The Brexit puzzle: polity attack and external rebordering

Abstract

Brexit constitutes a puzzle for integration theory. Functionalist analyses have not only failed to predict the UK’s exit but have also underestimated the disintegrative dynamics of the withdrawal negotiations. By contrast, postfunctionalism accounts for the disintegrative Brexit process but struggles to explain the unity and defence of supranational integration among the EU-27. This article tries to make sense of the Brexit puzzle. First, it argues that Brexit constitutes an attack on the EU polity rather than a policy failure – the type of crisis that functionalist theories explain best. Second, it complements the postfunctionalist account of domestic politicisation with an analysis of the reactions and strategies of the defenders of supranational integration. According to this expanded postfunctionalist analysis, the interaction of polity attack and polity defence has produced ‘external rebordering’: extreme disintegration of the UK from the EU, on the one hand, and strengthened integration of the EU-27, on the other.

Keywords:

• Brexit
• crisis
• European Union
• functionalism
• postfunctionalism

Brexit continues to puzzle integration theorists. The UK’s decision to exit from the EU and the disintegrative dynamics of the withdrawal negotiations contradict functionalist expectations about an integration process driven by economic interdependence, interest groups and path dependence. And whereas postfunctionalist assumptions about domestic politicisation and the quest for national self-determination match the UK vote and negotiating behaviour, they do not explain the cohesion that Brexit has produced among the EU-27.

The outcome of the June 2016 referendum not only surprised a large majority of policymakers and academics. It also ran counter to the expectations of the major functionalist theories of European integration – neofunctionalism and liberal intergovernmentalism – about the primacy of economic preferences and interest groups as well as the path dependence of integration. According to Paul Pierson’s historical-institutionalist version of neofunctionalism, ‘the constantly increasing costs of exit in a densely integrated polity have rendered this option virtually unthinkable’ (Pierson 1996: 146). Andrew Moravcsik, the leading proponent of liberal intergovernmentalism (Moravcsik 1998), viewed Brexit as mere ‘kabuki drama’ and ‘organized hypocrisy’, predicting in April 2016 that ‘under no circumstances will Britain leave Europe, regardless of the result of the referendum on June 23’ (Moravcsik 2016). The course of the withdrawal negotiations, and the concessions that the May government was willing to make, initially seemed to vindicate expectations, based on intergovernmentalist bargaining theory, of a soft exit on EU terms (Schimmelfennig 2018). In 2018, Moravcsik still expected that the UK would press ‘to retain substantive policies as close to the status quo as possible’ (Moravcsik 2018: 1664). But the negotiations ended with the most minimalist agreement that any Western European non-member state has with the EU, the almost complete renunciation of EU rules, and constant conflict over Northern Ireland, the one region, in which EU policies remain in place.

By contrast, Brexit fits well with the postfunctionalist narrative of European integration creating domestic backlash. Deep integration constrains national policies, threatens exclusive national identities, and opens the political space for Eurosceptic fringe parties. European issues shift from the arena of interest group politics to that of mass and identity politics. Referendums reinforce polarisation and risk disrupting the EU (Hooghe and Marks 2009). Whereas Liesbet Hooghe and Gary Marks justifiably claim that ‘postfunctionalism has the greatest leverage in explaining the origins, course and effects of the UK referendum on EU membership’ (Hooghe and Marks 2019: 1124), the theory has little to say on the remarkable ability of the EU-27 to defend the integrity of supranational integration against the Brexit challenge. For one, postfunctionalism focuses on domestic politics and how it constrains European integration; it is not a theory of EU-level decision making and institution building. Moreover, the postfunctionalist expectation of ‘downward pressure on the level and scope of integration’ and a limitation of governments’ room to manoeuvre (Hooghe and Marks 2009: 21–3) contrasts with assessments of Brexit as an ‘integration-friendly process’ (Hillion 2018), the ‘maturing of the EU polity’ (Laffan 2019) or ‘disintegration reversed’ (Chopin and Lequesne 2021).

This article proposes two answers to the Brexit puzzle. For one, it distinguishes two types of integration crises: policy failures and policy attacks. Whereas policy failures result from exogenous shocks that expose deficiencies of EU policies, polity attacks originate endogenously with Eurosceptic actors taking determined action against core institutions and principles of the EU polity. And whereas policy failures correspond best with the assumptions and expectations of functionalist integration theories, polity attacks are consistent with postfunctionalism. In other words, the type of crisis is a scope condition for the relative explanatory power of both integration theories. Because the Brexit crisis originated from a polity attack, functionalism struggles to account for its dynamics and outcomes, and postfunctionalism offers the best approach in principle.

In addition, and against the postfunctionalist focus on disintegration or stagnation, the article interprets Brexit as an instance of integration through ‘external rebordering’ (Schimmelfennig 2021). The disintegrative dynamics of polity attacks cause alienation, raise barriers, and reduce transactions between the attackers and defenders of the polity. At the same time, however, they promote integration inside. The defenders of the polity close their ranks, ideologise and institutionalise contested principles, strengthen the demarcation and control of external boundaries, and discourage future exits. A full postfunctionalist account needs to include the mobilisation of the defenders of the community and the integrative effects that the defence against polity attacks generates.

This article focuses on the theoretical interpretation of the Brexit crisis. Rather than reporting original empirical analysis, it builds on existing accounts to propose and assess a novel conceptual and theoretical framework. The theory section presents functionalist and postfunctionalist expectations on the crisis origins, responses, processes and outcomes – depending on crisis type. It furthermore expands postfunctionalist reasoning to include the reactions of the defenders of the supranational community and the integration-friendly rebordering effects of polity attacks. The article thus contributes to theory development and synthesis on European integration, and crises of integration specifically.

