Cooperative bargaining in the EU's common security and defence policy: EUNAVFOR Atalanta

ABSTRACT

This article examines the EU's decision-making on military operations to assess the contested nature of the EU as an international security actor. It investigates both the decision-making process on the launch and subsequent development of EUNAVFOR Atalanta, from 2008 to 2014. First, empirically, the article, based on policy documents and semi-structured interviews, unpacks the contestation concerning the use of military force in the EU by distinguishing between the justification of military force and the policy embeddedness of military force at the more operational level. Second, it shows how the strategic interaction among different advocacy coalitions, which is characterized by a process of cooperative bargaining, explains the particular development of EUNAVFOR Atalanta in terms of its justification and policy embeddedness.

KEYWORDS:

• EU Security and Defence
• cooperative bargaining
• advocacy coalitions

Introduction

In 2008 the EU launched EUNAVFOR Atalanta to fight piracy off the Somali coast. As the EU's first maritime operation, it attracted considerable attention from EU foreign policy scholars to assess whether this entailed a qualitative shift in the EU's Common Security and Defence Policy (CSDP). It has been argued that the EU kept its normative foreign policy as it did not prioritize the protection of European ships and bound itself to global law regarding the treatment of pirates (Riddervold, 2011). Others, however, point at Atalanta as a shift away from the EU's ‘normative’ legitimation of military force by privileging the maximization of (economic) security (Norheim-Martinsen, 2013, p.165; see also Weber, 2009).

This reflects a wider debate on the character of the EU's Security and Defence Policy (see Whitman, 2011). This article aims to make an empirical and theoretical contribution to this debate by answering the question: How to explain the changing character of EUNAVFOR Atalanta? Empirically, it unpacks the contestation concerning the use of military force in the EU and problematizes existing accounts of the normative shift EUNAVFOR Atalanta by highlighting its paradoxical nature. It does so by distinguishing between the justification of military force and the policy embeddedness of military force at the more operational level. While EUNAVFOR Atalanta indicates a shift away from the EU's ‘normative’ justification of military force, i.e. the notion of economic interests is explicated and becomes more prominent as the operation develops, at the same time the operation is increasingly embedded in an overarching political framework that aims to provide a more fundamental solution to piracy that moves away from a purely military approach. Theoretically, the article aims to show that cooperative bargaining among different advocacy coalitions explains this apparent contradictory development. As such the paper is able to reflect upon the extent to which EUNAVFOR Atalanta is the manifestation of an incidental or rather structural shift in the EU's Security and Defence Policy.

The article proceeds as follows. The first section introduces ‘cooperative bargaining’ as mechanism that could account for the interaction between actors in the decision-making process on the launch and development of EUNAVFOR Atalanta. It then proposes an advocacy coalition approach as a heuristic device to study this interaction. The empirical part of the paper consists of two parts. After a short description of the launch of Atalanta in 2008, the justification of different advocacy coalitions for launching Atalanta is discussed, and the relative influence of different coalitions on the character of Atalanta and the underlying decision-making mechanisms is examined. Subsequently the paradoxical development (until 2014) of Atalanta in relation to the coalition dynamics is discussed. Finally, the paper concludes with a discussion of the main findings and reflects on its wider implications for the study of the EU's multilevel foreign policy-making in a changing global political environment.

Decision-making mechanisms & advocacy coalitions

This article aims to identify the mechanisms that drive the activation and organization of military operations by the EU. To this end, three explanatory mechanisms are discussed, based on different institutionalist approaches: cooperative bargaining, hard bargaining and institutional learning.¹ Moreover, the EU's decision-making on military operations is conceptualized in terms of advocacy coalitions with considerable different ideas on the use of military force.

Well-established theories in both International Relations, EU-studies and Institutionalism emphasize the importance of power politics to account for the evolution of (EU) foreign policy. While realism, liberal intergovernmentalism and rational-choice institutionalism focus on different levels of analysis, they share the assumption of rational actors making cost-benefit calculations to achieve a fixed set of preferences and an emphasis on power differences among different actors (see Delreux, 2015). From this follows the expectation that decision-making will be characterized by hard bargaining. Since, it assumes that actors’ preferences and ideas are constant, i.e. not changed by the interaction with other actors, the decision-making processes would be dominated by the most powerful actors, which push for their particular preferences, or ideas, and are not afraid of blocking decisions that are not in line with these preferences (see Hyde-Price, 2013). This results in the lowest common denominator or a deadlock (non-action) (Ibid.).

In contrast, sociological institutionalism and constructivist approach highlight the dynamic nature of preferences. One of the key mechanisms that is able to account for policy changes as a result of changing preferences and ideas is institutional learning. Institutional learning focuses on the collective learning processes, across coalitions. Moreover, for the purpose of this paper, institutional learning is understood to go beyond procedural and organizational learning and refer to instrumental and ideational learning (see Hall, 1993). Central to institutional learning is the expectation that with increasing common experiences the divergent ideas of actors will converge (Levy, 1994; see Smith, 2012 for application to CSDP). It highlights the role of professional expertize and epistemic communities (see Cross, 2013). As such the decision-making dynamic is depoliticized and a consensus is expected.

In contrast to these two ‘extremes’, the expectation of this paper is that the EU's decision-making on Atalanta has been a process of cooperative bargaining. This mechanism is derived from Thomas’ normative institutionalist approach, in which EU actors are conceptualized as ‘thinly-socialized actors who are affected by substantive and procedural norms of the EU’ (Thomas, 2011, p. 5). From this perspective, the central mechanism that connects the diversity in actors’ preferences to a common outcome is cooperative bargaining, which results from both normative (substantive) and institutional (procedural) considerations. This cooperative bargaining is characterized by give-and-take and mutual compromises (Ibid., p. 19). Such an understanding points at a situation in which the actors involved prefer a compromise over defending their preferences with a veto (threat). Divergent preferences have not disappeared, but they are softened by the consensus norm and consultation reflex (Ibid.). In sum, cooperative bargaining concerns a decision-making process that is characterized by (continuing) different preferences, combined with the political will to go beyond the lowest common denominator. It refers to the situation in which no veto-threats are used despite divergent preferences.

