Power Sharing in Deeply Divided Societies: Consociationalism in Wonderland and the Northern Ireland Peace Process

Abstract

Consociationalists have claimed nearly 50 successful cases of conflict management. This article argues that Classic Consociationalism has been constantly revised so that the theory has become ambiguous, incoherent, and even contradictory. Consociationalists have entered Wonderland in the sense that their theory now appears to be anything that they want it to be. Paradoxically, theoretical incoherence has made Consociationalism academically powerful because the theory can be marketed as all things to all people. Consociational theorists have revised their theory to claim a successful Northern Ireland case study and used its academic prestige to advance an anti-unionist political agenda.

Consociationalism is the dominant theory of conflict management in both democratic and authoritarian states, its proponents having claimed success for their theory in approaching 50 cases. In 2019 Matthijs Bogaards, Ludger Helms and Arend Lijphart found, particularly since 2000, growing interest and support for Consociationalism and a literature that was dominated by the Northern Ireland case. Northern Ireland “has become a prism through which many of the facets of consociational democracy are studied.” In their survey of Consociationalism Bogaards et al identified very few critics reinforcing the sense of the theory’s dominance of the debate (Farag et al claim 69 cases of “powersharing” which they consider synonymous with Consociationalism).¹

Over 25 years ago, Ian Lustick argued in World Politics that Consociationalism had not triumphed because of the coherence and logic of the theory but because of its power within the academy. Consociationalism was a “degenerate” research programme and he expected to see its demise.² This means that Consociational academics can constantly (re)define the terms of debate (by claiming all powersharing is Consociational), exclude critics from debate and claim success for their theory. Paradoxically, this article will argue, the power of Consociationalism has increased precisely because the research programme has degenerated and entered Wonderland. Because the theory is so ambiguous, incoherent, and contradictory critics can be deflected by claiming that Consociationalism has been “wrongly interpreted.” But by being so ambiguous Consociationalism can also incorporate its critics, who may be happy to join a powerful paradigm.

The successful transition of Northern Ireland from an intense, violent conflict to a peaceful accommodation is claimed as the key case that demonstrates the effectiveness of Consociational theory.³ Leading Consociationalists, John McGarry and Brendan O’Leary, have been key advocates of the application of Arend Lijphart’s “Classic Consociationalism” to Northern Ireland. They argued that the “Belfast or Good Friday Agreement” was “Consociational” and should, therefore, be implemented according to Consociationalism’s segregationist principles. Claims of success in Northern Ireland gave Consociational academics the credibility to advise the Kurds on the 2005 Iraqi Constitution. The failure of the Belfast Agreement to sustain moderate powersharing led McGarry and O’Leary to welcome the electoral triumph of the “hardline” parties in 2007 and declare the conflict resolved. Subsequently, they abandoned the Belfast Agreement and endorsed an integrated united Ireland.⁴

This article is both a critique of Consociationalism and of leading Consociationalist’s interpretation of the Northern Ireland peace process. The article first, provides an interpretation of Arend Lijphart’s “Classical Consociationalism.”⁵ Second, it is argued that Consociationalism’s pessimistic primordialism or essentialism led the theory towards conservative realist conclusions. Third, by the turn of the century, the apparent successes of the South African and Northern Ireland peace processes led Consociationalists to transform their theory and develop a more optimistic liberal idealist and normative interpretation. Fourth, “Revisionist” and “Complex” Consociationalism were invented to incorporate liberal, integrationist, and Cosmopolitan critics within the theory and create the illusion that there is no alternative to Consociationalism. Fifth, constant revision led Consociationalists into Wonderland, where the theory was defined instrumentally and contextually to claim success and distance themselves from failure. Finally, the Consociational interpretation of the Northern Ireland case illustrates the ambiguity, incoherence, and contradictions of the theory. But also shows how Wonderland Consociationalism (often reflecting Classical Consociational assumptions) can be instrumentalised to pursue the nationalist-republican ideological goals of Consociationalists (or of the Kurds in Iraq).

Interpreting Classical Consociationalism

Consociationalism is a highly controversial and influential theory of conflict management because, critics argue, it is based on a discredited Primordialist or essentialist view of the subject and consequently prescribes “segmental autonomy,” or “voluntary apartheid” and authoritarian rule by “elite cartel.”⁶ The world-renowned Dutch political scientist, Arend Lijphart, based Consociational theory on his interpretation of how the Netherlands managed “plural conflict” between 1917 and 1967. In his seminal work, Democracy in Plural Societies (1977) he argued that the Dutch experience was a model for the global management of conflict. Consociationalism is a universal, one size fits all, analysis, and prescription to manage conflict in “plural societies,” suggested by their claims to approximately 50 highly diverse case studies.

