Differentiation in the European Union in Post‐Brexit and ‐Pandemic Times: Macro‐Level Developments with Meso‐Level Consequences
本文以英国脱欧和新冠疫情为例，论证差异化已成为欧盟政治体系的永久属性，并探讨宏观层面的差异化如何导致中观层面的制度变化，进而影响欧盟的长期可持续性与制度韧性。
2021 marked the official start of a post-Brexit era for the European Union (EU) since the United Kingdom (UK) had only fully withdrawn when the transition period lapsed on 31 December 2020. The transition period prevented the EU and the UK from falling off a cliff before a mutual Trade and Cooperation Agreement (TCA) was to be ratified and implemented into UK law by Parliament before 1 January 2021. This cumbersome process marked the pinnacle of a three-decades long period of European integration which saw the political architecture of the EU gradually shift from a system characterized by integration to one that is more succinctly marked by differentiation. Using Brexit and the Covid-19 pandemic as cases in point, we argue that differentiation has become the system property of the EU's institutional polyarchic architecture. While differentiation has sometimes been conceived as being mostly temporary in nature, we may think of a differentiation of the EU's political order as a permanent and systemic property. Differentiation is both an empirical and a theoretical phenomenon. Empirically, it has been on the rise ever since the ratification of the Maastricht Treaty in 1992. As Euroscepticism progressively became more mainstream, disintegrative tendencies feature more prominently in Europe's political order (Bátora and Fossum, 2020; Gänzle et al., 2020). Membership withdrawals or centrifugal moves away from the EU, such as Brexit or Switzerland calling off negotiations on a partnership agreement in 2021, have therefore been coined conceptually as differentiated disintegration (Schimmelfennig, 2018). Similarly, reverse processes of ‘de-differentiation’, such as the Danish people's decision in May 2022 to revert the former opt-out in EU defence (Politico, 2022), have occurred, too. Prior to the landmark process of Brexit, European integration generally followed the ‘ever closer union’ model – as famously stated in the preamble to the Treaty of Rome of 1957. Even so, European integration was never truly uniform but shaped by mechanisms of differentiated integration (Leuffen et al., 2022). In addition to policy-centred vertical differentiation (that is, the variation in the level of centralization across policies), such as most prominently in Economic and Monetary Union (EMU), the EU has encountered various forms of horizontal differentiation. This includes an internal dimension with EU countries opting out from certain policies both de jure and de facto (Hofelich, 2022), as exemplified by Denmark and Sweden's non-adoption of the common currency as well as an external dimension with non-EU countries, such as Norway and Switzerland, opting into selected policies (Leuffen et al., 2022). Today, the EU can therefore be considered as an institutionally polyarchic architecture of concentric circles with a core of club members surrounded by countries that have managed to negotiate opt-outs from specific policy areas; countries willing to join; and various neighbourhood agreements involving both European and non-European countries (for example, Gstöhl and Lannon, 2018). Studies of European differentiation have underlined that it is paramount to conceive of differentiation as a comprehensive phenomenon that is not tied to specific directions of either integration or disintegration (Gänzle et al., 2020; Leruth et al., 2022; Olsen, 2007). Yet, while differentiation was initially considered as a temporary phenomenon (through the concept of ‘multi-speed Europe’), we nowadays witness the emergence of permanent forms of differentiated political order(s) within the EU, for instance through the form of opt-outs from the Eurozone. This study suggests that forms of macro-level differentiated political order in the EU may lead to transformations at the meso-level, which in turn may affect the institutional robustness of the EU. Ultimately, differentiation may serve as a vehicle for the long-term sustainability and institutional robustness of the EU. This argument is supported by studies showing that the survival of international organizations (IOs) is associated with the size of their executive capacity (Debre and Dijkstra, 2020). This also suggests how institutional features of IOs may assume a paramount role of robust global governance. It is also necessary to emphasize the attractiveness of studying the robustness of a differentiated European political system in turbulent times (Ansell et al., 2017). Clearly, the robustness of institutions may be taken for granted during periods of stability (Olsen, 2007). During periods of turbulence or crisis, however, established institutions may become subject to contestation, requests for reform, and objects of scholarly studies. Rather than leading to fragmentation and dissolution of institutions, we maintain that manifestations of differentiation foster resilience of the political order of the EU. After a concise conceptualization of differentiation, we examine how the EU has dealt with both internal and external differentiation at the meso-level using Brexit and Covid-19 as cases in point. Second, in our view, differentiation is and will continue to be prominent in the years to come, hence we sketch out an institutionalist research agenda which suggests how macro-level differentiation – for example, with