Empirically, the article traces the Brexit crisis process from origins to outcomes. For each stage, it aims to show that the Brexit crisis corresponds to the (expanded) postfunctionalist analysis of polity attacks but contradicts functionalist expectations. First, the origins of the Brexit crisis match the postfunctionalist expectation that integration crises result from perceived threats to national self-determination in community-relevant policy domains and from the domestic politicisation of integration in nationalist societies. Second, in the attack scenario, crisis decision making is characterised by ideologised and disintegrative negotiations. Political actors prioritise polity at the expense of policy interests. They focus on the defence of the integrity and autonomy of political communities rather than seeking issue-specific benefits from cooperation. In the Brexit negotiations, red lines defined by the EU regarding the integrity of the single market and the UK on the restoration of national economic sovereignty minimised the room for economically beneficial agreement. Flexible arrangements and compromise proposals met with principled opposition in the EU and by UK hardliners, and whereas the EU’s superior bargaining power was effective in thwarting UK attempts at ‘cherry-picking’, it failed to induce the UK to remain in the EU or settle for a soft divorce. Finally, Brexit resulted in external rebordering. On the one hand, the withdrawal negotiations produced an unexpectedly hard version of exit from a functionalist perspective. On the other hand, the remaining 27 member states and societies rallied in defence against the Brexit threat, presented a united front and delegated negotiations to supranational actors. EU institutions and principles proved resilient and were strengthened in the process, and Brexit restricted both internal and external differentiation.

Theory: integration crises and integration theories

Specifying the scope of functionalism and postfunctionalism

In the social context, a crisis denotes a situation, in which ‘a social system – a community, an organisation, a policy sector, a country, or an entire region – experiences an urgent threat to its basic structures or fundamental values …’ (Boin et al. 2016: 5). In the EU system specifically, crisis is constituted by a manifest threat of disintegration. Disintegration does not necessarily imply the dissolution of the union. Manifest threats of the breakdown of individual integrated policy regimes, the repatriation of supranational competences or the exit of a member state constitute integration crises, too.

I distinguish two types of integration crises: policy failures and polity attacks. In policy failures, shocks exogenous to (i.e. not caused by) European integration expose flaws in the design, instruments and capacity of integrated policies. The euro, migration and Corona crises are cases in point. By contrast, in polity attacks, political actors opposed to supranational integration take determined action aimed at disintegration. Polity attacks are typically endogenous: they respond to policies of the EU that reduce the autonomy and go against the political preferences of the attackers, and they address core values and institutions of the EU polity rather than, or in addition to, specific policy regimes.

To be clear, the conceptual distinction between polity attacks and policy failures refers to the origins of the crisis. Whenever a crisis begins with an intentional act of a political actor, usually a government, that aims to dismantle basic structures and undermine fundamental values of the EU polity, it qualifies as a polity attack. If, by contrast, a threat of disintegration is initially generated by events and developments that overwhelm and overburden EU policies and institutions, without being triggered by political actors aiming at polity disintegration, it is a policy failure. To be sure, policy failures may subsequently trigger attacks – think of the southern European opposition against austerity in the euro crisis or the mainly eastern European opposition against the relocation of asylum-seekers in the migration crisis. First, however, these attacks did not start the crisis but responded to the EU’s management of policy failure. Second, they opposed crisis management policies rather than attacking the core structures and principles of the EU as a polity.

The distinction between polity attacks and policy failures is theoretically relevant. Depending on their origins, integration crises exhibit a dominant dynamic or logic that corresponds to one of the major strands of theorising on European integration. In line with postfunctionalist theorising (Hooghe and Marks 2009), crises originating in polity attacks are driven by the quest for national self-determination, clashes of identity and fundamental conflict about the values and principles of the European order. By contrast, crises originating in policy failures exhibit features typically associated with functionalist theorising of European integration (Moravcsik 1998): the issue-specific search for efficient policy solutions under conditions of international interdependence, in which conflict is predominantly distributional rather than ideological. Correspondingly, actors seeking policy benefits and striving for ‘functional’ solutions are likely to be crowded out in polity attacks, whereas massive problem pressure in policy failures tends to relegate Eurosceptic challengers to the sidelines. That the Greek Syriza-ANEL government accepted the conditions of the 2015 bailout despite having won the anti-austerity referendum is only the most dramatic incidence of a general pattern (Schimmelfennig 2014).

The distinction between crisis types offers a venue for theoretical synthesis. Specifically, it corresponds to the ‘domain of application’ model of theoretical dialogue (Jupille et al. 2003: 21–2). In line with this model, the type of crisis is a scope condition for the explanatory power of integration theories (see Ferrara and Kriesi (2022) for a similar attempt to situate theories depending on the nature of the crisis pressures). Whereas neither functionalism nor postfunctionalism provide a comprehensive and exclusive account of integration crises, they complement each other in explaining different types.

Expanding the scope of postfunctionalism

Distinguishing crisis types thus addresses the Brexit puzzle for functionalism: insofar as the Brexit crisis originated from a polity attack, functionalist integration theory is worse positioned to explain its dynamics and outcomes than postfunctionalism. In turn, however, postfunctionalism suffers from its focus on domestic Eurosceptic politicisation and the resulting ‘constraining dissensus’ on European integration (Hooghe and Marks 2009). By contrast, it neglects the supporters of integration and the reactions of EU-level actors (Bressanelli et al. 2020). A complete postfunctionalist explanation of European integration crises requires equal attention to the attackers and the defenders, and it needs to explain their motivations and behaviour based on the same premises of communal self-determination and politicisation. As much as European integration may mobilise nationally oriented citizens and elites in support of national self-determination, polity attacks threatening the EU with disintegration have the potential to galvanise transnationally oriented citizens and elites to defend their supranational community. Rather than just constraining integration, polity attacks may strengthen it.