While the mechanisms do not logically exclude each other, analytically they are distinctive; they relate to different empirical observations (see Table 1). From the logic of hard bargaining it is key to look at the power resources of different coalitions, and their use of it. Moreover, hard bargaining becomes visible in the actual, or threat to, use veto power, and in the case of military operations, the unilateral withdrawal of troops.

Table 1. Operationalization of mechanisms of change.

Mechanism	Policy outcome	Indicators
Hard bargaining	Lowest common denominator/deadlock	Power resources of different coalitions Actual us or threat of veto Unilateral withdrawal of troops
Institutional learning	Convergence of policy positions	Common assessment of shared experience Processes of lessons learned, evaluation and review Important role for (military) experts
Cooperative Bargaining	Compromise beyond the lowest common denominator	Expanding ‘the pie’ Explicit acknowledgement of the preferences of others Positive identification with the process

Empirical observations in line with institutional learning are a common assessment of shared experience, processes of lessons learned, evaluation and review, and the central role for (military) experts.

Evidence for cooperative bargaining is the explicit acknowledgement and accommodation of the preferences and ideas of other coalitions, a positive identification with the process and actions that expand ‘the pie’, i.e. strategies to include additional elements in a mandate to satisfy different coalitions.

Although we may find empirical observations related to all three mechanism, it is expected a particular mechanism stands out for driving that change – acknowledging, however, that different mechanisms may relate to different dimensions of change. For example, a change in justification may relate to the power politics among coalitions, while a more operational change in the level of policy-embeddedness of an operation may be driven by a process of cooperative bargaining.

To assess the relative value of those different mechanisms, this study triangulates interviews,² policy-documents, parliamentary debates, newspaper articles and secondary literature.

To study the decision-making process in the EU's Common Security and Defence Policy, this paper uses advocacy coalitions (inspired by Sabatier’s (1988) advocacy-coalition framework) as a heuristic device. An advocacy coalition framework allows to study the (strategic) interaction of different groups of actors which share a common belief system, i.e. a set of value priorities and causal assumptions about how to realize them. As such it serves as a tool to get grip on the logic of decision-making, as it incorporates power dynamics among coalitions and highlights the ideational structure on which coalitions are based.

The advocacy coalitions are situated in a context where resources (e.g. material capabilities, institutional and discursive resources) and (institutional) constraints impact on the potential for success of the different coalitions.

An advocacy coalition approach is open to a variety of actors and does not take the objectives of actors as given (Sabatier & Weible, 2007). In line with this more open approach, this study starts from the different belief systems that are out there concerning the use of military force in the EU-context. These different coalitions involve a range of national governments, but also include EU-institutions (EEAS, Commission, Parliament).

Building upon the work of Vennesson (2010) and Meyer (2006), four coalitions can be identified in the Common Security and Defence Policy, involving Member States and EU-institutions, which share policy core ideas concerning foreign policy and, particularly, the use of force: Global Power EU, Euro-Atlanticists, Human Security, Bystanders. These four coalitions are based on key dimensions related to the use of military force at the EU-level: institutional preference for either EU, UN or NATO, emphasis on the military instrument relative to other foreign policy instruments (policy embeddedness), purpose and justification of the use of military force.

The launch of EUNAVOR Atalanta

In meetings of the ministers of Foreign Affairs of November and December 2008 the upswing of violence in Congo was discussed and the Joint Action for EUNAVFOR Atalanta was adopted. The letter of the UN Secretary General requesting for an EU military force in Congo was ‘noted’, but not answered. Rather, the Council conclusions call for a rapid ‘technical, humanitarian and political’ response (Council of the EU, 2008d).

The fight against maritime piracy in the Indian Ocean originated from a concern with the pirate attacks on World Food Programme ships. Already in 2005 and 2006 the United Nations Security Council (UNSC) Presidential Statements called attention to the threat of piracy to humanitarian access (UNSC, 2005, 2006). In October 2007, after a ‘renewed rise in attacks’, the International Maritime Organisation (IMO) and the World Food Programme (WFP) urged for coordinated action to tackle piracy off the Somali coast (WFP, 2007). From November 2007, France (Operation Alcyon), Denmark, the Netherlands and Canada escorted aid ships as piracy increasingly threatened the delivery of food rations for 1.2 million people in Somalia. In May and June 2008, the UN Security Council adopted two resolutions calling for action to protect shipping involved in delivering humanitarian aid (1814; UNSC, 2008a) and commercial maritime routes (1816; UNSC, 2008b).

After the Spanish put the issue of maritime piracy on the EU-agenda in April, the Council Secretariat and the European Commission were requested to come up with a plan to contribute ‘to the implementation of UN Security Council resolution 1816 of 2 June’ (Council of the EU, 2008e).

The EU initially responded in September with a coordination cell (EU NAVCO) in support of UNSC-resolution 1816 (Council of the EU, 2008a). Two things are important to note here. First, this was not a full-scale CSDP-operation, but merely a coordination effort. Second, this coordination effort was about protecting maritime trade routes rather than the humanitarian aid delivery, as the mandate refers to UNSC-resolution 1816 instead of 1814. While the UN's initial concerns were about humanitarian aid delivery, when the agenda of fighting piracy was broadened to include commercial trade routes, the EU's first response (EUNAVCO) focused on the latter.

EU NAVCO operated alongside NATO's military operation Allied Provider that was also responsible for escorting WFP-vessels, i.e. its mandate was based on both UNSC-1816 and 1814. Both EUNAVCO and Allied Provider were completed with the launch of EUNAVFOR Atalanta in December 2008, which was ‘commended’ by the UNSC (2008c). Atalanta's initial mandate consisted of the following tasks:

1. the protection of vessels of the WFP delivering food aid to displaced persons in Somalia, in accordance with the mandate laid down in UNSC Resolution 1814 (2008),

2. the protection of vulnerable vessels cruising off the Somali coast, and the deterrence, prevention and repression of acts of piracy and armed robbery off the Somali coast, in accordance with the mandate laid down in UNSC Resolution 1816 (2008) (Council of EU, 2008b).

