Contesting sovereignty and borders: Northern Ireland, devolution and the Union

ABSTRACT Supported by UK and Irish membership of the European Union (EU), the 1998 Belfast/Good Friday Agreement’s post-sovereignist compromise helped to diminish the contested politics of the border in Ireland. However, by altering the status of the border, Brexit aroused and fomented politically charged divisions in Northern Ireland. We explore the confluence of four consequences of Brexit for Northern Ireland. Firstly, we detail how Brexit highlights the inadequacies and dysfunction of the UK’s territorial architecture. Second, we examine the specific structural, institutional and relational weaknesses of Northern Ireland’s devolved system. Third, we explain how Brexit further weakened the scaffolding that supports Northern Ireland’s devolved settlement. And fourth, we explore why Brexit has prompted profound reconsideration of the UK’s existing territorial set-up. Brexit’s challenge to the status quo in Northern Ireland, and by extension the UK’s constitutional and territorial integrity, is linked not just to internal political dynamics in Northern Ireland, but also to the ambiguity of the existing asymmetrical UK devolution settlement, its lack of embeddedness within the UK constitutional order and the absence of binding cultural narratives. Finally, we extend this analysis to posit that Brexit has revived the ‘Irish question’ and stirred a potentially destabilizing debate about Irish unity.


INTRODUCTION
In 1920s' Northern Ireland, the establishment of the devolved Stormont regime, and the 'Orange State' (Farrell, 1980) over which it came to preside, entrenched patterns of power and (sectarian) discrimination which, in the mid to late 1960s, catalysed a prolonged period of political violence. As Cramer (2006, p. 63) suggests, 'the euphemism "the Troubles", for the conflict in Northern Ireland from 1968 to 1998, captures some of the definitional ambiguity of that conflict', which hinged on a number of overlapping issues, including political economy, culture and ethnonational identity. Ultimately, however, the conflict was (and is) about the zero-sum question of sovereignty and of Northern Ireland's constitutional status. As Woodwell (2005, p. 167) asserts, '[w]hat started as civil rights based rioting was able to develop into a much larger conflict because of the fundamental political difference regarding the legitimacy of the Ireland-Northern Ireland border' (see also O'Callaghan, 2006).
After decades of conflict, the signing of the Belfast/Good Friday Agreement in 1998 signalled some prospect for peace by creating a post-sovereignist context for managing contested relationships and competing political aspirations. The roll-out of devolved power in Northern Ireland after 1998 was part of a broader constitutional reform programme which included the devolution of powers to Scotland and Wales. This process of advanced and asymmetric decentralization happened against the backdrop of shared UK and Irish membership of the European Union (EU). For Northern Ireland in particular, the joint EU framework and context judiciously lowered the stakes over matters of sovereignty while the European Union single market (SEM) complemented the terms of the Agreement (Hayward & Murphy, 2018). Although the border between Northern Ireland and the Republic of Ireland was retained, it was transformed and virtually disappeared as a physical barrier. Cross-border relationships and cooperation were buttressed by all-island institutions, and all-Ireland markets were encouraged. Social and political tensions were lowered. Despite this, Northern Ireland's devolution settlement has not been stable. A combination of residual conflict issues and structural weaknesses has interrupted the evolution of its devolved system of governance. Brexit's challenge to the status quo in Northern Ireland, however, is not just linked to internal political dynamics. These potent domestic forces also interplay with a tradition of ambivalence and ambiguity about the UK's constitutional arrangements. The shock of Brexit (further) exposed the limits of the UK's existing territorial arrangements, challenged the solidity of Northern Ireland's place within that territorial governance architecture, and heightened constitutional anxieties and agitation across the devolved territories.
In what follows, we start by providing a contextual overview of Northern Ireland's governance arrangements. We then explore the confluence of four consequences of Brexit for Northern Ireland. First, we detail how Brexit highlights the inadequacies and dysfunction of the UK's territorial architecture. Second, we examine the particular structural, institutional and relational weaknesses of Northern Ireland's devolved system. Third, we explain how Brexit has further weakened elements of the scaffolding which support Northern Ireland's devolved peace settlement. And fourth, we explore why Brexit has prompted profound reconsideration of the UK's existing constitutional and territorial set-up. We argue that Brexit's challenge to the constitutional status quo in Northern Ireland is linked to more than internal political dynamics. It is also connected to the ambiguity of the existing asymmetrical devolution settlement, its lack of embeddedness within the UK constitutional order and the absence of binding cultural narratives. Finally, we extend this analysis to posit that the resurgence of the 'Irish question' and the developing debate about Irish unity in the wake of Brexit may prove to be part of the undoing of the Union.