First, polity attacks lead the defenders to ‘rally round the flag’. Originally studied as a boost to the popularity of US presidents during international crises (Mueller 1970), it also stands to benefit a threatened EU. Under the impact of a manifest threat of disintegration, other conflicts are set aside and public approval increases. Second, the defenders engage in ‘othering’ the attacker. European identity formation has traditionally built as much on constructing ‘Europe’s others’ (Neumann 1999) as on invoking common values. The contestation of fundamental supranational principles marks the attackers as ‘Uneuropean’, placing them outside the community. Third, the defenders ideologise and institutionalise the attacked supranational principles. They turn them into articles of faith, legally codify or constitutionalise them, and create enforcement mechanisms. They may also strengthen the competence and capacity of supranational actors to deal with the attack and protect the supranational polity.

In the process, both sides come to prioritise the integrity and autonomy of their (supra)national community at the expense of materially beneficial functional cooperation. Their rallying, othering, ideologising and institutionalising strategies tend to mirror and mutually reinforce each other. The EU crisis management of polity attacks is thus likely to have a dual effect. It ostracises the attacker, severs ties and raises barriers at the same time as it strengthens the unity, collective identity, and institutional integration of the defenders.

This dual effect is captured by a ‘bordering’ analysis of European integration (Schimmelfennig 2021), which builds on Stein Rokkan’s seminal work on state building as boundary formation (Rokkan 1974), elaborated and applied to the EU by Stefano Bartolini (2005). In this analysis, integration has two major dimensions: the removal of internal boundaries and the cohesiveness of the member states (internal debordering), on the one hand, and the strengthening of external boundaries demarcating the polity from its international environment (external rebordering), on the other. Moreover, in the Rokkanian perspective, external rebordering is a prerequisite of effective integration. Whereas it restricts exits and cross-border transactions, it thereby also strengthens the internal cohesion of the community: the development of collective identity, political loyalty and solidarity. In sum, expanding the scope of postfunctionalism to include the defenders of the supranational polity and processes of both national and supranational community demarcation helps us see polity attacks not just as a ‘constraining dissensus’ and pathway to disintegration, but also as a process of external rebordering that potentially enables more effective integration. This theoretical elaboration addresses the Brexit puzzle for postfunctionalism.

Functionalism, postfunctionalism and integration crises

To conclude the theory section, Table 1 contrasts the main functionalist and postfunctionalist expectations about the process-level characteristics (origins, preferences, negotiations, and outcomes) of integration crises.

Table 1. Functionalist and postfunctionalist crisis expectations compared.

	Postfunctionalism	Functionalism
Crisis type	Polity attack	Policy failure
Crisis origins	Rejection of supranational polity formation; nationalist governments and societies; community-relevant policies	Shocks and dysfunctions of supranational policies; vulnerable societies; low-capacity policies
Crisis preferences	Ideological polity interests; disagreement on polity preservation	Material policy interests; agreement on policy preservation
Crisis negotiations	Ideological and disintegrative; defence of communities; bargaining power depends on ideological determination	Distributive and integrative; distribution of preservation costs; bargaining power depends on affectedness and indispensability
Crisis outcomes	Defence of polity; integration of defenders and disintegration of attackers	Preservation of policy; integration

Crisis origins

In line with postfunctionalist assumptions, polity attacks originate in perceived threats to national self-determination. Such threats are most likely to emerge in societies with pronounced exclusive national identities and nationalist governments. Attacks also generally have their origin in community-relevant policies. Such policies concern internal and external security, immigration, social sharing as well as cultural and morality issues. By contrast, policy failures result from exogenous shocks that expose or exacerbate flaws in the policy design and instruments. They typically originate in integrated policies characterised by weak supranational financial, administrative and coercive shock-absorbing capacity, and in those member states that are most vulnerable to exogenous shocks.

Crisis preferences

Postfunctionalism further assumes ideological polity interests that derive from general attitudes towards European integration. Attacks are characterised by fundamental disagreement on polity preservation between attackers and defenders. Disintegration is the explicit objective of the attacking governments, whereas the defenders seek to protect the supranational polity. According to functionalist assumptions, actors form their crisis management preferences based on their issue-specific risk from disintegration. They prioritise tangible, predominantly material, policy interests (Moravcsik 1998). Faced with policy failures, member states are likely to agree that policy preservation is desirable. For one, failures come as unanticipated shocks disrupting policies that member states had joined voluntarily. Moreover, functionalism assumes that path dependence and loss aversion bias actors in favour of preservation (Pierson 1996). By contrast, ideological polity interests reduce the actors’ sensitivity to policy costs and material losses.

Crisis negotiations

Correspondingly, crisis negotiations responding to attacks tend to be ideology laden and driven by the determination of both sides to defend the integrity and autonomy of their political communities. Actors regularly forgo policy benefits of cooperation for the sake of self-determination gains. Bargaining power is shaped by the relative ideological determination of both sides. A ‘constraining dissensus’ (Hooghe and Marks 2009) is obtained if the intergovernmental search for compromise is undermined by the domestic dominance of ideological Eurosceptics. Crisis negotiations thus tend to exhibit a polarising and disintegrative dynamic driving attackers and defenders further apart. In the functionalist scenario of policy failure, however, in which governments typically agree on policy preservation, conflict focuses on the distribution of preservation costs. In this distributional conflict, bargaining power reflects asymmetrical interdependence. The less states are negatively affected by the policy failure, the better unilateral or international alternatives they have to EU crisis management, and the more their consent and capacity is needed for preserving the policy, the more they can shape the crisis policy in their favour (Moravcsik 1998). Polarising effects of distributional conflict are constrained by the overriding common interest in avoiding disintegration.

Crisis outcomes

In the functionalist perspective, governments not only have a common interest in policy preservation but are also willing to strengthen integrated policies to the extent necessary to end policy failure. By contrast, political attacks typically have a dual outcome. On one hand, they run a high risk of escalating up to the disintegration of the attacker – typically producing a form of exit that is inefficient in a functional, policy-oriented perspective. On the other hand, attacks promote further integration. They induce unity and cohesion among the defenders, buttress the supranational capacity to respond to attacks and reassert the values and norms of integration.

