proposal that was on the table (Riddervold 2018b).

This is not to say that all aspects of Mare Nostrum were carried over. Mare Nostrum focused largely on SAR activities, while Operation Sophia had a legal basis in anti-smuggling agreements – though rescue was still a component as required by international maritime law (Gauci and Mallia 2017). However, while this was a legal shift, it followed an increasing use of the rhetoric of containment and securitisation, targeting boats carrying smugglers in the hopes of reducing the need for rescue (Tazzioli 2016). More importantly, the clearest parallel between Mare Nostrum and Sophia was the duplication of the operation’s command structure. Sophia was likewise structured as a naval operation, using larger vessels and extending well beyond domestic territorial waters (Boșilcă 2017, Moreno-Lax 2018, Nováky 2018).

4.3. From Operation Atalanta to Operation Sophia

The pre-existence of a naval SAR mission in the Mediterranean influenced the EU’s choice to launch a naval mission in the region. Once the Lampedusa events created a sense of urgency, the HR/VP and the Commission were quick to suggest a new mission based on the previous operation in the area. Sophia was carried out in the same waters, and some of the same ships joined the new EU mission.

At the same time, however, the operation and institutional setting of Sophia can be traced back to Atalanta. While Mare Nostrum created a precedent for a naval response to migration, Atalanta provided the institutional template that the EU drew on when designing the mission: when looking for a proposal, the EU institutions emulated past institutional designs and practices considered successful.

Fighting piracy, however, is operationally very different from countering smuggling: whereas pirates make their profit at sea, traffickers and smugglers make theirs on land, irrespective of the fate of the boats. Destroying the vessels once they set sail thus not only failed to disrupt smuggling networks or reduce the number of arrivals but also resulted in more dangerous crossings at sea (UKHL 2017). Unlike Atalanta, Operation Sophia also lacked a UN mandate to operate in the Libyan territorial waters; thus, its effectiveness was increasingly called into question.⁸ Another major difference between the two operations was the “legal finish”. EU personnel lacked legal means to transfer apprehended suspects to local authorities, as the legal mandate applied to piracy but not human smuggling.⁹ For these reasons, military experts and personnel were sceptical of the idea of deploying military means to deal with the migration crisis, with the EU Military Committee even opposing the HR/VP, Commission, and EEAS suggestion (Johansen 2020). As argued by one of our interviewed naval officials: “They looked at Atalanta as a model […] but there were big differences between the two operations which meant that the success of Atalanta could not be copied to Sophia”.¹⁰ As feared by military and other experts, Sophia encountered many different challenges, both in terms of refugees rights and efficiency (FCO 2017). A strategic review of the operation for example showed that Sophia had conducted “more than 1,700 hailings and more than 100 friendly approaches”, but had inspected only “three vessels under the provisions of UNSCR 2420, which led to seizures of prohibited arms”, suggesting rather modest results (EEAS 2018, p. 5). Nonetheless, and contrary to what one would expect if the decision to launch a naval mission were based on well-thought-out efficiency considerations, the institutional setting of Atalanta served as a model for Sophia. We see these similarities across all the main aspects of the mission – (i) the military/naval response, (ii) coordination and maritime security cooperation, and (iii) the training and capacity building dimension.

1. The naval/military response: Pirates, smugglers, and terrorists

First, in both missions, the military focus was on deterrence and prevention. In the national and EU discourses and policies, both piracy and irregular migration have increasingly been addressed in a transnational security frame, launched as actions developed to deal with particular challenges (Guilfoyle 2016, Roberts 2018). Piracy has been framed as a menace to the collective interest in upholding the freedom of navigation; in turn, the rising number of asylum seeker boat arrivals into Europe was seen to erode national security and state sovereignty (Guilfoyle 2016). Both phenomena were thus cast as security threats to the EU’s seaspace and maritime borders, requiring robust interventions and coercive actions to prevent and/or curb migratory influxes (Dombrowski and Reich 2019).