Consociationalism was originally built on a Primordialist foundation which describes “ethnic identities” as biological and deeply rooted in human nature.⁷ This gives Consociational theory a strong structuralist orientation suggesting that these primordial identities are “facts” to be accepted and worked around rather than, as Constructivists might argue, challenged and remade. Consequently, there is a preference for the segregation of groups through pillarisation or “voluntary apartheid.”

Consociational theory differs from other theories of integration not only in its refutation of the thesis that cultural fragmentation necessarily leads to conflict, but also in its insistence that distinct lines of cleavage among subcultures may actually help rather than hinder peaceful relations among them. Because good social fences may make good political neighbours, a kind of voluntary apartheid policy may be the most appropriate solution for a divided society. Political autonomy for the different subcultures is a crucially important element of a Consociational system, because it reduces contacts, and hence strain and hostility, among the subcultures at the mass level.⁸ (my emphasis)

Consociationalists favour the consolidation and reinforcement of these group identities to make them into the stable pillars on which communal elites can build a settlement. The image of pillars suggests strong but separate columns: “Consociational democracy results in the division of society into more homogenous and self-contained elements.” The territorial intermingling of nationalists and unionists in Northern Ireland, for example, was seen as a negative condition for Consociationalism.⁹

The “primary instrument” of Consociationalism is an “elite cartel” involving “…government by a grand coalition of the political leaders of all significant segments of the plural society.”¹⁰ While consensus did not exist at a mass level, it was achieved at the elite level and could be imposed on a deferential population. What matters for Lijphart is the consensual behaviour of elites and rejection of the adversarial style of politics.¹¹ Consociation, with its prescription of rule by elite consensus and party and popular deference, is not very democratic. The theory is applied to and may be more compatible with non-democratic states.

Consociationalists oppose social integration because they claim it is “unwanted” and likely to antagonise rather than ameliorate a conflict situation. McGarry and O’Leary conclude their attack on integrationist initiatives, “it must always be remembered… at best … [they] will fall short of what is required to resolve the conflict, and at worst they may even deflect attention and energy from the crucial political measures necessary to change the logic of the cruel game in which the participants are presently trapped.”¹² John McGarry argued, “… The problem with integrationist solutions is that they require a willingness to be integrated, and no such willingness exists in deeply divided societies. … Many blacks in the United States are now coming to realize, ironically, that the separate but equal doctrine in ‘Plessey v. Ferguson’ is more attractive than the separate means unequal doctrine of ‘Brown v. Board of Education’.”¹³

Consociationalism, therefore, prescribes the segregation of ethnic groups into ethnic pillars, or “verzuiling,” that will be dominated by their respective ethnic elites. This cartel of ethnic elites is assumed to be benign and to have the power to manage conflict through cooperation over the heads and often against the wishes of their ethnic pillar. Lijphart’s original theory contained seven “favourable conditions” or “factors” for Consociationalism which “are conducive to overarching elite cooperation and stable nonelite support.”¹⁴ According to Bogaards these favourable factors have varied over time and in all he lists 14 separate “favourable factors”:

1968: 6

1969: 8

1977: 9

1985: 8

In 1977 these 9 conditions were,

1. Segmental isolation and federalism

2. A multiple balance of power among the subcultures

3. External threats, small country size

4. A relatively low load on the system, small country size

5. Overarching loyalties

6. Moderate multiparty system

7. Representative party system

8. Crosscutting cleavages (in some instances)

9. Tradition of elite accommodation¹⁵

“Favourable conditions” reflect and provide further evidence of Consociationalistm’s “groupist” diagnosis of conflict which flow from its Primordialist/Ethnonationalist assumptions.¹⁶ The structuralism of these “favourable conditions” was contradicted by the agency orientation of Consociationalism’s prescriptions which suggested that the benign motivation of elites could create Consociation even in unfavourable conditions. Primordialism or Ethnonationalism emphasise such a powerful biological or cultural influence that political actors are determined by or at least powerfully influenced to perform in an ethnonationalist way.¹⁷

Consociational analysis, or description of conflict, leads logically to their prescriptions for managing conflict. They seek to achieve “segmental autonomy” so that well-motivated elites can manage conflict insulated from their “followers,” the masses are assumed to be “rather passive and apolitical.”¹⁸ Consociationalism’s four institutional prescriptions, therefore, should be interpreted as an attempt to achieve “segmental autonomy” and rule by well-motivated elites.

There are three key elements to Consociationalism:

1. Description, analysis or diagnosis of conflict: originally Primordial (or biological) but moving over time towards an ethnonationalist or culturalist interpretation.