THE GENEALOGY OF DEVOLUTION IN NORTHERN IRELAND
That Northern Ireland was founded as a 'Protestant state for a Protestant people' is a claim attributed to its first Prime Minister, James Craig. While this is something of a misquoting of Craig, it nonetheless captures the essential nature of Northern Ireland's early experiment in devolution from 1921 to 1972, which was predicated on forms of state-and nation-building which sought to institutionalize forms of Protestant and Unionist privilege, and effectively marginalized the Catholic and Nationalist minority (Farrell, 1980;Whyte, 1983). This reflects something of the way in which 'Unionism' in Northern Ireland has been, and remains, more 'ethnic' than 'civic': conceived of as a form of ethno-communal designation as much as an expression of constitutional preference (McAuley, 2010;Todd, 1987Todd, , 2020. Northern Ireland's first period of devolved government ended abruptly as peaceful civil rights protests against Unionist (mis)rule became entangled with and gave way to violent constitutional conflict. The outbreak of violence and serious civil unrest heralded the end of the devolved Stormont regime and the introduction of direct rule from London. The years that followed were scarred by a violent and intractable conflict based on differing Unionist versus Nationalist interpretations of the Northern Irish state and its legitimacy. The calling of paramilitary ceasefires in 1994, however, marked a historic breakthrough and paved the way for negotiations which led to the signing of the Belfast/Good Friday Agreement in April 1998. The agreement represented an attempt to de-escalate, deterritorialize and, crucially, remove the gun from this conflict (Morrow, 2012(Morrow, , 2017, and move it from the streets to the debating chamber. From 1972, devolution remained at the forefront of ongoing attempts to address the Northern Ireland conflict. The ill-fated Sunningdale Agreement (1973) introduced a new form of devolved government based on power-sharing between 'constitutional' nationalism, moderate unionism and the nascent cross-community Alliance Party (Eggins, 2015), as well as new cross-border institutions. It was opposed by Republicans (who had been excluded from its negotiation) and ultimately crushed by strong Unionist opposition to its cross-border provisions and a general strike underpinned by loyalist paramilitary muscle (McCann & McGrattan, 2017). The Northern Ireland Constitutional Convention of 1976 unsuccessfully sought to revive the prospect of devolved power-sharing, with Nationalists having refused to participate because the initiative lacked an 'Irish dimension' (Byrne, 2001). The British government again attempted to devolve power on a partial and rolling basis to a Northern Ireland Assembly following elections in 1982. Although Unionists supported the creation of this Assembly, Nationalist political parties boycotted it. Despite these unsuccessful attempts to restore devolution in Northern Ireland, 'the idea of devolved powersharing, located within a wider political framework, had not gone away' (Tonge, 2000, p. 46).
Following extensive political negotiations, a new devolution formula materialized in 1998 with the signing of the Belfast/Good Friday Agreement. Parity of esteem, power-sharing and the principle of consent sought to protect the group identities and interests of both communities in Northern Ireland. The consent principle requires that constitutional change in Northern Ireland is (1) a matter of democratic consent and (2) a matter for the people of Ireland alone to decide upon. It is the responsibility of the Secretary of State for Northern Ireland to trigger a border poll (or referendum on Irish unity) in Northern Ireland if s/he is persuaded that there is majority support for constitutional change. Historically, demographic trends, election results and opinion polling have pointed to ongoing majority support for the constitutional status quo. In this context, the inclusion of an Irish dimension in the Belfast/Good Friday Agreement was particularly significant for Nationalists as it explicitly recognized and gave expression to their political identity and aspirations. New East-West provisions sought to offset Unionist concerns about the institutionalization of North-South relations. Whereas Sunningdale had sought to build a new form of devolved power-sharing around a moderate centre through the exclusion of both Unionist and Nationalist 'extremes', the negotiations that led to the signing of the Belfast/Good Friday Agreement deliberately sought to bring these extremes into the fold. From the outset, the devolved institutions have actively involved both the Democratic Unionist Party (DUP) (Tonge et al., 2014) and Sinn Féin (Whiting, 2017) in the governance of Northern Ireland. In ways discussed further below, this has, in the end, been to the electoral detriment of the more 'moderate' Ulster Unionist Party (UUP) (Tonge et al., 2019) and Social Democratic and Labour Party (SDLP) (Farren, 2010;McLoughlin, 2010).
The 1998 Agreement includes territorial, cross-territorial and non-territorial elements. The non-territorial dimension contains provisions in relation to human rights, policing, prisoners, security and equality. The territorial dimensions, based on three separate strands, created a set of interlocking institutions: the power-sharing Northern Ireland Assembly and Executive, the North-South Ministerial Council (NSMC) and British-Irish Council (BIC) ( Table 1).

DEVOLUTION IN THE UK: AMBIVALENCE AND AMBIGUITIES
The signing of the 1998 Agreement was part of a wider process of constitutional reform, which included the asymmetrical devolution of powers to constituent parts of the UK. This ambitious process of state restructuring and rescaling is viewed by Jeffery (2009, p. 92) as 'partial disintegration' of the UK into different territorial communities with new and increasingly distinct forms of territorial politics. The process was based on a piecemeal approach that failed to identify 'an overall conception of the impact of devolution on the UK state' (p. 92). Sandford and Gormley-Heenan (2020) note that 'constructive ambiguity' has been the hallmark of the UK's territorial constitutional arrangements, and Wincott et al. (2020, p. 2) similarly demonstrate that ambiguities attached to devolution are in fact 'a long term feature of the UK's territorial constitution'. This ultimately reflects what Welsh First Minister Mark Drakeford (2019) has referred to as 'deep and profound ambivalence to devolution' at the centre. The Belfast Agreement stipulates a range of areas for North-South cooperation and implementation: agriculture; education; transport: environment; waterways; social security/social welfare; relevant EU programmes; inland fisheries; aquaculture and marine matters; health; and urban and rural development. The work of the NSMC is supported by a series of allisland implementation bodiesone such body is the Special EU Programmes Body (SEUPB), which oversees cross-border EU funding programmes. c The Belfast Agreement is less prescriptive in relation to areas of BIC cooperation, when compared with the NSMC. However, the Agreement does suggest that suitable areas for early discussion may include transport links, agricultural issues, environmental issues, cultural issues, health issues, education issues and approaches to EU issues. The work of the BIC has since expanded to 12 work-streams. Source: Adapted from Murphy (2018, p. 5).