The introduction of more restrictive measures and the use of military force gained increasing political traction after officials expressed concerns about an alleged nexus between Somali piracy and irregular migration on the one hand, and terrorism on the other.¹¹ Similar issue-linkage patterns in the EU response to the migration crisis emphasised migrant smugglers facilitating unlawful cross-border movements for profit and their potential involvement in other forms of transnational organised crime or terrorism. ‘The UNSC Resolution 2240 (2015), which authorised the interception of suspected vessels off Libya’s coast, flagged migrant smuggling and human trafficking as threats potentially supporting organised crime and terrorist activities in Libya. Similarly, UNSC Resolution 2292 (2016), which allowed for the inspection of suspected embargo-breaking vessels off Libya, highlighted the dangers of smuggled arms falling into the hands of terrorist groups. Against this backdrop, the British, French, and Spanish officials reinforced the notion of a “terrorism – arms proliferation – migrant smuggling” nexus and agreed to expand the mandate of Sophia (UN Press 2016). At the EU level, the six-month report of the operation called for a clear understanding of how smuggling and trafficking networks worked, including “[…] where they might interact with other illegal and terrorist organisations” (EEAS 2016, p. 17). To counter potential terrorist threats, states thus had to balance humanitarian tasks with beefed-up security measures (Sénat de la France 2015). These actions, however, targeted not only smugglers but extended de facto to the victims of smuggling or trafficking themselves (Moreno-Lax 2018, Stierl 2019).

Consequently, military efforts in both operations have focused on deterrence and disruption strategies aimed at preventing or containing the movement of pirates or irregular migrants. Atalanta’s ultimate rationale was to prevent and avert piracy as this impeded the free flow of humanitarian aid and merchant shipping. For this purpose, naval vessels patrolled major sea lanes and used “threats of military force, apprehension, and prosecution” (Gilmer 2014, p. 22). Disruption, aiming at interrupting piratical activity, occurred both at sea through the destruction of pirate vessels, the confiscation of ammunition, and the apprehension of suspects, as well as on land, through “over the beach” operations (Gilmer 2014).

In a similar vein, Operation Sophia mainly focused on disrupting the “business model” of migrant smuggling, seen both as a criminal enterprise and a push factor. As emphasised by the Deputy Commander (2019) of the operation: “All aspects of Sophia are devoted to fighting trafficking […]; we rescued lots of people, but not as part of the mandate”. The use of naval assets reflected this ranking of priorities: instead of dispatching coast guard vessels focusing on SAR, member states relied on frigates, which were better equipped for gathering intelligence and dismantling trafficking networks.¹² Moreover, this is also made evident in the operation’s strategic review, which summarised its primary results as “the apprehension of 148 suspects” and the “neutralization of some 550 assets” (EEAS 2018, p. 5).

• (ii) Coordination and maritime security cooperation

Another area in which the EU has visibly tapped into existing counter-piracy governance mechanisms and expertise to respond to challenges posed by migration is maritime security cooperation. Two policy forums in particular stand out: the Contact Group for Piracy off the Coast of Somalia (CGPCS) and the Shared Awareness and Deconfliction (SHADE) mechanism in the Indian Ocean, and their corresponding structures in the Mediterranean, respectively: the Central Mediterranean Contact Group (CMCG) and the SHADE MED.

The CGPCS was established in January 2009 as a voluntary, informal, and inclusive framework for consultation and coordination on counter-piracy policies among more than 50 states and international organisations. The open architecture of the format has enabled the EU, especially EUNAVFOR representatives, to play an important role in the forum. The EU has regularly taken part in the group’s biannual plenary where participants discussed and agreed on common strategies to combat piracy, as well as in its specialised and technical working groups (Mitcham 2018). Moreover, from January 2014 to January 2016, the EU was entrusted with the CGPCS chairmanship. This mandate, carried out jointly by the European Commission and the EEAS, offered unprecedented opportunities for the EU to steer global counter-piracy initiatives by setting the agenda, shaping negotiations, and drafting the group’s communiqué (Bueger 2016). This meant that the Union could now draw on its newly acquired experience to engage in some degree of learning about how to tackle maritime insecurity more effectively.¹³

By comparison, no inclusive and flexible mechanism or dedicated forum existed to systematically address these topics in the Mediterranean (Seyle and Madsen 2016). The initiative to establish a new venue for cooperation came from the Italian Interior Minister in March 2017, after the adoption of the Malta declaration on the external aspects of migration and the conclusion of a memorandum of understanding between Italy and Libya. Both documents called for intensified cooperation with Libya and neighbouring North African and sub-Saharan countries to facilitate the exchange of information and disrupt smuggling networks on the Central Mediterranean route (Council 2017). To this end, the Home Affairs ministers of six member states (Italy, Malta, France, Germany, Austria, and Slovenia), plus Switzerland, Libya, and Tunisia, decided to set up a Contact Group in the Central Mediterranean (CMCG 2017). The composition of the new forum included the ministers of interior of participating countries, as well as the HR/VP and the European Commissioner for Migration, Home Affairs and Citizenship.

