download in pdf


Quentin Loïez*


The Inclusion of Strategic Autonomy in the EU Law: Efficiency or Ambiguity?


The strategic autonomy (SA) of the European Union (EU) is a political concept that recently emerged in the field of the EU’s foreign policy. The concept was introduced by the High Representative of the Union for Foreign Affairs and Security Policy and Vice-President of the European Commission (HR/VP) in June 2016, when she presented the EU Global Strategy[1] (EUGS) to the European Council. This document, encompassing the EU’s external action guidelines, aims to reach an ‘appropriate level’[2] of SA in the field of defence and security.

Within the EU framework, the defence and security field is traditionally and primarily seen as a Member State’s prerogative. The Common Security and Defence Policy (CSDP) is a perfect example of this intergovernmental approach. This policy, established and organised in a section within Title V of the Treaty on EU, constitutes the operational arm of the EU. The section provides a strict framework, unanimity-based governance and a set of objectives for this policy.[3] It also includes the possibility for the Member States to launch a Permanent structured Cooperation[4] (PESCO), a member states-driven form of cooperation further described in a dedicated protocol.[5] Nevertheless, the European Commission also acted in this field, from an economic point of view, by adopting its two defence Directives of 2009[6] in order to establish a single European defence market.

Since the presentation of the EUGS, the activity of the Member States and the European Commission has increased considerably, launching a whole set of initiatives such as PESCO and the European Defence Fund (EDF). Each on its own contributes to the implementation of the EUGS and its new concept of SA.

However, there is no precise and clear definition of what SA means. The EUGS remains quite vague, if not ambiguous on the content and significance of this concept. Controversy arises when this concept starts to be used in legally binding documents, because its legal implications may be affected by the lack of clarity.

Therefore, one could wonder about the relevance of having introduced this vague concept in the framework of EU law. This paper aims to explain what SA is and why it emerged in EU law. It will also demonstrate that the inclusion of SA in the EU law will not be effective and fruitful if the concept remains unclear.


I A Jump from Strategic Thinking to Law

1 SA: What’s in the Vague Concept?

SA is a complex concept to grasp. It is used in different ways by different countries in different geopolitical contexts.[7] The presentation of SA, in this paper, will be limited to the field of defence and security in the European context.

The first thing to highlight is that there is no clear and unanimous definition of SA regarding defence and security, shared at EU level. However, on a theoretical level, the concept appears to suggest the will from a political entity to become autonomous in the field of defence and security, in other words to be able to conduct its defence and security policy without depending on any external actor(s).

Second, the SA concept shall be distinguished from the international public law notion of sovereignty. Indeed, the latter is related to the absence of legal submission of a political entity to any other power. SA does not correspond to such a definition, because one can be legally sovereign without being factually strategically autonomous. This helps us to understand that SA shouldn’t be understood as a legal status, but more as a factual situation.

Third, this de facto status is linked to the power and the freedom that a political entity has in the international arena. Thus, the political entity is evolving on a spectrum ranging from complete dependence on other partners to absolute independence from any partner. Absolute independence represents an unachievable objective, given the current interdependence that governs international relations.[8] Therefore, aiming to reach a certain degree of SA is the result of the impossibility of being totally independent and of the political will not to be fully dependent.

Finally, the expression ‘autonomy’ can imply a freedom to, which means the ability to do things on one’s own, and/or a freedom from, which is the ability to do things without being subject to external control.[9] These two sides of the same coin tend to confirm that SA is a factual situation that can be assessed internally (regarding one’s own capacity for action) and externally (regarding one’s relations with others).

In the EU, the Member State that has most developed the concept of SA is France.[10] It refers to this concept as one of the main principles of its national defence policy, considering it as a condition of its sovereignty (freedom from) and ability to play a full role in international relations (freedom to)[11]. France is also a promoter of the extension of its SA concept to the European level.[12] The French influence is apparent in the development of the concept in the current European strategic thinking,[13] as well as in the scientific literature. In its 2017 strategic review, France considers that ‘strategic autonomy rests on a political foundation comprised of two pillars: a high degree of industrial and technological autonomy on the one hand, and the means and resources to ensure operational autonomy on the other’[14]. This approach remains to this day the most agreed outline of SA in the scientific literature.[15] There is a relative convergence to present SA as the combination of Political autonomy, Operational autonomy and Industrial autonomy.

