Hegemonic constituent power: Fear of the people and lessons for Irish reunification

Abstract

This article introduces the concept of hegemonic constituent power to argue for a greater role for the people in the process of Irish reunification through the establishment of a constituent assembly. Hegemonic constituent power contends that, ideally, constituent power should be possessed by the people; however, descriptively, this is invariably not the case. Constituent power instead is best understood as the manifestation of hegemony – the dominant power base in a given legal order that legitimates and reinforces this power through institutions, prevailing ideas and culture. Hegemony performs an important function in descriptively explaining legitimacy formation while not necessarily conferring normative legitimacy on existing power structures or those who exercise constituent power. Legitimacy and illegitimacy are both embedded in this notion of hegemonic constituent power. This allows for constituent power to perform a legitimating function and its creative potential to be unleashed while still leaving space for critical contestation over how this power was exercised. In this way, hegemonic constituent power also seeks to address critiques of constituent power as enabling populism. The article then deploys this concept of hegemonic constituent power to argue that fears of invoking the will of the people in debates on Irish reunification are misplaced.

I. Introduction

For as long as democracy has existed, there has been a fear of the people. Plato famously denigrated the citizens of Athens for sentencing Socrates to death, and his model of five regimes of government has democracy decay into tyranny as excess liberty erodes virtue, paving the way for a champion of the people to appeal to and exploit their base interests. ¹ Today, fear of democracy manifests most clearly in the literature on populism – that ‘thin-centred ideology’ that juxtaposes the ‘pure people’ who ideally possess political power against a ‘corrupt elite’ who actually possess it. ² As constitutions both allocate and constrain public power, constitutional authorities and the principles of constitutionalism become key targets for the populist political project, particularly right-wing populist movements and their repugnant conceptualization of the pure people along ethnically homogenous lines. ³ While some have sought to muster constitutionalism against this populist challenge, others caution against over-estimating law’s role in this fight. ⁴ Despite these warnings, a flight to law and existing institutions can be seen in literature exploring the role of the people at the very limits of constitutionalism: in the act of constitutional creation itself. ⁵

‘Fear’ is also prominent in the nascent debates on Irish reunification – the issue of whether the six counties of Northern Ireland, a constituent part of the United Kingdom, should leave the United Kingdom and reunite with the independent sovereign state of the remaining 26 counties of the Republic of Ireland. ⁶ From fears of a return to the intense political violence seen from 1969 until 1998, to fears of the loss of identity of the British unionist minority, these anxieties shape the horizon of possibility regarding what both the process and result of Irish reunification should look like. Many of these concerns echo those levied at democracy in general and a similar flight to law and existing institutions can be seen in this emerging literature on Irish reunification. Despite these fears, democracy must remain a key value for the continued legitimacy of constitutionalism today and the process of Irish reunification. This is embodied in mainstream accounts of constituent power – the power to create a constitution – that is assumed to be vested in the people. As Irish reunification would require a new constitution, one would have expected constituent power to feature prominently in these debates; however, save for a few interjections, the opposite is the case. ⁷

This article argues that this flight to law is misplaced. Failure to consider constituent power risks ostracizing the people as the ideal possessors of constituent power from the reunification process. While vesting trust in elites instead may be the preferred course of constitution-making for some, a top-down constituent process dominated by existing constitutional actors runs the risk of collapsing the constituent power into the pre-existing constituted powers and diluting its creative potential. Fundamentally, it weakens the possibility of the constitution-making process acting as a moment through which the very idea of the people or peoples can be reconstructed, a conception of the people that moves beyond the existing dichotomy between Irish nationalists and British unionists.

To address this tension between the creative and legitimating potential of constituent power on the one hand and fear of the people on the other, this article delineates a novel conception of constituent power as hegemonic in nature. Ideally, constituent power should be possessed by the people; however, descriptively, this is invariably not the case. Instead, constituent power is exercised and tempered by those who claim to speak on behalf of the people. Constituent power therefore is best understood as the manifestation of hegemony – the dominant power base in a given legal order which legitimates and reinforces this power through institutions, prevailing ideas, and culture. Hegemony does this not just through force but also through active and passive consent and understanding how this consent is constructed and maintained is imperative. Hegemony performs an important function in descriptively explaining legitimacy formation while not necessarily conferring normative legitimacy on the exercise of constituent power. As legitimacy and illegitimacy are both embedded in this notion of hegemonic constituent power, this allows constituent power to perform a legitimating function and its creative potential to be unleashed while still leaving space for critical contestation over how this power was exercised.

Irish reunification provides an excellent opportunity to explore how fear of the people impacts upon constitution-making processes and how hegemonic constituent power can address many of constituent power’s anti-populist discontents. Part II argues that critiques of constituent power manifest a deep-rooted fear of the people. This is mirrored in much of the literature on populism. Part III then demonstrates that the absence of constituent power from debates on Irish reunification is likely deliberate, motivated by a latent fear of its destructive potential and an aversion towards invoking the idea of ‘the people’ when the very concept of a unitary Irish people is deeply contested. Nascent debates on Irish reunification have instead focused on existing institutions and legal texts for answers on how Irish reunification should proceed. No consideration of the people as constituent actors in and of themselves is demonstrated; instead, the processes envisaged place heavy emphasis on deliberative decision-making and consensus between elites.

Part IV confronts this fear of the people by advocating for an agonistic account of politics. As disagreement in politics is unavoidable, attempts to displace politics and to retreat to law risk ossifying rather than tackling the populist problem. As majority decision-making is inevitable in a democracy, there will always be those in power and without; however, such disagreement and conflict can be productive rather than simply negative. The concept of hegemony is then articulated, demonstrating how political power in a given state is structured and reinforced, influencing and shaping people’s consent towards those who claim to speak on their behalf. As constituent power is always a manifestation of hegemony, what matters normatively is how this hegemony is constructed. Consequently, this article argues for a pluralistic conception of ‘the people’, as this is necessary for an agonistic conception of politics and to ensure that the people can challenge those who claim to speak on their behalf.

Part V applies hegemonic constituent power to illuminate the path towards Irish reunification. The best way to ensure an agonistic and pluralistic exercise of constituent power is through the establishment of a constituent assembly elected directly by the people following an initial referendum in favour of reunification. The resulting draft constitution should then be put to the people in a second referendum, on the basis of an ‘all-Ireland franchise’. While aspects of the literature are sceptical of directly elected constituent assemblies, it will be shown that these concerns of abuse of power are not relevant to Irish reunification. Fundamentally, only a constituent assembly can open the requisite space necessary to allow for the formation of a pluralistic conception of the people of Ireland.

II. Constituent power, populism and fear of the people

To speak of constituent power, claims Negri, ‘is to speak of democracy’. ⁸ Negri’s confident assertion epitomizes the idea that the power to create a constitution should be vested in the people. Vesting constituent power in the people is done for normative reasons, conferring democratic legitimacy on the constitution. The origins of constituent power in the French Revolutionary pamphlet of Emmanuel Sieyès are illuminating regarding this function. Sieyès sought to vest constituent power in the Third Estate of the Estates General as only it, being composed of representatives of the common people, could speak for the nation as a whole. ⁹ Speaking for the nation, the Third Estate could challenge the divine right of the King upon which the Ancien Regime was based, destroying this old constitution and legitimating the creation of a new one grounded in the nation’s will.