In sum, the mandate consists of both the value-based objective of protecting the UN World Food Programme and the utility-based objective of protecting commercial trade routes.

Advocacy coalitions and the launch of EUNAVFOR Atalanta

The Global Power EU coalition, with France, Spain and High Representative Solana as key members, was the main driver behind the launch of EUNAVFOR Atalanta. In general, this coalition is keen on the military instrument as part of the foreign policy toolbox. According to this coalition European citizens should be willing to make a financial and human effort for the launch of military operations, which can serve a whole range of purposes, including more geostrategic and national interests.

As France and Spain had a considerable tuna fleet off the Somali coast and were lobbied by their fishing- and transport industry, this coalition had been keen to extend the UN's support for protecting humanitarian aid (UN Resolution 1814) to include commercial vessels as well, by initiating and largely authoring UN Resolution 1816 (Kouchner in: Clarens, 2008). France and Spain used this resolution for launching an EU military operation:

They [France] had pushed for passage of UNSCR 1816 to set the scene for an autonomous ESDP operation (US Embassy Brussels, 2008c).

It is with reference to the attacks on French and Spanish vessels that Spain put the issue of piracy off the Somali coast at the Council-agenda of April (Council of the EU, 2008c). France, using the advantages of having the Presidency from July onwards, kept the issue on the agenda and was keen to have the operation launched during its Presidency (former French official, interview March 1, 2016; Kouchner, 2009).

While humanitarian concerns were invoked, and France had been active with escorting the WFP, the members of this coalition emphasized that the operation is not ‘just’ about protecting WFP-shipping (PMG-member D, interview June 11, 2013). American diplomats observed that ‘the WFP might not be the highest priority’ (US Embassy Paris, 2008). With reference to the kidnapping of a Spanish fishing vessel, ‘Beach Bakio’ the Spanish Government justified the operation ‘as a matter of national interest that seriously jeopardizes the life and interests of Spanish citizens’ (Spanish Parliament, 2009). The Spanish Government even made its contribution to the EU-operation conditional upon a mandate that would not be restricted to supporting the WFP: ‘it would be very difficult for Spain to devote a ship to the WFP mission when it did not have sufficient resources on scene to protect its own fishing fleet’ (US Embassy Madrid, 2008).

This underlines the fact that for this coalition the pull-factor of Atalanta is that it is not restricted to humanitarian aid delivery. This position is nicely captured in a quote from Sarkozy, French President at that time:

Against the pirates in the Gulf of Aden attacking our ships, we launched the “Atalante” operation, the first EU naval operation, the first operation defending purely European interests and, what's more, the first EU military operation commanded by the United Kingdom (Sarkozy, 2009).

Notably, the Global Power EU coalition was one of the main providers of frigates in the period 2009–2014 (see Table 2).

Table 2. Main contributors (2009–2014)^a.

Year	Main contributors (>10%)
2009	Germany, Sweden, France and Spain
2010	France, Spain, Germany, Italy, Greece
2011	Spain, Germany, France
2012	France, Spain, Germany, Italy
2013	Spain, Germany, France, Italy, Netherlands
2014	Germany, Spain, Netherlands, France

^a Table 2 is based on own calculations based on newsarchive of EUNAVFOR Atalanta (http://eunavfor.eu/archives/). This overview only includes the deployment of frigates. Contributions of surveillance aircrafts are excluded.

In line with its ambition for EU power projection, this coalition emphasized the need for the NATO operation to complement Atalanta, and not to ‘dilute’ the EU's initiatives:

France does not oppose studying options at NATO, as long as they take into consideration existing EU efforts to act in complementarity (US Embassy Paris, 2008).

In sum, the Global Power EU coalition sees EU military operations as a show of EU power. Humanitarian challenges play a facilitating role. It valued Atalanta for being more than ‘just’ protecting the WFP. By UN-resolution 1816 it made sure that there was an international legal base for including the protection of commercial interests against piracy. Moreover, this coalition was supported by the active involvement of the shipping industry that put pressure on the other coalitions as well.

The Euro-Atlanticist coalition consists of actors that share the conviction that the military instrument performs an important role in foreign policy, primarily in the context of NATO. Key members of this coalition are the United Kingdom, the Netherlands, Portugal, Greece and Poland.

Initially the UK was reluctant to support an EU military operation to fight piracy. While it supported UNSC 1816, it referred to resource constraints and the legal issues as obstacles for active involvement in its implementation (US Embassy London, 2008). It also was open to the US’ critique of the French proposal for an autonomous EU operation (US State Government, 2008). Hence, the UK deliberately delayed the decision-making process in September by asking for more specific information regarding costs, coordination and command and control (US Embassy Brussels, 2008c). It expected that:

(…) should the EU see that the WFP ships are protected and perhaps, that the WFP is able to use ever larger ships in its operations due to the increased security provided by other actors, any support for an autonomous operation will wither (US Embassy Brussels, 2008c).

The UK preferred a small EU coordination cell over a full-scale EU naval operation (ibid.).

With their preference for NATO, this coalition did not want France to claim sole credit for an operation that would fit very well with the resources of its key member, i.e. the UK as maritime power (Germond & Smith, 2009). To prevent activation of an Operation Centre in Brussels, the UK offered their Operational Head Quarters (OHQ) in Northwood (former CSDP-official B, February 17, 2016).

