Combined differentiation in European defense: tailoring Permanent Structured Cooperation (PESCO) to strategic and political complexity

ABSTRACT

Sustaining meaningful defense cooperation in Europe is made difficult by defense-industrial fragmentation, a multiplicity of institutional frameworks, divergent strategic cultures and domestic opposition to integration. The European Union’s recent foray into defense integration incorporates multiple forms of differentiation to overcome these barriers, with Permanent Structured Cooperation (PESCO) characterized by selective membership, external participation, and project-based clustering. Such “combined differentiation” offers an instructive example of how EU practices and principles can contribute to meaningful defense collaboration, even though Brussels is often thought a weak actor externally. It also illustrates how distinct forms of differentiation can be embodied within a single structure to accommodate complexity in strategic preferences. Using the example of PESCO, this article shows how “combined differentiation” has emerged as a response to the nature of the European defense landscape and how debates between member states about how to respond to specific challenges have brought about further differentiation over time.

KEYWORDS:

• Combined differentiation
• PESCO
• European defense
• CSDP
• differentiated cooperation

The international environment has become more insecure in recent years, owing to a combination of new security challenges and the rise of multi-polar competition. States have responded to these new realities by increasing their defense spending and updating their strategic doctrines, but given the limitations of individual actions—especially by smaller states—renewed attention has been given to questions of security and defense cooperation and integration, which can bring about economies of scale, more efficient procurement, collective solidarity, and better coordination of national policies (e.g., Howorth, 2001). Such efforts at defense cooperation have historically been stymied by a number of issues, including divergent strategic cultures (Becker & Malesky, 2017), domestic opposition to integration (Henke & Maher, 2021), uncertainty over the preferences of key actors (Jones & Jenna, 2022), the fragmentation of the defense-industrial landscape (Calcara, 2019; Hoeffler, 2019), and the presence of multiple potential (and often overlapping) frameworks for cooperation (Cold-Ravnkilde & Jacobsen, 2020; Hofmann, 2011).

How states can overcome barriers to international cooperation is a question that has animated much research in the field of International Relations (IR) and security studies, where scholarship has focused on the degree of centralization and legalization in the design of international organizations (Koremenos et al., 2001) and the articulation of shared norms and values (Adler, 2008; Rathbun, 2011) as means of facilitating cooperation under such uncertain and complex conditions. Yet far less has been written in these disciplines about differentiation as an alternative means of overcoming obstacles to defense cooperation. The concept, which has its origins in the study of European integration, refers to the different ways in which selectivity can be introduced into existing institutional structures or patterns of cooperation in order to overcome political hurdles, bring about greater efficiencies, or accommodate diversity (Holzinger & Schimmelfennig, 2012). Differentiation has not traveled outside of European Union (EU) studies—with a small number of notable exceptions (Buzan & Albert, 2010; Donnelly, 2012)—because of its niche application to European integration and the EU’s lack of actorness in the defense field.

Yet the EU’s recent foray into defense issues has highlighted the utility of differentiation for overcoming hurdles to integration and cooperation in this area. These hurdles have included considerable divergence in strategic preferences between EU member states, opposition in some quarters to defense integration, the need to work alongside and avoid competition with non-EU institutional formats, uncertainty regarding the role of non-EU members (including the post-Brexit United Kingdom), and the fragmented nature of the European defense-industrial landscape in Europe (Béraud-Sudreau & Pannier, 2021; Tocci, 2017). In response, the EU’s principal defense initiative of recent years, Permanent Structured Cooperation (PESCO) has embodied differentiation in a myriad number of forms, including selectivity in membership, project-based clustering, compatibility with existing organizations, and mechanisms for third country participation.

What is most interesting about PESCO, aside from the EU’s entry into the defense field, is the combination of distinct forms of differentiation within a single initiative. While existing accounts of differentiation have often focused on discerning “types” of the concept, PESCO embodies multiple kinds simultaneously, including examples of both differentiated integration (Holzinger & Schimmelfennig, 2012; Stubb, 1996) and differentiated cooperation (Klose et al., 2022; Viceré & Sus, 2023). Moreover, in debates surrounding the design of PESCO, the choices between distinct variants of differentiation were foregrounded. While the policy was always intended to be an example of selectivity in membership, subsequent discussions over the eventual format for PESCO have added additional forms of differentiation to this in order to design a structure that could accommodate the complex realities of the European security and defense landscape.

This article explores these dynamics in detail, asking why distinct combinations of differentiation have emerged in PESCO, and what purposes such combinations serve. It articulates the concept of “combined differentiation” as a framework for understanding how initiatives in this domain can be tailored in response to specific hurdles to international defense cooperation. The article contributes to the broader literature on defense cooperation by showing how distinct forms of differentiation can be applied outside the EU’s core policy areas and contributes to the literature on differentiated integration within EU studies by showing how and why distinct combinations of differentiation can emerge. Our findings suggest that although the EU is regarded as a weak actor in the defense field, core EU principles such as differentiation can offer creative ways of realizing structured cooperation under conditions of considerable strategic complexity and political diversity.

Empirically, the article examines debates between member states on the nature of differentiation within PESCO from its incorporation into the Lisbon Treaty in 2009 to recent debates on the merits of third country access. It charts how German preferences for inclusivity produced a compromise agreement on a project-based, processual approach during the eventual launch of the framework in 2017, motivating a French recourse to non-EU forms of selectivity. It also highlights the effort by proponents of PESCO to bring around recalcitrant member states by demonstrating the framework’s compatibility with NATO, and examines the different patterns of differentiated cooperation which have emerged through the project-based framework, including the distinct dynamics of more and less intensive projects. And it shows how disagreement on the merits of third country participation in PESCO projects brought about a compromise position which allowed access to non-members on a selective basis, thereby further differentiating between distinct categories of external partners.