These origins also reveal constituent power’s constructive and destructive potential. While constitutional creation is the principal function of the exercise of constituent power, this new constitution can only come into existence if it fills a lacuna caused by the destruction of the old constitution. This destruction haunts the new constitutional order, a portent of what could befall it should constituent power be exercised again. Sieyès was acutely concerned with limiting constituent power’s destructive potential, with Rubinelli arguing that he intentionally vested constituent power in the nation rather than the people to temper the political power of the demos. ¹⁰ The people could not exercise constituent power directly; instead, it was channelled through their representatives in the Third Estate, now transformed into the National Assembly. ¹¹ The people can be controlled further by recognizing only certain groups of property-owning men as ‘active citizens’ and permitting them to participate in public life. ¹²

Fear of the people is thus evident in the genesis of constituent power, and today constituent power’s discontents continue this theme. Some theorists argue for a ‘closed’ model of constituent power, quarantining it solely to the moment of creation of the constitution, but then rejecting any future exercise of it. Instead, all future political and legal powers – including the power to amend the constitution itself – are conceptualized as ‘constituted’ powers, authorized and limited by the constitution. ¹³ If constituent power does have any further role, it is not to empower but to curtail and constrain the jurisdiction of the constituted powers. ¹⁴

‘Open’ models, in contrast, stress that the creative and legitimating dimensions of constituent power necessitate its continued existence, even after the constitution is in effect. Constituent power animates the constitutional order, shaping, changing and in some cases formally amending this constitution. This ensures the enduring democratic legitimation of the constitution. Open models of constituent power thus seek to avoid privileging one generation’s view of constitutional ordering over subsequent generations. ¹⁵ The people permanently possess constituent power, and the constituted powers should be acutely aware of this. Variations on this open concept argue for a ‘relational’ understanding of constituent power, viewing it as emerging from a dialectical relation between the rulers and the ruled and, in turn, producing and continually reshaping the people themselves. ¹⁶ Some proponents of open models turn the destructive potential of constituent power on its head, arguing instead that truly destructive power invariably lies in the hands of the constituted powers. For Negri, constituent power can be thought of as a ‘counterpower’ to the constituted powers and of capitalist institutions. ¹⁷ Subsuming of the constituent power into the constituted powers must therefore be resisted fiercely in order to preserve this fundamental function as a ‘counterpower’. ¹⁸

Nevertheless, several authors reject the notion of constituent power altogether, arguing that its destructive potential – particularly when vested in a fictive conception of the people – undermines the requisite stability required to found a legal order. This rejection is gaining traction in parts of the literature, with Verdugo cautioning against constituent power’s over-romanticization of the people. This excessive idealization makes constituent power a useful device for populists to exploit, deploying a ‘pure’ anti-pluralist conception of the people as possessors of the constituent power. ¹⁹ Doyle echoes this, noting that open theories of constituent power can lend support to populist claims of ‘a unitary and unchanging people’ with ‘an immanent but continuing role as a constitutional actor superior to the constitution itself’. ²⁰ Arato voices similar concerns, arguing instead for a ‘post-sovereign’ model of constitution-making, where existing institutions and procedures similar to constitutionalism are utilized during the constituent act. ²¹ These multi-stage constitution-making processes should bring together different actors and different representatives, placing a heavy emphasis on elite input and control. Arato’s paradigm of ‘post-sovereign’ constitution making seeks to de-dramatize ‘conceptions of constituent power, linked to mythological and dangerous notions of total rupture and the full embodiment of the will of the people’. ²² Similarly, in an Irish context, Doyle argues that the enactment of the 1937 Irish Constitution was not a ‘big-bang event in which constituent power willed itself into existence’; rather, it is an example of a momentary exercise of constituent power by an entity created by the previous legal system. The enactment of the 1937 Constitution is an example of how an existing legal system can pave the way towards its own unlawful replacement; recourse to a fictive people is unnecessary. ²³

Populism and constituent power

Critiques of constituent power are often directed at work derived from Carl Schmitt. ²⁴ For Schmitt, there must exist a priori a degree of political stability upon which a constitutional order can be founded. This stability is produced through the irrational distinction decided by the sovereign between friend and enemy, which in turn creates the people or volk. ²⁵ Consequently, the political must exist prior to the legal and so all state power cannot be constrained by law. Although Schmitt insists that the friend–enemy distinction is based upon an irrational decision, there must be some empirical basis grounding this distinction for it to be effective. ²⁶ This empirical distinction between groupings must be sufficiently strong or antagonistic to unify the people. Here, normatively problematic assumptions as to the necessary ethnic homogeneity of the state arise and the dangers are themselves revealed in a catastrophic fashion by Schmitt’s own personal history as the ‘Crown Jurist’ of the Third Reich. ²⁷

Vesting constituent power in the volk or people serves to raise further questions of who ‘the people’ are and how they come to possess it. Paradoxically, the people are only identifiable following the exercise of constituent power, so constituent power not only creates the constitution itself but also creates ‘the people’. ²⁸ Constituent power thus paves the way for claims to speak on behalf of the people, a concept that is more fictive than real. Schmitt’s conception of constituent power is often represented as enabling populism – a ‘meta concept’ referring to the promotion of the ‘true will of the people’ against the will of the so-called elites in power. ²⁹ While the definition of populism is vague, deeply contested and often cynically deployed to conflate left- and right-wing opposition to the status quo, ‘populist constitutionalism’ is a concrete, referring to the constitutional dimension of the populist project. ³⁰ This entails dual strategies of circumventing and commandeering the constitution to give effect to their ideological project. ³¹ Schmitt’s homogenous volk embodies the populist’s notion of the pure, true people, the heartland of the state that is being enfeebled by corrupt elites in power.

III. Fear and Irish reunification

The year 2021 marked the centenary of Northern Ireland. For some, this was a moment to be celebrated; ³² however, others viewed Irish partition through more sombre eyes as something to be regretted. ³³ This also marked the moment when Northern Ireland lost its protestant majority. ³⁴ The partition of Ireland has been described as a ‘sectarian head count’. ³⁵ Hatched in London during the Irish War of Independence from 1919 until 1921, Northern Ireland as a geo-political construct had no predecessor. It was created from six of the nine counties of the province of Ulster and was designed to have an in-built protestant majority, given the alignment between Protestantism and unionism. ³⁶ At the time of partition, Northern Ireland had a protestant majority of 65 per cent of its population, with a sizeable 35 per cent Catholic minority making up the remainder.

With the 2021 census results, Northern Ireland’s original raison d’être no longer exists. Catholics now make up 46 per cent of its population and 43 per cent identify as Protestant. ³⁷ This means a further 11 per cent fall outside these two categories. Calls for reunification have intensified, leading to an increase in studies on what the reunification process and resulting united Ireland constitution should look like. Ostensibly, the path to Irish reunification is legally straightforward. The Good Friday Agreement – the 1998 international treaty ratified by the United Kingdom and Ireland, which brought an end to the 30-year conflict in Northern Ireland – provides for reunification following a referendum in which a majority expresses that Northern Ireland no longer wishes to be a part of the United Kingdom. ³⁸ This would necessarily be accompanied by a referendum in the Republic of Ireland also expressing a desire for reunification. ³⁹ Two referendums; two simple majorities. However, this simplicity obscures what would, in essence, constitute an exercise in constituent power.

Fear, constituent power and the threat of violence

Despite what should be a quintessential moment for its consideration, constituent power is almost wholly absent from nascent debates on Irish reunification. This is not necessarily surprising as constituent power has long been overlooked in the United Kingdom – a state without a codified constitution and with a sovereign parliament lacking any ostensible limitation on its legislative power. ⁴⁰ The Republic of Ireland too has also not fully grappled with constituent power. This may be for several reasons, including the quotidian status of referendums to amend its constitution, which can be conceptualized as the exercise of a constituted constitutional power. ⁴¹ In debates regarding Irish reunification specifically, constituent power’s close connection with revolution and fear of the people is arguably a factor as to why the concept has not been broached. The Good Friday Agreement that articulates the conditions for reunification is essentially a peace treaty. For the most part, its ratification in 1998, brought an end to the three decades-long conflict colloquially known as ‘The Troubles’, which cost the lives of over 3,500 people. ⁴² In a post-conflict or dormant conflict society such as Northern Ireland, where no transitional justice project has been pursued effectively, the risk of a return to violence is of high political concern.