The sum of these ambiguities makes for what Gaskell et al. (2020, p. 3) categorize as 'an especially complex hybrid of … two multilevel governance models, with a strong emphasis on functional division combined with territorial autonomy granted in different forms to Scotland, Wales and Northern Ireland'. The devolved territories differ markedly in terms of the powers they enjoy. In practice, this has meant that when policy issues cut across devolved and reserved functions, difficulties and often failure can result. Moreover, following the onset of the global financial crisis in 2007, McKinnion (2015, p. 51) notes how episodes of policy disagreement and divergence were linked to ideological and party differences between the British government and devolved administrations. The functionality of the system was also hampered by the inability of informal and ad hoc intergovernmental structures to adequately moderate and manage this challenging policy landscape (Gallagher, 2012).
The 2016 UK vote in favour of leaving the EU (and more recently the coronavirus pandemic) has revealed these inherent tensions within the devolution settlement, and radically disrupted the process of 'devolve and forget' which had defined the politics of the UK's territorial constitution since 1998 (Wincott, 2018). The result of the referendum was determined by English voters (Henderson et al., 2017) and it conflicted with the majority support for Remain recorded in Scotland and Northern Ireland. In the aftermath of the vote, the character of devolution arrangements, coupled with an absence of effective intergovernmental mechanisms , limited the extent to which the UK's approach to the Brexit challenge could be managed collectively and concordantly by the British and devolved governments. In its totality, therefore, the UK's devolved system's inbuilt structural inadequacies militated against the emergence of shared or binding narratives about what Brexit should mean, prevented joined-up thinking and compounded the problem of managing cross-territorial tensions. In this way, Brexit served as a catalyst for bringing questions about the scope and limits of devolution to the surface (Wincott et al., 2020). This has perhaps been nowhere more apparent, or of greater consequence, than in Northern Ireland, where devolution has not only been characterized, but also been largely defined by sustained political contestation, structural weakness and systemic fragility. As outlined above, set against the other of the UK's constitutional arrangements, Northern Ireland's devolution settlement emerges as particularly complicated and delicate. All its interlocking parts are vulnerable to disruptive shocks (Cochrane, 2020). And before Brexit, Northern Ireland had already long been an unstable part of a more widely inconstant and fluctuating territorial constitution.

DEVOLUTION IN NORTHERN IRELAND: FITS AND STARTS
The devolution 'settlement' in Northern Ireland has consistently proven itself to be anything but settled. Since 1998, the Assembly and Executive have been suspended or gone into abeyance five times, for a cumulative total of some eight years: around one-third of the more than 20 years since its first meeting. The latest hiatus, in the wake of the 'Cash for Ash' scandal (McBride, 2019), endured between January 2017 and January 2020. Even when devolution has functioned, Assembly and Executive business has often been stalled by brinkmanship, boycotts and walk-outs. A pattern of political crisis, (near) collapse, (re)negotiation and renewed agreement (often underpinned by a new injection of cash from the British exchequer) has been a (and perhaps the) defining feature of devolution in Northern Ireland since 1998 (Birrell & Heenan, 2017). A cycle of crisis-talks-crisis has produced five further agreements since the Belfast/Good Friday Agreement was signed: St Andrews (2006), Hillsborough Castle (2010), Stormont House (2014), Fresh Start (2015) and New Decade, New Approach (2020), each of which has tweaked the devolution settlement in an attempt to put it on a more sustainable footing, with (self-evidently) somewhat mixed results.
This tinkering with devolution in Northern Ireland has sought to refine, rather than radically alter or undermine, the consociational model of governance enshrined in the Belfast/Good Friday Agreement. However, it is this modelwith its mandatory coalitions, communal designation and bloc voting, mutual vetoes, and other measures to promote 'parity of esteem' between unionism and nationalismthat is arguably at the root of instability in Northern Ireland's devolved institutions. While its proponents have argued that consociationalism is simply a realist response to, and means of managing, violent and deep division in Northern Ireland's body politic (McGarry & O'Leary, 2004), critics have identified the role that consociation has played in sustaining and even deepening this division (Guelke, 2003;Hall, 2018;Taylor, 2008).
Crucially, far from addressing the root cause of Northern Ireland's Troubles, that is, the fundamental conflict between divergent Nationalist and Unionist constitutional interpretations and aspirations, the Belfast/Good Friday Agreement has instead reinscribed this conflict as an organizing logic of governance in Northern Ireland. As Nagle (2018, p. 401) argues, 'rather than resolve the question of self-determination, the Belfast/Good Friday Agreement incentivized those who could successfully frame themselves as the best parties to either deliver Irish unity or secure the long-term future of the Union'. Those best able to represent themselves as either the staunchest defenders of the constitutional status quo or most likely to upend it, namely, the DUP and Sinn Féin, have benefited, to the detriment of their (allegedly) more moderate or conciliatory rivals. Recent shifts in patterns of identity and electoral trends (Hayward & McManus, 2019;Tonge, 2020) have not (yet) manifestly changed Northern Ireland's essentially zero-sum politics. This politics finds its expression in forms of cultural conflict which have often destabilized Northern Ireland's peace and political processes (Nagle, 2014;Nolan et al., 2014). But it also bleeds into other areas of social policy, which, on the face of it, have little or nothing to do with the constitutional question, including abortion and marriage equality, and 'bread and butter' issues such as welfare reform.