Business interests of insurance companies contributed to convince the UK-government to support the launch of the operation. Hence, similar to the Global Power EU-coalition, economic concerns take a prominent role in the UK's justification, though humanitarian concerns are not absent entirely. As UK FCO-Minister Flint argues:

It is strongly in the UK's interest to support this mission because piracy off the Horn of Africa is threatening a key global economic artery as well as regional trade, and because the UK remains an important centre of global international shipping (…) There is also the key driver that without escorts, vital humanitarian assistance (…) would cease, endangering the lives of thousands of Somali people (European Scrutiny Committee, 2008a, 2008b).

Greece also emphasizes the interests of its maritime industry (US Embassy Athens, 2008a, 2008b). The Netherlands refers to the humanitarian situation as the primary reason for participating in EUNAVFOR Atalanta, while acknowledging the ‘economic and strategic importance’ of free and secure global shipping routes and the international rule of law as well (Dutch Government, 2008).

The UK provided one frigate at the start and one in 2011, both for 2–4 months, which is considerably less than its contribution to the NATO-operation (see NATO, 2015). The Netherlands pulled out the Hr. Ms. Tromp from NATO to be deployed under EU-flag. However, in terms of financial costs (reflecting the size of its contributions), the Dutch contribution to Ocean Shield and EUNAVFOR (respectively 21,2 million euro and 9,5 million euro, cf. Dutch Ministry of Defence, 2013) indicates that the Dutch do not prioritize the EU-operation.

Although the contributions (financial or deployed units) of the Euro-Atlanticist coalition to Atalanta are limited (i.e. not signalling a shift away from their institutional preference for NATO), by providing the Operation Commander (Rear Admiral Philip Jones) and Operational Head Quarter in Northwood this coalition kept control (see Table 3; a similar strategy was employed with EUFOR Althea, see Palm, 2017) and reserved for itself an influential position:

When you want to have any impact, you should have the top ranking (…) With a UK commander and a UK staff, they are running it for so long, the biggest influence is through the UK. (Military Representative, interview May 27, 2015)

They [UK] could get the political benefit of commanding at little financial costs. (CSDP-official D, 2016)

In sum, just as the Global Power EU coalition, the Euro-Atlanticists preferred an EU military operation that would encompass humanitarian aid protection as well as the task to protect commercial vessels. By taking a powerful position in the chain of command of Atalanta its concerns of competition with NATO were accommodated.

Table 3. Overview nationality Atalanta force commanders.

2009-I	Greece	2011-I	Spain	2013-I	Spain
2009-II	Spain	2011-II	Portugal/Spain	2013-II	Portugal
2009-III	Netherlands	2011-III	Germany	2013-III	Netherlands
2010-I	Italy	2012-I	Spain	2014-I	France
2010-II	Sweden	2012-II	France	2014-II	Germany
2010-III	France	2012-III	Italy	2014-III	Italy

Note: Since the operation has been launched, the UK provided the Operation Commander.

The key characteristic of the, least cohesive, Bystander coalition is the unwillingness to take substantial risks to defend European values by the use of force, albeit for different reasons. Key members of this coalition are: Germany, the European Commission and High Representative Ashton.

In the case of Atalanta, the Bystander coalition was primarily mobilized by concerns over hostage taking of their own nationals and by economic concerns. The German government underlines the protection of the WFP and Germany's involvement with Somalia with non-military means, but also notes repeatedly that: ‘(…) as an export nation it has a particular interest in secure trade routes, especially since it is dependent on the import of raw materials that in large part come from sea’ (Bundestag, 2008a).³

Since Germany was one of the countries worst affected by piracy in the Gulf of Aden (cf. Weber, 2009) protecting ‘vulnerable vessels’ was increasingly accepted as a legitimate concern, allowing for the operation to go beyond protection of the WFP.

Recall how it started: protecting the WFP. This was a very limited and clearly defined task: easy to explain to the public, because otherwise the ships with food are not arriving at the right place. Then we realized how problematic piracy was, which raised the question: why only protect the WFP? (German diplomat, interview May 28, 2015)

Moreover, the repeated references to the shipping industry in the Bundestag-debates serves to underline the importance of economic concerns (Koenig, 2016, p. 102). While the German government acknowledges that a long-term approach needs to deal with the root causes of piracy, it is pessimistic about the options for doing so:

The long-term fight against piracy, however, must be targeted on eliminating the causes in Somalia itself (…) The possibilities to take action in Somalia, however, are very limited (Bundestag, 2008b).

In the end, most debate in Germany revolved around the legal issues involved with contributing to NATO or the EU. Germany exclusively deploys troops under EU-flag to fight piracy off the Somali coast (Bundestag, 2009), with a permanent contribution of at least one frigate. Its active participation is a quid pro quo for its force reduction in the Balkans (Weber, 2009). At the same time, in the case of Atalanta Germany's contributions are far from symbolic, as it contributes around 25% of the total contributions (frigates, by months deployment) of Atalanta. The European Commission's role should not be underestimated either; they had a big impact on the decision to launch the operation, as its competences and expertize helped the French Presidency to establish EU agreements with countries in the region on the transfer of suspected pirates (Riddervold, 2016).

In sum, the Bystander coalition was mobilized by hostage taking and economic concerns. While this coalition remains risk-averse, the range of legitimate concerns for military intervention was broadened – not so much under pressure of other coalitions but based on its own experience.

In the Human Security coalition, primarily consisting of the ‘neutrals’ (Ireland, Finland and Sweden) and the European Parliament, the use of military force is narrowly circumscribed, i.e. only legitimated for the protection of civilians. Hence, in contrast to the prominent presence of economic concerns in the justification of the coalitions mentioned above, the Human Security coalition legitimated the operation primarily with reference to humanitarian concerns. The Swedish Government (2009) justified its participation by pointing at the UNSC-resolutions and the dire humanitarian situation in Somalia. When it refers to the task to ‘facilitate’ commercial shipping it does so cautiously. Moreover, when addressing the European Parliament in the context of the Swedish EU-Presidency in 2009, Prime Minister Bildt highlighted the humanitarian concerns and played down the economic concerns by stating that Atalanta was about ‘imperative humanitarian and other reasons’ (European Parliament, 2009). This coalition was not deaf, however, to ‘a specific request from European companies’ as the protection of commercial shipping did not only have material aspects, but also was about the ‘safety of those people’ (Finnish diplomat, interview June 6, 2013).