In this context, aversion to constituent power with its revolutionary origins in the tumult of the French Revolution is predictable. ⁴³ Constituent power discourse raises the political stakes; its populist bent and lauding of the people as the legitimating basis for the constitutional order are also superficially unappealing where the identity of the people is deeply contested, with some manifestations of Britishness and Irishness defined in mutual opposition to each other. ⁴⁴ Irish reunification is haunted by the threat of violence that such a process may unleash, and concerns abound that pleas to ‘the people’ are counter-productive and exclusionary. References to ‘fears’ are replete in both in the literature and in the media. The most substantial work to date on the mechanics of a process for Irish reunification has been carried out by UCL’s Working Group on Unification Referendums on the Island of Ireland (WGURI). ⁴⁵ Published in May 2021, the document is an impressive technical analysis of the various issues pertaining to Irish reunification that require resolution. These entail, inter alia, the criteria for calling a referendum, the question to be asked, the referendum franchises, whether a draft constitution should be produced before or after the initial referendum, whether interim institutions would be required and what this constitution should look like. Throughout this incredibly detailed 302-page document, constituent power is not mentioned.

Instead, the WGURI report, like the broader debate on Irish reunification, centres on existing institutions and searching within existing legal texts for answers. The WGURI is careful not to recommend one particular path to Irish reunification; rather, it lays out a series of different configurations ranging from a ‘maximum plan’ in which the form of a united Ireland is spelled out in advance of the initial referendums, and ‘process plans’, in which the process for determining the final form of a united Ireland would be agreed upon in advance of the initial referendums but that this actual process of determination would not begin until after a successful vote for reunification. ⁴⁶ The WGURI is also agnostic about whether an entirely new Irish constitution would be required or whether unity could be completed through constitutional amendments to the 1937 Irish Constitution. There is acknowledgement of the value of a new constitution involving citizens of Northern Ireland but, conversely, concerns are documented about the loss of a large amount of accumulated jurisprudence in which a new constitution would result. ⁴⁷

While the WGURI does not spell out one specific preferred pathway for reunification, two common themes can be identified throughout: first, that there should be heavy elite control over the reunification process; and second that, while there are some references to ad hoc engagements with the people, there is an absence of any understanding of the people or peoples as constituent actors in their own right. Instead, the people are understood primarily in terms of the electorate for the initial referendums on reunification as established by pre-existing law, or as electors to pre-existing institutions. This way of understanding the people is also seen in other contributions to the debate on Irish reunification analysing the franchise that would participate in the reunification referendums. This is done by examining the existing electoral rules in both jurisdictions on the island of Ireland. McCrudden et al argue that the ‘most sensible approach’ to Irish reunification should be

a presumption against departure from the existing franchise model that operates in the area where a referendum will take place, tempered with a willingness to depart from that model for strong reasons, provided any changes are introduced well ahead of any referendum. ⁴⁸

This would entail the franchise for Irish constitutional amendment being used in the Republic and the franchise for Assembly elections used in Northern Ireland. They corroborate this conclusion with analysis of previous experiences of referendums in Ireland and the United Kingdom, and of existing international and European human rights obligations. While such approaches are indeed sensible, they are nevertheless based upon analysis of existing law. These approaches risk closing off any possible innovation in the franchise for Irish reunification and, in turn, inhibit different conceptualizations of the Irish people or peoples.

In addition to electoral franchises, other existing institutions are also examined to shed light on what Irish reunification would and should look like. Similar to the WGURI, O’Leary identifies two distinct approaches to reunification. First, the ‘modelling approach’ envisages that when the referendum on Irish reunification occurs, there is a clear model in place as to what this united Ireland would look like. ⁴⁹ Ideally, this model would have been

extensively deliberated through the work of citizens’ assemblies, deliberative forums, and after ministerial-led planning, preferably coordinated through a Ministry of National Reunification. ⁵⁰

The modelling approach places primacy on pre-existing institutions, albeit with some reference to ad hoc ‘citizens’ assemblies’ that have seen increased recourse to in the Republic of Ireland in recent years. These assemblies consist of a selected group of unelected but demographically representative citizens who debate a specific issue before them, guided by a chair with significant elite and expert control over the agenda and material presented to the assembly. ⁵¹ Indeed, elite control over the Irish citizens’ assemblies through their ‘close connections with parliamentarians and government’ is hailed by the WGURI as one of their strengths. ⁵² The advantage of the modelling approach is that people would have a high degree of certainty as to what a united Ireland would look like before voting in a referendum. The WGURI also discusses this possibility:

The decisions to be made would have major implications for the lives of people throughout the island of Ireland. It would therefore be important for these decisions to be shaped by a broad and diverse range of people. Elected representatives would be central at every stage. Enfranchised persons from multiple perspectives (such as different localities, ages, religion, ethnicity, sectors of employment, and education levels) should also have opportunities to contribute. A range of mechanisms of wide and substantial public engagement could be used. Citizens’ assemblies, for example, have become useful forums for enabling deeper public discussion of major constitutional or policy decisions before they are made. Other mechanisms could focus on bringing in the voices of disadvantaged or marginalized groups. ⁵³

O’Leary contrasts the ‘modelling approach’ with the ‘process approach’, which contends that concrete constitutional drafting should not occur until after the initial reunification referendums. This draft constitution would likely be the product of a constituent assembly, but the processes and mandate surrounding this constituent assembly could be articulated in advance of the vote. ⁵⁴ Both the WGURI and O’Leary caution that this approach would come with a high degree of uncertainty with one commentator referring to it as ‘voting for an engine without seeing the car’. ⁵⁵ Brexit features prominently in both the WGURI and O’Leary’s work as a textbook example of how not to conduct a referendum process owing to the uncertainty under which the people operated when making their decision. This uncertainty was masked by the crude binary nature of referendums, which over-simplified the question before the people. Avoiding uncertainty and political turmoil is therefore paramount and, to this end, the ‘modelling approach’ with its emphasis on the role of extant constituted powers in constitution-making is superficially attractive.

This fear of political uncertainty is particularly heightened in the context of Irish reunification. Brexit itself typifies this with fears of a return to violence frequently voiced in the context of where and whether any border security architecture should be placed in the event of customs checks being required. ⁵⁶ Fears of violence would be more acute in relation to Irish reunification with the WGURI going into substantial detail about the extent of these fears. ⁵⁷ This fear is not irrational, as paramilitarism is still the primary security-related threat in Northern Ireland. ⁵⁸

‘Physical security’ concerns are accompanied by additional fears surrounding ‘ontological security’. ⁵⁹ This ontological insecurity manifests as fear of the loss of British identity in a united Ireland, their sense of belonging and their links to the past, as well as their ‘kith and kin’ in Britain. These concerns, Todd suggests, ‘cannot be reassured by reasoned argument, pragmatic appeals or appeasement’. ⁶⁰ These are fears about the identity of a minority, and not only whether this identity will be adequately protected and capable of expression in a united Ireland but whether it will ultimately continue to exist. If constituent power is to be possessed by the people of Ireland, a fundamental opposition to this is that a large minority does not want to be part of the Irish people. Avoiding the language of constituent power, reaching instead for legal analysis of existing election franchises and institutions may be an attempt to allay these fears, side-stepping the polarizing ‘will of the people’. That stated, political discourse in Northern Ireland is often couched in the language of ‘concerns’ or ‘fears’ with Unionist political claims more likely to be framed in terms of fear than Irish nationalists. ⁶¹ This often masks the fact that such positions are actually political claims different from all others; albeit in relation to the concept of identity, these political claims may not necessarily be rational. Furthermore, the fact that many of these political claims may be grouped and juxtaposed against another group, giving rise to the ‘Two traditions paradigm’ of identities on the island of Ireland, does not mean that such paradigms must form the basis for Irish reunification – a point that we will return to in the next section. ⁶²

Ultimately, constitution-making is a high-stakes and often dangerous endeavour, and this article does not downplay the risk of a return to violence. Yet this risk is not unique to Irish reunification; constitutions are often drafted during or shortly after exceptional political tumult. ⁶³ Irish exceptionalism cannot be automatically invoked to explain why constitution-making processes used elsewhere may not be suitable to Irish reunification and why constituent power should not be used to illuminate the Irish reunification process. Moreover, there may be serious downsides to this eschewing of constituent power. Emphasizing only its destructive potential risks losing constituent power’s equally important creative potential. It risks Irish reunification embarking upon a hyper-legalized, technocratic constitution-making process that side-lines the people as the ideal possessors of constituent power, raising questions about the democratic legitimacy of the process. ⁶⁴ Most importantly, it may inhibit the emergence of different conceptions of the Irish people(s) beyond the nationalist–unionist dichotomy, a different conception of the Irish people(s) more resistant to antagonism between communities. Consequently, constituent power must not be avoided; it must be embraced, unpacked and utilized to shape and frame Irish reunification.