Although there is no clear definition and explanation of what SA means precisely in the EUGS, different elements contained in this document reflect the three dimensions of SA. The following extract tends to confirm this approach adopted in the EUGS: ‘The EU will systematically encourage defence cooperation and strive to create a solid European defence industry, which is critical for Europe’s autonomy of decision and action’[16].

The EUGS echoes operational autonomy, as the document states that ‘European security and defence efforts should enable the EU to act autonomously’[17]. The EUGS refers several times to the idea of autonomous action, which implicitly reflects operational autonomy. The EUGS also emphasises industrial autonomy, as we can see in this other statement: ‘A sustainable, innovative and competitive European defence industry is essential for Europe’s strategic autonomy and for a credible CSDP’[18]. The document emphasises many times on the importance of the industrial autonomy. Although it is less obvious, the EUGS does not exclude political autonomy. The document provides for the EU’s ‘decision-making autonomy’, in particular in relation to its potential defence and security partners, as we can see in the following extract: ‘The EU will deepen its cooperation with the North Atlantic Alliance in a complementary manner and with full respect for each other’s […] decision-making autonomy’[19].

Moreover, the EUGS entails both the freedom to, as it states that ‘We [Europeans] must be ready and able to deter, respond to, and protect ourselves against external threats’[20] and the freedom from, given the fact it provides that ‘While NATO exists’[21] Europeans must be able to act autonomously ‘if and when necessary’[22]. However, the document is vague regarding the exact degree of ‘freedom from’ it foresees. Indeed, the EUGS intentionally remains unclear on the content of SA it seeks for the EU and its Member States. In fact, the document uses the expression ‘appropriate level of […] strategic autonomy’[23]. This wording shows that the EUGS refuses to give details on the exact content of the SA. With such an expression, the EUGS leaves the opportunity to determine this appropriate level of SA to the Member States and the EU.


2 Why SA Has Been Included in EU Law?

One could wonder about the relevance of including this new concept in the EU law, and not keeping it in non-legal documents. To confine SA to the political domain would have seemed more appropriate given its lack of clarity. However, the EU and its Member States intended to include the concept in EU law.

The first mention of SA in EU law occurred before the EUGS, in a regulation on the Galileo programme in 2013.[24] However, this first inclusion was isolated and very limited. Its real emergence at the EU level should be dated from the EUGS, because, based on this crucial document, the concept will progressively enter in the EU law. In order to explain the willingness of the EU and its Member States to do so, it is important to present its main factors. We can distinguish three contextual drivers.

First was the intention to build a strategy to face new challenges in the field of defence and security. Indeed, outside the EU, an arc of instability, with lasting security troubles appeared from the Eastern border of the EU (crisis in Ukraine), to its Southern border (instability in the Sahel-Saharan area), passing through the war against Isis in the Middle East and the Syrian civil war. Inside the EU, Member States suffered from multiple terrorist attacks. It is because of the rise of these security challenges that the European Council asked the HR/VP to work on a new global strategy for the EU external action.[25] The SA concept ended up as one of the most significant outcomes of this documents, as evidenced by the foreword of the HR/VP.[26]

The second driver was the willingness to deliver results in the field of defence and security, after Brexit. On 23rd June 2016, the UK decided by referendum to leave the EU. On one hand, it is one of the strongest opponents of the emergence of a common defence policy within the EU that decided to leave this framework, but on the other hand, it is one of the most capable member states from a military perspective. The Bratislava summit, on 16th September 2016, clearly stressed the necessity for the remaining 27 Member States to provide results regarding different issues, including external security and defence.[27] Therefore, we can consider that Brexit helped and disinhibited Member States from developing the concept of SA in the field of defence and security at the European level.

The third driver was that the SA concept appeared to be an appropriate way to address concerns about the reliability on the US: on the 8th November 2016, Donald Trump, a candidate who repeatedly expressed his scepticism towards NATO,[28] was elected to the White House. This new administration created real concerns on the other side of the Atlantic. Although the long term trend of the US external policy was already to switch progressively towards the Pacific area, the arrival of Donald Trump at the White House and the ensuing destabilisation of the transatlantic link,[29] had a real catalytic effect on the emergence of the SA concept at the EU level.[30] This also explains why the debate on SA is often linked to transatlantic relations, highlighting at the same time the question of the degree of freedom from sought by the Europeans.

