Between two unions: UK devolution, European integration and Brexit

ABSTRACT

The UK in recent decades has been the site of radical reform in governing structures. There has been privatization, welfare retrenchment and repeated managerial reform. Redistributive regional policies have been run down. Devolved institutions have been established in Scotland, Wales and Northern Ireland, but the territorial governance of England remains unresolved. The nature of the territorial state remains disputed as some see it as a continuing unitary state while others see it as a pluri-national union or emerging federation. Devolution was sustained by membership of the European Union, which provided an external support system, ensured economic union and allowed issues of sovereignty to remain in abeyance. Brexit puts immense centrifugal pressures on the union, but neither fragmentation into separate states nor recomposition as a unitary state are straightforward alternatives. The UK thus presents a stark example of the pressures and conundrums of rescaling, legitimacy and sovereignty facing contemporary European states.

KEYWORDS:

• UK
• devolution
• rescaling
• cohesion
• unions
• sovereignty

STATE TRANSFORMATION AND THE UK

The UK over recent decades has experienced, in concentrated form, most of the challenges and transformations facing modern European states. From the 1980s a set of reforms aimed at ‘rolling back’ the interventionist state in favour of markets and competition. Public administration and services were shaken up on ‘new public management’ lines, with the introduction of privatization, contracting out and competition. The principles of social welfare were challenged, with the concept of social security giving way to a more selective benefits regime accompanied by punitive sanctions, although overall welfare spending remained high. Integrative spatial development policies were run down. The global position of the UK changed as it withdrew from empire and reduced its global role, belatedly joined the European Union (EU) (as it became) and then, nearly 50 years later, withdrew again.

These challenges form the backdrop to a major reconfiguration of the UK as a territorial polity. It represents a striking example of spatial rescaling, the process by which functional systems, citizen identities and demands, policy communities, and governing institutions migrate to new levels, above, below and across the territorial state (Keating, 2013). Successive strategies of territorial management have sought to reconcile national diversity with successive regimes of economic management and social welfare, from the post-war welfare settlement, through the neo-liberal (Crouch, 2011) phase of Conservative governments, to the revised social democracy of the New Labour era. Devolution reforms, which failed in the 1970s and were enacted at the end of the century, represent a major effort to stabilize territorial management, politics and policy. These reforms were partial and not accompanied by a wider programme of constitutional renewal, and they left in abeyance important issues about the nature of the state, sovereignty and territorial cohesion. Devolution occurred during UK membership of the EU, which provided an important external support system for the new territorial dispensation as well as opening the prospect of new modes of multiscalar politics. UK withdrawal from the EU, by contrast, represents an effort to reinvent the unitary, sovereign nation-state, which the UK never really was, and certainly has not been since the end of the 20th century. It thus destabilizes the devolution settlement and territorial politics across the UK and the divided island of Ireland.

The articles in this special issue are the product of research which most of the authors have been undertaking since devolution began, but specifically of a collaborative project, ‘Between Two Unions (2017–20)’, examining the interplay of devolution, European integration and Brexit. The papers address the various dimensions of union, the devolution settlement, and the continuing tug of centralizing and centrifugal forces. While they are all focused on the UK, they do pose critical questions about the meaning of political, economic and social union, sovereignty, and authority in a rescaling Europe.

UNDERSTANDING THE UK

At one time, the UK was widely seen in the social sciences as the paragon of the unitary nation-state, the apogee of the modernist paradigm which identifies modernity with territorial integration and functional differentiation (as in Durkheim, 1964). For Blondel (1974) it was the most homogeneous of industrial societies, while Finer (1974, p. 137) argued, ‘Britain, too, has had its “nationalities” problem, its “language” problem, its “religious” problem. These are problems no more.’ Most European states have a national historiography, and for the UK this was the influential ‘Whig’ school (Butterfield, 1931/1968), which presented the history of the UK as one of continual progress to liberty and democracy, starting from England. The peripheral nations of Wales, Scotland and Ireland had walk-on parts when they joined the union, after which progress proceeded uninterrupted. UK institutions were essentially English institutions, with the other nations invited to join in.

Yet, there has always been an alternative to this nationalizing teleology. This other perspective sees the UK not as a unitary nation-state but as a pluri-national union, forged in a series of constitutional deals, from the incorporation of Wales in the 16th century, to the unions with Scotland (1707) and Ireland (1801). Successive unions brought about parliamentary unification, but left much of the cultural and institutional infrastructure and civil society of the smaller nations intact, to develop on their own within union. The unions themselves were very different. Wales was legally incorporated into England, while Scotland and Ireland remained distinct jurisdictions. Ireland was governed in quasi-colonial mode, while in Scotland local elites had great sway. Rokkan and Urwin (1983) introduced the term ‘union state’ to capture this kind of arrangement, in which pre-union entities retain some of their historic institutions and prerogatives. It also has echoes in the ‘two nations’ doctrine of Canadian federalism (McRoberts, 2019) and Spanish ideas about the ‘historic rights’ of the component territories (Herrero de Miñon, 1998). In both Ireland and Scotland, there are doctrines of contractual unionism, which present union as a pact, the terms of which are binding on both sides. The idea has surfaced, however, only in times of tension, as between the 1880s and the First World War, the period from the 1970s to the end of the 20th century, and again in recent years. It is at these times that abstract doctrines are mobilized as instruments of power.