The 'culture war' that has defined post-conflict Northern Irish politics has thus seen almost all policy positions come to be identified as necessarily either 'Orange' or 'Green'. This has tended to marginalize the voices of the 'others'the increasing number of Northern Ireland voters whose politics are not determined primarily by their constitutional preference (Hayward & McManus, 2019). Somewhat paradoxically, this culture war has also provided cover for a considerable degree of convergence between Northern Ireland's parties, in general, and the DUP and Sinn Féin, in particular, in the realm of political economy (Murtagh & Shirlow, 2012;Nagle, 2009). Redistribution of the dividends of economic development (such as they have been) have tended to be subject to an ethno-sectarian carve-up: a quid pro quo politics of the 'pork-barrel' whereby resources are divvied up between the DUP and Sinn Féin at the centre and then conveyed back to their respective 'communities' through patronage networks (DeYoung, 2018). Thus, and as Barry (2017) argues: increasingly Northern Ireland looks like it's heading towards a One party Janus faced system, where each ethnic champion publicly appeals to its sectarian base for electoral power by blaming the 'Other' for all the Assembly's faults while privately collaborating with the very same 'Other' to ensure they remain the dominant power in the political process.
Building on long-running fatalistic trends in Unionist political culture (Farrington 2001;Finlay, 2001), the framing of contemporary Northern Irish politics in terms of a zero-sum 'war by other means' (Curtis, 2014) has been a particular and defining feature of post-conflict Unionist discourse and praxis (Evershed, 2018;McAuley, 2010). As Nagle (2018, p. 401) argues, the Belfast/Good Friday Agreement was framed by Nationalists as a victory, as representing 'the institutional expression of the "equality agenda", a positive process of redressing the historical experience of inequality and exclusion'. These Nationalist gains have been read by Unionists as necessarily implying Unionist losses, such that support for the Agreement has been weaker among Unionists than among Nationalists from the outset, and has declined steadily since 1998. Crucially, the Agreement's post-sovereignist recasting of Northern Ireland's politics has been viewed by Unionists as an intrinsic and existential threat to Northern Ireland's long-term position in the Union. Although the NSMC has no executive powers and the cross-border bodies have only a modest remit, there is Unionist antipathy to the institutionalization of the North-South relationship (Tonge et al., 2014, pp. 56-61; see also Hayes et al., 2006, pp. 155-156;Tannam, 2018). This is linked to the symbolic significance of institutionalized North-South cooperation as opposed to its practical outworking. The upshot has been that, like Strand 1which has arguably done as much to entrench as it has to overcome Northern Ireland's deep political divisionthe North-South dimension of the Agreement has never fully functioned as initially envisaged. Tannam (2018) has examined how the promises of Strand 3 have likewise been incompletely fulfilled. As argued by Todd (2015), although the Belfast/Good Friday Agreement sought to deepen the institutionalization of the British-Irish relationship, this relationship has remained informal and somewhat ad hoc: reliant as much on the personalities and priorities of relevant government ministers in both London and Dublin as on the institutions established by the Agreement. In this respect, the British-Irish relationship since 1998 has mirrored many of the weaknesses of intergovernmental relations within the UK since devolution . As Todd (2015, p. 64) notes, above all, the 'informal British-Irish mode of implementing the Agreement and of adjudicating on its principles was … dependent on the states' prioritization of Northern Ireland', which, in the face of the global financial and sovereign debt crises, the ascendance of the Scottish independence movement and the debate about the UK's future membership of the EU, has diminished steadily since 1998.
With both states distracted by other matters, and given the apparent success of devolution in Northern Ireland after St Andrews, Northern Ireland slipped off the political agenda. The British-Irish Intergovernmental Conference was allowed to go into abeyance, while the work of the BIC was largely 'tangential' (Tannam, 2018, p. 249) and garnered little interest or buy-in, particularly from the government in Westminster. The Brexit referendum, in general, and the surprise 'Leave' result, in particular, put severe and immediate strain on an East-West relationship that had been allowed to slacken, with knock-on implications for the other strands of the Belfast/ Good Friday Agreement (Murphy, 2018). Indeed, Brexit has more generally served to further destabilize Northern Ireland's already unstable governance arrangements.

BREXIT AND BORDERS
Enduring internal tensions in Northern Ireland's devolution settlement were complicated further by the British government's decision to hold a referendum on UK membership of the EU in 2016. In particular, the referendum and its aftermath served to refocus Northern Irish politics on the border. A key success of the 1998 Agreement had been the way in which it had diminished the salience of conflict about this border per se in day-to-day politics (albeit that, as noted above, the Agreement failed to finally resolve the constitutional question, and this conflict has found new expression in the 'culture war'). First, the 1998 Agreement guaranteed that constitutional change could only happen by peaceful means and with the consent of the majority. This responded directly to Nationalist ambitions for a united Ireland, whilst providing a guarantee for the Unionist majority that the prospect of such a change was unlikely in the context of existing demographic realities (where Unionists were in a majority) and a functioning devolved arrangement capable of quelling Nationalist dissatisfaction. Second, the Agreement provided for the expansion of cross-border cooperation, and the creation of North-South institutions. This was an especially vital component of the settlement for Irish Nationalists because such institutions gave expression to Nationalist identification with the rest of Ireland. This effectively de-dramatized the contested questions of identity and political self-determination by allowing people living in Northern Ireland to identify as British or Irish, or both. It also accommodates opposing political ambitions allowing Unionists and Nationalists to legitimately (and peacefully) aspire to different constitutional futures.