Differentiation in EU security and defense

Research on differentiated integration in EU studies has its origins in the 1990s in response to the national opt-outs which emerged as part of the ratification of the Treaty on European Union (TEU), although proposals for a differentiated Union can be traced back to at least the 1970s (Schimmelfennig et al., 2015; Stubb, 1996; Warleigh, 2002). Indeed, examples of differentiation proliferated so readily in this era that scholars had begun to describe the practice as “a permanent and ‘normal’ feature of European integration” (Leruth & Lord, 2015, p. 754) rather than an aberration of the integration process. The 2016 Brexit vote in the United Kingdom produced a further surge of interest in differentiation, given the significant number of opt-outs the United Kingdom had possessed while still a member, and the wealth of proposals that followed for reinvigorating the integration process inter alia through principles of differentiation (Martill, 2021). Partly in response to the Brexit vote, recent years have seen the publication of a multitude of journal special issues on differentiated integration, testifying to the considerable contemporary interest in the topic (e.g., Fabbrini & Schmidt, 2019; Leruth et al., 2019b; Pirozzi et al., 2022).

There are many forms that differentiation can take within the EU (Leruth & Lord, 2015, p. 75; Warleigh-Lack, 2015, p. 874). These include, but are not limited to: (i) selective participation in policies, through opt-ins, opt-outs, and derogations (Adler-Nissen, 2014); (ii) differences across policy areas in the Union’s competences (Holzinger & Schimmelfennig, 2012); (iii) action taken outside the EU framework by select coalitions of member states (Biermann, 2022; Siddi et al., 2022); (iv) project-based clustering, such as that adopted in PESCO (Blockmans & Crosson, 2021); (v) external participation in EU policies and adoption of EU standards (Gstöhl, 2015); (vi) distinct models of external association with the EU with specific arrangements for particular regions (e.g., Cyprus, Northern Ireland, San Marino, etc.) (Skoutaris, 2017); and (vii) the existence of multiple institutional actors establishing complex and overlapping divisions-of-labor between multiple organizations, such as characterizes the European security and defense field (Hofmann, 2011).

Moreover, the scholarship on differentiation has often sought to articulate a series of binaries which capture some of the kinds listed above, but which also serve to further sub-divide them. Thus, scholars have distinguished between internal and external differentiation (Egeberg & Trondal, 1999; Gstöhl, 2015; Gstöhl & Frommelt, 2017; Wachowiak & Zuleeg, 2022), vertical and horizontal differentiation (Holzinger & Schimmelfennig, 2012; Leruth & Lord, 2015; Sitter, 2021, p. 129), positive and negative differentiation (Howorth, 2019), formal and informal differentiation (Hoeffler, 2019; Svendsen, 2019), differentiated integration and differentiated disintegration (Cladi & Locatelli, 2020; Schimmelfennig, 2018), and differentiated integration versus differentiated cooperation (Klose et al., 2022; Siddi et al., 2022; Amadio Viceré & Sus, 2023).

Recently, scholarship has focused more specifically on differentiation in the domain of foreign, security and defense policy, including on the existing (Danish) security and defense opt-out (Schimmelfennig et al., 2015, p. 778), the rise of ad hoc coalitions in EU foreign policymaking (Grevi et al., 2020; Howorth, 2019; Siddi et al., 2022; Amadio Viceré, 2023), differentiation in member state security preferences (Ewers-Peters & Baciu, 2022; Onderco & Portela, 2023; Riddervold & Bosilca, 2021; Sitter, 2021), the place of third countries in the EU’s foreign and security policy architecture (Martill & Sus, 2018; Rieker, 2021a; Svendsen, 2019, pp. 529–531; Wachowiak & Zuleeg, 2022), experimental forms of differentiation (Leruth, 2023), the fragmentation of the defense-industrial landscape (Bátora, 2021; Bunde, 2021; Calcara, 2019; Faure & Smith, 2019; Hoeffler, 2019; Rieker, 2021b), the presence of multiple overlapping institutional frameworks in European security and defense (Celik, 2020; Cladi & Locatelli, 2020, p. 9; Duke, 2008; Svendsen & Adler-Nissen, 2019, p. 1425), and the differentiated nature of post-Brexit EU initiatives like PESCO (Biscop, 2017a; Blockmans, 2018; Blockmans & Crosson, 2021; Sitter, 2021). Recent research on differentiated integration has been complemented by work on differentiated cooperation in security and defense, which it has been argued better captures some of the dynamics in this less rules-based and more intergovernmental issue-area (Klose et al., 2022; Siddi et al., 2022; Amadio Viceré & Sus, 2023).

Toward combined differentiation

Typologies and binaries usually encourage us to think in terms of discrete categorizations, but many examples of differentiation within the EU sit uneasily with this task, embodying multiple and overlapping forms of differentiation. In the field of security and defense, scholarship has singled out PESCO as a framework embodying high levels of differentiation in its very design (Biscop, 2017a; Blockmans, 2018; Blockmans & Crosson, 2021). As we discuss in greater detail below, PESCO exhibits not only selectivity in membership but also project-based clustering and patterns of differentiated cooperation that result in external differentiation through the engagement of third countries, differentiation in the relationship with third countries, and a complex division-of-labor vis-à-vis non-EU institutions, including NATO, and the European Intervention Initiative (EI2). In other words, PESCO illustrates some of the limitations of existing typologies and binaries in practice, since it embodies variants of both internal and external differentiation, and forms of differentiated integration and differentiated cooperation.