The concepts of nation-state and union are, of course, ideal types and the UK can, at various times, be appreciated using one or the other. They might also be seen as points on a spectrum, denoting more of less unity. As such they provide distinct takes on three constitutive elements of any polity: demos, telos and ethos (Keating, 2021). Demos refers to the people (in the singular) in which sovereignty is ultimately vested, and is thus the foundation of legitimate authority. Traditionally, sovereignty in the UK has been vested in the monarch-in-parliament, so avoiding the tricky question of defining the people. At times of crisis, when parliamentary sovereignty is contested, however, even hardened supporters of the doctrine such as Dicey (1959) fall back on the idea that it is a proxy for popular sovereignty. This, in turn, implies the need for a unitary people. A union, on the other hand, can accommodate multiple demoi (Bellamy, 2019; Nicolaïdis, 2012, 2013). Telos has two dimensions. One looks to the past. State-focused historiographies link integration to a linear progress; union-based historiographies focus on accommodation, compromise and entrenched rights. The second dimension looks to the present and future and rests on instrumental arguments about how union contributes to desirable substantive goals. Ethos refers to shared values, which may underpin citizenship and belonging. While nation-state constitutions often claim common normative foundations, in unions it these become more contentious. This may be because the values differ but that is not necessarily the case. The argument in modern times is often about the level to which these (often universal) values should be attached. We have seen all these issues argued in the debates about the need for, or content, of a European constitution.

While the ideal-type unitary state unites polity, economy and social welfare within the same territorial boundaries, in unions these do not coincide so easily. Political union refers to governing institutions, their composition and the scope of their authority. There is an affinity between the idea of union and federalism, but unions (like the pre-1998 UK) may have unitary constitutions, while not all federations are unions in the sense used here. Rokkan and Urwin (1983) distinguish between ‘mechanical federalism’, which is uniform and symmetrical, and ‘organic federalism, built from below by contract among territories.

Before devolution, the UK squared the circle by combining a single, unitary Parliament with symbolic recognition of the component nations and extensive administrative devolution, with most domestic policy entrusted to territorial Secretaries of State, whose task was to adapt policy to the distinct conditions of Scotland, Wales and (before 1922) Ireland. From the First World War, the system worked smoothly when there were concurrent electoral majorities in England, Scotland and Wales, but came under strain when the Conservatives governed the UK by virtue of their English majority, but Labour won in the smaller nations. Northern Ireland worked in a different mode, governed by a devolved Parliament from 1922 until 1972 and thereafter by a Secretary of State, who was always an MP sent over from Great Britain.

Economic union refers to market integration and has changed and expanded in meaning as the interventionist role of the state has waxed and waned. The union legislation for Wales, Scotland and Ireland provided for economic union within a laissez-faire regime. The currencies were unified, tariffs were abolished and there were some rather general clauses about freedom of trade and establishment. In the modern era, economic union has taken on two additional meanings. The first is ‘territorial cohesion’ (a term introduced by the EU) meaning active measures to economic conditions across the union. The second is, again using EU terminology, an internal market, which refers to the removal of regulatory barriers to trade, to fair competition across the territory and to government interventions including procurement.

‘Social union’ is also a debateable concept. The term has sometimes been used to refer to the private domain of family and civil society. In that sense, important elements of the union survived the independence of the larger part of Ireland in 1922. It is also used to refer to what in European parlance is called ‘social cohesion’, the counterpart to the competitive market. The terms ‘welfare union’ or ‘sharing union’ are used in the UK debate. The welfare state is recognized as having had a centralizing influence in most countries (Banting, 2006; Greer, 2006). One reason is that it draws on ideas of affective solidarity located in the unitary demos. Another is that the bounded territorial state locks in both capital and labour, preventing ‘partial exit’ (Bartolini, 2005) and so encouraging positive-sum social compromises. A third is that decentralization of redistributive policies might encourage competitive regionalism, in which substate governments bid for mobile capital and wealthy taxpayers through low taxation and spending and lax regulation. The welfare state in the UK, starting in a small way in the early 20th century and coming to fruition after the Second World War, was a centralized affair, consciously linked to ideas of common social citizenship. Even Northern Ireland, where most competences were devolved, followed British welfare policies, with British taxpayers meeting the bill.

The closest the UK has come to the unitary ideal was in the post-war decades of the ‘Keynesian welfare state’. The political union was largely unquestioned in Scotland and Wales. Violations of civil rights in Northern Ireland were mostly ignored in Westminster, where a convention was established that questions about matters devolved to the province could not be entertained. This was, perhaps for the first time, a territorially delineated UK national economy and active state (Edgerton, 2018). The welfare compromise was broadly accepted and applied in a largely undifferentiated way. From the 1960s, active regional policy sought to divert investment to the lagging areas of the country, mostly located at the periphery, and to restrain development in the economic core, the South East of England. This was accompanied by regional economic planning machinery including representatives of local business and civil society. Yet even during these years, elements of differentiation remained. The Scottish Office (founded in 1885) expanded administrative devolution, and a Welsh Office was established on similar lines in 1965. After the collapse of the Stormont regime, a Northern Ireland Office tasked with the administration of the province.

Strains in this mode of territorial management were apparent from the late 1960s, and in the 1970s the UK government responded with a round of devolutionary reforms. The Northern Ireland regime collapsed as civil protest, initially demanding the same rights in the province as in the rest of the UK, transmuted into more traditional nationalist claims and into political violence. This triggered an attempt at a new settlement based on devolution and consociational power sharing (the Sunningdale deal). Devolved assemblies were proposed for Scotland and Wales, with the Scottish one having extensive legislative powers. Some reformers within government hoped that the regional planning machinery in England could evolve into elected regional government (as happened in Italy and France). None of these was successful. The Sunningdale deal was brought down by industrial militancy on the part of the Protestant working class. Devolution for Scotland and Wales was undermined by opposition within the governing Labour Party and referendum requirements. Welsh devolution was massively rejected by the electors of Wales and the Scottish proposals, while gaining majority support, failed to meet a threshold, imposed by Parliament, of 40% of the entire electorate.