While the European Parliament with its ‘distinct’ foreign policy identity with an emphasis on human rights and democracy promotion (Zanon, 2005), is part of the Human Security coalition, in the case of Atalanta, the position of the European Parliament is close to that of the Global Power EU. As the EP's first resolution, in October 2008, on the EU's action to tackle piracy was an initiative from the Transport Committee (European Parliament, 2008), it reflected a concern with shipping off the Somali coast:

[The European Parliament] welcomes the progress made by the European Council in the preparation of a European Union naval operation against piracy, aimed at guaranteeing the safety of cargo ships transiting a shipping channel in the Gulf of Aden.

Although several MEP's from the Security and Defence-committee voiced severe criticism of the operation⁴ in February 2009 the EP ‘welcomes’ EUNAVOR Atalanta (European Parliament, 2009).

Their initial reluctance towards Atalanta did not prevent the Human Security coalition from contributing to Atalanta once the operation was launched. Although the absolute contribution of this coalition to Atalanta is limited, their contributions cannot be set aside as just symbolic (e.g. a few staff members at Northwood). Sweden even provided the Force Commander for the second period in 2010.

In sum, this coalition was the most cautious one concerning the launch. However, since the operation would ensure the save delivery of humanitarian aid, it did support the operation. Moreover, its support was backed up by the fact that UN-resolution 1816 provided an international legal base for including the protection of commercial vessels, framed as ‘vulnerable vessels’.

Coalition dynamics contributing to the launch of EUNAVFOR Atalanta

Obviously, the driving advocacy coalition behind Atalanta has been the Global Power EU coalition, which dominated the decision-making process. Under its Presidency, France used its institutional resources to get Atalanta off the ground. The Global Power EU coalition valued Atalanta for being more than ‘just’ protecting the WFP, particularly because some of its members had key economic interests that were threatened by piracy. By UN-resolution 1816 this coalition engineered that there was an international legal base for including the protection of commercial interests against piracy.

Like the Global Power EU coalition, the Bystanders and Euro-Atlanticists were also mobilized in favour of Atalanta by the economic concerns, while they did acknowledge the importance of the humanitarian dimension. In the case of the Euro-Atlanticists their initial reservations had to do with their preference for NATO as first option. By taking a powerful position in the chain of command of Atalanta, its concerns of competition with NATO were, however, accommodated.

The Human Security coalition got convinced by the guarantee that protecting commercial vessels would not come at the expense of protecting the WFP and was not expected to play a big role in the operation's ‘actual’ work (see Dutch Government, 2008). Moreover, the UN-resolution 1816 provided an international legal base for including the protection of commercial vessels, framed as ‘vulnerable vessels’.

This strategic interaction among competing advocacy is reflected in the mandate. The Council Decision of Atalanta prioritized the humanitarian aid element, but this was firmly complemented with the task to protect ‘vulnerable vessels’ (i.e. commercial vessels).

The claim that the EU has come to give greater priority to its economic security in its military operations is reinforced in light of the alternative decisions the EU Member States might have taken. First, the EU could have limited itself to protecting the WFP-programme vessels, while leaving it to NATO to launch an anti-piracy operation ‘to improve the safety of commercial trade routes’. This was far from a theoretical alternative as NATO subsequently launched two anti-piracy operations to improve the safety of commercial trade routes: Allied Protector (March-August 2009) and Ocean Shield (since August 2009). In fact, the relationship between the EU and NATO turned out to be one of the most contentious issues of launching Atalanta (former PMG-chair, interview May 29, 2015; former DG EUMS, interview November 18, 2014).

Second, in the same period that EUNAVFOR Atalanta was decided upon, the EU did reject the UN's request for a third military operation in Congo (á la Artemis, 2003). At the GAERC meeting of 10 November the Joint Action launching Atalanta was adopted and the situation in Congo was discussed. As the EUobserver put it: ‘Europe to tackle Somali pirates but not Congo rebels’ (Pop, 2008b). This fact underlines that the launch of EUNAVFOR Atalanta did not take place in a vacuum, but was a particular choice excluding possible other alternative military operations.

In sum, the Global Power EU coalition was successful in mobilizing the other coalitions to support the EU's military anti-piracy efforts. Prioritizing the humanitarian task, but including a broader range of tasks, was key to get all actors on board.

Protecting the WFP is a prominent part of Atalanta's mandate. It is a political construction to satisfy some Member States (…) Once in a while little things are added to the mandate to satisfy particular countries (former CSDP-official B, February 17, 2016).

Commercial aspects cannot be the only reason – with only the commercial aspects I don't know whether this would have been possible. I think there would have been a deployment anyway, but probably it would have been a national deployment – I’m not sure whether the EU-angle would have been used. The conjunction of UN-request plus commercial aspects made the process much smoother (former French official, interview March 1, 2016).

Moreover, in terms of operational tasks the specific interests of actors were taken into account as well: ‘once the Swedes deployed a corvet, we let them do the WFP-tasking, knowing that they are very interested in the WFP’ (CSDP-official D, 2016). Hence, EUNAVFOR Atalanta can be seen as a case of cooperative bargaining.

The paradoxical development of Atalanta: utility creep and increasing embeddedness

After Atalanta was launched in 2008, several changes can be observed that reflect Atalanta's paradoxical development (see Table 4). On the one hand, the actual balance between the two key tasks shifts towards an emphasis on protecting commercial vessels. On the other hand, Atalanta is increasingly embedded in an overarching political framework. Again, these developments are assessed against the backdrop of the interaction between the different coalitions.

Table 4. Main changes in EUNAVFOR Atalanta.