Why does combined differentiation emerge? To understand this, it is helpful to consider the specific rationales for differentiation. Existing works have highlighted numerous reasons for the adoption of differentiation across policy areas. Differentiation can allow some member states to proceed with integration, whilst acknowledging the difficulties others have in specific areas (Chopin & Lequesne, 2016, p. 534; Schimmelfennig, 2018, p. 1158). It can also produce “catch-up” effects which can incentivize laggards, thereby pushing forward the integration process (Leruth et al., 2019a, p. 1385). Moreover, differentiation can lead to more efficient outcomes by excluding those without adequate capacity for particular policy areas. By drawing on the capabilities of select groupings or of non-EU states in some fields (Svendsen, 2021, pp. 529–532) it can also enhance the EU’s capacity to shape behavior externally through “milieu shaping” (Nielsen & Vilson, 2014). On the flip side, differentiation comes with risks, including the creation of in- and out-groups and a resulting decline in intra-EU cohesion (Chopin & Lequesne, 2016), lowest-common-denominator problems in integration as member states opt-out of specific policies (Blockmans & Crosson, 2019, p. 4), moral hazard as laggards fall further behind (Martill, 2021, pp. 979–980), vulnerability to the interests of non-EU members alongside legitimacy problems in third countries (Leruth et al., 2019a, p. 1386; Lord, 2015), and increased complexity within the EU system (Cardwell, 2019, p. 1412).

One reason combined differentiation arises is that different forms of selectivity perform different functions and therefore are the product of distinct rationales (as well as bringing their own potential pitfalls). For example, while selectivity in membership can produce greater efficiencies, it is also more vulnerable to the charge that it undermines overall cohesion. Moreover, while external differentiation can increase the credibility of EU policies and programs, it comes with a greater risk of creating problematic dependencies and legitimacy problems. Because different forms of differentiation perform different functions, they interact with member state preferences in different ways. Thus, some member states will not only have a preference for a greater or lesser degree of differentiation tout court, but will also have priorities which are served better (or worse) by particular kinds (or combinations) of differentiation.

Moreover, the need in European integration to reach agreement across the board—the so-called “joint decision trap” (Scharpf, 1988)—means that these distinct preferences play out in decisions over the nature of differentiation in any given area. This not only places distinct forms of differentiation on the agenda (as with more inclusive designs on PESCO and proposals for third country participation) but also establishes the starting positions for negotiations (and where successful, compromise) on the nature of differentiation to be adopted. There is thus always a politics of differentiation, even when member states are keen to seek the same end goal (like efficiencies in monetary or defense policy) because of the different ways member states view these issues and the different ways their societies are exposed to integration in these domains.

Beyond the interplay of member state preferences—the “bread and butter” of much of EU integration theory—combined differentiation is also the product of path dependent dynamics stemming from earlier decisions on differentiation; in other words, differentiation often begets more differentiation. There are three respects in which alternative forms of differentiation can be shaped by previous decisions. First, alternative forms of differentiation can act as the basis for a meaningful compromise by allowing both sides in complex negotiations to have their way. For instance, the adoption of modular formats can allow for the simultaneous existence of inclusive and exclusive formats (Blockmans & Crosson, 2021) while differentiation in relations between external actors can help accommodate diverse preferences on third country access to EU policies (Globsec, 2020). Second, alternative forms of differentiation can substitute for one another, compensating for options not adopted at a particular time. For example, states can adopt forms of external differentiation in order to compensate for the deficiencies of internal EU policy areas (de France, 2019). Third, alternative forms of differentiation can serve to increase the viability of other forms of differentiation being adopted, insofar as they either compliment—or do not actively conflict—with the initial format adopted. For example, the more open the policy area to member states, the more likely it is that commitments will need to be reduced and the greater the viability of modular formats will be (Biermann, 2022).

Methodology: PESCO as a natural laboratory

The empirical research for this article is drawn from the recent launch of PESCO. Built into the Lisbon Treaty in 2009 but never successfully operationalized until recently, PESCO emerged as a process through which EU member states could bind themselves into achieving more ambitious defense capabilities against the backdrop of a less secure external environment. The launch of PESCO in 2017 owed much to the realization that Europe inhabited a less hospitable international environment than had previously been the case along with parallel fears of American disengagement and a lag in European defense spending. But it was also an attempt to bring about a form of cooperation better suited to the fragmented nature of the defense-industrial landscape in Europe and better suited to dealing with the challenges of Brexit and demands for European states to do more in NATO.

PESCO is also an interesting example of differentiation, which manifests itself in various forms that are not only applicable to the specific EU context. Participation in PESCO is selective, with three member states declining to take part, for various reasons, leaving an inclusive core of 25 participants. The initiative is designed around project-based clustering, with participating member states taking part in at least one project, and many taking part in multiple projects. PESCO is also open to the participation of third countries, so long as they meet specific criteria, and the initiative has been explicitly crafted to feed into parallel non-EU processes, including the commitments of NATO members.

These distinct forms of differentiation have also been subject of contestation between the member states over time, with prominent debates on the extent to which the PESCO framework should be inclusive or exclusive, be complementary with NATO and non-EU initiatives, and be open to third country participation. Moreover, these debates and the choices made are linked, with combinations of differentiated practices representing political compromises that seek to balance competing rationales whilst taking into account previous aspects of the design process. Thus, PESCO is a useful laboratory for studying combined differentiation as a response to strategic and political complexity.

In our case study we focus on the key milestones in the articulation of the PESCO framework, the rationales put forward by different actors to justify the differentiated elements, and how these shaped the resulting outcomes. It covers the period from the articulation of initial designs on PESCO in the Lisbon Treaty and the decision to re-launch the PESCO process in 2016 and the resulting Franco-German debate on inclusivity, through debates on membership composition and the relationship with NATO, up to the more recent decision to allow third country participation in PESCO. The research draws on several different sources, including policy papers, government documents, media reports, official data on the individual projects, and the available secondary literature. Because PESCO embodies both differentiated integration and differentiated cooperation, we refer to both concepts in our analysis.