2. Prescription: “Segmental autonomy” and rule by elite cartel. This persists but there has been a shift away from such explicit language.¹⁹

3. Four institutional prescriptions:

4. Grand coalition – the political leaders of all significant elements should share power in a consensual or cooperative coalition: “Elite cooperation is the primary distinguishing feature of consociational democracy.”²⁰

5. Proportional representation – in the distribution of government resources and List system of PR in elections since it tends to give elites control over selection of candidates and does not provide incentives for cross-community voting.²¹

6. Mutual veto – a minority veto on vital rights and autonomy.

7. Autonomy – maximise each pillar’s self-government.

Cosmopolitan critics of Consociationalism are not necessarily opposed to the four institutional prescriptions as such. It is the implementation of these prescriptions to achieve “segmental autonomy,” rather than “integration,” and rule by elite cartel (authoritarianism) rather than popular empowerment (the quality of democracy) that are at issue. Consociationalists say little about if and how an integrated society will be created in the future beyond an expressed hope that the pillars will naturally erode or “biodegrade.”²²

Consociationalists claim to have a universal solution to conflict (claiming approximately 50 cases) which suggests that their prescriptions are to be implemented with little regard to context (see Table 2). These cases are as diverse as the Soviet Union, Switzerland, Lebanon, South Africa and Sri Lanka. Consociationalists are positivists who assert that they are neutral observers of the “facts.” They are “realists” because they are prepared to see and accept the world “as it is” and define themselves against more optimistic, integrationist advocates of idealist Cosmopolitanism. Consociationalists are committed to “objectivity” and dismiss the influence of their values or the normative implications of Consociationalism and its impact on the world.

Conservative Realism: Pessimism and Triage

Consociationalism takes a positivist approach to the social sciences, and this explains some of the drawbacks of this theory. Positivism applies the methods of the natural sciences to the social sciences to uncover useful objective knowledge, developing general, universal laws from which hypotheses can be made to predict the future. The positivist sees herself standing outside the context in which she pursues research and is, therefore, able to find scientific truth and produce facts that are free from value judgements and personal subjective elements. Positivists, therefore, see themselves as precise in their scientific use of concepts, presenting themselves as empiricists who see normative work as separate.²³

Post-positivists reject this approach and argue that all descriptive and normative theory contains elements of both types of theorising. There is no neutral point from which to observe the world so social science cannot be “objective” and “value free.” Theory and experiment are not separable; rather theory affects both the facts we focus on and how we interpret them. Consociationalism’s Primordialist (or essentialist) assumption is incompatible with positivism because it suggests that the Consociational analyst herself will be determined or strongly influenced by their ethnic background and this will affect their research. The existence of diverse interpretations of the Northern Ireland conflict and of Consociational theory itself lead Consociationalists to try and explain this away by claiming that they are the only “objective” academics and all other academics are subjective.²⁴

Post-positivists may see the assertion of scientific neutrality and empiricism, therefore, as rhetoric that disguises the normative position of Consociationalists. Positivism tends to be structuralist because of its emphasis on explaining reality rather than what could and should be. The conservative assumption of regularity underpins structuralism, and this creates problems in explaining political change, such as the shift – for example in South Africa and Northern Ireland – from armed conflict to peace processes.

Arend Lijphart, since his classic formulation in 1977, has shifted from Consociationalism as descriptive theory (realism) to Consociationalism as normative theory (idealism). It is no longer necessary, he argues, to have an analysis of conflict to prescribe for that conflict.²⁵ There is an affinity between Positivism and Conservative Realism. They portray themselves as reluctantly concluding that the “fact” of Primordialism or Ethnonationalism must be accepted and worked around. The spectre of barbaric “ethnic” war is used to justify the implementation of Consociationalism’s prescriptions as emergency triage. This world-weary realism sees a gap between popular, publicly proclaimed idealism and the reality of private primordialism and antagonism. The choice posed by Consociationalists is first, between a highly restrictive Consociational Democracy (that may act against the interests of ethnic followers and favours a deferential people) and no democracy at all.²⁶ Second, there is a choice between segregation and elite control on the one hand and primordial slaughter on the other. This rhetorical strategy allows Consociationalists to claim credit where conflict becomes less violent (Northern Ireland after 1994 or Lebanon 1943–75) and to relinquish responsibility when conflict is more violent (Northern Ireland 1975–94, Lebanon 1975–90).

The problem for Consociationalists is that their pessimistic primordialism led them, in the early nineties, to predict that the conflict was getting worse when there were developments behind the scenes that led to the peace process. Consequently, Consociationalists could not convincingly account for the success of the peace process and the Belfast Agreement, which exceeded Consociationalism’s limited and conservative institutional prescriptions and was explicitly integrationist (rather than segregationist). Their over-optimistic claims for Consociationalism’s success were then confounded by the collapse of powersharing in 2017 whereupon O’Leary advocated for an integrated (non-Consociational) united Ireland.