Year	Changes
2009	Justification of extension: protecting trade routes Tasks: Monitoring illegal fishing is added to mandate
2010	Other instruments: launch of EU Training Mission Somalia
2011	Tasks – Autonomous Vessel Protection Detachment (AVPD)
2012	Area of operation to include land Other instruments: launch of EUCAP Nestor Other instruments: Horn of Africa-strategy Other instruments: EU Special Representative appointed (already proposed in 2009)

Discussions on the mandate continued after the first year (2008–2009) of the operation. Whereas in the first year, due to the (limited) capabilities at Atalanta's disposal, the number of escorting non-WFP frigates was limited (Dutch Government, 2008), after the first year the operation expanded, coinciding with a shifting balance concerning its key tasks: one or two frigates are used for the protection of the WFP and AMISOM, while five or six other frigates are deployed to protect commercial vessels (Dutch Parliament, 2009b, 2010). The focus on protecting trade routes does not come at the expense of the protection of the WFP, but does reflect that the main reason for expanding EUNAVFOR Atalanta has more to do with economic concerns than humanitarian ones. As a representative of the shipping industry (2016) put it: ‘it [Atalanta] is still is about protecting the WFP, but protecting shipping lanes became more of a priority.’

Another change that underlines this shifting balance is the deployment of Autonomous Vessel Protection Detachments (AVPD), from the end of 2011 onwards. While in 2009 it was not an option to put military security guards on vessels in both the EU and NATO-context, because of the operational risks and judicial complexity, and the fear of branch organizations IMO and BIMCO for the escalation of violence (Dutch Government, 2009), it becomes increasingly accepted as a matter of efficiency: ‘(..) It was an operational decision how to better provide security’ (CSDP-official D, 2016). These AVPDs on board of WFP vessels allow the frigates of Atalanta to engage in other activities, i.e. protecting trade routes.

By means of using the AVPDs, naval ships that would otherwise be used to escort WFP transports can perform other tasks of Atalanta and the effectiveness of the operation as a whole increases. The deployment of AVPDs is of special importance because operation Atalanta is facing a shortage of naval vessels for the next period (Dutch Parliament, 2011).

So, the use of AVPD's are legitimated with reference to the need to unlock frigates for protecting trade routes (former CSDP-official B, February 17, 2016). The use of AVPDs coincides with reducing the fleet from 8 to 2 á 3 vessels (European Parliament, 2012).

The decision to extend the area of operation to include land territory, contributing to the ‘robustness of the operation’ (Council of EU, 2012b), can also be seen as a departure from the initial focus on humanitarian aid delivery (Nielsen, 2012). It entails another activity that may divert the attention from the priority of escorting WFP-ships.

In addition to the increased focus on economic concerns, in response to the concerns of the Transitional Federal Government (TFG) of Somalia about the industrial fishing of large tuna fleets in the Indian Ocean (Council of the EU, 2009a; see Hudson in House of Lords, 2009), the task of monitoring illegal fishing activities was included in the mandate. This indicates an increasing awareness in the CSDP-domain that tackling the root causes of piracy relates to economic attractiveness of subsistence fishing. At the same time, the task of ‘monitoring’ is a rather watered down compromise; it was emphasized that this activity would not come at the costs of the key tasks (Hudson in Dutch Parliament, 2009a).

In addition to the justification of the military operation, its relationship with other foreign policy instruments is key to its character. While Atalanta was initially quite dominant in the EU's involvement with Somalia, over time it provided a trigger for boosting the EU's engagement with the country, i.e. the EU's activities in relation to Somalia expanded both within the CSDP-domain and CFSP more broadly.

Since 1994 the European Commission has been engaged with Somalia by providing humanitarian aid.⁵ Moreover, the Council of Ministers agreed upon an arms embargo in 2002 (Council of the EU, 2002). From 2006 onwards the EU's involvement with Somalia became part of the ‘EU partnership for peace, security and development in the Horn of Africa’ (European Commission, 2006a), and was further developed in the Joint Strategy for Somalia of 2008–2013 (European Commission, 2008; see also Weber, 2009).

However, EUNAVFOR Atalanta was quite an isolated exercise initially. The first Council Joint Action of Atalanta (Council of the EU, 2008b) does not refer to other EU foreign policy instruments – i.e. the Commission's activities listed above are not mentioned.⁶ Moreover, Solana's proposal to appoint an EU Special Representative (US Embassy EU, 2009), which would have strengthened the link between CSDP and non-CSDP instruments, was only realized in 2012.

After the launch of Atalanta, however, the scope of the EU's activities is broadened. As EEAS-official E (2016) put it:

The first response was military, but once it was contained and political structures started to emerge, then you could start with tackling other issues (e.g. democratization, rule of law).

The increasing policy-embeddedness of Atalanta happened first within CSDP and subsequently in connection to the EU's external policies. The EU's CSDP activities were extended to include a military training mission to strengthen the Somali security forces (EUTM Somalia, Council of the EU, 2010) and a civilian mission to enhance the maritime capacities of Somalia (EUCAP Nestor, Council of the EU, 2012b).

Moreover, while the first Council Joint Action of Atalanta does not refer to other EU foreign policy instruments (Council of the EU, 2008b), the amended mandate of 2012 explicitly refers to the Strategic Framework for the Horn of Africa (launched in 2011),⁷ in which the fight against piracy is part of a comprehensive approach (Council of the EU, 2011a, 2012a). It is acknowledged that the root causes of piracy cannot be solved by the military instrument. This led to a closer involvement of non-military instruments of the European Commission. For example, the Instrument for Stability⁸ has been used ‘to support in Kenya and the Seychelles to assist with trials of piracy suspects detained by the EU military mission EUNAVFOR Atalanta’ (European Commission, 2010). Moreover, available EDF-money for Somalia was doubled to over 400 million under the 10th EDF (2008–2013) (European Commission, 2015). Since 2012 an EU Special Representative is embodying the idea to coordinate all policy instruments (see Council of the EU, 2011b). Also, in 2014 the mandate of EUFOR Atalanta was revised to include a new secondary, non-executive, task:

In addition, Atalanta may contribute, as a non–executive secondary task, within existing means and capabilities and upon request, to the EU's integrated approach to Somalia and the relevant activities of the international community, thereby helping to address the root causes of piracy and its networks (Council of the EU, 2014).