The following sections assess these competing theoretical perspectives in the Brexit crisis. In a within-case process-tracing design (Beach and Pedersen 2019), the analysis aims to demonstrate that at each stage of the process from crisis origins to outcomes, Brexit unfolded in line with postfunctionalist but contradicted functionalist expectations. As the description will also show, the Brexit process was not completely devoid of functionalist elements. In the end, however, they were crowded out by the postfunctionalist dynamics. The analysis thus corroborates the ‘domain of application’ argument of this article. Because the Brexit crisis constituted a polity attack, it is best explained by (expanded) postfunctionalism rather than functionalism.

Crisis origins: a moderate attack on ever closer union and the freedom of movement

The origins of the Brexit crisis correspond to the postfunctionalist scenario. The crisis started in a society with a highly exclusive national identity and in response to underestimated negative effects of a community-relevant EU policy. Domestic politicisation pressures caused the government to launch a modest attack against supranational principles of the EU and authorise a referendum on EU membership.

According to the Eurobarometer surveys, the British had by far the strongest exclusive national identity among all member societies of the EU. On average over 60 percent of British respondents identified with their ‘nationality only’. The UK consistently sought and obtained opt-outs from the supranational integration of community-relevant policies: the Schengen free-travel regime, Economic and Monetary Union, the Charter of Fundamental Rights and parts of the Justice and Home Affairs acquis.

Apart from principled concerns about national sovereignty, the attack responded to unanticipated negative effects of the one community-relevant EU policy, from which the UK had not opted out: internal migration. Both Conservative and Labour UK governments strongly supported the internal market and its enlargement to new member states. The UK was one of only four member states opening their labour markets to the ten new member states in 2004. Subsequently, and then again after the Great Recession starting in 2008, the immigration of EU citizens to the UK increased considerably, reaching an all-time high of 216,000 (net of emigration) in 2015.¹ In a 2015 survey, 63 percent of the respondents named immigration as the most pressing issue facing Britain (Clarke et al. 2017: 11).

The United Kingdom Independence Party (UKIP) stood ready to exploit this salient issue and linked immigration to its traditional anti-EU stance. This formula resonated well with voters and carried UKIP to electoral successes in 2013 and 2014 (Clarke et al. 2017: 111–45; Evans and Menon 2017: 18–19, 42). To deflect the UKIP challenge and appease the sizable Eurosceptic group within his Conservative party, Prime Minister Cameron promised in January 2013 to renegotiate UK membership and hold an in-out referendum on a ‘new settlement’.² Yet Cameron had no intention to lead the UK out of the EU. In his November 2015 letter to Donald Tusk, President of the European Council, on a ‘New Settlement for the United Kingdom’, Cameron mainly proposed reforms that amounted to a general confirmation and moderate expansion of Britain’s differentiated integration in the EU.³

Two issues presented a bigger challenge. First, Cameron sought a ‘formal, legally-binding and irreversible’ opt-out from the treaty obligation of ‘ever closer union of the peoples of Europe’. Yet, this treaty commitment had not prevented the UK from opting out of the deepening of European integration in the past. Moreover, the European Council had already conceded, in its Conclusions of June 2014 that ‘the concept of ever closer union allows for different paths of integration for different countries, allowing those that want to deepen integration to move ahead, while respecting the wish of those who do not want to deepen any further’.⁴ The second issue was immigration. In this area, he put forward measures to limit the free movement for citizens of future new member states and the access of current EU citizens to UK welfare benefits. However, these proposals aimed at indirect and temporary measures rather than an opt-out from the free movement of workers.

In sum, the Brexit crisis started with only a moderate attack on two principles of the EU’s supranational polity: ever closer union and free movement. Clearly, however, it originated in the rejection rather than the failure of supranational policies. This was true for the supranational deepening of European integration and the legally binding harmonisation of policies in general, and for the freedom of movement specifically, which had enabled millions of EU citizens to settle and work in the UK.

Disintegrative negotiations

Despite its modest beginnings, the Brexit crisis was fuelled by community-relevant concerns and characterised by increasingly disintegrative negotiations from the ‘new settlement’ to the trade and cooperation agreement. Under domestic pressures of a constraining dissensus, consecutive UK governments from Cameron to Johnson intensified their attack on the European integration of the UK, thereby neutralising the EU’s superior material bargaining power. At the same time, the EU converged around an agenda of defending the integrity of the internal market. While both the UK and the EU initially sought to maintain mutually beneficial and functionally efficient cooperation after the Brexit vote, the primacy of self-determination concerns restricted the room for agreement severely. Polity interests trumped policy interests. These negotiation dynamics are in line with the theoretical expectations about the response to polity attacks.

The new settlement and the referendum

The initial negotiations on the British demand for a ‘new settlement’ were still conducted in an integrative mode. It was clear that Cameron preferred the UK to be in the EU, and so did the other member states. Cameron asked for moderate concessions, and the other member states were willing to give him something to show at home (Weiss and Blockmans 2016). In February 2016, the European Council reiterated its 2014 conclusions on ‘ever closer union’, underlined that ‘the references to ever closer union do not apply to the United Kingdom’ and clarified that the principle did not provide a legal basis for the informal expansion of EU competences.⁵ On free movement, the EU agreed to a temporary ‘emergency brake’ for access to non-contributary in-work benefits and the possibility to index child benefit exports. But these measures would only apply to future migrants and require authorisation by the EU.

These limited concessions failed to appease the opponents of membership and those concerned about immigration. A formal UK opt-out from the freedom of movement, or even the possibility to introduce immigration quota under exceptional conditions, might well have changed the outcome of the referendum given the centrality of the issue and the defection of leading Conservatives to the Leave camp after the renegotiation (Evans and Menon 2017: 50). Yet the other member states defended the integrity of the internal market, including the free movement of workers, as a non-negotiable principle of the EU (Weiss and Blockmans 2016: 9).