Given this tense geopolitical context, the EU and its Member States nurtured the will to develop the concept of SA. However, although the EUGS gave them the opportunity, the member States decided to avoid a complicated debate over the SA content. It is because of their great diversity (there are differences in national defence expenditure, strategic culture, threat perception, perception of the role of the EU in the field of defence, relation with the US, etc.) that Member States didn’t want to confront each other on this question. The content of the SA is a sensitive political issue, on which a compromise for the 27 remaining Member States after Brexit would be difficult to reach. Up to this day, the perception of the concrete meaning of SA among Member States is very heterogeneous, as evidenced by a recent study by the ECFR.[31] This is because such a compromise is made on a ranging scale between total dependence and complete independence, having potential major and structural implications for the defence policies of Member States. The difficulty of conducting such a debate may explain why the EU and Member States have bypassed the political obstacle by using the legal path.

Indeed, the EU and the Member States decided to adopt a pragmatic approach. They avoided a tedious and uncertain debate on the exact meaning of this concept. Instead, they preferred to reach a quicker compromise on the tools and structures, based on which the SA would be further developed and defined in practice. This approach is quite common for the EU. Margriet Drent (former senior research fellow at Clingendael Institute), when talking about SA, explained that ‘Leaving all kinds of ambiguity is of course not a new phenomenon for processes related to European integration. It serves a purpose to leave a concept somewhat vague, as it relieves the Member States of the obligation to address the differences of opinion on the matter’[32].

These new tools and structures were created through EU secondary law. EU law appeared to be a pragmatic way to ensure outcomes, as it can either encourage (in the case of Soft law) or force (in the case of Hard law) Member States to comply with the rule of law, and thus guides their behaviour. In this way, the inclusion of the concept in EU law could contribute to ensuring the development and durability of this concept.

In order to understand whether this bypassing strategy can contribute to the SA, it is important to analyse how the concept progressively entered secondary law.


II A Fruitful Inclusion of SA in the EU Law?

1 The Progressive Inclusion of SA in the EU Law

To take a closer look to the inclusion of SA in EU law, we will take the examples of two recent initiatives launched after the EUGS and that refer to this new concept in their legal corpus.

The first example is PESCO, a treaty-based cooperation, created by the Council of the EU (Council)[33] to foster cooperation between its participating Member States (pMS) in the field of defence. The second example is the European Defence Industrial Development Programme (EDIDP), an EU regulation[34] proposed by the European Commission to support the competitiveness of Europe’s defence technological and industrial base (EDTIB). This pilot programme is envisaged as the first step before a more ambitious regulation called the ‘European defence fund’ following the same objective.

The inclusion of SA in the legal provision of these two initiatives has been progressive, by first becoming part of Soft law, and secondly being extended to EU Hard law. In 2016, the concept was introduced into Soft law, understood here as non-binding law, in other words legal provisions that encourage rather than impose. We include, for example, the Conclusions of the Council and the Communication of the European Commission in this category, as they fulfil the complementary function of preparatory work prior to the adoption of binding law/Hard law.[35] In the case of PESCO, the Council conclusions on implementing the EUGS in the area of Security and Defence include some references to the SA as an objective to follow.[36] In the case of EDIDP, The European defence action plan includes some references to SA as well.[37] For both initiatives Soft law was an antechamber between the EUGS (in the political domain) and Hard law.

Since 2017, the concept has entered Hard law. Legally binding provisions officially mention and entail the objective of SA. In the case of PESCO, the Council decision of December 2017 encompasses elements referring to the SA concept. Given that a ‘decision shall be binding in its entirety’,[38] there is no doubt about the effect of these different references. In this document, Member States agreed on 20 binding commitments to foster their cooperation in the field of defence. The Council’s Decision states that the list of commitments must ‘strengthen the strategic autonomy of both Europeans and the EU’[39]. Thus, all commitment should be implemented in the light of this objective. One specific commitment even provides that the capability projects adopted in the framework of PESCO ‘shall increase Europe’s strategic autonomy and strengthen the European Defence Technological and Industrial Base (EDTIB)’[40]. Therefore, the action carried out within the PESCO framework is legally oriented towards the achievement of the SA objective.