The case shows how the differentiated aspects of PESCO relate to one another, charting the shift from a vanguardist model to an inclusive format with effective opt-outs at the behest of German designs, the subsequent decision to adopt voluntarist project-based clustering and the parallel articulation of the EI2 by France, efforts to emphasize complementarity with NATO in order to bring around potential laggards, and the adoption of restrictive rules on third country participation in order to differentiate between external partners. It further illustrates how diverse preferences on the rationale of differentiated can bring about complex configurations of distinct kinds of differentiation. While PESCO is unusually complex in this respect, and not representative of all forms of differentiation, the dynamics the case highlights are broader ones and the lessons remain theoretically generalizable and, we would argue, substantive and relevant.

Our central claim in this article is that different kinds of differentiation may be combined in the design of security and defense structures in response to underlying hurdles to defense integration, including the diversity of preferences and presence of competing institutional frameworks. We therefore expect to find evidence that different forms of differentiated solutions are considered by policymakers, that these are presented as responses to strategic or political problems, and that the eventual outcome reflects a balance of different kinds of differentiation which is calibrated so as to overcome these issues. We also expect, as a by-product of this process, that temporal dynamics play a role, with earlier decisions to adopt specific forms of differentiation contributing to later decisions by narrowing the range of options or necessitating further differentiation down the line.

The empirical narrative presented below provides evidence to support our claims: Strategic divergence between France and Germany over the extent to which inclusive or exclusive variants of differentiation should be adopted for PESCO produce a compromise position in which inclusivity is privileged at the price of an underlying modular framework. Concern over the division-of-labor between the EU and NATO among Atlanticist states brings about important reassurances that the design of PESCO projects will not prejudice the primacy of the Atlantic alliance in defense (Fiott, 2017). Divisions over the extent to which third countries might contribute to PESCO are resolved through a compromise formulation which differentiates between different kinds of third countries, in order to prevent potential spoilers from joining.

Case study—Combined differentiation in PESCO

Origins in the Lisbon Treaty

PESCO has its origins in the Lisbon Treaty (Art. 42(6) TEU) although it remained one of the “sleeping beauties” of the Treaty, never activated until November 2017, despite several false starts. PESCO was envisioned as a form of enhanced cooperation in defense through which a group of at least nine member states could commit themselves to go further in pursuing collaborative defense projects. Five specific commitments for PESCO were envisioned—commitment to increased defense investment, harmonization of military standards, enhancement of interoperability, addressing capability gaps, and participation in EDA projects—and were annexed to the Treaty in a separate protocol (Biscop, 2017b, p. 2). Because PESCO was always intended to be a selective endeavor, it represents a rare example of formal differentiation in EU security and defense. What differentiation in PESCO would look like, however, was contested from the outset. Echoing later Franco-German disagreements, the discussion on PESCO began in response to a Franco-British desire to set out shared criteria for high-intensity force projection which, finding little favor in Germany, morphed into a more project-based framework (de France, 2019, p. 11).

PESCO suffered from several false starts, including an initial “exploratory seminar” convened by the Spanish Council Presidency credited with keeping PESCO on the agenda even as the British and French delegations had seemingly lost interest (Biscop & Coelmont, 2011, p. 153). Next came the tabling of a Belgian-Hungarian-Polish discussion paper during the Belgian Council Presidency in June 2010 which promoted the idea of an “inclusive” PESCO as a means of overcoming inefficiencies in defense spending and which explicitly rejected the idea of a “two-speed Europe” in defense but which also claimed the initiative must be ambitious enough not simply to reflect existing priorities (Belgian Presidency, 2010). The Belgian initiative took place against the backdrop of the financial crisis, which created fiscal pressures to locate efficiencies in defense spending, and the 2010 Franco-British Lancaster House summit, but other than facilitating debate among member states, did not produce tangible progress toward PESCO (Biscop & Coelmont, 2011, p. 159). Even at this stage, it was feared exclusive designs might undermine solidarity within the Union, and proposals for PESCO morphed from the initial Franco-British idea of a platform for major defense players into a more inclusive project-based framework (Biscop & Coelmont, 2011, p. 154; de France, 2019).

A further request from Italy and Spain to the High Representative in May 2011 to keep PESCO on the Council’s agenda failed to produce sufficient momentum to bring about agreement on its activation (Blockmans & Crosson, 2019, p. 3). In the end, it was the combination of geopolitical change, especially the annexation of Crimea in 2014 (e.g., Duke & Gebhard, 2017), and the Brexit vote in 2016, which provided the catalyst for security and defense reform, creating a window of opportunity which policymakers could exploit and thereby enabling internal changes that might otherwise have been blocked by the United Kingdom (Béraud-Sudreau & Pannier, 2021). Following the referendum result and the publication of the EU Global Strategy (European External Action Service, 2016) in June 2016, the EU’s High Representative for Foreign and Security Policy (HR/VP) prompted member states to revisit PESCO as a potential platform for solving their collective defense concerns (Blockmans & Crosson, 2019, p. 3). In December 2016 the member states instructed the HR/VP to “present proposals in the coming months as regards the development of … elements and options for an inclusive Permanent Structured Cooperation based on a modular approach and outlining possible projects” (European Council, 2016). The election of Donald Trump in November 2016 as the incoming leader of the United States acted as a further factor pushing leaders to explore new avenues in European defense collaboration, given Trump’s bombastic rhetoric on NATO and European defense spending.

Over the course of 2017 the HR/VP, Federica Mogherini, and the European External Action Service (EEAS) would support member states to reach consensus over the design and governance of PESCO (Blockmans, 2018, p. 1809). PESCO was also accompanied by a host of other initiatives aimed at improving the coordination of defense activities among EU member states. Following the publication in 2017 of the first EUGS implementation report and the Commission’s own “reflection paper” on defense (European Commission, 2017), a stream of new initiatives was introduced at EU-level (Martill & Sus, 2019; Sweeney & Winn, 2020; Tocci, 2017). These included, alongside PESCO, the Coordinated Annual Review of Defense (CARD) through which the military capabilities of member states would be subject to a coordinated review process, the European Defense Fund (EDF), through which €1.5bn seed funding from the EU budget would eventually be made available annually for defense research, the European Defense Industrial Development Program (EDIDP), which would allocate EU-funding biannually to support the development of defense equipment and technologies with an aim to boost competitiveness and innovation capacity, and new defense-related governance structures within the European External Action Service (the Military Planning and Conduct Capability, MPCC) and the European Commission (DG Defense Industry and Space) (see e.g., Béraud-Sudreau & Pannier, 2021; Martill & Sus, 2019; Sweeney & Winn, 2020; Tocci, 2017).