IV. Hegemonic constituent power

This avoidance of constituent power in debates on Irish reunification is accompanied by a current of analysis favouring modes of democratic deliberation or consensus-based decision-making as a means by which fundamental constitutional issues may be settled. Consensus is the fundamental principle at the heart of the Good Friday Agreement. ⁶⁵ Consequently, the WGURI argues that it ‘would be far preferable for the governments to take a consensual approach … drawing in voices from all communities and all parts of society’ when deciding the constitutional form of a united Ireland. ⁶⁶ Failure to adopt a consensual approach, they conclude, ‘could cause serious tensions, leading to concerns for legitimacy and stability’. ⁶⁷

Consensus also features prominently in the Irish Government’s ‘Shared Island Initiative’, a newly established unit within the Department of the Taoiseach (Prime Minister) with the objective of harnessing:

the full potential of the Good Friday Agreement to enhance cooperation, connection and mutual understanding on the island and engage with all communities and traditions to build consensus around a shared future. ⁶⁸

The Initiative is backed by a ‘Shared Island Fund’, providing finance for research, the arts, communities, cross-border projects and capital infrastructure. ⁶⁹ It also seeks to establish forums for dialogue in various communities on this shared island. ⁷⁰ Yet, despite this emphasis on consensus in the Good Friday Agreement, on the ultimate question as to the constitutional status of Northern Ireland, this decision will decided by simple majorities in two separate referendums held in Northern Ireland and the Republic. These consensus-based approaches invariably reach for existing law and institutions as the fora for creating consensus. Notwithstanding some modest attempts of the Shared Island initiative at producing some novel forums for dialogue, it is inevitable that it will be consensus between elites that matters.

Reaching for consensus to shape Irish reunification is to seek refuge in rationalism and deliberation as the exclusive means through which Irish reunification can occur. The literature on Irish reunification is not naïve regarding the difficulties this approach faces. There is a clear understanding that some will be vehemently opposed to Irish reunification and will, regardless of the process, be deeply unhappy with the result. ⁷¹ In short, it is unavoidable that there will be losers. Attempts at consensus and deliberative modes of constitution-making with a strong role for elites and existing constitutional institutions acknowledge this but strive to make this losing group as small as possible. This is laudable; however, these attempts risk replacing politics with rationalism and essentially de-politicizing the highly political question of Irish reunification. This is most notable regarding the ignoring of constituent power, in particular a Schmittian conception of constituent power vested in a homogenous people created by the Sovereign’s irrational decision that distinguishes ‘friend’ from ‘enemy’. While the temptation to de-politicize issues through consensus is understandable, these irrational disagreements cannot be resolved in the marketplace of ideas. ⁷²

However, one need not endorse Schmitt’s intensely antagonistic conception of politics to argue for a constitution-making process that deeply involves the people. Schmitt’s critique of liberalism as reversing the priority of the political over the constitution is helpful in demonstrating that the flight to law as a means of depoliticizing disagreements is of limited effectiveness. Having said that, this article favours an ‘agonistic’ rather than a Schmittian ‘antagonistic’ conception of politics. ⁷³ An understanding and appreciation of agonism can result in a constituent process capable of legitimate decision-making, which is equipped to confront irreconcilable issues that a deliberative and consensus-based conception of politics may seek to avoid.

Agonism

The fundamental tenet of agonism is that there will always be disagreement in politics. It is also unavoidable that there will always be those in power and those not in power. Consequently, Mouffe argues that we should not pursue a consensus that is impossible; rather, we should recognize and acknowledge adversary positions but, importantly, simultaneously recognize adversaries as legitimate, respecting their right to defend their own interests and values. ⁷⁴ The possibility of antagonism must still be taken seriously, but this can only be done through an agonistic conception of politics that allows space for political conflict. ⁷⁵ Invariably, liberal conceptions of politics strive for closure, or what Honig refers to as the displacement of politics with ‘virtue’ theories of politics. ⁷⁶ These virtue theories identify politics with administration and treat juridical settlement as the task of politics and political theory. ⁷⁷ Wenman further argues that:

Where liberals and deliberative democrats typically seek to overcome or transcend conflict by bringing it under a set of regulative principles … the agonists insist that these responses actually serve to exacerbate the problem.’ ⁷⁸

Rather than ensure a stable settlement, agonists content that as more and more of politics is displaced, this may actually risk increasing populism as avenues for disagreement are also closed off. ⁷⁹ Opportunistic actors may then claim that power has been corrupted and wrested away from the ‘true people’; instead it is now vested in the hands of a corrupt elite. Only fundamental constitutional change – the populist constitutional project – can wrest it back and speak for the true people.

In contrast, agonism responds to the right-wing populist challenge, not by closing off politics but by stressing the idea of creative conflict. Although disagreement is unavoidable, this is not necessarily a negative; on the contrary, disagreement drives imagination and innovation. For Mouffe, liberal conceptions of democracy fail to take account of this value of political conflict and the necessity of the potentiality of antagonism. Instead, liberalism stresses the priority of procedural mechanisms for mediating competing political perspectives of the good. ⁸⁰ Agonism thus sees value in conflict or disagreement and this can only be ensured by simultaneously valuing and embracing pluralism. ⁸¹ Pluralism, Mouffe argues, is not merely a fact but an axiological principle. ⁸² What is key to agonism is that disagreeing parties are not considered to be enemies; the legitimacy of the adversary must always be acknowledged and respected. This does not mean, however, that everything is up for grabs; rather, citizens can agree on the importance of principles like liberty and equality while sharply disagreeing on their content. ⁸³ Consequently, political views that are hostile towards pluralism are inimical to agonism. Agonism therefore does not equate human dignity with rational consensus; therefore, reflecting Schmitt, consensus is impossible. Instead, agonism opens up space for disagreement and political contestation while still maintaining adversarial respect for one another. Agonism embraces the disagreements that liberalism eschews; however, when doing so, the asymmetry between disagreeing parties – between the powerful and the powerless – must not be ignored. Hegemonic power must be acknowledged and, in a democracy, hegemony reaches for a majority for its legitimation. Democracy should be pluralistic, but this pluralism will always be asymmetrical. As such, political power will always be hegemonic; what matters is how this hegemony is constructed.

Agonism and hegemony

Hegemony is about who possesses political power, and also how this possession of power is reinforced and exercised. Mouffe’s idea of hegemony builds upon the work of Antonio Gramsci – an Italian communist who wrote much of his life’s work while imprisoned by the Italian Fascist regime until his death. ⁸⁴ Gramsci contended that ‘man is not ruled by force alone but also by ideas’. ⁸⁵ Ideas perform the vital function of preserving the ideological unity of a whole social bloc. ⁸⁶ According to Bates, Gramsci’s conception of hegemony is

really a very simple one. It means political leadership based on the consent of the led, a consent which is secured by the diffusion and popularization of the world view of the ruling class. ⁸⁷

In Gramsci’s understanding of hegemony, consent is key as it is not simply through force alone that the ruling class holds power. Hegemony therefore requires an understanding of how this consent is created and how the ruling class’s worldview is popularized and diffused to garner the support of the ruled. Gramsci thus emphasizes the role of civil and political society in consent formation. ⁸⁸