In the case of the EDIDP, the regulation emphasises the industrial autonomy of the EU by stating that ‘actions proposed for funding under the Programme shall be evaluated on the basis of each of the following criteria: […] contribution to the industrial autonomy of the European defence industry and to the security and defence interests of the Union’[41]. This means that the EDIDP will only fund actions which contribute to the industrial dimension of the EU. Furthermore, the regulation refers to the SA as one of the objectives of the programme, as evidenced in the following statement: ‘The Programme shall have the following objectives: […] to foster the competitiveness, efficiency and innovation capacity of the defence industry throughout the Union, which contributes to the Union’s strategic autonomy’[42]. This first objective of the EDIDP shows that the SA concept, as provided for in the EUGS, underlies the spirit of this regulation.

This progressive inclusion of the SA concept creates a form of continuity and coherence between the EUGS’ SA approach and the one entailed in the two initiatives’ legal provisions. It ensures a minimum of understanding regarding the concept’s outlines. As regards the content of SA, the EU law has two effects that can address this issue over time.

First, the inclusion of SA in EU law forces Member States to implement the SA. The purpose of the EU law is to ensure the development and definition of the concept through practice. Indeed, by setting it as an objective in PESCO, or in the EDIDP, we can expect the Member States to seek to achieve it. Given the absence of an initial political compromise on its content, it is the implementation by Member States of these legal provisions that could bring about the observable content of SA that is eventually sought.

Second, it forces Member States to develop and structure the legal provisions of SA. The inclusion of SA in EU law is still work in progress, hence it forces the legislative actors to agree on new political compromises in order to set some new rules and limits in order to have functional and applicable legalisation. In the case of the EDIDP, it was necessary for the Member States to find a political agreement on the question of the eligible entities, even though it was a sensitive issue given its implication on the content of SA, which was eventually decided. The EDIDP regulation now allows external entities to benefit from grants but with some restrictions, in order to protect SA.[43] More flexible rules on this issue could have meant a less ambitious degree of SA. In a very similar way, the participation of third states to projects adopted in the PESCO framework is an important issue, which the pMS are politically forced to agree on.[44] Restrictive or flexible rules will give an insight into the degree of SA in this framework. In this way, the inclusion of SA in EU law forces the legislative actors to reach political compromise in order to develop the concept.


2 A Political Obstacle Bypassed or Postponed?

The bypassing strategy doesn’t erase the political issue. The EU legislative actors cannot escape indefinitely from the political question on the content of SA. This issue will have to be solved if the EU and its Member States wish to ensure an effective implementation of the legal provisions referring to SA. Indeed, some criticism of this inclusion of SA in EU law must be highlighted. The benefits of such an inclusion shouldn’t hide the problem of uncertainty it faces in the long run.

It starts with semantic uncertainty. The legal provisions that refer to the SA concept are usually unsatisfactory because they are ambiguous. They use, in almost undifferentiated ways, the expressions ‘European strategic autonomy’, the ‘European Union’s strategic autonomy’ and the ‘strategic autonomy of Europeans’. These expressions refer to different States (Member States of the EU, European states outside of the EU) and thereby have completely different legal implications. These various expressions can be used in political documents but as soon as they enter the EU law the legislative actors must be a lot more rigorous.

Moreover, the inclusion of SA in EU law is uncertain in its implementation. Indeed, the control competence of the Court of justice of the EU (CJEU) is limited regarding the provisions relating to the common foreign and security policy and to the acts adopted on the basis of those provisions.[45] Thus, even if the CJEU can potentially control and interpret the legal provisions referring to SA in the EDIDP regulation, it can’t do the same for PESCO. The legal provisions of the decision creating PESCO remain beyond the legal control of the Court of Justice of the EU. On the contrary, this competence is in the hand of the pMS of PESCO. This framework only has a mechanism of evaluating the implementation of the commitments to control the implementation of the SA objective. This mechanism, led by the HR/VP and supported by the PESCO secretariat, is ultimately under the political control of the Council. Thereby, the control of the implementation of the SA objective in the PESCO framework becomes a political issue again. Given the difficulties that the pMS of PESCO have to politically agree on regarding the content of SA, the PESCO evaluation mechanism might fail over time. A recent study assessing the first results of the PESCO contribution to SA tend to demonstrate this.[46]

This situation also raises the potential difficulty of implementing a common understanding of SA between PESCO and the EDIDP. Indeed, the involvement of different actors (the Council for PESCO and the European Commission for the EDIDP), different purposes (a broad purpose of deepening defence cooperation for PESCO and a more economic purpose of supporting EDTIB for the EDIDP), served by the two initiatives might contribute to divergent concepts or implementations of the SA concept. However, a common understanding of the SA concept is essential if the EU wants to ensure a consistent action in the defence field.