While Consociationalists claim to be Realists they make “unrealistic” assumptions about politics. There is an important contradiction between Consociationalism’s Primordialist, or essentialist, assumption, and its elite prescriptions. Primordialism suggests powerful, unstoppable, structural forces coming from the people below and logically this strongly constrains the agency of political actors. Yet, simultaneously, Consociationalism attributes considerable agency to political actors in managing conflict. Illogically, Consociational theory both suggests political actors have little agency and considerable agency. Consociationalists make heroic or utopian idealist assumptions that political elites:

1. Will be motivated to engage in consensual, conflict management.

2. Will be able to lead their parties and voters to a settlement that may be against their wishes.

3. Will engage in a cooperative, power sharing executive.²⁷

Presumably, if these conditions of elite dominance do not exist then Consociationalists would favour the engineering of these authoritarian conditions. When Consociation was seen to have failed in Northern Ireland, Consociationalists prescribed solutions which lacked consensus and had little party or popular support, such as repartition and joint authority (see below).

Political elites (as the peace process showed) are not necessarily benign (or malign), and they enjoy a complex range of relations with supporters and the people they are supposed to represent as well as other powerful groups and interests. They can rarely be sure of the intentions of their followers let alone those of rival politicians, whether within or outside their party.

Consociationalists support a “separate but equal” policy and play down the role of materialism in managing conflict. For example, Lijphart did not see as significant inequality between or among “Catholics” and “Protestants” in Northern Ireland.²⁸ There is ample evidence that the promotion of contact in the right context and of the right quality can have a beneficial effect in reducing prejudice and conflict. Despite Consociational claims, there is also strong evidence of a desire for integration in Northern Ireland. Certainly political parties do not openly advocate segregation and opinion polls show considerable support for integration, people prefer to work in mixed-religion workplaces, live in mixed religion areas, do not mind if people marry people from the “other” religion and send their children to mixed religion schools (http://www.ark.ac.uk/nilt/results/comrel.html#contact).²⁹ Although there is some truth in the claim that people tend to conceal some of their beliefs from pollsters, but even allowing for this the evidence of support for integration is consistent and substantial.

Liberal Idealism and Normative Theory: From Pessimism to Optimism

Consociationalism is ambiguous and this has led to attempts to incongruously combine Conservative Realism with Liberal Idealism. Lijphart had argued that Consociationalism was a descriptive theory that described the world as it really is and, based on that analysis, arrived at its theory of conflict and four institutional prescriptions. Increasingly, he argued that Consociationalism was not descriptive but normative and by the mid-80s he was arguing for its democratic merits.³⁰ Bogaards argues that by the early nineties, “Consociationalism no longer derives its usefulness from the accuracy with which it describes existing political systems in relations to their divided societies, but from its usefulness as a democratic solution for divided societies such as South Africa. The main implication is that Consociationalism as a normative type becomes immune from empirical criticism.”³¹ In 2015 Lijphart argued that it is not necessary to diagnose conflict to prescribe for it,

In a way I am more like an “irrelevantist”: we do not try to figure out what the causes of the ethnic conflict are, we just look at the ethnic conflict and see how it can be solved.³²

This statement assumes that there is an analytic consensus that a particular conflict is “ethnic” rather than considering an “ethnic” explanation as one of several possible interpretations of conflict. There are always conflicting interpretations of conflict, and these are often fiercely contested because of their implications for prescription.

Consociational theory, because it claims about 50 successful cases, and makes the same prescriptions in every case is a universal theory which pays little attention to context. Consociationalism has adopted a single “primordialist” or “ethnonationalist” interpretation of “plural conflicts” that leads logically to Consociationalism’s segregationist and elitist conclusions. Lijphart’s “irrelevantist” position side-steps the complexity and contestability of interpretation by assuming that the conflict is “ethnic,” without defining what this means. The result is that the Consociational prescriptions are advocated for every conflict regardless of the context.³³

Revisionist Consociationalism: Incorporating Critics

Powerful critiques of Lijphart’s Classical Consociationalism led to McGarry and O’Leary inventing “Revisionist Consociationalism.”³⁴ First, in keeping with Consociationalism’s primordialist assumptions and pessimistic realism, Consociationalists believed the conflicts in South Africa and Northern Ireland were deteriorating just as they shifted into peace processes. The relatively benign outcomes of these two “deeply divided societies” forced Consociationalists to revise their theory to consider developments in these cases. If Consociationalism couldn’t convincingly explain influential cases of conflict management then their theory could be fatally undermined. Second, the Primordialist (biological) assumptions which underpinned Classical Consociationalism were discredited, not least through their association with violent, right wing ideologies, such as the Apartheid system in South Africa. Rupert Taylor has argued that Lijphart’s statements on apartheid in South Africa appeared to offer a defence of National Party policy by assuming that ethnic differences were “an unalterable fact” and that the apartheid government’s policies had succeed not in manufacturing differences but in counteracting and softening them.³⁵ Consociationalists replaced Primordialism with Ethnonationalism, which emphasised a deeply rooted cultural basis for identity rather than biology. The implications of the two positions were, however, very similar.