Gramsci understood politics as a constituent power capable of transforming social identities. ⁸⁹ Mouffe describes this understanding of hegemony as ‘the articulation of the interests of the fundamental class to those of its allies in order to form a collective will, a unified political subject’. ⁹⁰ In this way, the ‘dominant group becomes the interest of other subordinate groups’. ⁹¹ For Gramsci, the ‘intellectual’ was pivotal in creating hegemony to the extent that they ‘extend the world view of the rulers to the ruled and, in so doing, secure the “free” consent of the masses to the law and order of the land’. ⁹² One single class of intellectuals does not exist; rather, each class has its own group of intellectuals and the intellectuals of the ‘historically progressive class’ exercise such a power of attraction that they influence and subordinate the intellectuals of the other classes. Through this web of solidarity, hegemony reinforces the ideas of the ruling class and creates consent. Behind this, the coercive apparatus of the state lurks. ⁹³ Hegemony therefore stresses the importance of cultural and intellectual factors in historical development and will formation. Bates argues that there are Machiavellian tendencies in Gramsci’s conception of hegemony as, ‘For all modern Machiavellians, the fundamental categories of power are force and consensus, and these are not mutually exclusive but interdependent realities.’ ⁹⁴ Gramsci thus believed that there could be no consensus without force and no liberty without authority. ⁹⁵ From the factory to the church, the school to university, the media to our social circles, these are all pivotal factors in shaping and framing the consent of individuals. Consent is therefore created beyond the official organs of the state, and it may be both active and passive. ⁹⁶ Indeed, passive consent can play a fundamental role as people feel that they have no choice but to defer or acquiesce to existing ideologies or modes of power.

Hegemony has a powerful explanatory force, demonstrating the reality of where political power lies and how it is reinforced. However, hegemony does not automatically perform a legitimating function; to the contrary, it illustrates that there is an inherent illegitimacy at the basis of authority. One can see echoes of Gramsci’s conception of hegemony in other thinkers on constituent power. For instance, Hannah Arendt argues that while the French nation-state was saved from imminent collapse by Napoleon’s ability to manipulate the national will and direct this towards declaring himself the ‘pouvoir constituant’, in reality it was the solid structure of class society that ensured the perpetuation of stability over long periods of time. ⁹⁷ Relatedly, Kalyvas argues that Schmitt’s concept of sovereignty and Gramsci’s notion of hegemony are two distinct variations on a single theme: the idea of the political as the original instituting moment of society. ⁹⁸ The concepts share greater similarities, including the mechanisms through which the will of the people is formed. For instance, Schmitt states that:

In every democracy there are parties, speakers, and demagogues from the προτάται of the Athenians up to the bosses in American democracy. Moreover, there are the press, film and other methods of psycho-technical handling of great masses of people. All that escapes a comprehensive set of norms. ⁹⁹

Here, Schmitt highlights the various forces that may influence the construction of ‘the will of the people’ in a manner similar to Gramsci’s idea of hegemony. Recently, Gilbert and Williams have argued that so-called ‘Big Tech’ and private capital play a fundamental role in reinforcing the existing hegemony. ¹⁰⁰ With digital public spaces in the hands of private actors, an illusion of free speech and independent will-formation is presented; simultaneously, algorithms amplify specific content that increases engagement and (and therefore financial rewards) while de-amplifying or blocking other content altogether. Meanwhile, the financial power of capital is such that it wields a vastly disproportionate influence on politics. This is most notable in the United States, particularly in the aftermath of the Supreme Court judgment in Citizens United on campaign financing with some commentators now describing the United States as a plutocracy rather than a democracy or republic. ¹⁰¹

Gilbert and Williams are acutely concerned with how hegemony produces the ‘atomized’ individual under capitalism, and these themes are also evident in Gramsci, Schmitt and even de Tocqueville. Schmitt’s understanding of democracy is intensely public, and his critique of parliamentary democracy is such that it would be almost impossible for large states to exist owing to the emphasis he places on public modes of democratic expression. In Constitutional Theory, Schmitt rejects the secret ballot as democratic because ‘it removes the individual state citizen from the public sphere and transforms him into a private man’. ¹⁰² Instead, the secret ballot is ‘an expression of individual liberalism’. ¹⁰³ For Schmitt, true democracy had to have a public element; he argued that public mass assemblies such as those seen in antiquity are better manifestations of the people than elections. Liberalism’s distinction between the public and private, prioritizing the latter leads to de-politicized individuals. Similar themes are seen in de Tocqueville’s tyranny of the majority. Rather than simply being about the concept of the minority’s rights being at the mercy of an abusive majority, de Tocqueville was concerned about the difficulty in capitalist democracies for people to have the time and energy for political participation. Instead, in a liberal capitalist democracy, people are more likely to prioritize their material needs. When it comes to acting politically, they take short-cuts to the correct answer, treating politics as fashion and assuming that the correct decision that they must make is the popular one. Although voting may be a conscious act, the reasoning underpinning it may look much more like passive consent or contracted-out consent as the will of the majority becomes a self-fulfilling prophesy. The idea that the ‘will of the people’ is something created in the marketplace of ideas, that it is something over which each individual has control, is thus overly simplistic.

Reflecting these views, hegemonic constituent power is the idea that the exercise of constituent power is the product of hegemony in a given legal order. This statement should be understood as descriptive rather than normative. Importantly, hegemony is not immune from change and, for Gramsci, the articulation of hegemony was necessary in order to better understand political power and, in turn, how to seize it. Mouffe and Laclau build upon this, arguing that hegemony must be understood as a ‘contingent strategic effort to construct a collective emancipatory project’. ¹⁰⁴ Politics is therefore ‘a constituent power capable of shaping social, cultural, and economic forms of identity’. ¹⁰⁵ For this reason, Gramsci embraced the importance of action or praxis to effect change in the hegemony.

Agonistic hegemony and Irish reunification

Agonism may prima facie appear as hopelessly idealistic, conceptualizing political conflict as mere disagreement between respectful interlocutors who recognize their opponent’s legitimacy. This conclusion would be mistaken. As agonism does not view such political conflicts as capable of being resolved rationally, agonism, in contrast, is inherently anti-utopian, understanding that disagreement is unavoidable. Drives for consensus decision-making may risk replicating liberalism’s mistakes of de-politicizing disputes, displacing politics and fuelling right-wing populist disaffection. ¹⁰⁶

This is not to completely reject attempts at reaching consensus or deliberation; there must be some forum for deliberation, otherwise disagreements could not emerge. Hence, Honig argues that agonism presents citizens with opportunities for political activity other than revolution by committing modern liberal democracies institutionally to continual world-building. ¹⁰⁷ Ultimately, though, a political decision will have to be taken and it is here that the concept of hegemonic constituent power reveals both its explanatory force and its normative value. Asserting that constituent power is hegemonic avoids the automatic legitimation problem that arises by claiming that the people possess constituent power. Instead, hegemonic constituent power insists that, ideally, constituent power should be possessed by the people but in reality the fictive imaginary of the people is the product of hegemony, with the locus of constituent power lying somewhere between the people and those who claim to speak on their behalf. The legitimating function of hegemonic constituent power can only arise, therefore, if this locus is close to the people. The automatic legitimation problem that can give rise to ethno-national populism can be further avoided if ‘the people’ is constructed not through an antagonistic process of distinguishing friend from enemy, but in an agonistic way – between agonistic adversaries rather than enemies. Agonism is comfortable with majority decision-making while insisting on pluralism in society; without pluralism, there is no possibility for creative disagreement. By requiring open spheres of contestation, agonistic conceptions of democracy seek to construct a valve allowing the release of such pressures through the medium of democratic debate while still facilitating majority decision-making.

Turning to Ireland, this concept of hegemony may run counter to the consensus at the heart of the Good Friday Agreement. However, hegemony reveals a fundamental truth: there are always opposing sides in political disputes and one side always loses. To that end, hegemony already exists in both the United Kingdom and Ireland, and a constitution-making process heavily reliant upon prior existing constitutional constraints or prior existing representatives elected to already existing institutions is simply an embodiment of this hegemony, rather than a repudiation of hegemony per se. If hegemony is unavoidable, then the key question is how hegemony is constructed. The constituent process producing a reunited Ireland and its constitution will not be the product of the will of the people; instead, it will be the product of hegemony. This hegemony will itself be revealed and constructed by the constitution-making process and this must embody a pluralistic people. The best way to achieve this, it is submitted, is through a constituent assembly.