The referendum paved the way for the further domestic politicisation of British European integration and for the intensification of the British attack. It turned EU membership, which had been a low-salience issue among British voters, into a polarising and identity-forming issue on par with, and cutting across, party identification (Hobolt et al. 2021). It also delivered a narrow popular majority in favour of leaving the EU, for which there had not been a majority in Parliament. Cameron’s moderate attack turned into a vote for full exit.

According to functionalist theory, the preferences of domestic economic interest groups and the economic costs of leaving the EU should have carried the day in the referendum. Indeed, the Remain campaign focused on the negative economic consequences of Brexit. Economists from international organisations such as the OECD and the IMF, the British Treasury, think tanks, and private consultancies were in almost full agreement about the harm that Brexit would do to the UK economy. In addition, the UK business community and its major interest groups were overwhelmingly in favour of remaining in the EU (Jensen and Snaith 2016: 1304–5). By contrast, the Leave campaign focused on self-determination and identity issues, above all immigration, which resonated strongly with the major concerns of the Leave voters (Clarke et al. 2017: 161–5). The success of this campaign corroborates the postfunctionalist logic of the Brexit crisis.

The vote did not prescribe any specific form of exit, however. In a functionalist perspective, the UK and the EU would have been expected to make the best of the ‘referendum accident’ and strive for a ‘soft’ solution that would keep the UK in the internal market and customs union and thereby minimise the economic and other policy costs of Brexit (Hix 2018). Yet Theresa May, the new prime minister, came out in favour of ‘hard Brexit’. She excluded continued membership in the internal market, which would have meant accepting EU legislation, jurisdiction, and the freedom of movement for labour. She also vowed to pull out of the customs union that prevented the UK from concluding its own trade agreements. This position gave clear priority to the polity goals of enhancing sovereignty and curbing immigration. In her withdrawal letter to the EU, Theresa May called the referendum ‘a vote to restore […] our national self-determination’. At the same time, the UK sought ‘the greatest possible access’ to the internal market through a ‘new, comprehensive, bold and ambitious free-trade arrangement’. May spoke of a ‘deep and special partnership’, including ‘both economic and security cooperation’ and covering areas ‘crucial to our linked economies such as financial services and network industries’ – the two sectors that would suffer most from the UK’s exit from the internal market.⁶

This ‘cakeist’ attempt to preserve a maximum of policy benefits from European integration while minimising the polity costs of supranational regulation, which continued throughout the negotiations with the EU (Grey 2021), met with principled EU resistance. In its Guidelines for the Brexit negotiations, the European Council prioritised the ‘integrity of the Single Market’, ruled out a ‘sector-by-sector approach’ and called for a level playing field in terms of state aid and competition. To a large extent, the EU reaction is compatible with functionalist expectations: the EU had an interest to prevent ‘cherry picking’ and contain demand for exemptions in the remaining member states as well as from non-members in the European Economic Area and Switzerland whose access to the EU market is conditional on the free movement of workers.⁷ At least for some EU leaders, a tough EU bargaining position also served to deter potential emulators. Hard Eurosceptic parties across the EU celebrated the Brexit vote, considered it a boost for their cause and called for EU referendums in their own countries. To prevent that ‘other countries or other parties will consider leaving the European Union to obtain the alleged advantages’, French President Francois Hollande insisted that ‘(t)here must be a threat, there must be a risk, there must be a price to pay’ (Hollande 2016: 7).

Yet the EU-27 went on to ideologise the integrity of the internal market and the indivisibility of the four freedoms and turn them into non-negotiable principles and ‘key components of the EU constitutional identity’ (Hillion 2018: 50). Both economic theory and institutional practice question the functional rationale for indivisibility. In theory, even though the removal of all types of barriers to free movement offers the highest benefits of market integration, individual freedoms are largely substitutable and not so interdependent that the limitation of any one freedom would eliminate the benefits of the others. Currency areas are considered an exception, but the UK was not a member of the Eurozone (Felbermayr et al. 2019: 9; Kohler and Müller 2017). In practice, the EU has not implemented the four freedoms equally and simultaneously either (Barnard 2017: 203–4). Moreover, Art. 63 TFEU prohibits ‘all restrictions on the movement of capital … between Member States and third countries’, regardless of the state of liberalisation of the other market freedoms. In a functional perspective, an integrated EU-UK market without an unlimited freedom of movement for persons would thus have been both technically feasible and economically preferable to a mere free-trade agreement. Rather, the EU stance reflected its general polity interest to tolerate differentiated integration if it helps moving supranational integration forward but oppose it if it undermines the achieved level of integration.

Moreover, the Brexit challenge triggered a general urge among EU actors to close their ranks, demonstrate their unity and accentuate the dividing line between EU members and non-members. Michel Barnier, the EU’s Chief Negotiator, defined several ‘first principles’ before the start of the negotiations that found strong support among the member governments, namely, that negotiating success will depend on ‘very strong unity between the twenty-seven member states’, ‘no EU country should find itself in a position where it has less say than a country outside the Union’ and ‘no country outside the Union should be given a veto on, or even the right to intervene, in the decision-making process of the twenty-seven’ (Barnier 2021: 22). The EU based its negotiation stance and narrative on the principles of integrity, unity, membership, and autonomy, while it distanced and othered the UK as a ‘third country’ (Hillion 2018; Laffan 2019). Procedurally, the EU rejected separate, bilateral negotiations with the UK in favour of a ‘single package’ and a ‘single channel of communication’. The member states conferred an exclusive horizontal competence to the EU and its institutions for dealing with all matters of withdrawal (Hillion 2018). Thus, in response to the Brexit vote, the EU immediately turned to rallying the 27, othering the UK, ideologising the indivisibility of the market freedoms, and centralising the withdrawal negotiations.