Third, “Cosmopolitan Liberal” and “Integrationist” critiques had a sympathetic audience within the academy because they exposed Consociationalism as a conservative nationalist ideology that opposed liberalism and supported segregation (with all its negative connotations). Such attacks limited Consociational theory’s appeal so the theory was revised so that Cosmopolitan Liberals and Integrationists could be incorporated within Consociational theory, as “Revisionist” or “Complex” Consociationalism. By incorporating all powersharing opposition, it can be claimed that there is no alternative system of powersharing to Consociationalism.

McGarry and O’Leary’s more optimistic Liberal Idealism was reinforced by their support for the invasion of Iraq in 2003 and the project of transforming that country through Consociationalism into a liberal democracy or a Western protectorate. Shortly after the invasion O’Leary stated: “viable consociations that address ethno-national disputes may have to be the de facto or de jure protectorates of external powers.” He points out “High commissioners appointed by great powers, as in Bosnia and Herzegovina, are indistinguishable from the prefects of protectorates.”³⁶ Another Consociationalist states: “Consociation has become a tool favoured by the west for intervention in regions where its interests are threatened.”³⁷ McGarry and O’Leary’s Liberal Idealist and optimistic attitude towards the transformation of Iraq contrasted with their prior advocacy of Consociationalism as “a form of pragmatic realism” and “political triage.”³⁸ Consociationalism’s essentialist assumptions might have suggested that the invasion of Iraq was unwise in removing the authoritarian power that prevented “ethnoreligious” civil war. Academic experts on Iraq did warn policymakers of the consequences of the invasion.³⁹

Consociationalism in Wonderland

In Lewis Carroll’s Alice Through the Looking Glass (1871) Alice is perplexed,

“When I use a word,” Humpty Dumpty said in rather a scornful tone, “it means just what I choose it to mean – neither more nor less.”

“The question is,” said Alice, “whether you can make words mean different things–that’s all.”

“The question is,” said Humpty Dumpty, “which is to be master–that’s all.”

Consociationalism has been constantly defined and redefined in an ambiguous and elastic way that allows it to be marketed as “all things to all people” and to claim favourable cases of conflict management and reject unfavourable ones. Donald Horowitz stated, “The tendency to shift the goal posts and to claim countries for the theory is palpable. Whenever a divided society seems to be more or less democratic and more or less lacking in the most severe forms of conflict, the reason must be that it is consociational.” Such ambiguity had been apparent in the eighties and nineties (see Table 1).⁴⁰ But since the 1990s “Revisionist” and later “Complex” Consociationalism have created a proliferation of definitions and it can now be: non-ethnic, democratic, non-democratic, regional, central, weak, ambivalent, complete, pluritarian, traditional, revisionist, corporate, liberal, rigid, concurrent, complete, semi, quasi, formal, informal, flexible, “low fat,” hybrid, “fully-fledged,” light and intra-party rather than inter-party. While McGarry and O’Leary assert that consociationalism should be falsifiable it is not clear how this can be done.

Table 1. The elasticity of consociationalism.

Consociationalism full definition	Full definition	Conditions and prescriptions	4 Prescriptions	2 Primary characteristics or instruments (grand coalition, autonomy)	1 Primary instrument (consensual, grand coalition)	½ Primary instrument (participation, grand coalition)
Theory	✓	✗	✗	✗	✗	✗
Primordialism	✓	✗	✗	✗	✗	✗
Prescription	✓	✗	✗	✗	✗	✗
Segregation	✓	✗	✗	✗	✗	✗
Elite Cartel	✓	✗	✗	✗	✗	✗
7 Conditions	✓	✓	✗	✗	✗	✗
4 Institutional prescriptions	✓	✓	✓	✗	✗	✗
4. Minority veto	✓	✓	✓	✗	✗	✗
3. Proportional	✓	✓	✓	✗	✗	✗
2. Autonomy	✓	✓	✓	✓	✗	✗
1. Grand coalition	✓	✓	✓	✓	✓	✓
Source	Lijphart 1977	Lijphart 2015 (irrelevantist)	MCGarry and O’Leary 2009: 350	Lijphart 1995: 859	Lijphart 1977: 25	Lijphart 1977: 31

The definition of Consociationalism can be shrunk from a full definition (column 2) to a definition that leaves out the theory of conflict but keeps in the “7 conditions” and “4 institutional prescriptions” (column 3) to the 4 institutional prescriptions along (column 4), to 2 “primary instruments” (column 5) to 1 primary instrument (column 6) to ½ a “primary instrument” (column 7). Over the years Consociationalism has generally shifted from maximal towards more minimalist definitions. McGarry and O’Leary 2009: 350 select different combinations of 3 or 2 institutional prescriptions to Lijphart.