V. Hegemonic constituent power and Irish reunification

The main challenge of both agonistic democracy and Irish reunification is to keep antagonism at bay. ¹⁰⁸ As noted, this may be a latent reason why constituent power has not featured prominently in debates on reunification to date. Injecting constituent power into Irish reunification raises questions of how constituent power should shape Irish reunification and, further, whether processes designed to facilitate the exercise of constituent power would actually make a difference to the legitimacy and stability of the resulting constitutional order. One may assume that the more democratic the constituent process, the closer the locus of constituent power is located to the people and the greater legitimacy the resulting constitutional order has. However, this assumption has been challenged alongside the very notion of whether the constitution-making process matters at all. ¹⁰⁹ The opposite has been claimed by some, with Holmes and Sunstein arguing that ‘the greater role granted to popular referenda and extra-parliamentary authorities, the less constitutionalism matters as a political force’. ¹¹⁰ This view captures the motivation behind the ‘flight to law’ seen in many of the critiques of constituent power.

Writing in 2009, Ginsburg et al. lamented the lack of research on constitution-making, with the field replete with ‘much speculation but relatively little evidence about the impact of different design processes on constitutional outcomes’. ¹¹¹ They concluded that the theoretical side of the debate points towards a consensus about the importance of public involvement in constitution-making. This is complemented by a requisite empirical trend in this direction in practice. However, several of the assumptions underpinning this remained untested. ¹¹² Since the publication of Ginsburg et al.’s conclusions, the literature has advanced considerably, and many of the conclusions reached do not bode well for popular participation. Much of this literature focuses on critiquing what may be considered the archetypal model of constitution-making that embeds public participation to the greatest degree possible: constituent assemblies coupled with a referendum on the draft constitution produced by this assembly.

Constituent assemblies

Constituent assemblies are extraordinary bodies convened for the sole task of drafting a constitution. Delegates to the assembly are elected exclusively for this specific purpose and the assembly should exercise no other power – legislative, executive or judicial. Once this task is completed, the assembly is dissolved and plays no further part in the resulting constitutional order. While the people are the ideal possessors of constituent power, in modern constitutional theory there is an assumption that there is a limit to direct democracy and some form of representation is necessary. ¹¹³ Constituent assemblies are therefore often conceptualized as exercising derived constituent power which, in Sieyès’ original iteration made it possible for him to empower the Third Estate while simultaneously limiting the people’s direct role. ¹¹⁴ Constituent assemblies may thus be conceptualized as the archetypal institutions that exercise constituent power, deriving this power from the people.

This trend towards constituent assemblies has prompted a theoretical and empirical push-back. Theoretically, the concept of unlimited constituent power vested in constituent assemblies has been critiqued by arguing that there should be limits on their derived constituent power. Fasel suggests that these limits are inherent in Sieyès’ original account, arguing that Sieyès proposed that the constituent power of the nation was limited both by the mandate from the people and by natural law. ¹¹⁵ This is an even narrower reading of Sieyès’ constituent power than that proffered by Rubinelli’s aforementioned analysis of Sieyès’ vesting of constituent power in the nation rather than the people. ¹¹⁶ The democratic legitimacy of constituent assemblies have also been questioned on the basis that it is unclear whether there is a distinction between ordinary and higher or constitutional politics, and therefore between ordinary legislatures and constituent assemblies. ¹¹⁷ As such, there is no clear reason why a specially established constituent assembly should be prioritized over the constitutional authority possessing the ordinary legislative power.

This second point is problematic. When voters elect an ordinary legislature, they do so for a variety of reasons – to enact legislation, to hold the government to account and, in the case of parliamentary democracies, to voice support for a specific programme for government through the indirect election of the executive. In this way, the mandate for legislatures to amend a constitution is only tangential at best, even if a promise of constitutional amendment is contained in a party’s manifesto. This mandate is even weaker when tasking it with drafting a completely new constitution. In the context of Irish reunification, a further difficulty arises in that it would be unclear as to which existing legislature should be tasked with drafting the constitution as there is currently no single legislature that represents the people of Ireland as a whole. For this reason, these theoretical objections to constituent power can be rejected in the context of Irish reunification.

Parallel to these theoretical critiques of constituent assemblies run arguments for procedural alternatives owing to what is claimed is the propensity of constituent assemblies to exceed their mandates. Fear of the people is once again prevalent in many of these critiques, which argue that constituent assemblies are prone to exceeding any mandate, or any procedural or temporal limitation placed on them. For example, studies of constitution-making in eastern Europe after the fall of communism and post-dictatorship in Latin America suggest that extra-constitutional constituent assemblies have fared worse at constitution-making than those that have utilized pre-existing constitutional structures as the means through which to effect constitutional change. ¹¹⁸ Under this line of analysis, constituent assemblies are prone to facilitating the domination of the victors over the losers and reducing buy-in from elites under the previous regime. Constituent assemblies thus risk articulating an untrammelled notion of constituent power to legitimate and produce an authoritarian constitutional order. This is possible by virtue of the mandate the constituent assembly claims to possess from the people who elected it and therefore it is the body best placed above all others to speak for the people. This idea of untrammelled constituent power possessed by the constituent assembly has even been given judicial approval in Colombia. ¹¹⁹

To counter constituent power’s unlimited and populist potential, several authors advocate for pre-existing institutions as the means through which a new constitution, if it is needed, should be drafted. ¹²⁰ They attempt to reject victor’s justice and ensure buy-in from existing political elites, many of which are likely to lose power following the enactment of the new constitution. Arato’s ‘Round Table’ constitution-making format is the most in-depth study of this ‘post-sovereign’ model and is an attempt to prevent a runaway assembly by applying the theory of the separation of powers to the constitution-making process. ¹²¹ If constitution-making is separated into different processes with different actors at each stage, the consolidation of unitary constituent power in a single entity is avoided. In essence, it is an attempt at injecting key principles of constitutionalism into the constituent process.

Irish reunification and fear of a runaway assembly?

While these critiques should not be ignored, their applicability to Irish reunification is questionable. Most jurisdictions covered in these studies centre on constituent assemblies in the context of transitions from authoritarian regimes to democratic constitutional orders. ¹²² These jurisdictions are instances of what Gardbaum calls ‘revolutionary constitutionalism’: where the constitution-making process is an attempt to institutionalize and bring to a successful conclusion a political revolution; moreover, it attempts to entrench the political revolution utilizing the tools of constitutionalism. ¹²³ Constitutionalism to these states is revolutionary. Additionally, as is the case with Venezuela’s much-maligned constituent assembly, constituent assemblies may be convened to catalyse a political revolution. ¹²⁴ In contrast, Irish reunification would be undertaken in two jurisdictions that have a tradition of constitutionalism, espousing principles such as democracy, the rule of law, respect for human rights and the separation of powers. These values themselves therefore are not revolutionary. Moreover, the constituent assembly would not be seeking to catalyse Irish reunification as it would not be established until after the initial vote on reunification. The catalysing event therefore is the referendum, not the assembly.

Consequently, studies on constituent assemblies drawn from these jurisdictions are not necessarily applicable to Irish reunification. Nevertheless, the importance of a constraining mandate on a constituent assembly illustrated by many of these studies is an important lesson for Irish reunification. Here again, Irish reunification has a distinct advantage in that any such mandate can be clearly articulated in advance of the initial referendum on reunification with commensurate safeguards also ex-ante prescribed to avoid the possibility of a ‘runaway assembly’. These limits can be illustrated by returning to the parallels drawn between Irish reunification and Brexit. Importantly, Brexit was never in the sole hands of the United Kingdom to deliver. Rather, it required negotiations with entities outside of the United Kingdom, such as the European Union and its 27 remaining states. Much of the dreaded uncertainty surrounding Brexit was the result of politicians promising something that they were unable to deliver unilaterally.