THE MILLENNIUM SETTLEMENT

The failure of the devolution schemes did not put an end to territorial unrest. The governments of Margaret Thatcher after 1979 combined a strident British nationalism with measures that eroded the political and economic bases of union. Privatization and the embrace of globalization undermined the concept of the managed national economy. The welfare state was not dismantled, but the language of social citizenship was replaced with the old idea of the deserving and undeserving poor. Regional policy based on cohesion was replaced by the language of territorial competitiveness and the regional planning and development machinery abolished. Spending on regional grants of all sorts fell by over 90% (in constant prices) between its peak in the mid-1970s and the beginning of the 2000s (Wren, 2005). Spatial disparities, which had been declining since the 1960s, increased steeply (McCann, 2016). Promises made in 1979 to revisit Scottish devolution (Heald & Keating, 1980) were shelved, although the search for an agreed devolved solution for Northern Ireland was continued. The 18 years between 1979 and 1997 represented the longest period that Scotland had been governed by a party for which it had not voted, while Conservative fortunes in the periphery declined to the point that, in 1997, it held no seats at all outside England and scarcely any in the cities of northern England. The territorial division of the vote had returned to patterns familiar from the 19th century, reversing a pattern of convergence during the rise of Labour and the interventionist state (Field, 1997). A strong devolution movement emerged in Scotland and a weaker one in Wales based not on revolutionary new values but on defence of the British post-war welfare settlement. It was not even that the peripheral nations were strongly attached to these values while England had abandoned them. Surveys have consistently shown that Scots and Welsh are only on average slightly more to the left than the English, and that social democratic values survived Thatcherism in England. It is rather that, in the face of declining class attachment, territorial identity had become a new vehicle for articulating these values.

These circumstances ensured that devolution would be a priority for the incoming Labour government in 1997. The new proposals represented slightly stronger versions of the failed ones in the 1970s, but the politics had changed. Scottish devolution was supported in a referendum by three-quarters of those voting. The Welsh proposals passed a referendum by slightly over 50% on a 50% turnout, but this was an advance on 1979. In Northern Ireland, there was a large majority in favour among nationalists and a slim majority among unionists for the proposals embodied in the Good Friday Agreement (GFA). Scotland was given a parliament, with legislative powers over all matters not expressly reserved to the centre, and a cabinet form of government. Wales initially got an assembly with only secondary legislative powers over a list of defined competences and no separate executive although this model rapidly proved unworkable and there has been a steady progression towards the Scottish model. Northern Ireland was granted a legislative assembly on the same ‘reserved powers’ model as Scotland but with the intention that powers would be transferred over time. In line with consociational power-sharing theory, the executive must be drawn from the two communities (unionist and nationalist). The Northern Ireland one has an explicit external dimension to accommodate aspirations for an all-Ireland expression for politics. Citizens are encouraged to identify variously as Irish, British, Northern Irish of any combination, and these are given expression through new institutions. The Republic of Ireland is given a role in safeguarding and implementing the settlement. There are provisions for unification of the two parts of Ireland subject to referendums in each jurisdiction. Initially, none of the devolved bodies had significant fiscal powers but relied on a block allocation from Westminster under a non-statutory formula known as the Barnett Formula. The question of England was not dealt with in the initial settlement, but a section of the Labour government reopened the issue of regional devolution there. The first moves were made in the north of England, where demand was felt to be greatest, but a referendum for a weak assembly in North East England was lost by 80% in 2004, terminating the debate. Attention then moved to the idea of city-regions.

This millennium settlement was the culmination of over a century of intermittent debate about the territorial constitution of the UK. Hitherto, opponents had argued strongly that subordinate ‘national’ parliaments could not work in a unitary state based on the principle of parliamentary sovereignty and supremacy. It was not that they did not recognize the nationality of Scotland, Wales and, successively, Ireland and Northern Ireland. On the contrary, it was precisely the national basis of these territories that made it inevitable that such institutions would arrogate themselves the status of demos and claim sovereign rights (Dicey, 1913). Now this objection was put aside at a stroke and constitutional order did not collapse. Indeed, unionism in Scotland and Wales rapidly accepted the new dispensation and redefined itself in opposition to secession. Nationalism, which had hitherto spanned a wide range from autonomy to separation and gradualism to rupture, now defined itself as for independence. It was not long, however, before a new middle ground opened and various ‘third way’ solutions were being canvassed.

AMBIGUITIES, ABEYANCES AND ANOMALIES

Some constitutional settlements are based on shared foundational principles, including definition of demos, telos and ethos. Others are practical compromises, offering institutional mechanisms to work together as a union in the absence of foundational agreement. The UK’s territorial constitution must be seen as an example of the latter. There is no agreed teleological account of the devolution settlement since it must accommodate different long-term aims. Scottish devolution was introduced as a measure to stabilize union. Nationalists see it as a steppingstone to independence – a view shared with alarm by diehard unionists. The Northern Ireland settlement was deliberately designed to satisfy competing teleologies, allowing both continuing union and Irish reunification and was swathed in ‘constructive ambiguity’ (Powell, 2008). All constitutions include silences and abeyances (Foley, 2012), but the new UK settlement has more than most and is not underpinned by a written or codified document.

So several contrasting interpretations are available. For Westminster parties, it is a way of managing territorial diversity within a unitary state, a deeper form of the old territorial management. It does not in any way threaten the sovereignty and supremacy of Westminster, which has merely lent power to the smaller nations and can take it back at any time. A second interpretation is that this is a new stage in the politics of union, in which the component parts have negotiated a new compromise without surrendering their constituent rights as demoi. They are free to change matters again, including to secede, while recognizing that this would have to be negotiated. Although only the Northern Ireland settlement makes this explicit, few unionists have denied that Scotland could legitimately secede, although they argue endlessly about the conditions in which it might do so. There is a similar attitude with regard to Wales, although it is rather muted given the limited strength of the independence movement there. A third interpretation is that the new UK constitution represents a form of federalism, introducing new rules and practices that should be respected, not abolishing the principle of Westminster sovereignty but curtailing its practice. In Northern Ireland, there is yet another reading, which emphasizes that the GFA was part of an international agreement, key parts of which were deposited as a treaty at the United Nations. Yet another interpretation is that the UK settlement is a harbinger of a new post-sovereign order (MacCormick, 1999; Keating, 2001) in which sovereignty itself is shared and divided at multiple levels.