Inclusive or exclusive designs

The decision to re-launch PESCO reignited intense discussion over the most suitable format for the initiative, especially between France and Germany, the states which together would lend the necessary capacity and credibility to the initiative. France favored an exclusive design for PESCO that would see a smaller “avant garde” of significant defense actors agree to stringent targets for defense cooperation that could be delivered rapidly, while Germany preferred an inclusive approach that would keep as many member states as possible on-board and thus strengthen EU unity (Quencez, 2017). France argued that a higher bar for entry and a more restrictive membership would produce a more ambitious PESCO that would act as an enabler for European “strategic autonomy,” with strong and binding commitments facilitating joint European action in high-intensity conflict regions, including the Sahel (Fabry et al., 2017, p. 2). Germany, in contrast, erred on the side of inclusiveness, seeing PESCO “through the prism of EU integration” and therefore aiming to keep as many member states as possible on-side (Fabry et al., 2017, p. 2). Berlin worried that a high threshold would alienate member states and create new divisions within the Union (Billon-Galland & Quencez, 2017).

A preliminary compromise emerged between both sides in July 2017 on the criteria for admission to PESCO and the commitments entailed, supported by Belgium, Czechia, Finland, and the Netherlands (Fabry et al., 2017, p. 2). It was on the basis of this initial compromise that the initial joint letter from the French, German, Italian and Spanish Defense Ministers was issued on July 21, 2017, indicating to the HR/VP their support for launching PESCO. The final compromise position which emerged later in September 2017 would see both sides converge on flexible criteria for entry combined with a view that PESCO should aim to bring about convergence between member states (Fabry et al., 2017, p. 2). In place of strict targets, the compromise position would see a voluntary approach through which member states would specify their own timeline through which to meet the deliverables (Billon-Galland & Quencez, 2017). Instead of the initially restrictive approach, the aim would be to keep PESCO open to as many member states as possible; to do this, the previously strict targets were reformulated as outcomes to be achieved through PESCO, rather than as conditions of entry into the club (Fiott et al., 2017).

This compromise was closer to the German position than the French one, and Paris feared accordingly that early designs on PESCO from 2009 had been watered down and that the initiative would produce “more bureaucracy than defense output” (de France, 2019, p. 9). But phasing still represented a compromise, of sorts, between both positions. Instead of meeting ambitious requirements, member states would commit to meeting them, turning PESCO into a process while maintaining (ostensibly) the same high level of ambition. Moreover, the processual approach would obviate the need to resolve broader underlying disagreements over the direction of European defense, while still allowing for capability generation projects to move forward (Billon-Galland & Quencez, 2017). Nonetheless, the failure for PESCO to live up to French ambitions motivated both skepticism and hedging, with Paris slow to engage in the initial tranche of PESCO projects (de France, 2019). In his Sorbonne speech on September 26, 2017, President Macron announced his intention to launch the EI2 (Macron, 2017) which amounted, effectively, to the articulation of the operational platform aspect of PESCO outside of the EU framework, given the underlying similarity of the goals of both initiatives (Koenig, 2018; Lebrun, 2018). The move was widely regarded as a response to the failure of France’s more ambitious designs for PESCO, though it was not publicly justified in these terms (Blockmans, 2018, p. 1814).

The debate over inclusive vs. exclusive formats, while settled quickly, showed that even where initiatives were differentiated by design, member states harbored significant disagreements over what differentiation should look like in practice. The divergence between France and Germany—finding themselves at each end of the trade-off between inclusivity and ambition—was rooted in their distinct politico-strategic backgrounds. This includes France’s commitment to a strategic Europe capable of expeditionary action (Lebrun, 2018), its “Europeanist” stance on continental security, and its longstanding designs on a multi-speed Europe (de France, 2019, p. 6), as well as Germany’s commitment to EU solidarity, its inherent caution in matters of hard security, and Berlin’s reticence to move too far away from NATO.

Though underlying political concerns are evident in both positions, it is not possible to de-link these from the strategic postures of both capitals. The outcome was not only to bring about a particular variant of differentiation—the inclusive, modular, process-oriented format—but also to promote forms of external differentiation through the launch of the EI2, as a distinct (non-EU) platform which also counted non-PESCO members Denmark and the United Kingdom among its membership. In line with our theoretical expectations, then, not only was differentiation advocated by both sides in the debate, but the eventual outcome brought about a new combination of differentiation in which inclusivity was achieved through the adoption of a modular structure and France’s designs for a more flexible structure were transposed into a distinct format.

Compatibility with NATO

The compromise which produced the decision on the inclusive format signaled that membership was likely to be more diverse than in earlier iterations of PESCO, but membership would in any case be selective owing to the distinct profiles of member states. Denmark, for instance, has maintained until recently an opt-out from the defense aspects of European integration, and would remain outside PESCO along with neutral Malta and the United Kingdom, which by mid-2017 had notified the European Council of its intention to leave the EU. Because of the design of PESCO, it is the clustering that has received the most attention. But the fact of selectivity remains interesting, both for its own sake—the effective opting-out of Denmark and Malta are important in their own right—and for the impact which efforts to reduce those left on the outside would have on the way PESCO was articulated.