respect to polity-level differentiation – may lead to meso-level institutional responses. Finally, we close by discussing implications for the post-Brexit and post-pandemic future of differentiation. As a strategy of integration, the roots of differentiation can be traced back to a report produced under the direction of Leo Tindemans (1975), which laid the conceptual foundations of a ‘multi-speed Europe’ (Stubb, 1996). Yet, scholarly debates on differentiated integration only started in the early 1990s – against the backdrop of the legally enshrined British and Danish opt-outs in terms of currency union and defence for example. With subsequent discussions on further constitutionalizing forms of differentiated integration, the enhanced cooperation mechanism was introduced in the Treaty of Amsterdam. Considering these developments, Stubb (1996, p. 283) defined differentiated integration as ‘the general mode of integration strategies which try to reconcile heterogeneity within the European Union’ in terms of time, space, and matter. From a theoretical perspective, other scholars also focused on the scope and limits of differentiated integration in the EU (for example, de Neve, 2007; Holzinger and Schimmelfennig, 2012; Leruth and Lord, 2015; Schimmelfennig et al., 2015; Sitter and Andersen, 2006). The financial and economic turmoil of 2007–2008, subsequently followed by socio-economic and political unrests of many kinds, brought the EU into a new multi-faceted poly-crisis (Riddervold et al., 2021). The future of European integration became increasingly questioned, as domestic Eurosceptic actors pushed for reducing the scope of core EU powers, opt-outs or even withdrawing from the Union as a whole. Brexit eventually opened a new chapter in the study of differentiation (Leruth et al., 2019). Some of the recent literature on crisis, disintegration and differentiation in Europe also combines explanations based on collective actors' cost–benefit calculations – such as the promotion of equality of opportunity among EU members (Jones, 2018) and institutionalist explanations focusing on how crises are channeled through and mediated by pre-existing institutional frameworks and resources (for example, Bátora and Fossum, 2020). Brexit thus largely reinvigorated differentiation as a central focus of research in EU studies. In contrast to previous iterations of differentiation – which were exclusively used to qualify forms and grades of integration – the term has now come to be applied in instances of both integration and disintegration. Both growth and reduction of the level, scope and membership of IOs remain unequal and therefore differentiated. Differentiation applies to the institutional, policy and territorial dimensions of such processes. With regards to the institutional dimension, differentiation is relevant for grasping the settlement of both emergent organizations and organized systems. Institutions and organized systems are temporary sets of rules and procedures, demography, locations, beliefs and norms. Different approaches to differentiation are particularly helpful in conceptualizing the settlement of emergent political orders such as the EU that involve multiple tiers of authoritative decision-making. The EU has been depicted in a two-pronged way both as ‘a conceptual battleground and an institutional building site’ (Olsen, 2010, p. 81) with a varied mix of organizational forms, governance patterns and ideas about legitimate forms and speeds of integration. Institutional differentiation was eventually understood as ‘new institutional spheres [that] have split off from older ones and developed their own identities’ (Olsen, 2010, p. 142). From an institutional theory perspective, differentiation ultimately ensures a form of interdependence that ties institutional spheres together. In such a highly dynamic process of institutionalization, de-institutionalization and re-institutionalization, European integration is constantly being reconfigured. Varieties of organizational structures and institutional norms in the EU contribute to institutional change within EU member states. It has been observed that both intergovernmental, supranational as well as non-majoritarian institutions tend to encourage patterns of institutional differentiation. Egeberg and Trondal (2017) demonstrate that EU agencies have become vehicles for integrating regulatory bodies also from non-EU member states. In addition, they document an acceleration in the creation of new decentralized agencies over time not only in terms of numbers but also in terms of powers and quality. The ‘agencification’ of both EU (non)member-states and the EU administrations serve as an institutional infrastructure for differentiated administrative integration (Trondal, 2014) through a differentiated inclusion of administrative bodies from member and non-member states in EU governance. Central agencies in Switzerland, EEA/EFTA countries, candidate and neighbouring countries are offered access to EU decentralized agencies' management boards, leading the latter to become privileged fora for external differentiated integration (Lavenex, 2022). During the Covid-19 pandemic, EEA countries coordinated their response strategies with the European Centre for Disease Prevention and Control (ECDC). Since agency officials generally enjoy a considerable degree of independence from their states' central administration, and that their involvement in EU networks is thus relatively insulated