Appealing to constitutional law to resolve these ambiguities merely results in restating them since, as Wincott et al. (2022, in this issue) show. They unpack the debates on sovereignty in a historical context, emphasizing the return to the absolutist ideas associated with Dicey, although alternative narratives are available. The UK Supreme Court has consistently taken the Diceyean view that the Westminster Parliament is sovereign and supreme and that the devolved bodies are no more than creatures of statute. Yet this competes with other interpretations by academics and even (in writings and obiter dicta rather than judgements) of judges, which employ a wider epistemic, drawing on theories of union, practice and convention (Keating, 2021; Wincott et al., 2022, in this issue).

In practice, the matter of supremacy is dealt with by a legislative consent mechanism (the Sewel Convention), according to which Westminster will legislate in devolved areas (and by extension, alter the devolution settlement itself) with the consent of the relevant legislature. It was then extended to legislation changing the powers of the devolved bodies. There are other consent mechanisms governing the transfer of powers by statutory order, the powers being brought back from the EU and the UK Internal Market. One reading of the consent mechanisms is that they are a convenient measure to allow the two levels to deal with issues on which they are agreed. Another is that it is intended as a real restraint upon Westminster.

As long as these foundational issues could be left in abeyance, it might have been reasonable to conclude that, following successive revisions of devolution up to 2016, the UK was set on a federalizing trajectory, albeit of a highly asymmetrical kind. Yet, while the devolution settlement changed matters radically at the periphery, it changed almost nothing at the centre. The operation of Whitehall (government) and Westminster (legislation) carry on as before without any federal element. It is as though devolution is no more than a series of exceptions to the normal operation of politics. Indeed, it is easy for politicians and officials at the centre to forget that devolution exists. Intergovernmental relations are under-developed and ad hoc and there is little territorial input to policy at the centre.

This in turn has a lot to do with the question of England, discussed by Kenny (2022, in this issue). On the one hand, England is the effective centre of the emerging federal system, with a preponderant influence over the whole. On the other hand, England is one of the constituent nations of the union and a grievance has been articulated that it is being short-changed by not having its own institutions. It crystallized in the West Lothian Question, named after anti-devolution Scottish Labour MP Tam Dalyell who in the 1970s asked why he, as a Scottish MP, would be able to vote on domestic policies in England, but neither he nor English MPs would be able to vote on the same subjects in a devolved Scotland. The issue is of limited practical importance but has been used to demonstrate the incoherence of devolution. The English question will not go away as long as there is no consensus within England on what the question actually is, let alone the answer. A range of options have been canvassed for over a century, including federalism, regional government and altering parliamentary procedures. This provides a sharp contrast with Spain, where devolution to the ‘national regions’ was followed quickly by demands for equality elsewhere until the whole state was devolved, although that still did not satisfy the demands of the nationalities for special recognition. There is resistance in England to formal federalism since that would tie it down to common rules in the same way as applies to Scotland, Northern Ireland and Wales.

THE EUROPEAN FRAMEWORK

The devolution settlement was reached during the UK’s membership of the EU and was deeply embedded in it. For Brexiters, EU membership represents a contradiction with the unitary vision of the state based on parliamentary supremacy. If we see the UK as a union, however, there is a rather good fit between the two. Both the UK and the EU are asymmetrical, quasi-federal unions in which demos, telos and ethos are continually contested and renegotiated and where the locus of sovereignty is disputed. In this way, they can be seen as post-sovereign polities defying the usual constitutional categories. EU membership provides a discursive space in which post-sovereign ideas of sovereignty can be entertained, although institutional opportunities for substate actors have disappointed enthusiasts for a Europe of the Regions or the Peoples. It also lowered the stakes for secession in both Scotland and Northern Ireland since the assumption was that the seceding territories would share an economic and regulatory space with England. Rescaling upwards to Europe and downwards to the substate level were thus largely in harmony.

In a more concrete way, EU membership allowed for a more extensive devolution settlement than would otherwise have been possible.¹ It is the EU that provides for the internal market within the UK itself, ensuring regulatory harmonization across the state. There are few framework laws and, in matters that are both devolved and Europeanized, the devolved territories had the same discretion in applying European norms that is available elsewhere to national governments. This includes agriculture, fisheries, environment, state aids and (except in Wales) justice and home affairs. It was the combination of the EU Single Market with the GFA that enabled the removal of the physical border structures between the two parts of Ireland and the free movement of goods and people. For the Republic of Ireland, EU membership had enhanced rather than curtailed sovereignty, giving it the status of a member state of a powerful union on equal terms with the UK and allowing it to reduce economic dependency on the UK. This coincided with a ‘post-nationalist turn’ in Irish political thought and its reinvention as a small state in a globalized world. Politicians and officials frequently note that participation in European meetings brought them together in a way that had rarely happened since the 1940s when Ireland left the Commonwealth.

Another important European influence was the European Convention on Human Rights (ECHR), which comes under the Council of Europe rather than the EU. This was incorporated into UK law in the Human Rights Act (1998) under which the courts are able to rule on whether laws or government action contravene the Convention. They cannot strike down non-conforming laws, but Parliament can use an expedited procedure to bring them in line. The Convention is entrenched in the devolution acts in a stronger way so that any court can disallow laws in devolved matters. This had the important effect of separating human rights from nationality or the idea of a unitary demos. A British Charter of Rights, long championed by the Conservative Party as a Eurosceptic and British nationalist alternative to the Human Rights Act, would link rights to Britishness. This would not work in Scotland or Northern Ireland, where substantial parts of the population do not feel British – indeed, the right not to be British is recognized in the Northern Ireland settlement. In fact, most of the challenges to devolved legislation have come under the ECHR rather than the devolution statutes and rulings have generally received broad assent. While the ECHR is not affected by Brexit, is part of the Eurosceptic agenda and is likely to return as an issue.