Like Brexit, Irish reunification will require negotiations with parties outside of the island of Ireland – namely, the United Kingdom. In particular, the calling of the initial referendum in Northern Ireland is in the hands of the Secretary of State for Northern Ireland. ¹²⁵ In light of the Brexit experience, it makes sense that such negotiations take place before the initial referendum on Irish reunification takes place. This will give the people of Ireland the requisite degree of certainty required to make an informed and legitimate decision. Further, while in theory everything is ‘up for grabs’ at the moment of the exercise of constituent power, emphasizing that the derived nature of this constituent power can be used to ensure that this mandate is clearly prescribed and to justify certain enforceable and legitimate limitations to its exercise. Colón-Ríos has shown that competing perspectives on constituent power and its relation to law is replete in the historiography of the term. Constituent power vested in a constituent assembly can be understood as being required to be exercised in a manner prescribed by the popular mandate conferred on it. There can therefore be legal obligations on the constituent assembly to produce a constitution, the content of which aligns with this mandate. A constituent assembly on Irish reunification established as such would be in line with this understanding of constituent power. ¹²⁶

Such legitimate limitations placed on the constituent assembly when exercising its derived constituent power could be based on retaining Ireland’s commitments under international law. Namely, these limitations could stipulate that the state will remain a member of the United Nations, the Council of Europe and ECHR, and the European Union. The European Union has already confirmed that any negotiations regarding a united Ireland’s accession to the European Union would take place while Ireland remained within the European Union, mirroring German reunification in the 1990s. ¹²⁷ Relatedly, both Ireland and the United Kingdom (at the time of writing) are members of the Council of Europe, meaning accession to or retained membership of this international organization should also be relatively straightforward. This also has substantial ramifications for the degree and manner of protection to be afforded to human rights within a united Ireland. There is already a principle of equivalence on human rights protections in both jurisdictions embedded in the Good Friday Agreement. ¹²⁸ As both states have incorporated the ECHR into domestic law, the protection of these rights in a united Ireland can be confirmed in advance of a referendum. ¹²⁹ This is further underlined by the fact that both parties have ratified the International Covenant on Civil and Political Rights and similar international human rights instruments pertaining to socio-economic rights.

An initial referendum on Irish reunification could be conducted on the basis that if the referendum were to be affirmed, the resulting Irish state would, at a minimum, recognize the judicial protection of human rights as contained in the ECHR, would be a member state of the United Nations, European Union and would use the Euro. On this latter point, this necessarily entails a substantial degree of this state’s monetary policy being decided at the Eurozone level rather than by internal state actors. Such external negotiations could also be completed in advance of any initial referendum on Irish reunification. This would mean that fundamental political questions pertaining to the liberal identity of the state and the capitalist basis of its economy – for better or worse – would effectively be settled, meaning that a more limited horizon of possibility would be open to a subsequent Constituent Assembly exercising derived constituent power.

If these international negotiations are settled in advance of an initial referendum on reunification, then Brexit is no longer a useful comparator. Any remaining decisions to be taken would not necessarily require negotiations with external entities and so the people of Ireland would be free to decide as they see fit – that is, through a pluralistic and democratic process. That stated, existing judicial institutions could be empowered to ensure that the constituent assembly stays within the bounds of its mandate. The possibility of this constituent assembly receiving juridical support for unfettered constituent power would be low, as the initial referendum supporting Irish reunification could make it clear that the mandate accorded to any subsequently established constituent assembly is derived from the people and limited in the manner prescribed by the people.

The people and the constituent assembly

Critiques of constituent assemblies often assume that the resulting constitution will be in effect in the same polity. They express concern that a single constituent assembly claiming to possess the constituent power of the people is at risk of transforming itself into a ‘runaway assembly’ and abuse its power. In contrast, Irish reunification will result in the creation of an entirely new state, with new geographical borders and a new population of citizens that, prior to reunification, were part of two separate states. Trust cannot be placed in pre-existing institutions to deliver constitution-making as there are no specific pre-existing institutions for this entity. Instead, there are separate institutions in the Republic of Ireland and Northern Ireland and, while there are some cross-border institutions established by the Good Friday Agreement, these are unlikely to form the basis for a constituent assembly. Consequently, existing bodies are not suitable to be tasked with constitutional design for a United Ireland.

This issue of two jurisdictions becoming one raises the further, more existential question of two peoples potentially becoming one. On this point, in attempting to rebut the criticisms of constituent assemblies in the literature, a rather conservative and limited role for a constituent assembly for Irish reunification may inadvertently have been presented; one stressing that the derived constituent power of the constituent assembly must be exercised within the constraints of its prescribed mandate. This model proposed may sound little different to the two-stage ‘roundtable model’ of constitution-making with the pre-referendum stage dominated by existing elites. Far from unleashing constituent power’s radical and innovate potential, the proposed model may lend weight to more radical objections to elected constituent assemblies that contend that the old regimes have a propensity to dominate them still. ¹³⁰ There is certainly a risk of this; however, the concept of hegemonic constituent power confronts this by expressly acknowledging that the actual exercise of constituent power is the manifestation of hegemony, of which residues of the old regimes may certainly comprise part. Normative legitimation of this order is not axiomatic; rather, it is dependent upon the degree to which the locus of constituent power can be considered as located near the people. Here, a constituent assembly on Irish reunification can have a truly innovative and positive role regarding the construction of a pluralistic conception of the people through an agonistic process. The constituent assembly is fundamentally necessary to ensure that the descriptive possessor of hegemonic constituent power is as closely aligned as possible with this agonistic, pluralistic people – the ideal possessors of this constituent power. Key hegemonic factors affecting this process will invariably be the pre-existing constitutional authorities and pre-existing law in both the Republic of Ireland and the United Kingdom. These will have a pivotal role in setting the parameters and mandate of the constituent assembly in advance of the referendum. However, once this role has been completed, the constituent assembly must be allowed to do its work and exercise the derived constituent power delegated to it.

Murray and O’Donoghue argue that,

while the identity of constituent power holders is often taken for granted, radical changes such as the merger of two polities necessarily redefine constituent actors. In a process to decide whether there should be a united Ireland, there are two existing groups of constituent power holders. ¹³¹

Elections to the constituent assembly, the constituent assembly process and the subsequent referendum on the draft constitution itself will open up spaces for this redefinition of constituent actors, not least the people themselves. These elections should be based on a system of proportional representation as alternative methods such as first past the post artificially reduce the plurality of voices. Holding the Constituent Assembly after the vote for reunification also reduces the possibility of a unionist boycott. Reliance on pre-existing institutions alone such as parliaments with their standardized rules for procedure, strict whipping system and domination by existing political parties necessarily erect barriers to true innovation. This is further compounded by the specific lack of an all-Ireland assembly comprising elected representatives from the entirety of the island.

Ultimately, ‘Northern Ireland’s “constitutional question” is, in essence, about constituent power’. ¹³² The exercise of constituent power creates a unity out of a multitude; it creates a ‘people’ out of peoples. At the heart of a vote in favour of reunification lies the reality that such an outcome is irreconcilable with the identity-giving belief of unionists that Northern Ireland remain part of the United Kingdom. This may sound obvious, but it reveals the fundamental question of identity that lies at the heart of Irish unification: the depth of political ramification that will flow from a referendum on Irish unity. It also reveals the limitations of a deliberative consensual politics. Only an agonistic conception of politics can broach this issue in a manner that simultaneously rejects any notion of triumphalism or victor’s justice – a rejection of the idea of the triumph of the friend over the enemy. This agonism must be directed towards ensuring the construction of a new hegemony and that hegemonic constituent power is exercised in the name of a pluralistic conception of the people. Here again, a constituent assembly is vital in delivering this pluralism.