from political intervention form their governments, they are highly sensitive to governance processes at different levels of government (Egeberg and Trondal, 2009). The result is a differentiated and semi-autonomous polycentric patchwork of memberships and associations in EU administrative networks. In March 2020, the global Covid-19 pandemic eventually hit Europe; it affected most policy domains of the EU and substantially put solidarity amongst member states to a test. Although the EU has only limited competence on public health, which remains largely under the control of member states (European Court of Auditors, 2021), the EU was criticized for its lack of coordination, for not acting immediately concerning medical equipment for Italy and not reintroducing internal borders (Wolff and Ladi, 2020). The EU's immediate health policy responses mostly rested on the use of pre-existing resources and institutions: the ECDC, the health emergencies system, the Joint Procurement Program (JPP) and the RescEU civil protection mechanism. The initial EU response to the crisis was differentiated partly because the ECDC was dwarfed by the many national agencies it was supposed to coordinate, the lack of an independent budget line, and its sole support by an administrative agency with no independent legal basis (Consumers, Health, Agriculture and Food Executive Agency, CHAFEA). As a result, member states responded quite differently to the crisis, largely reflecting different government systems, governance traditions, policy styles, attention structures and pre-existing crisis routines within each country (Askim and Bergström, 2021; Zahariadis et al., 2021). On the issue of fiscal support to struggling EU member states, an existing political division lingering from the Eurozone crisis between so-called ‘frugal’ Northern countries and the Southern member states reemerged, yet with one crucial difference: Germany appeared to have changed its stance on the matter (Schulz and Henökl, 2020). This resulted in the approval of the unprecedented €750 billion NextGeneration EU recovery plan. With a view to violations of the rule of law in Poland and Hungary, the recovery plan included access clauses tied to the fundamental principles of the EU via the Rule of Law Conditionality Regulation, which acted as preventative mechanisms against further democratic backsliding in both countries, thereby implying that differentiation cannot be applied when it comes to these core values. Legal challenges were unsuccessfully launched by the governments of both countries, yet there were still major concerns over the inaction of European institutions to tackle democratic backsliding in Poland and Hungary. However, the Commission's decision to endorse Poland's recovery and resilience plan in June 2022 was met with strong skepticism, given the government's refusal to loosen control over the judicial system (The Guardian, 2022). Existing patterns of territorial differentiation led to institutional differentiation among national health agencies. Health agencies in Norway and Sweden, for example, activated different strategies first to combat the virus nationally and second in their adoption of advice and services from the ECDC. Contrary to the general observation of tight agency networks among the Nordic countries (Schrama et al., 2020; Stie and Trondal, 2020), the Norwegian and the Swedish health agencies adopted opposite strategies in both regards, for example regarding the implementation of lockdowns and with regards to their referral to the ECDC. Reflecting different domestic institutional designs and past choices, the Norwegian and Swedish governments differed in how they framed and managed the crisis – as a healthcare crisis (in Norway) and a public health crisis (in Sweden; Askim and Bergström, 2021). This observation moreover reflected general patterns of EU differentiation, in which administrative differentiation of the EU contributed to provide institutional architectures for meso-level differentiation (Leruth et al., 2022). One effect was that a non-EU member state (Norway) coordinated Covid-19 responses vis-à-vis the EU more extensively than an EU member state (Sweden). Subsequent internal differentiation occurred once vaccines were deemed safe by peer-reviewed studies: as the EU and the European Medicines Agency were criticized for initially slow-paced decisions to approve vaccines and organizing an effective rollout program, some countries such as Hungary and Slovakia relied on non-European Medicines Agency-approved vaccines such as the Russian Sputnik V (Politico, 2021). Brexit is a second illustrative case providing on how administrative networks in the EU resources for meso-level institutional differentiation. polity-level transformations as with the produced by Brexit, actors their in EU administrative while at the time some variation to institutional and 2021). The scope of meso-level in terms of policy differentiation in the way the post-Brexit EU policy Although the has not yet produced disintegration in one of the most of UK public policy et al., 2022), the institutional and the new agreement (for example, access to and of a strong degree of regulatory and institutional This may even for the to more to the EU in the may be In the process of Brexit has the independence agenda in and the to long-term of the EU may provide further for the and other to for further The implementation of the on has on the agenda in 2021. the of the and thereby a