The case of Poland is instructive here, since Warsaw was initially skeptical about the benefits of joining PESCO, seeing in the initiative a threat to NATO’s role in defense matters on the European continent (Billon-Galland & Quencez, 2017) as well as a threat from the major European defense industries to Poland’s smaller national industry (Quencez, 2017). But Poland was, in a way, only the tip of the iceberg and the most visible sceptic as regards NATO’s role, since other Central and East European member states shared similar views about the potential risk posed to the alliance. Gitanas Nausėda, President of Lithuania, argued that strategic autonomy and PESCO both “should be developed together with our allies in NATO which are not member[s] of the EU … It would be the worst scenario if there was competition between the EU and NATO” (Politico, 2019a). For its part, the United States, while perhaps most concerned that PESCO would penalize American companies and deny them access to lucrative contracts in Europe, also saw in the framework a potential threat to NATO’s pre-eminence in the defense sector.

The French and German proposals anticipated such a reaction from Atlanticist states such as Poland, and went to considerable effort in their initial proposal to reassure these countries that there was no threat to the Atlantic alliance, and that the initiative would complement NATO (Politico, 2019b). The July 2017 joint proposal confirmed that territorial defense would remain the preserve of Atlantic alliance and pointed out that the capability projects would directly contribute towards the collective defense effort through NATO. They also signaled that PESCO projects would not contribute solely to the missions of the CSDP by noting they would be put in service of a “full-spectrum” force package that surpassed the operational requirements of even the most high-stakes EU missions (Billon-Galland & Quencez, 2017).

The inclusive, modular format and the list of 25 participating member states was codified in the Council Decision of December 11, 2017 to formally launch PESCO (Council of the EU, 2017). Although Warsaw’s decision to join PESCO was taken at the last minute and somewhat reluctantly—the country opted to join so as not to be left on the side-lines or lose the ability to shape PESCO (Blockmans, 2013; Blockmans, 2018, p. 1813; Quencez, 2017)—the articulation of PESCO as complementary to NATO, coupled with the voluntarist approach, not only laid the groundwork for Polish participation, but also made it less likely that lukewarm commitment would damage the credibility of the initiative. In terms of our theoretical perspective, (external) differentiation again became a solution to ensuring that PESCO would remain a broad church, with collective efforts to reinforce the message PESCO would contribute to—and not compete with—NATO facilitating a more inclusive format and laying the groundwork for the future involvement of third countries.

Project clustering

The articulation of a project-based format for PESCO produced differentiation in other respects, also, since it facilitated the emergence of specific constellations of member states across different projects, so long as they met the (low) threshold of participating in at least one project. Thus, a core component of PESCO is the presence of differentiated cooperation through selective participation by smaller coalitions of member states in the project-based framework. The decision to adopt a modular framework was intended both to make PESCO more politically acceptable to wavering member states (Biscop, 2017b, p. 7; Blockmans & Crosson, 2019, p. 4) but also reflected in the move to regarding PESCO as a process for defense-industrial collaboration, since this domain is highly fragmented and characterized by ad hoc collaboration between select member states (Calcara, 2020; DeVore, 2013; Fiott, 2019; Hoeffler, 2012; Kluth, 2017).

Member states had defended the modular approach as a means of combining inclusivity with ambition (Blockmans, 2018, p. 1812) and proponents argued such a combination could facilitate positive differentiation by allowing for the sharing of best practices—an “open method of coordination” in defense (Sweeney & Winn, 2020, p. 226)—and by mapping member state capabilities onto areas of specialization and comparative advantage more efficiently (Fabry et al., 2017). Critics contended, however, that clustering would provide a basis for permanent opt-outs (Blockmans & Crosson, 2019, p. 4) and that, far from producing a race to the top, now “the speed of European defense cooperation and integration is determined by the slowest wagon in the train” (Blockmans, 2018, p. 1813).

Since the initial wave was announced in March 2018, there have so far been 60 PESCO projects launched, released in four distinct tranches, with 17 projects in the initial two waves, 13 in the third, 14 in the fourth, and the remaining project closed (European External Action Service, 2021, p. 2). Existing mapping exercises (e.g., Blockmans & Crosson, 2019, p. 7; Fiott, 2021) highlight considerable variation within and between these different tranches. The projects themselves cover a number of different sectors, including joint capabilities, training and facilities, land formations, joint enabling, cyber capabilities, air systems, and intelligence, surveillance and reconnaissance (Fiott & Zeiss, 2021, p. 168). While initial projects focused principally on joint capabilities, enabling, maritime and cyber capabilities, second wave projects added air systems and training facilities to the focus on enabling and joint capabilities, and the third wave saw a further increase in these latter two project types (Blockmans & Crosson, 2019, p. 10). The fourth wave comprised a combination of air (6), land (2), maritime (2), space (2) and intelligence, surveillance and reconnaissance (2) projects (EDA, 2021).

The initial waves of projects contained a number of pre-existing initiatives that were folded into the PESCO framework, including those on maritime surveillance, the military disaster relief capability package, the network of logistics hubs (NetLogHubs), the Eurodrone, and Military Mobility, all of which were existing EDA or NATO projects (Blockmans & Crosson, 2019, p. 6; Fabry et al., 2017). The prominence of existing projects highlights the limitations of PESCO in some respects, since the projects would have occurred outside the PESCO framework in any case (de France, 2019, p. 10), although to the extent these projects are boosted by inclusion in PESCO—with its binding commitments—they also highlight areas of value-add (Billon-Galland & Quencez, 2017).

The projects vary also in their intensity, with some, like the Eurodrone, the amphibious assault vehicle, the Tiger Mk III helicopter and the European Patrol Corvette (EPC) filling significant capability gaps, and with others—including training centers and supporting capabilities—rather complementing other initiatives (Fabry et al., 2017, p. 3). Unsurprisingly, the more ambitious projects also tend to be those with the longest expected returns, with such projects as the Eurodrone, patrol corvette, attack helicopter, remote piloting systems and armored vehicles not expected to bear fruit until the second half of the 2020s (Fiott & Zeiss, 2021, p. 168). The second tranche of projects was more ambitious than the initial one, not least given the number of pre-existing projects in the first wave. Moreover, later waves witnessed fewer participants per project on average (Blockmans & Crosson, 2021, p. 92).