Membership of the SEM organically supported both the cross-border elements of the 1998 Agreement and the post-sovereignist compromise on which it was based. The SEM's four 'freedoms of movement' permitted the virtual disappearance of not just the physical, but also the metaphorical, border between North and South. Gormley-Heenan and Aughey (2017) note how EU membership helped 'to contextualise being either British or Irish or both, mainly for Nationalists but not only for them. In other words, it was yet another way of not talking about the border' (p. 502). Crucially, the UK decision to leave the EU in general, and the SEM in particular, represented a change in the status of the border between Northern Ireland and the Republic of Ireland (Hayward, 2018). Brexit, therefore, aroused and fomented the politically charged contestedness of the border in Irish politics and created conditions which provoked political division and constitutional agitation in Northern Ireland.
Significantly, any acknowledgement of the contested politics of the border in Ireland was absent from the British referendum campaign narrative. Little, if any, consideration was given by the British political establishment to the potentially risky impact of a Brexit referendum, or its outcome, on Northern Ireland, the border with the Republic of Ireland or the Belfast/ Good Friday Agreement (Cochrane, 2020;Murphy, 2018). The British campaign narrative was heavily focused on the supposed burden of EU migration and on the notion of 'taking back control'. These issues differed markedly from the more localized economic and political concerns of the Northern Ireland electorate, for whom political identity was a decisive factor in determining positions on the Leave/Remain referendum question. Where Nationalists were wholeheartedly opposed to a UK exit from the EU, Unionists tended, by a factor of 2:1, to favour it (Garry & Coakley, 2016). Northern Ireland's 56% vote in favour of remaining in the EU did not facilitate any coalescing of Unionist and Nationalist positions. Instead, it precipitated the sharpening and hardening of ethno-national dividing lines (Murphy, 2018). Unionists of all shades moved to support the Leave position, while Nationalist opposition to Brexit crystallized around calls for 'Special Status' for Northern Ireland. As with other issues in Northern Ireland politics, Brexit quickly became starkly 'Orange' versus 'Green'.
The full implications of Brexit for Northern Ireland became increasingly apparent in the period after the referendum. Independent analyses highlighted the severity of Brexit's economic consequences for Northern Ireland (Budd, 2015;Oxford Economics, 2016;Springford, 2015) and concerns about the politically destabilizing impact of Brexit became more pronounced (Connolly & Doyle, 2019; Gormley-Heenan & Aughey, 2017; Teague, 2019). The latter centred on concerns, voiced predominantly by Irish Nationalists, but also shared by the Irish government, business and sectoral interests, and security services in Northern Ireland, about the damaging impact of a hard border between North and South on both the spirit and the practical operation of the provisions and institutions of the Belfast/Good Friday Agreement. In August 2016, the First Minister and Deputy First Minister reflected some of these concerns in a joint letter to Prime Minister Theresa May.
Shared concerns around Brexit's potentially damaging consequences for Northern Ireland, however, did not override existing political tensions. Sinn Féin's focus on the achievement of 'Special Status' as a formula for protecting (as far as possible) the political-legal status quo in Northern Ireland after Brexit was interpreted by the DUP and its political fellow-travellers as an affront to the constitutional integrity of the UK and as an attempt to institute Irish unity by the backdoor (Murphy & Evershed, 2020). Brexit was thus layered over, and arguably consolidated, existing divisions in Northern Ireland by becoming entangled in the elementals of the constitutional question. By early 2017, these Brexit tensions were intermingling with existing and unresolved 'culture war' pressure points and eventually culminated in the 'Cash for Ash' scandal which precipitated the resignation of Sinn Féin Deputy First Minister Martin McGuinness in January 2017. This move collapsed the Northern Ireland Assembly, triggered an Assembly election and resulted in a prolonged hiatus for Northern Ireland's devolved government.
Just over one week after the collapse of the devolved settlement, May's Lancaster House Speech committed the UK to leaving the SEM and customs union. This stoked further divisions in Northern Ireland because it jarred with concomitant assurances that there would be no return to 'the borders of the past' on the island of Ireland. Reconciling the objective of combining a seamless Northern Ireland-Ireland border with UK exclusion from the SEM and customs union was to remain a problem throughout May's period as Prime Minister. Her ability to overcome this problem, and deliver her preferred form of Brexit, was negatively impacted by her decision to call a general election in May 2017. The election failed to return a Conservative Party majority and handed the DUP an unprecedented political opportunity in the form of a confidence-and-supply deal with the Conservatives. The arrangement included a generous financial package for Northern Ireland, and some policy gains for the DUP (Tonge & Evans, 2018), but an expectation of British neutrality on the 'Irish question' was shattered (Tannam, 2018, p. 86) and this served to further antagonize already precarious political relations. Brexit already represented the overriding of the express wishes of the majority in Northern Ireland, and seemed to fly in the face of the spirit, if not the letter, of the 'principle of consent'. May's tryst with the DUP served to entrench perceptions that Brexit was a functional reassertion of British sovereignty in and over Northern Ireland. It called into question the UK government's commitment to key pillars of its devolution settlement (Evershed, 2021).