“Complex” consociation, it is argued, allows the combination of consociationalism with “one other additional strategy” such as integration or partition.⁴¹ So Classic Consociationalism’s prescription for “voluntary apartheid” or segregation could, inconsistently, be combined with a prescription for integration (often considered to be Consociationalism’s “other”). Consociationalism, it would appear, means whatever leading Consociationalists currently choose it to mean, and in the case of Lijphart, McGarry and O’Leary and “New Generation” Consociationalists, they disagree with each other. Consociationalism can, therefore, be redefined according to context claiming whatever appears to be successful and distancing their theory from failure. Even Consociationalism’s “other,” the British majoritarian system, can be declared as Consociational because the “post-war consensus” and the alternation of Labour and Conservative governments could qualify the UK as a “diachronic grand coalition” and, therefore, Consociational.⁴² Or else these parties could be considered “Consociational parties” with powersharing existing within rather than between them.

Consociationalism, like any theory, should be judged by its ability to analyse or describe conflict and, following from that, to make appropriate prescriptions. Yet some Consociationalists have tried to claim that their theory is only prescriptive and not analytic. Debates around the “technical” aspects of the institutional prescriptions have diverted attention from the importance of the theoretical framework that informs Consociationalism’s prescriptions for conflict management. For example, Consociationalists and their critics may all favour “Grand Coalition,” proportional representation, mutual veto, and (some degree of) autonomy. But Consociationalists advocate these instruments to consolidate ethnic pillars while critics – Cosmopolitans or Integrationists – could use these constitutional devices to remake groups in less antagonistic forms and create a more integrated (or intercultural) society.

The full definition, set out in Democracy in Plural Societies (1977), requires the case study to conform to Consociationalism’s primordial or ethnonational analysis of conflict, its 7 favourable conditions and the prescriptions (segmental autonomy and elite cartel) pursued through the four institutional prescriptions (Table 1 column 2). The full definition restricts the number of case studies that can be claimed because fewer cases are likely to conform to such extensive conditions. Consociationalists may choose a more restricted or thinner definition of their theory as they shift along the columns from column 2 to 7 in table 1. If Consociationalism is defined as half a “primary instrument” in column 7, then it means merely participation in a Grand Coalition or “powersharing.”⁴³ Such a thin definition enables Consociationalists to claim more case studies than the full definition in column 2. Reducing Consociationalism to 4 or less institutional prescriptions means that the theory suffers from what Brubaker calls the “architectonic illusion,” that the right constitutional architecture will “solve” conflicts.⁴⁴

Consociationalists can, by reducing the definition of their theory, stretch their claim to more favourable cases, and they have claimed between 4 and 31 cases and overall approaching 50 cases. The more ambiguous the definition the more cases that can be claimed for Consociationalism and the more universal and impressive the theory appears to be. The more precise and detailed the defintion, the fewer cases can be claimed convincingly (see Table 2). The shifting claims to case studies over time, and the disagreements in categorisation between Lijphart and McGarry and O’Leary (particularly over South Africa and India), provide further evidence of the conceptual elasticity and ambiguity of Consociational theory (see Table 2).

Table 2. The case studies claimed by leading Consociationalists over time.