Notwithstanding various differences in terms of size, diversity of membership, or complexity of format, the purpose and functioning of the CMCG were largely inspired by the logic of the CGPCS.¹⁴ First, like the CGPCS, the CMCG sought to coordinate responses to a common problem characterised by complexity and urgency. Through its coordinating role, the CGPCS had an essential contribution to the success of counter-piracy actions and thus provided a familiar crisis-coping mechanism.¹⁵ Second, the Contact Group preserved the CGPCS’ flexible membership and agenda: interested EU states and partner countries were welcome to join the group at any point, while priorities for common action were addressed on an ad hoc basis. Third, the CMCG functioned through a series of flexible high-level meetings, during which participants set the agenda. Fourth, as in the case of the CGPCS, the EU secured an important position in agenda setting and consultations in the CMCG through the participation of its high-level officials. Nonetheless, despite early enthusiasm, the CMCG never equalled the same longevity or success as the CGPCS, and after only three meetings, discussions were moved to other forums.¹⁶

Another telling example of institutional “copying” is offered by the SHADE mechanism of maritime security regional cooperation. Originally established in early 2009, SHADE conferences were organised biannually in Bahrain to coordinate naval counter-piracy forces and develop a common understanding of maritime (in)security. The command of EUNAVFOR Atalanta in particular had a key role in the establishment and functioning of SHADE, co-chairing the forum’s meetings on a rotational basis. EU officials were also active in mobilising support for enhanced cooperation and information sharing among various independent deployers (Forbes 2018).

SHADE was widely seen as a “major success” and one of the most important outgrowths of the CGPCS (Houben 2014, p. 28). Three main factors accounted for its success. First, the informal and apolitical character of the venue eschewed cumbersome processes and political sensitivities, allowing participants to focus entirely on building maritime awareness. Second, SHADE was inclusive, bringing together numerous participants, including force-providing states, multi-national military operations, international and regional organisations, and maritime industry representatives. Third, actors focused on the exchange of tactical and operational information, leading to concrete and tangible results (Houben 2014, p. 29).

Building on this example, SHADE MED was launched in November 2015 at the initiative of the Italian Navy and has been held biannually since then. It acts as a forum where national representatives and organisations affected by the migratory phenomenon in the Mediterranean meet to de-conflict and coordinate their efforts by sharing situational awareness, trend assessment, and best practices (EUNAVFOR MED 2016). The key role of the Italian Navy in SHADE MED is further indicative of the additional importance of Mare Nostrum as an antecedent. The decision to draw on existing structures was based on two main factors. First, the SHADE mechanism provided an easy fix that could be imported to respond to the migration crisis. As stated by a high-ranking naval official: “one of the lessons from Atalanta was to have a stakeholders meeting where people from all the operations involved come together and share information, and that happened a year later for Sophia”.¹⁷ Second, the idea of SHADE MED travelled fast via the Italian command of both operations. As summarised by an EEAS official: “it is impossible for Atalanta not to be a model for Sophia because almost every naval officer in Europe has been in Atalanta […] Atalanta is the benchmark”.¹⁸ The operational commander of Sophia further explained:

I first saw this concept when I worked in Operation Atalanta as the Force Commander at the European Naval Forces for the Counter Piracy mission off the coast of Somalia. I saw the need to bring the different organizations that operated in the Mediterranean, along with those who have a broader interest in irregular migration to discuss issues. (Credendino cited in Cassidy 2017, p. 17)