This perception of a renewed and more muscularly Unionist interpretation of Northern Ireland's place in the Union ) came into sharp focusbut also into questionfrom December 2017 when negotiations between the UK and the EU were derailed by DUP objections to a proposed formulation for accommodating Northern Ireland's unique circumstances (Connelly, 2018). Prime Minister May's draft deal involved a limited degree of regulatory divergence between Northern Ireland and Great Britain, but within the context of a wider UK-EU customs partnership. Nationalists were largely supportive of the proposal, but the DUP were staunchly opposed, arguing that the deal diminished the constitutional link between Northern Ireland and the rest of the UK. Viewed through this prism, May's proposal was seen to facilitate a united Ireland by degrees. Tweaking of the withdrawal deal and what became known as the Irish border backstop did not assuage DUP objections. The party's pledge to support the government's Brexit policy was abandoned. On three occasions, the DUP failed to lend parliamentary support in the House of Commons to May's Withdrawal Agreement.
The party's anti-backstop stance was bolstered by Brexit hardliners in the Conservatives who alleged not to support the proposed differential treatment of a part of the UK, and who also voted to reject May's Brexit withdrawal formulation. The relationship between the DUP and Tory Brexiteers, however, ultimately proved highly contingent. When faced with the prospect of a withdrawal deal not passing through Parliament, formerly supportive Conservative MPs broke with the DUP. This followed a change of UK political leadership and a December 2019 general election which altered the political mood music in favour of finalizing the UK's exit from the EU. Prime Minister Boris Johnson's self-proclaimed commitment to the Unionist cause was shortlived and nominal (Kenny & Sheldon, 2020, p. 15). His support for the so-called 'frontstop' broke the Brexit stalemate and provided a means for the UK to agree a withdrawal deal and formally leave the EU on 31 January 2020. The Protocol on Ireland/Northern Ireland attached to the deal allows Northern Ireland to remain part of the UK customs territory. However, to ensure there is no return to a hard border in Ireland, Northern Ireland will de facto remain part of the SEM. This creates a necessity for some customs and regulatory checks between Northern Ireland and the rest of the UK. The arrangement is utterly opposed by Unionists for its supposed impact on the constitutional integrity of the UK (Wilson, 2020). British government positioning on Brexit and its management of the negotiations did little to quell political division in Northern Ireland.
Additionally, Brexit also sullied (already weakly institutionalized) North-South and East-West relations. The Irish government's persistent opposition to Brexit, and support for the EU negotiating position, led to a souring of relations with unionism in general, and the pro-Brexit DUP in particular (Murphy & Evershed, 2020). The collapse of Stormont in January 2017 precipitated the suspension of Strand 2 institutions, a move which eliminated opportunities for any form of structured dialogueincluding on Brexit and its implicationsbetween representatives of the Northern Ireland Executive and Irish government. The only functioning institutional strand of the Belfast/Good Friday Agreement during the withdrawal negotiation period was the BIC, but this had no substantive ability to confront the challenges thrown up by Brexit and was hamstrung by operational issues including irregular meetings and limited British government buy-in. In effect, the council did not materialize as a constructive space for dialogue or consultation on issues related to Brexit. Its evident impotence diminished a basis for bilateral cooperation at a critical turning point in the British-Irish relationship.

EXISTENTIAL CHALLENGES
The Belfast/Good Friday Agreement created a post-sovereignist context for managing contested relationships and political aspirations while simultaneously removing the disputed Irish border from the political frontline. The Brexit referendum and the process of the UK's withdrawal from the EU placed enormous strain on all these dimensions of the Agreement. According to the House of Lords European Union Committee (HL EC) (2020, p. 86) in its June 2020 report on the Protocol on Ireland/Northern Ireland: The process of the UK withdrawal from the EU, and the negotiations leading to the agreement of the revised protocol on Ireland/Northern Ireland have regrettably placed [British-Irish] relations and [Northern Ireland's] stability under considerable strain, with a concomitant diminution of trust on all sides.
In its final iteration, the Brexit Withdrawal Agreement, in general, and its Protocol on Ireland/ Northern Ireland, in particular, is replete with new burdens for Northern Ireland's devolution settlement.
The redrafted Brexit deal concluded between UK and EU negotiators in October 2019 replaced the all-UK 'backstop' negotiated by May with a new set of measures which will mark Northern Ireland as all-the-more enduringly distinct from the rest of the UK. As Hayward et al. (2020, p. 20) have argued, this 'frontstop' requires dynamic alignment between Northern Ireland and the EU, and is: no mere tokenism; the relationship [between Northern Ireland and the EU] will be substantive, with Northern Ireland set to follow EU regulations on goods and the Union Customs Code … the Protocol can be expected to determine the conditions of Northern Ireland's economic, regulatory, rights, etc. environment for the foreseeable future.
Particularly with Johnson seeking as distant a future relationship with the EU as possibleand although the Withdrawal Agreement also emphasizes Northern Ireland's place in the UK's internal marketin practice, this necessarily implies new economic barriers (both tariff and non-tariff) between Great Britain and Northern Ireland.