Country	1977 Lijphart	1985 Lijphart	1995 McGarry and O’Leary	1996 Lijphart	2002 Lijphart	2005 O’Leary	2007 Lijphart
Afghanistan						✓	✓
Antilles (NL)	✓				✓		✓
Austria		✓		✓	✓		✓
Belgium	✓	✓		✓	✓	✓	✓
Bosnia						✓	✓
Burundi							✓
Canada	✓	✓	✓	✓		✓	✓
Chile		✓
Colombia		✓		✓	✓		✓
Cyprus	✓	✓		✓		✓	✓
Czechoslovakia				✓	✓		✓
European Union
Fiji		✓	✓		✓		✓
Gabon		✓
Gambia		✓		✓
Guyana		✓
India				✓	✓		✓
Indonesia	✓	✓
Iraq
Israel	✓	✓					✓
Italy		✓		✓
Ivory Coast		✓
Kenya		✓		✓
Kosovo							✓
Lebanon	✓	✓	✓	✓	✓	✓	✓
Liechtenstein		✓
Luxembourg					✓		✓
Malaysia	✓	✓	✓	✓	✓		✓
Macedonia						✓	✓
Netherlands	✓	✓		✓	✓	✓	✓
Nigeria	✓	✓		✓			✓
Northern Ireland					✓	✓	✓
Portugal		✓
South Africa				✓	✓		✓
Spain		✓
Sri Lanka		✓		✓		✓
Sudan		✓		✓
Surinam	✓	✓		✓	✓		✓
Switzerland	✓	✓			✓	✓	✓
Tanzania		✓
Uruguay	✓	✓					✓
USSR		✓		✓
Venezuela		✓		✓
Yugoslavia		✓		✓
Total 44 cases (Lijphart is underlined)	13 (or 9)	31	4	21	15	11	25

Lijphart 1985: 84, “applied” also to European Community, Manitoba, New Brunswick and Northern Ireland; McGarry and O’Leary 1995: 338; Lijphart 2002: 40–41; O’Leary 2005a: 3; Lijphart 2007: 5. In Lijphart (2013: 3) he states that he increased his claim to case studies over the years: 1968, 1; 1977, 9; 1984, 21; 1994, 27; 1999, 36.

The contradictions of Wonderland Consociationalism mean that,

• First, it is difficult to know what a Consociational approach is to managing conflict?

• Second, whether and where this has been implemented?

• Third, if it is claimed that “Consociationalism” has been implemented whether or not it has been successful?

Consociationalists have claimed from a pessimistic 5 successful cases in 1995 to an ambitious 50 cases in 2019 (Table 2). The claim to fifty cases requires a “thin” definition if it is to include cases as diverse as the Soviet Union, Switzerland, Lebanon, Iraq, and the European Union.

Leading Consociationalists strongly disagree over the definition of Consociationalism and the ambition of its claims, although the “New Generation” of Consociationalists do not mention the extent of these differences.⁴⁵ In 1995 McGarry and O’Leary argued that the problem with Consociationalism was that “it has not worked” and identified just 5 cases of Consociationalism to a high of 11 in 2005, although they sympathise with the extension of the literature.⁴⁶ Lijphart has been much more “optimistic” and claimed a minimum of 13 cases (1977) and maximum 31 cases (1985) but does not seem to have rejected any of the 48 cases in Table 2. McGarry and O’Leary (2005) disagreed with Lijphart (2007) on whether 16 out of 26 cases (62%) were Consociational and only agreed on 10 (38%) cases. They disagree on two key case studies those of apparently majoritarian India and South Africa. But even if we take the 48 cases of “Consociational success,” and apparently zero failure, then Lijphart has, roughly, claimed from 13 to 31 cases of success which approximates to a 27–65% rate of success. By contrast, McGarry and O’Leary have claimed from 5 to 11 cases, which approximates to a 10%-23% rate of success.

Definitional confusion means that we cannot identify in which cases “Consociationalism” has been applied, where they believe it has been successful and, more importantly, where it has failed. By defining only successful cases as Consociational its advocates promote the impression that their theory has a 100% success rate. The Wonderland definition (all things to all people) means that any apparently successful case of conflict management can be claimed for Consociationalism, while the theory can be distanced from any apparent failure. Consociationalists are notably reticent about identifying failures, such as the failure of Iraq’s 2005 constitution to end the conflict in Iraq.⁴⁷ The Lebanese case is Consociational when less violent (1943–75), was not blameable for the subsequent civil war (1975–90) but then responsible for the relatively peaceful post-Civil War period (1990–). The Arusha Accords (1993) could be defined as Consociational, because they proposed powersharing, but the subsequent genocide in 1994 is not identified as a Consociational failure.

After the invasion of Iraq, 2003, McGarry and O’Leary acted as advisers to Kurdish parties in Iraq. Consociationalism was redefined this time to reject Consociationalism’s primary instrument of “Grand Coalition” or powersharing. This allowed Consociationalists to claim the Iraqi Constitution 2005 even though it permitted the exclusion of “Sunnis” from government (believed to be approximately 20% of the population). This redefinition meant that “Consociationalism” no longer required grand coalition or powersharing which was supposed to have been at the heart of any definition of Consociationalism (Table 1). In 2005, McGarry and O’Leary explicitly rejected powersharing by arguing that “consociational practices may also prevail without the participation of one or more ethnic segment that is demographically, electorally, or politically significant.”⁴⁸ Their rejection of Consociationalism’s “primary instrument” was restated, “… consociations do not as a matter of conceptual precision require grand coalitions” what is important is “joint consent across all significant communities.” Again in 2008, McGarry and O’Leary declared that Consociation required only “some element of jointness” and “does not require every community to be represented in government….”⁴⁹ Consociationalism’s “groupist” analysis led McGarry and O’Leary to hold all “Sunnis” collectively responsible for the crimes perpetrated by the Baathists.⁵⁰ By 2022, O’Leary had dropped the “primary instrument” of Grand Coalition and replaced it with the more vague “parity” which “implies equality in status and recognition.”⁵¹