on the of the further in the implementation of the in the of and health, and with Northern on 2021. while Brexit an unprecedented process of differentiated disintegration in the European it also the of the United Kingdom that lead to its own disintegration. In the UK on the with the EU, which further lead to between both and have at the both from the empirical and the of differentiation as a system property of the EU, we that there is a research agenda on differentiation that may be is a general of literature on institutional capacity which to how institutions and systems to and and which may studies of differentiation in the EU et al., 2020; 2021). Studies also provide on how different institutional institutions to to change (for example, and Although institutional are a conceived of as to organizational change and et al., et al., p. institutional theory may be used as a theory of differentiated governance responses. an institutional EU to the of European not with how but how be 2010, p. certain such as institutional resilience and robustness of the EU, be of The role of organizations and organizational however, is in the that institutional has largely the study of long-term and (for example, we as a to differentiation that for March and Olsen, March and famously that long-term of tend to be characterized by to in which tend to be applied to new to a in with an and as in they have not Yet, and of the EU and its institutional system, as with most institutions and is by many the of existing organizational institutional and and limited capacity to and a of previous that at time (for example, sometimes are sometimes characterized by the of that have taken and Olsen, p. or that processes involve and based on of and Olsen, The long-term of differentiated political such as the EU, has been observed to into (Bátora and Fossum, 2020). It thus be that the of a differentiated EU political order is not as a result of processes but subject to the of an institutional not that a political order in an and of the EU, each of which is may be by and may to form a of that therefore on intervention and change through the of organizational an institutional how decision processes and to a of organizational routines that may be and for are past in rules and in the organizational of a government (Olsen, 2017). are thus to and differentiation, some organizational more than – for example, – the organizational structures of the EU, the attention or the architecture of EU is also This that polity-level (that is, the macro-level of the are to the of to actors at the level of institutions (that is, the meso-level of EU institutions and organizational is on the that organizational may how institutions such as the EU and It that organizational are not an of and but in and processes by and and attention certain and certain more than and 2009). are given for it has been that the and effect of on and on studies of the European (for example, and Egeberg and Trondal, Trondal, 2006). Second, to other that in the policy is more subject to change and may thus be an (Egeberg and Trondal, and 2009). organizational therefore from the of EU thus conceive of as a which that of differentiation may be institutionally This implications that may be for in the EU. It may for example be that the institutional of a the of institutional at for institutional of the EU as a differentiated and we argue that the polity-level robustness of the EU is associated with of meso-level institutional differentiation, for instance among EU agencies and administrative networks In other the growth of EU agencies over time has the EU's internal robustness and opportunity to both member and countries in terms of administrative governance. has that the EU has in the of past such as and Brexit, by through established institutions, and Differentiation may as a for the Union to through While the of differentiation has been used and across these cases demonstrate p. quite most and even is not in with The years may the to which differentiation eventually to further the EU in the of future crises continue to a process a differentiated or are we to witness that may the process an The in initially discussions over the creation of new such as political which of a the of political a new dimension to debates over the of differentiation in the time, the Russian in since has a process of in as the Danish in of the defence opt-out in a on 1 June studies on differentiation examine mechanisms that may the emergence of and of differentiation. In we on studies to how differentiated EU for and crisis management et have established that crisis governance in the and within certain policy domains may long-term institutional long-term implications of crisis governance calling for further scholarly across policy and time as well as challenges to the EU's democratic crisis management tend to have that is as observed in the Covid-19 crisis that within one policy – such as public health – may in other policy or across systems 2021). the EU's crisis responses may have long-term and democratic implications and in when the EU to crisis through differentiated institutional and policy the EU's democratic may in the long since crisis responses tend to executive institutions and contribute to the of government decision This central institutions and institutions in processes are from crisis responses and at EU level may therefore a systemic executive and with long-term challenges not met with the support of and et al., 2021).