The two SHADE mechanisms, however, are not identical: whereas SHADE Bahrain is “focused at the tactical and operational level […] and promotes de-confliction between actors that are controlling units in the region”, SHADE MED “has an outreach and strategic level focus, and promotes cooperation between disparate organisations, rather than the physical de-confliction of units operating off Libya”.¹⁹ Despite these differences, SHADE has grown into a strong “brand” representing “a useful convening factor [that] has become synonymous with a collaborative approach to address regional issues”.²⁰

The structure and functioning of SHADE MED were heavily inspired by the anti-piracy policy forums in at least three respects. First, SHADE MED preserved the voluntary and non-binding character of the initial SHADE setting, which lacked legal authority but offered a venue for shared understanding and practical harmonisation (EUNAVFOR MED 2016). Second, just like SHADE Bahrain, SHADE MED encouraged the inclusive participation of national representatives, regional organisations, shipping and industry groups, law enforcement agencies, and military delegates, amongst others, in its meetings (EUNAVFOR MED 2016). Third, similar to the CGPCS, SHADE MED’s activity was organised around plenary sessions and thematic work groups. For example, a communications and information systems group inside SHADE MED was formed to develop a SMART system designed to facilitate communication between military, shipping, and NGO rescue boats, based on the model provided by Atalanta's Mercury information sharing system.

• (iii) Training and maritime security capacity building

Lastly, the EU has increasingly sought to emulate its comprehensive approach to better deal with migration issues, explicitly drawing on operation Atalanta. One of the most frequently cited factors contributing to the success of Atalanta was the EU’s so-called “comprehensive approach” in the Horn of Africa, which combined civil and military strategies to address the structural causes of instability, rather than only addressing superficial symptoms (UKHL 2017). Building on this experience, there were expectations that the EU could import a similar model into Libya.²¹

Indeed, both operations adopted similar “failed/failing” state narratives, which portrayed maritime insecurity as a result of insufficient regulatory and law enforcement capacity. Just as the root causes of piracy reflected Somalia’s protracted instability, irregular migration in the Mediterranean ultimately stemmed from the inability of the Libyan state to control its maritime borders. As outlined by a naval officer at the OHQ in Rome:

Both operations were initially established to address the symptoms of wider problems, which were presenting themselves at sea. They were both originally maritime operations, which then evolved in an attempt to address the roots causes of the problems: failed country, lack of governance, and the inability to cope with internal insecurity […]. Determining how to apply military force to disrupt that model was a challenge but the overall mission concept was the same in both operations, as was the desired end state.²²

In Somalia, the EU supplemented Operation Atalanta with two follow-up missions: EUTM Somalia, focused on training, mentoring, and advising activities, and EUCAP Nestor/Somalia, a civilian maritime capacity building mission. Unsurprisingly, Sophia reproduced this tripartite blueprint almost identically by incorporating military training and capacity building into its mandate via two supporting tasks, namely the training of the LNCG, and implementing the UN arms embargo on the high seas off Libya (Tomic 2017). Just as EUTM in Somalia, the training of the LNCG was a crucial part of the EU’s “exit strategy” as revealed by the operation’s six-month report (EEAS 2016, p. 22). Joint training exercises were expected to build confidence and thus enable a swift move to the next phases of the operation, which required the consent of Libyan authorities (EEAS 2016).

5. Conclusion

This article set out to explain how we can understand the launch of an EU military operation in response to the migration crisis. Empirically, we have shown that the EU’s policies, practices, and discourses surrounding its fight against migrant smuggling were inspired by its previous anti-piracy experience and by the Italian naval mission, Mare Nostrum. While the critical juncture resulting from the April 2015 shipwrecks near the island of Lampedusa galvanised member states into action and facilitated consensus on the need to take action, EU policy-makers had little time to devise new solutions and turned instead to readily available templates for action on which member states could agree in record time. EU institutions like the Commission, the HR/VP, and the EEAS – together with member states such as Italy – played a critical role in tabling a proposal for an EU-led mission. As shown, their choice of response and their rapid agreement on launching a naval operation can be explained through the process of copying. While Operation Mare Nostrum opened the path for a naval response to migration in the Mediterranean, Operation Atalanta provided a template on which Operation Sophia was planned and deployed.