In sum, there had been an impetus for integrating the EU's different foreign policy instruments from the side of the European Commission, but in the initial years of Atalanta the interaction between different initiatives was lacking (see also Ehrhart & Petretto, 2012). After the launch of Atalanta, a process got started that led to an increasing embeddedness of the EU's military operation in an overall foreign policy. Moreover, the operation is increasingly centred on protecting trade routes, though the humanitarian task is not neglected.

Coalition dynamics in the implementation of Atalanta

These changes reflect different decision-making mechanisms (see Table 5). Whereas the Global Power EU-coalition kept emphasizing the need for protecting its fishing fleet, the Human Security coalition was anxious about keeping a narrow mandate focussing on the protection of UN humanitarian aid:

The discussion on Atalanta will be on the form of its mandate: should it be to protect UN humanitarian deliveries, as Sweden would like, or be cast wider to protect EU fishing fleets, as Spain would like to see (US Embassy Stockholm, 2009).

While the contestation about the balance between humanitarian and non-humanitarian tasks was decided in favour of the latter, this was softened by an increasing embeddedness of the operation into a wider political approach, which does include a strong humanitarian dimension.

Table 5. Overview position coalitions on Atalanta.

Global Power EU Economic tasks: yes Comprehensive approach: hesitant Monitoring fishing: no	Euro-Atlanticists Economic tasks: yes Comprehensive approach: yes (learning)
Bystanders Economic tasks: yes Comprehensive approach: yes Operation-area to include land: yes (learning)	Human Security Economic tasks: no, but safety of people Comprehensive approach: yes

The emergence of the ‘comprehensive approach’ has some characteristics of a collective learning process. As a former UK Military Representative (interview December 19, 2013) stated: ‘[Atalanta] got PSC [Political Security Committee] really thinking about the military in a wider context as opposed to Chad.’ The Human Security coalition took an active role in strengthening the EU's comprehensive approach. The European Commission highlighted the ‘significant effort’ of the Swedish Presidency ‘to ensure the inter-pillar coherence of EU maritime policy’ (European Parliament, 2009). Sweden's contribution to Atalanta in 2013 is justified as a way to ‘strengthen Sweden's role in the implementation of the EU Strategy for the Horn of Africa and in the discussion of the EU's broader approach to meeting the challenges in the region’ (Swedish Government, 2012). This coalition was supported in its effort by the military commanders, ‘who quickly had realized that the real problem was not on sea (…). When they realized their own operation had its limits, they were not shy to say so’ (MEP S&D, interview February 23 2016).

Nevertheless, the continuing contestation over its content makes the comprehensive approach a case of cooperative bargaining. Whereas the Global Power EU-coalition had had a prominent role during the launch of Atalanta, it had to compromise on its subsequent development. The coalition kept highlighting the importance of the military instrument in the EU's involvement with Somalia – as was eloquently put by a PMG-member (D 2013) of this coalition:

It has been the development of pirate attacks which relates to commercial interests and the security of EU citizens. This has led to the launch of Atalanta. This has been the starting point of policy towards the Horn of Africa with several building blocks: Atalanta, EUTM Somalia, the appointment of an EUSR (diplomatic track) and the whole effort on EDF to bring more money to Somalia. We have gradually moved to a current policy involving all EU interest – with the starting point: Atalanta.

Moreover, in line with the emphasis on the importance of the military instrument, the Global Power EU coalition pushed for broadening the EU's military presence in Somalia with EUTM Somalia; Spain provided the first Commander, succeeded by Ireland and Italy.

Another element on which the Global Power EU coalition had to give in was the expansion of the mandate to include the monitoring of illegal fishing. This change was very much against the wishes of the Global Power EU coalition, as their fishing fleet were active in this area (former PMG-chair, interview 29-5-2015), some of which were allegedly involved in those illegal activities (see Swedish Government, 2014). The military staff at Atalanta was not in favour either. However, this change served to meet the preferences of DG-MARE of the Commission. As former CSDP-official B (2016) put it: ‘When you want to have interaction and synergy, everyone will have to give something to get more than the sum of its parts.’

While collective institutional learning is rather limited, the operational experience of Atalanta led to important instances of learning within coalitions. First, despite its initial scepticism, as the operation progressed the Euro-Atlanticist coalition came to recognize the comparative advantage of the EU over NATO, being able to use a range of non-military instruments, including the ability to get pirates prosecuted in the region (PMG-member B, interview June 6, 2013; CSDP-official D, 2016). The comprehensive approach is seen as making the EU more effective in tackling piracy than NATO's Ocean Shield. One could call Atalanta a ‘Berlin Plus Operation in reverse’, i.e. the EU's operation gained functional primacy over NATO's operation (Gebhard & Smith, 2015). Since it is the actual experience with NATO and the EU in fighting piracy that led to increasing enthusiasm for the EU's engagement in Somalia, this is an instance of learning.⁹ Another advantage of the EU over NATO concerned the coordination with other actors in the area. At the time of the conflict between Georgia and Russia, the EU was seen as more neutral than NATO (former French official, interview March 1, 2016).

Second, the increasing scope of the operation to include land was pushed for by the military (former CSDP-official B, February 17, 2016). While politically very sensitive to the Bystander coalition, this proposal was accepted by Germany as a response to the shifting activities of the pirates (Bundestag, 2012) and by the Commission as they realized that aid would not end up well otherwise (cf. former DG EUMS, interview May 27, 2013). Since this change in Germany's position is justified in terms of their operational experience with Atalanta and the willingness to accept the military expertize in this regard, this change is an instance of learning.