Since the more ambitious projects are also those with the smaller number of participants—given the costs involved in obtaining higher-end military technologies (Fabry et al., 2017, p. 3)—it makes sense that greater ambition overall is followed by a general trend towards fewer participants. This pattern of member state allocation is witnessed across the different kinds of projects, with land formations and enabling projects more inclusive in terms of participation than the more ambitious air systems and maritime projects. Indeed, the most powerful member states—France, Germany, Italy, and Spain—participate in the greatest number of projects overall and lead on the most intensive projects (Fiott, 2021), with the more intensive among these, like the patrol corvette and Tiger Mk III, featuring none of the smaller member states (Blockmans & Crosson, 2019, p. 11).

The adoption of a modular format, then, resulted from the initial compromise between inclusive and exclusive designs, and while the German position won-out in this overall debate, the available evidence suggests that the various project-based clusters offer examples of differentiated cooperation which is both inclusive and exclusive, depending on the nature and intensity of the project (Blockmans & Crosson, 2019). Thus, member state divergence on the nature of differentiation contributed to bringing about a further distinct form of differentiation within PESCO.

While the jury is still out on how effective this compromise will be (e.g., Biermann, 2022), there are elements of what might be termed positive and negative differentiation in the pattern of participation, with some member states keen to increase their role (e.g., France) and other member states aiming higher than their capabilities might allow (positive), but with many cases in which projects merely map onto existing interests and priorities (negative). In terms of the combined differentiation framework, the evidence from project-based clustering shows not only how this distinct form of differentiation operates in practice, but also that this particular solution to solving the inclusivity/exclusivity question has further produced specific modes of differentiation within the project-based clusters, some of which are more exclusive than others.

Third country participation

The question of third country participation loomed large even in initial discussions on PESCO membership. Third countries, such as Norway and the United States, already participate in CSDP missions through Framework Participation Agreements (FPAs), and since the architecture of European security and defense is so institutionally variegated (e.g., Hofmann, 2011), external differentiation is salient in this policy area. The first proposals on third country participation emerged in May 2018 in a position paper led by the Benelux states promoting external engagement in PESCO, partly because this offered a way to keep the United Kingdom joined up to EU security and defense post-Brexit (Politico, 2018b). British participation mattered to many member states because of its historical security commitment to the region and its twin status as one of Europe’s two major military powers and its most advanced weapons manufacturer (Valášek, 2018).

The proposals were published the same month as British government calls for a “deep and comprehensive” security and defense partnership, through which London intimated it would be keen to participate in select PESCO projects (Martill & Sus, 2021). The text was backed by the Baltic states, the Vicegrad states (minus Hungary), Bulgaria, Portugal, Finland and Sweden (Politico, 2018b). Although it would take until October 2020 to reach agreement on the matter, an initial compromise position among the member states on third country engagement began to emerge in September 2018, with proposals for case-by-case participation and the requirement to share “EU values” that would effectively exclude China and Turkey from participation. Given the extent of divisions between member states, a senior EU diplomat described the compromise as “the only solution” (Politico, 2018a).

Though a majority of member states were in favor of third country participation in PESCO (de France, 2019, p. 12; Globsec, 2020, p. 8), initial proposals did not receive unanimous support, with concerns about external access to EU defense contracts and potential admission of states such as Turkey and China deemed not to share the EU’s liberal values (Politico, 2018a). A mapping of preferences by Globsec (2020) found that while a significant majority of member states favored third country participation, three were opposed to it (Greece, Cyprus, and Austria) and several were either hesitant or keen to attach conditions, the latter including France, Germany, Italy, and Spain, the major initial promoters of the re-launched PESCO (Globsec, 2020, p. 9). Proponents of third country participation tended to regard the practice as a continuation of existing patterns of security collaboration, especially vis-à-vis the United Kingdom and the United States, and among the Nordic states.

More pragmatically inclined supporters of European defense and smaller states—including Benelux and Scandinavian member states—regarded transatlantic defense collaboration as a necessity alongside smaller European initiatives (Globsec, 2020, p. 9). In contrast, the initial architects of PESCO—France, Germany, Italy, and Spain, all major defense-industrial actors—were understandably more reticent to allow American firms access to PESCO and the EDF, but also more dependent on existing defense ties with the United Kingdom and Norway (Globsec, 2020, p. 10). Berlin was reportedly keener than Paris to see third country access approved, and it was Germany that set out to broker an agreement (Politico, 2020). Meanwhile, Greece, Cyprus, and Austria specifically opposed third country participation on the basis it might lead to Turkish participation (Globsec, 2020, p. 11).

In November 2019 the Finnish Presidency presented a draft text aiming to overcome these disagreements, the aim of which was to spell out ways the United States and the United Kingdom could partake in selective PESCO projects—in order to address demand from states with close military ties to these countries—while assuaging concerns of member states keen to exclude China and Turkey (Politico, 2019b). Member states reached agreement on the conditions for third country participation on October 27, 2020, confirming that non-EU states could be invited to join select projects where they could add substantial value and only on the basis that they abide by the values of the Community (Euractiv, 2020).

To overcome reticence in some quarters, the final compromise enshrined ideas circulating over the previous few years as to the value of “differentiat[ing] the shape and form of relations that the EU establishes with different partners” (Globsec, 2020, p. 8). The agreement paved the way for the announcement in May 2021 that the United States, Canada and Norway, all non-EU NATO members, were set to take part in the “Military Mobility” project, the aim of which is to reduce internal barriers to the transportation of troops and military hardware within the EU (Politico, 2021). Turkey also applied to participate in the project, but has not been admitted, with Austria objecting on the basis Ankara does not fulfill the “community values” requirement. On October 6, 2022, EU member states voted to admit the United Kingdom to the Military Mobility initiative (Politico, 2022).