INSTITUTIONAL DYNAMICS

It has become a cliché to describe devolution as a process rather than an event, and all three settlements have evolved over time. Perhaps the most striking case is that of Wales, if only because it started with such a limited set of powers. As Rawlings (2022, in this issue) charts, the main drivers here were a growing consensus on the desirability of devolution, in spite of the weak initial mandate and the dysfunctions of the original system. In successive Acts, the National Assembly for Wales gained primary legislative powers, then moved to the Scottish system of reserved powers and received tax-raising competences. In the process, it changed its name to Welsh Parliament/Senedd. In Scotland, the main driver was the challenge of nationalism. The arrival of the Scottish National Party (SNP) as a minority government in 2007 led the unionist parties to set up a commission and concede more powers, notably over taxation. Further powers over tax and welfare were granted in the aftermath of the independence referendum of 2014, following the inter-party negotiations under the aegis of the Smith Commission. Devolution in Northern Ireland was dependent on inter-party agreement, with justice and policing not devolved until 2010. It was not until the St Andrews Agreement in 2009 that all parties were fully signed up There were two extended periods (2002–07 and 2017–20) in which, because of the breakdown of relations between unionism and nationalism, the institutions were out of operation. Murphy and Evershed (2021, in this issue) chart the broad acceptance of the GFA, which before Brexit had gained a plurality of support in both communities.

In Scotland and Wales, the devolved institutions have, over time, become a prime focus of political attention and loyalty. The Covid-19 epidemic of 2020–21 provided dramatic evidence of this, as the first ministers in the devolved territories took the lead and became the public face of government response, emerging as more trusted than their UK counterparts. Yet at the same time devolution appeared to stabilize the union as the different devolved settlements gained broad support.

Even the victory of the SNP at the elections of 2007 and its absolute majority in 2011 was not the result in an upsurge in support for independence (Keating & McEwen, 2017). Once the UK government had agreed to a referendum and the campaign was engaged in earnest, however, a different dynamic came into play. Support for independence grew dramatically and eventually reached 45% of the vote. Demand for constitutional change, then, is not independent of the supply of options available, something that was also seen in the Brexit vote. In the succeeding years, support for Scottish independence did not fall below this level. Yet, as McEwen (2022, in this issue) shows, the concept of independence remains rather fluid as nationalists have absorbed ideas of post-sovereignty and shared authority. The pro-independence side campaigned on what was dubbed ‘independence-lite’, promising to leave the UK political union but to remain within five other unions with its neighbours – the monarchical, the defence, the monetary, the European and the social. The No side promised more devolution. So rather than evidence of fragmentation, the project of Scottish independence might better be seen as an adaptation to rescaling.

BREXIT

The different perspectives were reflected in the fact that all the nationalist parties of the periphery favoured Remain in the EU referendum of 2016. This pro-Europeanism contrasted with a British nationalism on the Leave side postulated on a unitary demos. One of the striking effects of the referendum was that the pro-Leave side, faced with a pro-Remain Parliament, abandoned the doctrine of parliamentary sovereignty on which it had campaigned in favour of a doctrine of popular sovereignty. Yet the outcome, in which both England and Wales voted narrowly (53%) to leave, while Remain won in Scotland (62%) and Northern Ireland (56%), put this conception of the demos in question and reignited the issues of Scottish independence and Irish reunification. The response of the UK government was to insist that ‘the British people have voted for Brexit’ (May, 2019a, 2019b) and that the whole of the UK would leave together in a ‘red, white and blue Brexit’. Brexit thus proved highly destabilizing for the millennium settlement, although this was little appreciated in political circles in London.

In the immediate aftermath of the 2016 referendum, the SNP First Minister of Scotland called for a new independence referendum, and Sinn Féin demanded a ‘border poll’ (referendum on Irish reunification). Brexit did not produce an immediate rise in support for Scottish independence and, indeed, the SNP fell back in the snap UK General Election of 2017. Surveys showed little relationship between mass support for remaining in Europe and supporting independence (Prosser & Fieldhouse, 2017); indeed the relationship had always been weak (Keating, 2009, 2021). Brexit might have given the nationalists a new argument for independence, but it made independence more difficult in practice as, with Scotland inside the EU and the rest of the UK outside, there would be a hard border between the two. Nor did Brexit induce an immediate rise in support for Irish reunification, which for some years had been a minority preference even among the Catholic community.

In the longer term, however, centrifugal forces were strengthened as a link formed between Scottish independence and being in the EU (Curtice & Montagu, 2018; McCrone & Keating, 2021). Support for independence rose to over 50%, mostly because of the movement of Remain voters towards. Irish reunification once again became the preferred choice within the Catholic community and political circles in Dublin began to talk of it as real, and in many ways, alarming possibility. Even in Wales, which had narrowly voted to leave, there was talk of independence or at least growing ‘curiosity’ about it (Trystan, 2019).

UNION AND UNIONISM

Unionism in the UK is a capacious term, often unstated or banal (Kidd, 2008). One strand, found on both the right and the left, is assimilationist, looking towards the creation of a unitary nation and state. Others are ‘patriotic’, cherishing the traditions of the nations while some are contractual, providing conditional loyalty (Keating, 2021). Yet all are attached to the principle of the sovereignty and supremacy of Westminster, with no room for subordinate legislatures. However, much as devolutionists, from William Ewart Gladstone in the 1880s to Tony Blair in the 1990s, might insist that devolution was a way of cementing union, traditional unionists refused to believe it.