The SA inclusion in EU law is also uncertain in its further development of the legal structure. Indeed, the legal provisions on SA are the result of political compromise between the legislative actors. Therefore, they can be vulnerable to political changes at the European level. A good example is the current negotiations on the EDF (the following step after the EDIDP programme). Some may wonder about the impact of recent American interference[47] on the negotiation of the new Regulation (specifically concerning the degree of industrial autonomy to be sought in this programme). The same thing can be said about the PESCO negotiations on third states’ participation in projects.

Furthermore, some might consider that the current blocking of negotiations on third states’ participation in PESCO projects demonstrates the limits of the inclusion of SA in the EU law. Although the pMS are forced to find a political agreement on this specific issue, so far they have failed to do so. As a result, the PESCO legislation is incomplete and resolving this problem’s has been constantly postponed since the end of 2018.

As we can see, the bypassing strategy chosen by the EU and its Member States can potentially lead to a chaotic and uncertain implementation and development of the SA concept. The fact that the inclusion of the SA concept could be hampered in the long run by the lack of political compromise over the content of SA demonstrates that sooner or later the Member States have to put this issue at the top of their political agenda. The inclusion of SA in EU law can facilitate the maturation of this debate and highlight the concrete questions to be answered, but at the end the solution can only be a political one. In other words, the inclusion of the SA concept in EU law can help to develop this concept at the EU level, if its legislative actors, in particular the Member States, agree to set the political content of SA.

* Quentin Loïez is a PhD Candidate at Université de Liège (Cité) and Université de Lille (CRDP). He is a PhD fellow of the European Doctoral School on the Common Security and Defence Policy of the European Security and Defence College, e-mail:

[1] High Representative of the Union for Foreign Affairs and Security Policy and Vice-President of the European Commission, Shared Vision, Common Action: A Stronger Europe [2016] (EUGS).

[2] Ibid, 19.

[3] Consolidated Version of the Treaty on European Union [2016] OJ C 202/38–39 art 42–46 (Treaty on European Union).

[4] Treaty on European Union, OJ C202/38–39 art 42.6, OJ C202/40–41 art 46.

[5] Treaty on European Union, Protocol (No 10) on Permanent Structured Cooperation established by Article 42 of the Treaty on European Union [2016] OJ C202/275–277.

[6] Directive 2009/43/EC of the European Parliament and of the Council of 6 May 2009 simplifying terms and conditions of transfers of defence-related products within the Community [2009] OJ L 146/1–36; Directive 2009/81/EC of the European Parliament and of the Council of 13 July 2009 on the coordination of procedures for the award of certain works contracts, supply contracts and service contracts by contracting authorities or entities in the fields of defence and security, and amending Directives 2004/17/EC and 2004/18/EC [2009] OJ L 216/76–136.

[7] Chilamkuri Raja Mohan, ‘Raja Mandala: Two discourses on strategic autonomy’ The Indian Express (18 September 2018) <> accessed 15 November 2019.

[8] Bastian Giegerich, Eva Gross, ‘Squaring the Circle? Leadership and Legitimacy in European Security and Defence Cooperation’ (2006) 43 (4) International politics 500–509.

[9] Daniel Fiott, ‘Strategic autonomy: towards “European sovereignty” in defence?’ [2018] European Union Institute for Security Studies, Brief <> accessed 15 November 2019.

[10] Frederic Mauro, ‘Strategic Autonomy under the Spotlight: The New Holy Grail of European Defence’ [2018] Group for Research and Information on Peace and Security, GRIP Report <> accessed 15 November 2019; Corentin Brustlein, ‘European Strategic Autonomy: Balancing Ambition and Responsibility’ [2018] Institut français des relations internationales, Éditoriaux de l’Ifri, 2–3 <> accessed 15 November 2019.

[11] Ministère de la Défense/SGA/SPAC, ‘French White Paper on Defence and National Security’ [2013] 19–22 <> accessed 15 November 2019.

[12] La Délégation à l’information et à la communication de la défense – Bureau des éditions, ‘Defence and National Security Strategic Review’ [2017] 56–59 (DNSSR) <> accessed 15 November 2019.