Discussion and conclusion

As already established by others, Atalanta indicated a shift by explicit reference to economic concerns compared to earlier EU military operations. This paper adds to this, by showing how the development within the operation reinforced this shift. Accompanied, however, by a higher embeddedness in the EU's overarching political framework. Central to development is a process of cooperative bargaining (see Table 6).

Table 6. Main observable implications.

Hard bargaining (power resources of different coalitions)	Institutional Learning (common experience)	Cooperative bargaining (give-and-take)
Linking troop contribution to tasks of mandate (Spain) Delaying decision-making (UK)	Allowing AVPD's Within Coalition Mandate – include land territory (Germany) Comparative advantage of EU over NATO (UK)	Mandate– protecting both humanitarian aid delivery and commercial vessels Mandate – add ‘illegal fishing’ Comprehensive approach

The support for EUNAVFOR Atalanta was largely uncontested, although different coalitions emphasized different dimensions of the operation and had different concerns. Whereas especially the Human Security coalition justified the operation in terms of the WFP, other coalitions emphasized that it was not just about that, highlighting economic concerns. The protection of commercial vessels did not come at the expense of escorting the WFP which allowed for the support and actual contribution of the Human Security coalition.

After its launch, EUNAVFOR becomes increasingly embedded within an overall political framework, while at the same time the utility-based justification for the operation becomes more pronounced. Those coalitions most hesitant to justify military involvement by economic interests, were keen to see the operation becoming integrated in the EU's so-called ‘comprehensive approach.’ In particular, the Human Security-coalition has been a driver of enhancing the embeddedness of the EU's military operations and of ensuring its complementarity to other, civilian forms of EU engagement.

Both the launch and subsequent development of Atalanta were neither dominated by hard bargaining nor learning. Rather what seems to have prevailed in driving the operation forward is a willingness to compromise which lifted the operation beyond the lowest common denominator. ‘Atalanta gave every Member State something’ (member CSDP-structures EEAS D, March 1, 2016).

In terms of the more general implications of these findings, this article, first, substantiates and nuances previous observations on the EU's distinctive use of military force. On the one hand, the explicit inclusion of protecting commercial vessels, even though not at the expense of escorting the WFP, signals a broadening of the EU's security objectives. The simultaneous rejection of the UN's request for a third military operation in Congo to secure humanitarian aid delivery and protecting civilians underlines this argument. On the other hand, EUNAVFOR Atalanta triggered the EU's involvement with Somalia at large, realizing that the naval operation would not solve the ‘root causes’ of piracy. Atalanta is best seen as an operation that is increasingly embedded in a broader engagement to tackle both the causes and the symptoms of piracy.

Second, this paper has introduced an actor-oriented framework for analyzing foreign policy-making in the EU's multi-level governance structures. An advocacy coalition framework provides the theoretical tools tot critically consider the role of a wider array of actors focusing on their substantive ideas to explain continuity and change in the outcome of policy-making processes. As such it can be of use to the study of foreign policy making in other multilateral contexts as well.

Third, the mechanism of cooperative bargaining points at the strong political will to act together, despite continuing differences on the use of military force. The ‘comprehensive approach’ is a case in this point and seems to refer a degree of ‘constructive ambiguity’ (Heisbourg, 2000; Jegen & Mérand, 2014). A discursive institutionalist approach could shed further light on this notion of constructive ambiguity, by showing whether this is restricted to the coordinative discourse among policy elites or used for communicative discourses aimed at the larger audience.

Acknowledgement

Earlier drafts of this paper were presented at Politicologenetmaal Maastricht (May 2015) and BISA, London (June 2015). I would like to thank Ben Crum for his helpful suggestions.

Disclosure statement

No potential conflict of interest was reported by the author.

Note on the contributor

Trineke Palm is postdoctoral researcher at Utrecht University.

ORCID

Trineke Palm http://orcid.org/0000-0003-3886-996X

Additional information

Funding

This research was supported by a Research Talent Grant from the Netherlands Organisation for Scientific Research (NWO) [Grant Number 406-11-027].

Notes

1 For a more comprehensive discussion of institutionalist approaches, see Peters (2012).

2 50 Interviews were conducted on the EU's eleven military operations in three interview rounds (May/June 2013, May 2015; February/March 2016) with EEAS-officials, Military Representatives, PMG-members, Commission-officials and MEPs.

3 This notion is not uncontested as in 2010 Bundespräsident Köhler had to resign after arguing that military deployment may be necessary ‘to protect our interests, for example, free trade routes’ (Die Welt, 2010).

4 EPP-member Dimitrakopoulous spoke about the setup of a ‘global armada’, and S&D-member Gomes referred to the operation as being only about ‘protecting oil tankers’ (Pop, 2008a).

5 Somalia gained access to the European Development Fund (EDF) with the Cotonou Agreement; 199 million euro was earmarked for Somalia for the period 2002–2007 (European Commission, 2002).

6 While some highlight the role of the European Commission in the planning stage of EUNAVFOR Atalanta (e.g. Chou & Riddervold, 2015), the fact that the Commission's existing activities in the country are not visible in the actual mandate suggests that the French Presidency made use of the European Commission because of its legal and development instruments, rather than the European Commission being able to push its agenda in the launch of Atalanta.

7 This 2011-framework refers to the ESS, the Joint Africa and EU Strategies and the 2009 EU Policy on the Horn of Africa (Council of the European Union, 2009b) as its guiding documents.

8 The Instrument for Stability was launched in 2007 to contribute (a) in situations of (emerging) crisis and (b) in stable situations for capacity building to address pre- and post-crisis situations (European Commission, 2006b).

9 This increasing enthusiasm is not shared across the board. Within the Euro-Atlanticist coalition there are critical voices that push for terminating the operation:

It is a very efficient operation in terms of bringing down pirate attacks and in its relation with other actions – it is the best example of the comprehensive approach. Concerning its sustainability, it is still critical; it is not addressing the root causes. We will push for termination. (Polish diplomat, interview June 11, 2013)