This makes it all the more remarkable that unionist opposition to devolution disappeared almost immediately once the spell was broken and the millennium settlement was introduced. Having defined itself against home rule or devolution for over a century, unionism had to redefine itself by opposition to secession. Devolution appeared to have clarified constitutional options by incorporating devolution into the definition of union, on the one hand, and, on the other, uniting nationalists around the next stage: independence.

Having embraced devolution, unionism needed to articulate a new conception of union. Unable to rely on banal assumptions about unity or tradition, neo-unionism has sought modern foundations encompassing demos, telos and ethos. Cetrà and Brown Swan (2021, in this issue) show how the unionist parties have struggled with the issue, falling back on instrumentalist justifications of union in the absence of deeper ideological resources.

More broadly, unionism has sought to attach itself to a higher ethos of shared values and norms, frequently referred to as ‘British values’. They include democracy, freedom, the rule of law and respect for difference. These are mostly universal values, but that is true of many national catalogues of values. The real difficulty is that these are precisely the values deployed by nationalists on the periphery.² There is a particular problem in Northern Ireland, where even the Democratic Unionist Party (DUP) has laid claim to the liberal teleology (Foster, 2018) but has defended socially conservative policies that contrast with those in the formerly deeply conservative Republic of Ireland. At the same time, unionists have engaged in a form of historical revisionism, arguing that unionism always sought to defend the nationalities and largely passing over its resistance to home rule (e.g., Robinson, 2020).

Following the Scottish independence referendum, the UK government elaborated a more muscular unionism (Kenny & Sheldon, 2022). One strand was to establish greater visibility in the devolved territories and to take more credit for its spending there. City and Region Deals, a programme pioneered in England to engage city leaders in regeneration projects, were rolled out in the devolved territories, allowing UK ministers to spend there and link with local governments. The Internal Market Act gave general powers for UK ministers to spend on devolved matters in Scotland, Wales and Northern Ireland. Moneys returned to UK control from the EU Structural and Cohesion Funds were to be spent by the centre and not distributed to the devolved governments. UK government ‘hubs’ were set up in prominent locations in Edinburgh, Belfast and Cardiff, as well as English regional cities, to provide a visible presence.

UK government responses to Brexit provided further signs of a unitary interpretation of the Constitution. Regarding this as a matter of foreign policy, it initially refused Parliament and, by extension, the devolved bodies, a role in triggering the process of withdrawal under Article 50 of the Treaty on European Union. The UK Supreme Court (2017) ruled that the Westminster Parliament must approve, but played down the role of the devolved legislatures. Instead of deciding the issue on narrow grounds, such as that EU withdrawal was not a devolved matter or that the situation was not ‘normal’, it chose to undercut the Sewel Convention altogether by declaring that it was merely a ‘political’ device. There followed a series of arguments over devolved powers and the lack of a role for the devolved administrations in the Brexit negotiations. The European Union Withdrawal Bill in 2017 initially proposed in effect that all competences coming back from the EU would go to Westminster, irrespective of whether they were devolved or reserved in the devolution settlements. Any that proved not to be needed might later be ‘released’ back. The argument was that as these competences had been subject to EU law, they had never been devolved anyway and that the UK framework would simply replace the EU ones. This idea was rejected in the devolved territories by nationalist and unionist parties alike, and eventually the UK government backed down from the blanket reservation while retaining the power to take back competences as necessary.

A broader issue concerns the UK internal market. The Internal Market Act of 2020 provides broad powers for UK ministers unilaterally to define the internal market, cutting across devolved competences, with only the usual weak and non-binding consent mechanisms. The White Paper introducing the proposals argued that the UK was a ‘unitary state’ (UK Government, 2020).

RECONFIGURATION

Devolution was intended as a ‘third way’ between secessionist nationalism and unionism, and it certainly corresponded to what most people on the peripheral territories wanted. The constitutional choice appeared to be between the new devolutionist unionism and separation. Yet a new middle ground emerged in Scotland and Wales, focused on more devolution or ‘devo max’, a vague term suggesting that only foreign affairs, defence and macro-economic stabilization should be reserved, with everything else devolved. This would mean that the key welfare compromises and decisions about the size of the public sector would be made at the devolved level. Polls consistently found this was the most popular option. More powers were indeed conceded to all three devolved bodies, but this appears to have had little effect on popular demands, while the question of England remains unanswered. There is no shortage of ideas for constitutional reform. Some unionists have returned to the idea of federalism, but, remarkably, nobody has presented an actual scheme. The Labour government in Wales, which is unionist, pro-devolution and pro-European, has been developing ideas about cooperative federalism, but these have not found an echo elsewhere. Some have suggested a Constitutional Convention and the Labour Party has taken up the idea. All these ideas, however, presuppose and fundamental agreement on the demos and telos underpinning the UK Constitution. They may indeed be agreement on social and political values, or ethos, but this does not answer the question about the level at which these should be realized.

Brexit poses a severe challenge to proponents of a ‘third way’ between the unitary state and secession, since the UK would either have to be in the EU or outside it. There were, however, efforts to square that circle. The Scottish Government (2016) put forward a proposal for Scotland to remain within the European Single Market on European Economic Area terms, while the rest of the UK came out. This would require it to gain all the competences required to manage Single Market matters and there would be elaborate measures to deal with goods passing between Scotland and the rest of the UK. The proposals were never tested or put to the EU because they were dismissed outright by the UK government on the grounds that there could not be any territorial differentiation in Brexit. The Scottish government did, however, bring in a ‘Continuity Act’ giving ministers powers to keep pace with EU regulations in devolved fields.