The conclusion of the Withdrawal Agreement and the redrafting of the Protocol on Ireland/ Northern Ireland was followed by uncertainty about its practical implications, and concerns about whether and how it will be fully implemented (Connelly, 2020;Rice, 2020). This stemmed in large part from confusing and contradictory messaging from the British government about implementation of the provisions of the Protocol, which alarmed its EU interlocutors and, at times, come close to suggesting an intention to resile from the legal commitments enshrined within it (Connelly, 2020;Rice, 2020). The net result was a diminution in trust in Brussels of the UK government and its intentions. Critically, UK government equivocation, vagueness and ambiguity vis-à-vis the Protocol created a climate of grave uncertainty for government, businesses and civil society actors in Northern Ireland.
Although final agreement between the UK government and the EU on implementation of the Protocol was reached, it is significant that this was ultimately without the express consent of the Northern Ireland Assembly, which voted to withhold its consent for the Withdrawal Agreement Bill passed by the UK Parliament in January 2020. Indeed, as Hayward et al. (2020) have noted, under the terms of the Withdrawal Agreement, scope for Northern Irish input into the mechanisms that will govern post-Brexit UK-EU relations is severely constrained, despite that Northern Ireland is arguably the region of the UK with the most at stake in the functioning of this relationship. 'In a real way', as Hayward et al. (2020, p. 20) argue (see also HL EC, 2020, p. 83): Northern Ireland is at risk of being subject to legislation coming from both Brussels and London without full sight or scrutiny of it … there is no automatic means by which Northern Ireland's devolved institutions will either be able to effectively scrutinise and shape EU law they need to download or be able to upload views to the EU. This democratic deficit is juxtaposed to the consent mechanism included in the Protocol, which gives the Northern Ireland Assembly the power to vote on its renewal every four (or, in the case that such a vote passes with cross-community consent, every eight) years. Given the politicization of the issue of Northern Ireland's relationship with the EU, and the manner in which this has been mapped on to the constitutional question at the heart of the political conflict in and about Northern Ireland, this too has the potential to be fractious and divisive. As noted by the HL EU (2020, p. 79): the democratic consent mechanism could exacerbate political division in Northern Ireland in the lead-up to each vote. It guarantees that the Assembly will be required repeatedly to debate the arrangements for trade within the UK and across the island of Ireland. The mechanism also creates the potential for significant economic instability and dislocation every four or eight years.
The form that Brexit has taken in Northern Ireland means that the question of its relationship with the EUwhich, viewed through the lens of the constitutional conflict which defines Northern Ireland politics, is necessarily zero-sumis now undivorcable from its territorial politics, and will remain so in perpetuity. The ultimate effect of this, and the other severe strains, that Brexit has placed on Northern Ireland's political settlement has been to raise, in stark terms, the question as to whether this settlement remains viable. In other words, it has stoked dissatisfaction with the functionality of devolved governance, re-enlivened the debate about the constitutional future of the island of Ireland, and imbued the question of Irish unity with an urgency it has not had since the height of the Troubles.

THE UNDOING OF THE UNION?
In the same way that devolution 'was not a one-time enactment of constitutional change, but rather a dynamic whose trajectory was open and whose endpoint was unclear' (Jeffery, 2009, p. 291), the Belfast/Good Friday Agreement has not provided a stable or final constitutional settlement for Northern Ireland. Instead, it has functioned as a 'a first, although as yet unclear, step toward a range of future constitutional changes' (Shirlow, 2001, p. 743). Until 2016, it was assumed that constitutional reform or transformation would take place in the context of UK membership of the EU.
The removal of EU scaffolding has produced adverse consequences for devolution and for the UK's territorial stability, and this is felt profoundly in Northern Ireland. Sandford and Gormley-Heenan (2020, p. 108) note that 'the technical requirements of Brexit will mandate the need for exact decisions where "constructive ambiguity" has existed up to now'. Nowhere has this been more apparent than when it has come to questions about borders and bordering on the island of Ireland, which Brexit has mandated can no longer be 'fudged' in the way they have been since 1998. Irish and UK membership of the EU accommodated and reinforced the Belfast/ Good Friday Agreement's conception of the border as fluid. This blurring of identities and sovereignty is what makes the border acceptable to the two traditions in Northern Ireland. In altering the status of the border between Northern Ireland and the Republic of Ireland, Brexit has undermined the extent to which critically important characteristics like fluidity and 'fuzziness', as applied to both the practical and symbolic meaning and operation of the border, can be maintained. The process of moderating Brexit's impact on the Northern Ireland border was complicated and restricted by UK governance arrangements which failed to adequately foster political and cross-territorial inclusivity.
When Brexit collided with the UK's ambiguous and ambivalent constitutional arrangements, in general, and the idiosyncrasies of Northern Ireland's particular post-1998 governmental architecture, in particular, it produced conditions capable of fatally challenging the resilience and durability of this infrastructure. The referendum result and the ensuing UK-EU negotiations revealed opposing interpretative differences about the nature of parliamentary sovereignty (prioritized by the centre) versus popular sovereignty (variously relied upon by the peripheries) (Wincott et al., 2020). Vulnerable to suspension, collapse and stalemate on select policy issues, Northern Ireland's devolved system was unable to engage effectively with the wider UK Brexit conversation. The absence of a functioning devolved administration in Northern Ireland between January 2017 and January 2020 was a consequence of both bitter political infighting and the internal structural weaknesses of the power-sharing arrangement. Notably however, even where devolved government was operational (as in Scotland and Wales), an ability to substantially shape and influence UK Brexit policy was subject to constraints and limits (McEwen, 2020). This lack of embeddedness of facilities to engage, include and accommodate devolved interests in the wider UK constitutional order has been exposed by Brexit and has revealed the extent of division, disagreement and divergence between the UK centre and the periphery.