Analytically, we argue that this decision can best be described by a revised version of the concept of “copying” (Verdun 2015), whereby actors in crisis situations, or other situations characterised by a feeling of urgency, use contingent learning (Kamkhaji and Radaelli 2017) to quickly evaluate potential policy responses and institutional reforms, leading them to establish new institutions or policy actions by borrowing in some way from already existing institutional configurations, policies, or practices.

By teasing out the importance of copying for understanding the EU’s puzzling decision to launch a military mission in response to a civilian and humanitarian crisis, the article also contributes to a better understanding of the mechanisms underlying policy developments in the area of CSDP. Rather than suggesting that the EU selects the most efficient means to deal with a particular crisis, the case of Operation Sophia illustrates how the EU’s choice of policy instrument was influenced by path dependence in connected fields, where contingent learning led to copying previous decisions and institutional structures. While further study is needed, our findings also suggest that such copying may be especially likely when the EU is faced with situations characterised by crisis or a sense of urgency, which is often the case in the EU’s largely reactive security policies; while critical junctures may allow for new possibilities of action, the urgency of crisis can limit the scope of what those actions may be. In times of crisis, decision-makers thus draw on tried and tested institutional structures and past knowledge to formulate solutions to current problems. Copying may be an understudied mechanism that, further explored, can yield important insights into new actions and integration in EU security policies.

Acknowledgements

We want to thank Chris Ansell and three anonymous reviewers for valuable and helpful questions and suggestions. A previous version of this paper was presented at the NEON conference in Lillehammer, Norway, in November 2019. Thank you to the participants for comments and questions. A special thanks also to our interviewees for sharing their insights and ideas.

Disclosure statement

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

Notes

1 We are grateful to an anonymous reviewer for reminding us to make this clarification and linking to this debate.

2 Even in the Mediterranean we see examples of this more conventional learning in response to crisis. Cusumano (2019a) describes how NGOs developed two separate strategies through selective emulation to respond to restrictions placed on SAR missions.

3 The analysis draws on three main sources. First, it examines EU policy documents such as EEAS strategic reviews, EUNAVFOR mission factsheets, European Council decisions, High Representative of the EU/Vice-President of the Commission (HR/VP) official declarations, and, whenever available, leaked working documents on sensitive or less publicly debated aspects of Operation Sophia from 2015 until 2020. Second, it builds on the existing secondary literature on the CSDP, the migration crisis, and the EU response to piracy and migrant smuggling in particular. Third, the article draws on 17 in-depth, semi-structured interviews with EU officials involved in the planning and conduct of operations Atalanta and Sophia, conducted in person in Brussels or via e-mail/Skype between September 2017 and February 2020.

4 Interview (17) with official from OHQ of Operation Sophia, e-mail, February 2020; interview (14) with naval officer, Skype, October 2019; interview (15) with naval officer, Skype, October 2019.

5 Interview (17) with official from OHQ of Operation Sophia, e-mail, February 2020.

6 Interview (1) with EEAS official, Skype, September 2017; interview (2) with EU diplomat, Brussels, September 2017.

7 Interview (17) with official from OHQ of Operation Sophia, e-mail, February 2020.

8 Interview (15) with naval officer, Skype, October 2019.

9 Interview (9) with EEAS official, Brussels, December 2019.

10 Interview (16) with naval officer, conducted by Dr. John Sherwood, Naval History and Heritage Command, Skype, October 2019.

11 Interview (2) with EU diplomat, Brussels, September 2017.

12 Interview (16) with naval officer, Skype, September 2019.

13 Interview (9) with EEAS official, Brussels, December 2019.

14 Interview (12) with European Commission official, Brussels, June 2019; interview (13) with European Commission official, Brussels, July 2019.

15 Interviews (10-11) with EEAS official, Brussels, December 2019.

16 Interview (5) with EU official, Skype, December 2018.

17 Interview (16) with naval official, Skype, October 2019.

18 Interview (4) with EEAS official, Brussels, June 2018.

19 Interview (17) with official from OHQ of Operation Sophia, e-mail, February 2020.

20 Interview (17) with official from OHQ of Operation Sophia, e-mail, February 2020

21 Interview (3) with EEAS official, Brussels, June 2018.

22 Interview (17) with official from OHQ of Operation Sophia, e-mail, February 2020.