Differentiation was a more urgent matter in Northern Ireland, given the importance of the border question (Connelly, 2017; Murphy, 2018). On a narrow interpretation, there is little in the GFA about Europe. On a broader interpretation, it is Europe that has allowed the GFA to develop over time. Its most striking affect is the disappearance of the physical border and almost complete disappearance of the border issue (as opposed to internal tension) as a matter of contention. Its disappearance has transformed attitudes and relationships in border communities themselves (Hayward, 2018; Todd, 2018). More widely, the open border and European context have allowed the emergence of all-Ireland markets fields, including in energy and agriculture, and shared public services. Brexit then gave rise to a series of arguments about the nature and significance of borders in general, and in Ireland in particular. Murphy and Evershed (2021, in this issue) show how the internal settlement within Northern Ireland and the question of the border are intimately related, and how Brexit has destabilized both. In the event, the need to reach an agreement acceptable both to the Republic of Ireland and the EU meant that, in effect, a new economic border has been drawn within the UK itself, between Great Britain, on the one hand, and Northern Ireland on the other.

Brexit, which might have provided a focus for a new British nationalism in opposition to Europe, has thus deepened the territorial fissures within the UK and between the UK and Ireland. There is no obvious territorial fix that would restore the old spatial hierarchies and the UK state lacks the financial resources, administrative capacity and political capital to undertake a major nationalizing project.

BRITISH EXCEPTIONALISM?

It is easy to see the UK case as exceptional and, indeed, it is the only country so far that has voted to leave the EU. Yet it might more profitably been seen as a test case of rescaling as the old governing formula has proved unable to resolve either internal economic and political tensions or the relationship with the supra-national order. Welfare state restructuring has played out across territorial lines as social democratic forces at the periphery (including some English regions) have faced a central state intent on recasting it in more individualistic and less solidaristic lines. Changing patterns of economic geography have called for new policy responses, although these do not map neatly onto the divisions among the nations. The breakdown of historic voting patterns means that the nations have distinct party systems.³ Shifting class and territorial loyalties mean that the Labour Party faces massive challenges in gaining a UK majority without regaining Scotland. The Conservatives, having won seats in post-industrial England on the back of Brexit, face the formidable task of keeping them.

In the 20th century, political union was underpinned by economic union, including policies for territorial cohesion. Following Brexit, the UK government stressed the need to retain a unified internal market, overriding where necessary, different regulatory choices made in the devolved territories. There was also the talk of ‘levelling up’, a slogan used to promise to uplift the deprived communities in northern England and Wales without taking anything from the prosperous constituencies in the south.

On the other hand, a neat disintegration into discrete national units is equally difficult to envisage. The millennium settlement embraced a high degree of ambiguity and was made possible by the lack of a codified constitution such as has proved an obstacle to redefining the state in other cases, like Spain. Yet the basis of the EU as a legal order and the concern of both EU institutions and member states about the integrity of their project make such flexibility and ambiguity more difficult in the context of Brexit. As the relationship of the UK to the EU will not be settled by the exit achieved in 2020, it will continue to exert a centrifugal influence across the islands.

THE ARTICLES

These issues are analysed from a theoretical and empirical perspective in this special issue. Cetrà and Brown Swan (2021, in this issue) examine the arguments made for the maintenance of the union by unionist parties of the left and right, using oral and written statements. They find that unionism struggles to articulate itself as a state ideology, as it lacks the unified message and thrust which one might expect. Instead, since devolution, and particularly following the Brexit referendum, it amounts to a series of instrumental arguments around economics and welfare, supplemented with the invocation of ‘British values’. The weakness is that instrumental arguments can be played both ways, depending on the circumstances and the times. ‘British values’ is rather a hollow idea because these are really universal values shared by unionists and nationalists alike. With this broad frame, Labour places more emphasis on common social rights, while the Conservatives emphasize a shared past. Both stress the risks of separation as much as the benefits of staying together.

The Northern Ireland settlement represents the most differentiated of the territorial dispensations in the UK combining, as it does, territorial devolution with consociational power sharing and a transnational dimension. Murphy and Evershed (2021, in this issue) trace the origins of the Belfast/GFA of 1998 and its turbulent history, marked by periods in which the institutions were out of operation. They argue that the GFA was always open to the criticism that it entrenched, rather than overcame, ethnic identity politics. There was, moreover, no institutional embedding or ‘binding cultural narrative’. Membership of the EU did provide an external support system and largely succeeded in containing the issue of the border by the joint membership of both Ireland and the UK in the European Single Market. Brexit removes this support and makes issues that had previously been dealt with by ‘constructive ambiguity’ a matter for exact decisions. The outcome has highlighted the need for a reconsideration of the UK territorial constitution as a whole.

Kenny (2022, in this issue) addresses the question of England and its place in the union. There has been much talk in recent years about a rising English nationalism, detached from the wider British identity. This has often been linked to social conservatism, hostility to immigration and Euroscepticism. Kenny argues that matters are more complex. If nationalism is associated with secession or even a desire for self-government for the territory concerned, that hardly applies to England. There has been, rather, an emerging sense of Englishness, not around a single imaginary, but around multiple imaginaries. Economic, cultural and political elements are used in the discursive construction of Englishness and its detachment from the old Anglo-British imaginary. Stereotypical interpretations of this as a manifestation of inward-looking and conservative populism, however, are misleading. As well as the Englishness of post-industrial towns, there is a metropolitan imaginary and rural imaginaries. To these cultural factors, reinforced by economic grievance, can be added institutional developments. Devolution to the smaller nations, while arousing little initial opposition in England, has over time fed into a view that England is politically (as well as financially) deprived. Yet there is little demand for an English Parliament and the concession of English Votes for English Laws (since abandoned) did nothing to assuage discontent. Metropolitan mayors, the latest in a long history of changes in the territorial government of England, were not accompanied by the devolution of real power and resources, leaving England highly centralized. Opposition to the EU was another motif and voting for Brexit was highly correlated with feeling English rather than British. None of this, then, has led to a territorial settlement for England, which might have helped stabilized the union as a whole.