Withdrawal negotiations

In the functionalist analysis, the EU possessed superior material and institutional bargaining power and was therefore expected to get its way in the withdrawal negotiations (Schimmelfennig 2018). Even though the UK was one of the largest, wealthiest and most globally oriented member states, in 2015, 44 percent of UK exports (in goods and services) went to the EU and 53 percent of UK imports came from the EU. By contrast, the UK only accounted for six to seven percent of EU exports and four to five percent of EU imports (Office for National Statistics 2016). Given its size and geographical proximity, the EU was bound to remain the UK’s most important foreign market. Projections of the consequences of Brexit generally found that, while both the UK and the EU would lose from market disintegration, British losses would be higher than those of any EU member state or region (Chen et al. 2018; Latorre et al. 2019) and would be the more severe the harder the Brexit turned out to be. In the functionalist analysis, then, UK threats of ‘no deal’ or outside agreements were irrational or lacked credibility. The UK was expected to soften its hard Brexit stance and compromise with the EU to reduce the economic costs of leaving.

Institutional factors reinforced the EU’s material bargaining power. First, any treaty with the UK required a consensus among the member states. This decision-making rule accords veto power to the most status quo-oriented member state. In the Brexit crisis, the state that was potentially most negatively affected by the exit of the UK defined the limits of EU concessions. This state was the Republic of Ireland, which received a strong EU backing on making the openness of borders on the island of Ireland a key condition of an orderly UK withdrawal and a follow-up agreement with the EU (O’Rourke 2018: 236–7, 248–57). Second, any agreement replacing the UK’s membership in the EU would be negotiated by the European Commission and needed the consent of the European Parliament – two supranational bodies with a vested interest in protecting and defending supranational integration. Finally, Article 50 limits withdrawal negotiations to two years, after which the exiting state crashes out of the EU without a follow-up agreement.

The initial negotiation process appeared to corroborate functionalist expectations. The May government made increasing concessions on the negotiating procedure, the honouring of financial obligations, the rights of EU citizens in the UK, and the Irish border question, pledging to ‘maintain full alignment with the rules of the Internal Market and the Customs Union’ for Northern Ireland.⁸ To accommodate the Northern Irish unionists, on which her parliamentary majority depended after the 2017 elections, May further envisaged keeping the entire UK in a customs union with the EU rather than establishing an economic border between Great Britain and Northern Ireland. The Chequers Plan of July 2018 was a culmination and turning point.⁹ It proposed setting up a free-trade area for goods that not only included a ‘facilitated customs arrangement’, but also the adoption of EU market rules, UK participation in the standard-setting EU agencies and ‘due regard’ to EU court decisions. If it had come to fruition, it would have significantly softened Brexit and mitigated the Irish border problem.

The Plan, however, crossed or blurred several of the UK’s red lines and came under fire from Brexit hardliners in the Conservative party. Brexit minister Davis and foreign minister Johnson resigned from their posts, and May was forced to accept several amendments. Donald Tusk duly dismissed it as unworkable at the September 2018 EU summit, ‘not least because it risks undermining the single market’ (O’Rourke 2018: 270–6). Notwithstanding the technical uncertainties of the Chequers Plan and other pragmatic and flexible solutions proposed during the negotiations, it proved politically impossible to find an agreement that would preserve the integrity and autonomy of both the EU’s and the UK’s internal markets together with open borders on the island of Ireland. The negative domestic and European reactions to the Chequers Plan effectively ended the softening of Brexit.

The draft withdrawal agreement of November 2018 provided for a transition period until the end of 2020, during which the UK would leave the institutions of the EU but keep the material rights and obligations of a member state. Afterwards, failure to agree on the post-transition EU-UK relationship would activate the ‘Irish backstop’, keeping the entire UK in a customs union, and Northern Ireland in regulatory alignment, with the EU. The backstop could only end by consensus and thus accorded the EU and, by extension, the Republic of Ireland veto power over changes to the status quo. This provision again demonstrated the EU’s bargaining power but triggered the resignation of Brexit minister Dominic Raab and other cabinet members.

Squeezed between Brexit Ultras and opponents of hard Brexit, Prime Minister May failed three times in 2019 to obtain a majority for the withdrawal agreement in the House of Commons. In July 2019, May was replaced by Boris Johnson who stood for a harder line on Brexit and won a snap election in December 2019 with a landslide on the promise to ‘get Brexit done’. In January 2020, the new House of Commons accepted the withdrawal agreement. The main difference to the November 2018 draft was Johnson’s agreement to an ‘Irish backstop’ that would place the customs border in the Irish Sea rather than keeping the entire UK in a customs union with the EU. This change, which corresponded to what the EU had initially proposed, was possible because the new government majority did not depend on the Northern Irish unionists anymore, and it helped win the support of English hardliners.

In line with functionalist expectations, the EU appeared to shape the intergovernmental withdrawal negotiations thanks to its superior material and institutional bargaining power. But because each concession of the UK government to the defenders of supranational integrity and integration in the EU eroded its support among the defenders of national integrity and sovereignty in Parliament, the EU’s bargaining power ultimately made little impact. In line with postfunctionalist expectations, the constraining domestic dissensus about the autonomy price to pay for economic gain undermined efficient intergovernmental bargaining (Biermann and Jagdhuber 2022; König 2018). If Cameron’s referendum and May’s hard Brexit threats were conceived as strategic moves to extract a maximum of concessions from the EU, they backfired spectacularly. They not only failed to impress the EU, but their failure also undermined the domestic standing of the UK government and its ability to sell compromises at home (Martill 2022).