Thus, debates over the nature of PESCO also concerned the extent of external differentiation which should be built into the initiative, with member states representing a range of opinions—some open, some skeptical—on the question of third country access. Mixed up in these preferences on external differentiation were questions of defense-industrial autonomy, especially the desire of major defense players to exclude United States industry from lucrative seed funding, as well as geopolitical rationales, including the desire of some members for closer cooperation with NATO members, and concerns about strategic effectiveness, notably in the debate over value-add versus vulnerabilities in the admission of non-EU members. The outcome of this complex interplay of preferences was a compromise in which selective third countries were invited to participate based on the twin criteria of value-add and commitment to EU values, a solution which not only circumscribed the extent of external differentiation, but which also succeeded in creating a new category of differentiation in relations with third countries by including the United Kingdom, United States, and Norway and excluding Turkey and China.

While the decision to allow third country participation was not directly related to earlier debates on the extent of inclusivity, it was facilitated by earlier decisions and declarations, including the desire to elaborate a PESCO format which would not challenge NATO and the articulation of specific projects within the framework where non-EU NATO members could play a significant role. As anticipated within our theoretical framework, the solution to political divergence over whether to open PESCO up to third countries resulted not only in a new variant of (external) differentiation, but one that itself involved further differentiation—between different categories of third countries—in order to satisfy both sides.

Conclusion

Differentiation has become a core concept in EU studies in recent decades. But the focus on different kinds of differentiation, the construction of typologies around these, and the articulation of rationales for specific forms of differentiation belie the complexity of forms in practice, the way different kinds of differentiation bleed into one another, and the interaction between different aspects of the concept. Nowhere is this more evident than in PESCO, a natural laboratory for differentiation that combines several distinct forms, including selectivity in membership, project-based clustering, compatibility with non-EU organizations, and (selective) access for third countries. While PESCO was designed as a differentiated policy area, how differentiation has been enacted in practice is not only highly complex—embodying multiple interacting forms—but also the product of a complex process of bargaining in which distinct rationales for, and designs on, differentiation between member states have played out.

By tracing the debates over the specific format PESCO was to take, from its origins in the Lisbon Treaty to more recent debates over third country access, we showed the beginnings of these different forms of differentiation. Though initially held as a means of bolstering cooperation between major defense partners, initial Franco-British designs were not realized, and successive efforts to activate PESCO by other member states sought to articulate a more inclusive vision, albeit unsuccessfully. Efforts in 2017 to re-launch PESCO reignited these debates, leading eventually to a Franco-German compromise through project-based clustering, aimed at combining inclusivity with effectiveness, in the process pushing Paris to explore non-EU formats for defense cooperation. Parallel efforts to convince recalcitrant member states to sign up to PESCO involved concerted efforts to portray the framework as fully compatible with NATO initiatives. The adoption of clustering paved the way for forms of differentiated cooperation to emerge through complex coalitions of member states across different projects, some of which reproduced the more inclusive model, especially in less ambitious projects. And, shaped by the inclusive format and roster of existing projects, member states subsequently agreed a basis for select third country participation which established not only a basis for external differentiation, but further differentiated between different categories of external partners.

The findings contribute to our understanding of PESCO and EU defense cooperation as well as to our conceptualization of differentiation more generally. In relation to PESCO, while we are certainly not the first to highlight the significance of differentiation in the framework (e.g., Biermann, 2022; Biscop, 2017a; Blockmans & Crosson, 2021), our findings move the debate forward by showing how the different forms of differentiation within PESCO emerged, the compromises they embodied, and how they influenced successive forms of differentiation. Our case study also shows that, far from being a failed example of differentiation and selectivity (Biermann, 2022), PESCO embodies compromises that have facilitated selectivity—and efficiencies—in different ways, as well as promoting parallel initiatives (like the EI2) to compensate.

Theoretically, our findings help explain how different kinds of differentiation relate to one another, taking us beyond typologies to show how different forms interact with one another and how differentiation often begets more differentiation by influencing subsequent debates and bringing about alternative formats which embody difference in other ways. They also illustrate how the principle of differentiation can be productively applied in the domains of security and defense as a means of facilitating effective forms of cooperation under conditions of strategic and political complexity often held to be inimical to denser forms of institutionalization. In policy terms, our findings suggest that although the EU is often considered a comparably weak actor in the security and defense domain, principles drawn from European integration, such as differentiation, can offer creative solutions to advance security cooperation under the many difficult conditions characterizing the contemporary security environment.

Acknowledgements

We would like to thank Thomas Christiansen, Hylke Dijkstra, Catherine Hoeffler, Stephanie Hofmann, Michal Onderco, Monika Sus, and Maria-Giulia Amadio Viceré for helpful conversations and comments on an earlier version of this manuscript.

Disclosure statement

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

Additional information

Notes on contributors

Benjamin Martill

Benjamin Martill is Lecturer in Politics and International Relations and Associate Director of the Europa Institute at the University of Edinburgh. His research focuses on European security and defense, the UK-EU relationship and the Brexit negotiations. He is an Associate of LSE IDEAS, the foreign policy think-tank of the London School of Economics, and co-editor of the UACES/Routledge Contemporary European Studies book series. He has previously taught at Canterbury Christ Church University, University College London, and the University of Oxford.

Carmen Gebhard

Carmen Gebhard is Senior Lecturer in Politics and International Relations at the School of Social and Political Science of the University of Edinburgh. She held academic positions at the University of Nottingham (UK), the European Institute of Public Administration (NL) and the Institute of Advanced Studies (AT). She has published in European Foreign Affairs Review, Cooperation and Conflict, European Security, and the Swiss Political Science Review. Her research focuses on European security and defense, inter-organizationalism as well as small states and regional autonomies. In 2017–2020 she was Co-Editor-in-Chief of Defence Studies.