Wincott et al. (2022, in this issue) review the contested interpretations of the UK constitution from a historical and legal perspective. They highlight the Anglo-British imaginary, in which constitutional development is entirely English, with the peripheral nations being absorbed into an expanded English state, with some minor recognition of distinctiveness. During the 19th century, this was elaborated in the ‘textbook tradition’, based on the absolute sovereignty of the monarch-in-parliament. Its best known and most consistent exponent was A. V. Dicey, Professor of Law at Oxford. Texts in this tradition did mention the various unions, but failed consistently to analyse the concept of union itself or, indeed, to engage in much conceptual thinking at all. It was not unchallenged but consistently recovered to dominate constitutional thinking in the 20th century. Membership of the European Economic Community from the 1970s encouraged fresh challenges to the textbook tradition and were followed by the programme of constitutional reform of the new Labour government from 1999, including the Human Rights Act and devolution in Scotland, Wales and Northern Ireland. While all these reforms ostensibly maintained parliamentary sovereignty, practising academic lawyers increasingly queried whether the old doctrine now corresponded with reality. Yet, constitutional law was bound to the Diceyean paradigm. After the EU referendum of 2016, Brexit revealed the extent to which the external (EU membership) and internal (devolution) aspects of sovereignty were interconnected, with both pointing to the need for new a reflection. Brexit, however, was pursued as though the UK were a unitary state. Although efforts to roll back competences of the devolved bodies after Brexit were rebuffed, the UK Internal Market Act achieves much of the same effect by providing a system of mutual recognition of products and services that allows much less room for differing regulations in the devolved territories than was available under the EU Internal Market regime. The UK government even returned to using the expression ‘unitary state’. Wincott et al. suggest that such a conception of the UK is unsustainable in the long term.

Rawlings (2022, in this issue) addresses the case of Wales, which he describes as the ‘work in progress’ of a substate polity. Initially established with limited powers and an unwieldy structure, the National Assembly for Wales evolved in successive reforms into a legislature (Senedd Cymru/Welsh Parliament) on the lines of its Scottish equivalent, from which a Welsh government is chosen. Labour has been in office continuously, albeit sometimes in coalition or other types of partnership with Plaid Cymru or the Liberal Democrats. Since 2011 it has been the only wholly pro-union government among the devolved territories, but also the only one wholly committed to devolution as Welsh Labour has developed its own constitutional agenda. This represents a commitment to pursuing further devolution, going beyond the limits of the present model. The final point of evolution remains vague, but concepts include ‘new union’, ‘strong’ or ‘entrenched’ devolution, ‘mutual respect’ and federalism. It is driven by what Rawlings calls a ‘soft nationalism’, as Plaid Cymru has moved to supporting independence for Wales. A critical step would be the creation of a Welsh legal jurisdiction to replace the current England–Wales ‘conjoined’ legal system, although this has been resisted by the UK government. The various measures to recentralise power following Brexit have provided further opportunities for Welsh Labour to speak for the nation and defend a stronger model of devolution. It has proposed a UK Council of Ministers to replace the current system of intergovernmental relations, which give the final say on disputed matters to London, including over repatriated EU competences. On the other hand, the Welsh government remains a pro-EU force in a nation which voted Leave. The gravitational pull of England remains stronger than is the case for Scotland.

McEwen (2022, in this issue) notes that the Westminster model of parliamentary sovereignty has long been challenged in Scotland. The competing interpretations have been highlighted by Brexit, which exposed latent issues in the Constitution and posed the issue of competing popular mandates. The issue is explored through an empirical analysis of claims made by UK and Scottish political leaders. The broader context the emergence in recent decades of new thinking about sovereignty, in which it is pooled and shared. Post-sovereignty ideas have been widely adopted by minority nationalist movements, usually with reference to the wider European framework. Yet this never entirely displaced an insistence on the ultimate sovereignty of nations, which could then agree freely to share it. Sovereignty claims have also been a feature of populist nationalist movements. In the UK, a traditional, monist vision of sovereignty underlay the Brexiters’ claim to ‘take back control’, albeit with a slippage from the sovereignty of Parliament to that of the people. In the years following the Brexit referendum of 2016, and in transition from prime ministers Theresa May to Boris Johnson, there was a hardening of the sovereignty rhetoric. At the same time, the devices that had allowed the sovereignty argument to be side-stepped in the UK devolution settlement came under pressure. These included the Sewel Convention, downgraded by the Supreme Court, while measures such as the EU Withdrawal Bill and the UK Internal Market Bill were based on the assumption that the UK is a unitary state. Yet, in practice, the UK government held back from pushing sovereignty claims to the limit and in practice has sought to find devices to manage the devolved territories without provoking outright confrontation; hence, measures such as the Common Frameworks programme to manage repatriated EU competences.

The SNP, for its part, picked up old Home Rule ideas and merged them with elements of the post-sovereigntist argument to argue in the 2014 independence referendum for a form of independence that would allow it to maintain elements of both the UK and European unions. Since Brexit, this package is no longer available, and independence means choosing to align with one neighbour or the other. The outcome is that unionism and nationalism no longer have a middle ground on which to reconstruct themselves, leading to an increased polarization around constitutional issues.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was supported by the Economic and Social Research Council’s (ESRC) project ‘Between Two Unions: The Constitutional Future of the Islands after Brexit’ [grant number ES/P009441/1].

Notes

1 More extensive than the proposals in the 1970s, before the European Single Market programme.

2 Unionist intellectual Ali Ansari accuses the nationalists of ‘appropriating’ these unionist values (Ansari, 2020).

3 Northern Ireland always did, but the old Ulster Unionist Party of Stormont times was allied with the British Conservatives, whose whip the Unionist MPs took in Parliament.