This scenario of dysfunction is further fuelled by the absence of binding cultural narratives across the UK. The UK referendum on membership of the EU was in fact an amalgamation of different territorial referendum campaigns. The Northern Ireland campaign generated territorially specific concerns about the border, EU Structural Fund support, the agriculture industry, among others. These differed markedly from those issues which dominated the wider British referendum campaign and exposed a gulf of misunderstanding between the periphery and the centre. This pattern of misunderstanding continued and intensified during the Brexit negotiation phase when, as a consequence of electoral arithmetic, Northern Ireland concerns were (at least initially) filtered through the DUP. Wincott et al. (2020, p. 2) succinctly note that in relation to May's premiership: 'Little here suggested a willingness to address the competing demands from the DGs [devolved governments] directly; their positions had been seen, not fully heard.' A Conservative victory in December 2019 meant that the British government could dispense with the confidence and supply arrangement. As Kenny and Sheldon (2020, p. 15) note: 'Tory MPs were compelled to choose between the competing priorities of achieving a negotiated Brexit and treating Northern Ireland as an integral part of the UK.' In the event, they chose Brexit and a withdrawal deal, which was not supported by a single political party in Northern Ireland. This absence of a binding and unifying cultural narrative around Brexit, grounded in an understanding of competing territorial priorities, undermines the merits of the devolved system and has spurred support for alternative constitutional futures.
In Northern Ireland, this is characterized by increased agitation for advancing the achievement of a united Ireland. The Irish unity agenda is most closely identified with Sinn Féin, a party which after the 2017 Assembly election was only one seat short of becoming the single biggest party in the Northern Ireland Assembly, and in 2020 was the most popular electorally in the Republic of Ireland. In the aftermath of the Brexit vote, Sinn Féin has spearheaded a renascent debate on Irish unity. This is premised on an increase in support for a border poll and for Irish reunification in the short to medium term among voters on both sides of the Irish border (though this increase has not been steady or consistent) (Donaghy, 2019(Donaghy, , 2020. The SDLP has also moved to strengthen its united Ireland credentials by proposing the creation of a New Ireland Commission (Irish Times, 2020). Other parties, including Fine Gael and Fianna Fáilwhose coming together in a 'grand' coalition in June 2020 itself marks significant shifts in Ireland's post-Brexit politicshave also now begun to take an interest in the concept of a 'shared island' (Tannam, 2020).
The prospects for constitutional change, however, are based on the principle of consent as set out in the Belfast/Good Friday Agreement. The Unionist community's resolute opposition to constitutional change is supported by those who identify as neither Unionist nor Nationalist. The so-called 'neithers' is the largest single segment of the Northern Ireland population and a majority of this group consistently favour the constitutional option 'devolution within the UK' (Hayward & McManus, 2019). This suggests that overall support for the constitutional status quo persists (for now). This is despite unionism having lost its inbuilt electoral majority, failing to win a majority of either votes cast or seats available in any election in Northern Ireland since March 2017. To some extent, the Unionist (and 'neithers') position is supported by successive Irish governments who are not persuaded by a simple majoritarian calculation in favour of a border poll and have tempered and nuanced their position: initially favouring the status quo, and latterly committing to 'build on the foundations laid in the Belfast/Good Friday Agreement to deepen peace and prosperity in Northern Ireland' (Irish Government, 2020, p. 104). In contrast, the British government's Brexit strategy is perceived to have 'sold out' Unionists and sidelined the Unionist plea for Northern Ireland to leave the EU on the same terms as the rest of the UK. Unionist anxiety is further fuelled by concerns that Conservative Unionist discourse does not include 'an authentic commitment to Northern Ireland' (Kenny & Sheldon, 2020, p. 15) and is aggravated by a British public which only flimsily supports Northern Ireland remaining in the UK (YouGov, 2019, 2020). The appetite for constitutional change is also influenced by the growing number of non-aligned voters in Northern Ireland who tend to be more inclined to 'don't know' in terms of having a constitutional preference (Hayward & McManus, 2019). Because the underlying dynamics that impact on Northern Ireland's constitutional future are less settled and more mutable in the aftermath of the UK decision to leave the EU, this adds to the layered tension felt by a perennially insecure Ulster unionism in its mission to maintain the constitutional status quo.
Northern Ireland's constitutional future lies at the intersection of dynamic, complex and clashing forces. Brexit's challenge to the status quo in Northern Ireland, and by extension the UK's constitutional and territorial integrity, is linked not just to internal political dynamics in Northern Ireland, but also to the ambiguity of the existing asymmetrical UK devolution settlement, its lack of embeddedness within the UK constitutional order, and the absence of binding cultural narratives. Insofar as it illuminates the underlying ambiguities of the UK's system of devolved powers, Matthews (2017, p. 608) characterizes the Brexit crisis as 'a "perfect storm", dramatically exposing the hollowing-out of the constitution'. The sum of these developments has produced tension, instability, and some contemplation of new constitutional ideals and opportunities on the island of Ireland. This is consequential for the future of the UK's devolution settlement, and for the future of the Union. It has the potential to portend the end of Northern Ireland's devolved political settlement, as part of a wider possible disintegration of the UK.

DISCLOSURE STATEMENT
No potential conflict of interest was reported by the authors.