Developments in public opinion paralleled and supported the disintegrative intergovernmental bargaining. In the UK, the 2016 vote crystallised into stable issue-specific political identities separating Remainers and Leavers (Hobolt et al. 2021). Remainers not only saw their economic concerns vindicated, but also cultivated their European identity after the vote (Sczepanski 2022). And even though Leave voters came to understand that the EU wielded superior bargaining power and was less accommodating than they had been let to believe, they remained steadfast in their support for Brexit (Grynberg et al. 2020; Walter 2021a). In sum, motivated reasoning and social identification dynamics immunised ‘hard Brexit’ against the experience of political and economic difficulties.

In the EU, a rise in public support was immediately visible before negotiations had even started (De Vries 2017), and it manifested itself most clearly as the difficulties of the UK became obvious (Glencross 2019). The absence of negative domestic politicisation in the EU-27 and overwhelming citizen support for the EU’s hard bargaining strategy reinforced intergovernmental unity (Altiparmakis et al. 2022; Walter 2021b). Correspondingly, most Eurosceptic parties that had celebrated the Brexit vote and hoped to benefit from a transnational demonstration effect saw their expectations disappointed and ceased to advocate exit from the Eurozone or the EU (Chopin and Lequesne 2021: 428; Miró et al. 2022).

Trade negotiations

The negotiations on the future relationship between the UK and the EU in 2020 continued the pattern set by the withdrawal negotiations. Both sides agreed on achieving free trade on goods, which was already a minimalist objective. In addition, however, the UK government sought to maximise sovereignty, whereas the EU strove to protect the integrity of its internal market. The negotiating mandates of both sides, published in February 2020, displayed a wide gap in preferences, particularly on regulatory alignment and governance arrangements.¹⁰

The EU wanted the UK to keep its market regulation – e.g. for labour rights, state-aid rules, and environmental protection – aligned to the EU to preserve a post-Brexit level playing field. It also envisaged a dispute settlement system involving the Court of Justice of the EU (CJEU) and sanctions in case of UK non-compliance. On the other hand, the UK claimed its ‘right to diverge’ and rejected regulatory alignment, unilateral sanctions, and any role of the CJEU. Yet it sought EU recognition of equivalence on numerous issues ranging from manufacturing via professional qualifications to financial services.

In the Trade and Cooperation Agreement (TCA) concluded in December 2020, the UK government largely achieved its goal of regaining formal sovereignty. The UK insisted successfully on detaching its domestic laws and the TCA from EU law and jurisdiction. In return, however, the UK paid a price on the ease and extent of market access.¹¹ The EU accepted UK demands for mutual recognition only to a limited extent. As a result, non-tariff barriers to trade and services and regulatory costs rose. In particular, the agreement excluded financial services, in which the UK is highly competitive and has benefitted from a large surplus in the past. UK-based financial services lost their unfettered access to the EU market and came to depend on unilateral and revocable EU decisions. So far, the EU has held back any decision of equivalence. In sum, where the UK regained economic sovereignty, the EU maintained market integrity. Northern Ireland is the only major area, in which these two principles continue to clash.

Crisis outcome: external rebordering

The Brexit crisis has resulted in a major external rebordering of the EU. It consists of far-reaching EU-UK boundary closure, a steep reduction of cross-boundary transactions, a moderate strengthening of EU boundary control, and increasing boundary congruence.¹²

As for boundary closure, Brexit has ended the customs union and the freedom of movement for goods, services, and workers (but not for capital). Whereas the TCA provides for zero tariffs and quota on goods, it raises non-tariff barriers and requires border controls to check compliance with rules of origin and product standards. Correspondingly, the UK’s trade in goods with the EU was depressed by 11–16% in 2021 (Springford 2021); trade in services suffered even more.¹³ Since the Brexit vote, EU net migration to the UK has decreased to pre-Eastern enlargement levels.¹⁴ Moreover, Brexit ended the representation of the UK in EU institutions, UK participation in EU decision making and – with a few exceptions – its obligations under EU law and the jurisdiction of EU courts. After having been a member state for 47 years, the UK’s boundaries with the EU are now more closed than those of any other Western European country.

Moreover, Brexit has moderately strengthened the supranational control of the EU’s external borders. Whereas exit remains a fundamental right of individual member states, the management of the exit process was delegated to the EU and its institutions in what Hillion describes as a ‘quasi-communautaire procedural arrangement’ (Hillion 2018: 31). The institutional elaboration of the Article 50 withdrawal process has affirmed that exit negotiations are an exclusive competence of the Union and are to be conducted by the European Commission.

Finally, Brexit has restricted the differentiated integration of the EU and thereby increased the congruence of EU boundaries. As for internal differentiation, the number of treaty-based differentiations dropped by 13% when the UK left (Schimmelfennig and Winzen 2020: 20); the exemptions from ‘ever closer union’ it had negotiated became null and void; and the opt-out countries lost political clout. As for external differentiation, Brexit has led the EU to rethink and consolidate its relationships with third countries. During the Brexit negotiations, the EU team produced a ranking of EU-third country arrangements, dubbed ‘Barnier’s staircase’, that was modelled on existing relationships but reduced both their number and their actual flexibility. It suggested that the UK would have to choose from a fixed set of arrangements rather than being able to negotiate a bespoke agreement. The staircase model is likely to guide future negotiations – and limit flexibility – with other third countries. The same applies to the principles of the integrity of the single market and the indivisibility of market freedoms that the EU reaffirmed in response to Brexit. The Brexit process thus exemplifies the expected ambivalent integration outcome of polity attacks. The disintegration of the attacker went together with the strengthened integration of the defenders.

Conclusions

This article has addressed the dual theoretical puzzle of Brexit. Whereas functionalist theories struggle with the disintegrative outcome for the UK, postfunctionalism neglects its integrative outcome for the EU-27. I have argued that the problems of functionalist theories result from the fact that Brexit represents a polity attack – a type of integration crisis that postfunctionalism captures better than functionalism. At the same time, a complete postfunctionalist explanation needs to include the reactions of the defenders of the supranational polity. Community concerns are not an exclusive domain of domestic politics and nationalists. Among the supranational defenders of the community, polity attacks trigger social processes such as rallying and othering and institutional processes such as ideologisation, institutionalisation and centralisation. In the Brexit case, they have led to the external rebordering of the supranational community: the exclusion of the attacker and the increased closure, control, and congruence of its external boundaries.