[13] Mauro (n 10) 4–16.

[14] DNSSR, 52.

[15] Felix Arteaga, ‘Strategic Autonomy and European Defence’ [2017] Real Instituto Elcano; Ronja. Kempin, Barbara. Kunz, ‘France, Germany, and the Quest for European Strategic Autonomy: Franco-German Defence Cooperation in A New Era’ [2017] Notes de l’Ifri, Notes du CERFA 141, Stiftung Wissenschaft und Politik; Mauro (n 10); European Political Strategy Center, ‘Rethinking Strategic Autonomy in the Digital Age’ [2019] EPSC Strategic Notes, Issue 30.

[16] EUGS, 11.

[17] Ibid, 20.

[18] Ibid, 46.

[19] Ibid, 20.

[20] Ibid, 19.

[21] Ibid.

[22] Ibid.

[23] Ibid.

[24] Regulation (EU) 1285/2013 of the European Parliament and of the Council of 11 December 2013 on the implementation and exploitation of European satellite navigation systems and repealing Council Regulation (EC) No 876/2002 and Regulation (EC) No 683/2008 of the European Parliament and of the Council [2013] OJ L347/1–24.

[25] European Council, ‘European Council meeting (25 and 26 June 2015) – Conclusions’ [2015] 5 <> accessed 15 November 2019.

[26] EUGS, 4.

[27] European Council, ‘The Bratislava Declaration’ [2016] <> accessed 15 November 2019.

[28] Ashley Parker, ‘Donald Trump Says NATO is “Obsolete”, UN is “Political Game”’ The New York Times (2 April 2016) <> accessed 15 November 2019.

[29] Cat Contiguglia, ‘Trump: EU is one of United States’ biggest foes’ Politico (15 July 2018) <> accessed 15 November 2019.

[30] Ulrike Franke, Tara Varma, ‘Independence Play: Europe’s Pursuit of Strategic Autonomy’ [2019] European Council on Foreign Relations, 6 <> accessed 15 November 2019.

[31] Franke, Varma (n 30).

[32] Margriet Drent, ‘European strategic autonomy: Going it alone?’ [2018] Clingendael Netherlands Institute of International Relations, Policy Brief, 5 <> accessed 15 November 2019.

[33] Council Decision (CFSP) 2017/2315 of 11 December 2017 establishing permanent structured cooperation (PESCO) and determining the list of participating Member States [2017] OJ L331/57-77 (Decision establishing PESCO).

[34] Regulation (EU) 2018/1092 of the European Parliament and of the Council of 18 July 2018 establishing the European Defence Industrial Development Programme aiming at supporting the competitiveness and innovation capacity of the Union’s defence industry [2018] OJ L200/30-43 (Regulation Union’s defence industry).

[35] Stamatina Xefteri, ‘La soft law dans l’ordre juridique de l’Union européenne, Revekka-Emmanouela Papadopoulou – Interview, part. 3’ [2019] Blog droit européen <> accessed 15 November 2019.

[36] Council of the European Union, ‘Council conclusions on implementing the EU Global Strategy in the area of Security and Defence – Council conclusions (14 November 2016)’ [2016] <> accessed 15 November 2019.

[37] Commission, ‘European Defence Action Plan’ (Communication) COM(2016) 950 final, 3 <> accessed 15 November 2019.

[38] Consolidated version of the Treaty on the Functioning of the European Union [2012] OJ C326/171–172, art 288.

[39] Decision establishing PESCO, OJ L331/71.

[40] Ibid, OJ L331/63.

[41] Regulation Union’s defence industry, art 10(d).

[42] Ibid, art 3(a).

[43] Ibid, art 7.

[44] Decision establishing PESCO, art 9.1.

[45] Treaty on European Union [2012] OJ C326/30–31 art 24; Consolidated version of the Treaty on the Functioning of the European Union [2012] OJ C326/166–166 art 275.

[46] Lucie Béraud-Sudreau, Yvonni-Stefania Efstathiou, Conor Hannigan, ‘Keeping the momentum in European defence collaboration: an early assessment of PESCO implementation’ [2019] The International Institute for Strategic Studies 10–13.

[47] Letter from Ellen M. Lord, Andrea L. Thompson to Federica Mogherini (1 May 2019).