The importance of voices beyond the existing hegemony and allowing space for this re-contestation is underlined by Houghton and O’Donoghue, who stress the need to understand manifestoes as claims to constituent power. They argue that it is unclear why some manifestos such as that of the US Declaration of Independence become fundamental constituent documents of a state, inherently enmeshed in this state’s constitutional identity; in contrast, feminist manifestos issued in similar conditions do not. Here, hegemonic constituent power algins with these feminist critiques to argue that this is due to the dominant ideas and ideology of the ruling classes shaping and influencing the consent of the other groups necessary to maintain a regime. As the grievances outlined in manifestos ‘underscore the limitations of a legal order that responds only to the concerns, needs and interests of the “constructed” constituent power-holders’, space is needed whereby these grievances can be articulated. ¹³³ Understanding the importance of manifestos as claims to constituent power aligns with the Gramscian emphasis on praxis as effecting change and challenging those who claim to speak on our behalf. ¹³⁴

A constituent assembly and elections to it would provide a forum for these voices to help shape the hegemonic order that possesses constituent power. This must be done in a manner not wholly dependent upon pre-existing institutions and political parties due to the dominance of existing hegemonic forces in these institutions. A constituent assembly paves the way for a pluralistic conception of the people, avoiding the bifurcation of identities on the island of Ireland into mutually exclusive and antagonistic camps. Instead, the constituent assembly can create a forum for people to express their multiple, overlapping identities – including but certainly not limited to race, religion, local and global identities, gender, sexuality and class. This would provide better realization of a pluralistic, agonistic conception of the people than attempts at constitution-making that would entrench nationalist and unionist traditions in the constituent process or based on pre-existing institutions’ currently entrenched norms. Such bifurcated entrenchment of Irish nationalist and unionist identities is more likely to arise from a constitution-making process that is heavily dependent upon pre-existing institutions and that views constitution-making as ‘bargaining’ or ‘consensus’ between elites. ¹³⁵ The public space is ‘as much a space for domination as it is for debate’, and this must be borne in mind when considering fora for constitution-making, not least those prioritizing pre-existing institutions and groups. ¹³⁶

A process of constitution-making heavily dependent upon pre-existing institutions or viewed as a ‘bargaining process between elites’ is more than likely to entrench existing identities, rather than allowing for the emergence of new ones. Such enfeeblement of the people would, in turn, feed what has been described as the ‘discouragement of active citizenship’ that plagues democracies today. ¹³⁷ This, in turn, could feed the atomized individual, disempowered and disinterested in realizing an agonistic, pluralistic politics. Consequently, a Constituent Assembly on Irish reunification is absolutely essential. Mere citizens’ assemblies that are based upon randomly drawn but demographically representative individuals who produce advisory guidance for constitution-makers are insufficient. This is not to say that they cannot or should not have a role in Irish reunification and certainly citizens’ assemblies could play an agenda-setting role for the constituent assembly, but this role should not be overstated. Other mechanisms could also be utilized to ensure a concrete connection between delegates to the constituent assembly and those who elected them. Partlett, for example, demonstrates the importance of the concept of ‘delegation’ as distinct from ‘representation’ in American understandings of constituent power. Delegation is not synonymous with representation. This is important, as the concept of delegation confers on the elected delegate a much narrower authority, and therefore discretion, than a representative has. A representative would have more freedom to exercise their conscious in contrast to delegates who would be required to stay within the bounds of the instructions given to them by their delegators. Embracing delegation rather than representation could further ensure that the constituent assembly stays within the bounds of its mandate conferred upon it by the initial referendum on reunification. Conceptualizing constituent assembly members as delegates would also point towards procedures and for a to facilitate communication and instruction between delegates and their electors. Local ‘town halls’ or other innovative public spaces could be utilized. While there has been some acknowledgement of these forums in the Irish reunification debate and movement, their importance in constructing this new understanding of the people and being necessarily connected to the people’s role as constituent actors has not been appreciated fully. Hegemonic constituent power and an insistence on an agonistic, pluralistic understanding of the people reveals that such fora are not simple additional niceties; they are essential.

It follows that hegemonic constituent power possessed by a pluralist conception of the people is therefore something that may take considerable time an effort to achieve. And it would not be achievable solely through a constituent assembly enacted after a vote for reunification. It would point in the direction of a constituent process that was not solely reliant upon a simple majority vote for reunification; rather, it may suggest that such a vote, while legally necessary, should be almost a foregone conclusion or one in which there is clear indications that it would pass. In this regard, hegemonic constituent power may echo some of the concepts contained in Lerner’s notions of incremental constitutionalism. For Lerner, constitution-making in ‘deeply divided societies’ may be more successful if the concepts of revolution are eschewed in favour of viewing constitution-making as simply ‘one stage in a long-term evolutionary process of collective redefinition’. ¹³⁸ Lerner expressly refers to the Irish Free State Constitution in 1922 as an example of a constitution that avoided particular contentious issues around many of the symbolic aspects of the new state. This bears strong similarities with Doyle’s aforementioned analysis of the subsequent enactment of the 1937 Constitution without a ‘big-bang event’ owing to the institutions established under the 1922 Constitution it was replacing. ¹³⁹ As such, a constituent assembly should be viewed as a stage in the process of constructing a pluralist conception of the people; moreover, it is not necessarily the first stage – and, indeed, the changing nature of identity in Northern Ireland, not least the increasing emergence of individuals who identify as neither Irish nationalist or unionist, suggests that this process is already ongoing. ¹⁴⁰ That acknowledged, the drive towards universal consensus must not be over-stated, and ultimately, as the process should be agonistic, it must be dependent upon the will of the majority. Most importantly, it points in the direction of a constitution-making process that cannot and should not be dominated by elites.

VI. Conclusions

Discussions on Irish reunification have so far ignored the question of constituent power. ¹⁴¹ While this may be understandable owing to the history of political violence on these islands, coupled with the lack of historical experience of framing constitutional creation in terms of constituent power, this explanation does not provide a justification or defence of such side-stepping. Ignoring constituent power does not somehow mean that Irish reunification will not constitute an exercise of constituent power; it will still amount to an exercise of constituent power in the descriptive sense. However, such failure may run the risk of a constitution-making process that shifts the locus of constituent power away from the people and, in turn, prevent a democratic process that newly imagines the people themselves.

This debate on Irish reunification and constituent power more generally should not be dominated by fear. After all, according to Arendt, politics is about action understood as moments of innovation. ¹⁴² By emphasizing fear, we run the risk of stifling innovation. This, in turn, may be ossifying, entrenching existing power structures and existing antagonisms. Irish reunification presents the possibility of a new Ireland and new collective identities. It also presents the possibility for new understandings of constituent power, acknowledging its hegemonic nature while insisting on its ideal possession by an agonistic, pluralistic people. A constituent assembly is, it is submitted, the best way to deliver this. A second referendum to affirm this constitution could further corroborate its attribution to the people.

Agonism stresses the somewhat paradoxical position that politics must be structured in such a way that decisions are possible, yet all questions must remain open to debate. While the literature on what an agonistic constitution may look like is sparse or almost non-existent, ¹⁴³ it is submitted that it ought to correlate closely with an open concept of constituent power that insists the people must be able to challenge those who claim to speak on their behalf. The resulting united Ireland constitution must therefore decide certain fundamental issues, while nevertheless allowing such decisions to be questioned, challenged and amended through democratic processes and, where necessary, involve the people. It follows that hegemonic constituent power must be an open concept of constituent power as it insists that the pluralistic people must be able to challenge those who speak on their behalf. It is not merely exhausted at the moment of the constitution’s foundation; it will continue to animate and drive change. Consequently, a hegemonic constituent power that insists on a pluralistic conception of the people exercised through an agonistic conception of politics must have implications for what the resulting constitutional order should look like. ¹⁴⁴ A detailed examination of this is beyond the scope of this article; it is up to the constituent assembly to decide.

Ultimately, constitutional theory and practice must take hegemony seriously. It allows us to insist on the importance of democratic participation while realizing that any claim to represent the true will of the people will always be flawed. There will always be the propensity for hegemony to both shape and interpret both the will of the people and who this fictive people is/are in the direction it wishes. As such, claims by the hegemonic power to represent the will of the people are always both legitimate and illegitimate. It is only by holding on to these two conflicting ideas that we can both enable democracy and constrain its excesses.