In the sociological analysis of polity formation going to back to Rokkan (1974) and elaborated by Bartolini (2005), such external rebordering is an important condition for political development. External rebordering strengthens in-group identity and conformity as well as incentives for in-group social cooperation and the production of collective goods (Bartolini 2005: 40–7). In this perspective, Brexit may provide several integrative effects to the EU. First, it has removed an identity outlier, a powerful opponent of centralisation and a strong advocate of differentiated integration. This is likely to facilitate uniform decisions in favour of supranational integration in the future. Second, the reaction of the EU to the Brexit challenge has accentuated the core identity and principles of the Union and reduced incentives for exit among the remaining members. This is likely to deter dissatisfied member states and citizens from considering exit as a remedy. It will also make it more difficult for non-member states to share in the benefits of European integration without accepting quasi-member state obligations. Future research on the long-term effects of Brexit should address these expectations.

The Brexit case study further calls for comparative research on the EU’s crises. The development of the Brexit crisis provides initial support for a synthesis of functionalist and postfunctionalist explanations of integration crises that takes the type of crisis as the scope condition for the application of theories. Integration crises come in two major varieties, and whereas postfunctionalist theorising fits polity attacks, functionalism explains policy failures. However, Brexit is just a single case representing only one of the crisis types. It remains to be seen to what extent the distinction of polity attacks and policy failures can be fruitfully applied to other crises of integration and help explain similarities and differences in the crisis processes and outcomes. Obviously, Brexit is not the only case of a polity-attack crisis of European integration. Like Brexit, the Russia crisis and the rule of law crisis originated in the resistance of nationalist and Eurosceptic actors against core principles of European integration and the loss of national control resulting from supranational integration. In the Russia crisis, Russia’s imperial and autocratic nationalism clashes with EU democracy support and the gradual expansion of European integration in Eastern Europe; and the rule of law crisis resulted from an attack of nationalist backsliding governments in Hungary and Poland on institutional checks and balances, in particular the independence of the judiciary, a fundamental value of the EU and a prerequisite for the EU’s ‘integration through law’. Clearly, none of these crises resulted from a failure of integrated policies to achieve their goals, but from the perception of the attackers that European integration had gone too far.

According to the postfunctionalist scenario for polity attacks, we should observe similar processes and outcomes in all three crises – the prioritisation of polity interests and community defence at the expense of material policy benefits and a disintegrative dynamic producing external rebordering. At the same time, the attackers pursue different aims. In contrast to the UK, the rule of law backsliders do not demand a formal renegotiation of the treaties (let alone exit), and Russia intends to roll back the European integration of states in its vicinity. This variation in the goals of the attacks – and the threats of disintegration they pose – may well introduce variation in the crisis management and response of the EU.

Acknowledgements

I gratefully acknowledge financial support from the Jacques Delors Centre, Hertie School of Governance, Berlin (Michael Endres Visiting Professorship) and the European Research Council (ERC Advanced Grant No. 101018300, EUROBORD), and I thank Brigid Laffan, Waltraud Schelkle, Stefanie Walter as well as the anonymous reviewers of the journal for their helpful comments.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Frank Schimmelfennig

Frank Schimmelfennig is Professor of European Politics at ETH Zurich, Centre for Comparative and International Studies. His recent books include Ever Looser Union? Differentiated European Integration (OUP, 2020, with Thomas Winzen) and Integration and Differentiation in the European Union (Palgrave 2022, with Dirk Leuffen and Berthold Rittberger). [[email protected]]

Notes

1 See the migration statistics at https://www.migrationwatchuk.org/.

2 https://www.gov.uk/government/speeches/eu-speech-at-bloomberg.

3 PM letter to President of the European Council Donald Tusk (publishing.service.gov.uk).

4 143478.pdf (europa.eu).

5 European Council meeting (18 and 19 February 2016) – Conclusions’ (EUCO 1/16).

6 Prime Minister’s letter to Donald Tusk triggering Article 50, https://www.gov.uk/government/publications/prime-ministers-letter-to-donald-tusk-triggering-article-50.

7 See Angela Merkel’s address to the congress of the Bundesverband für Gross- und Aussenhandel, The Guardian, 5 October 2016, https://www.theguardian.com/politics/2016/oct/05/angela-merkel-takes-significantly-tougher-brexit-stance.

8 ‘Joint report from the negotiators of the European Union and the United Kingdom government on progress during phase 1 of negotiations under Article 50 TEU on the United Kingdom’s orderly withdrawal from the European Union. TF50 (2017) 19 – Commission to EU 27’, https://ec.europa.eu/commission/sites/beta-political/files/joint_report.pdf.

9 The future relationship between the United Kingdom and the European Union (publishing.service.gov.uk)

10 ‘Directives for the Negotiation of a new partnership with the United Kingdom of Great Britain and Northern Ireland’, 25 February 2020, https://www.consilium.europa.eu/media/42736/st05870-ad01re03-en20.pdf; ‘Our approach to the Future Relationship with the EU’, 27 February 2020, https://www.gov.uk/government/publications/our-approach-to-the-future-relationship-with-the-eu.

11 ‘Brexit trade deal means “freedom”, but at a cost: the arguments will be far from over’, The Guardian, 27 December 2020, https://www.theguardian.com/politics/commentisfree/2020/dec/27/britain-trade-deal-eu-arguments-brexit-renegotiation.

12 See Schimmelfennig (2021) on these dimensions of bordering.

13 ‘How a year outside the EU’s legal and trading arrangements has changed Britain’, The Economist, 1 January 2022.

14 Statistics-net-migration-statistics | Migration Watch UK.