This theoretical confusion allows Consociational theory to be marketed as “all things to all people” but obscures the persistence of Classical Consociational theory among its leading advocates which lead logically to elitist and segregationist prescriptions, tempered by their normative orientation. On Iraq, Consociationalists have tended to argue for the reinforcement of ethnoreligious identities and maximum autonomy for ethnoreligious regions, or even the break-up of Iraq because it is an “artificial” state.⁵²

Consociationalism and the “Real Politics” in Northern Ireland

Wonderland Consociationalism, it is argued, despite its flexibility and ambiguity (with its liberal, revisionist and complex variants), has often concealed prescriptions associated with Classical Consociational theory. Consociationalists distanced themselves from the Northern Ireland case (1975–94) but as the peace process developed Consociationalists reinterpreted and redefined their theory to chase Northern Irish political practice and an apparently successful case of powersharing (1994–). Consociationalism can be interpreted to support a unionist perspective but may be more compatible with a nationalist/republican one because of its ostensible emphasis on power sharing and minority rights. Consociationalism’s institutional prescriptions may focus on maintaining the unity of the state, but its elitist and segregationist prescriptions have led Consociationalists to advocate repartition, joint authority, protectorates and, most recently, an integrated united Ireland. These reflect Classic Consociationalism’s orientation towards rule by elites, lack of concern for political polarisation, advocacy of segregation and with little concern for integration and reconciliation. Indicated by their argument that the Belfast/St Andrews Agreement (Belfast Agreement 1998, StA 2006) had resolved the conflict. Arguably, Consociationalism’s technocratic rhetoric also concealed the neo-nationalist/republican normative preferences of leading Consociationalists on Northern Ireland.

Consociational Pessimistic Realism: From the First Peace Process, 1972–74, to the Second Peace Process 1994

The Belfast or Good Friday Agreement (1998) (Belfast Agreement) was described by Seamus Mallon as “Sunningdale for Slow Learners” because there are considerable similarities between the powersharing agreement achieved in 1973 and the Belfast Agreement 1998. Those unionists and republicans who rejected powersharing in 1973–74, he suggests, bear considerable responsibility for the subsequent violence.

Leading Consociationalists, Lijphart and McGarry and O’Leary, all argued that Northern Ireland’s first peace process (1972–74) was an attempt at Consociational powersharing.⁵³ In 1975 Arend Lijphart prescribed repartition for Northern Ireland – an involuntary apartheid – despite the lack of popular or political support for that option.⁵⁴ In 1977, he described the prospects for Consociationalism as “overwhelmingly unfavourable” in “the basically unfavourable environment of Northern Ireland.” Consociationalists found the seven “favourable conditions” to be absent from Northern Ireland and argued that this demonstrated the “Limits of Consociationalism.”⁵⁵ By 1991, Lijphart “almost agree[d]” that “the problem in Northern Ireland is that there is no solution.”⁵⁶ By 1989 McGarry and O’Leary were also arguing for repartition.

New Generation Consociationalist, Allison McCulloch disagrees with Lijphart and McGarry and O’Leary. She claims that powersharing (1974) was not Consociational and represented a rival “Centripetalist” approach. The implication is that “Centripetalism” is responsible for failure in 1974, contrasting with the success of Consociationalism during the successful second peace process (1994–). She argues that the logic of the Belfast Agreement is both “Corporate” and “Liberal” Consociationalism and that the Iraq constitution is a “Liberal” Consociation.⁵⁷

McGarry and O’Leary’s prescriptions for Northern Ireland were informed by Consociational theory’s pessimistic essentialist theory of conflict which leads logically to more segregationist and elitist prescriptions. They supported the Anglo-Irish Agreement, 1985, to coerce Northern Ireland into Consociationalism. Just four years after the failure of the Anglo-Irish Agreement (AIA) to coerce Consociationalism McGarry and O’Leary argued that, echoing Lijphart, “Repartition is the drastic but logical solution to Consociational failures.”⁵⁸ The “essential conditions” for Consociationalism were not present and the AIA had not worked to develop these. By 2004, McGarry and O’Leary were rejecting partition because its threat can be “self-fulfilling and encourages pre-emptive ethnic expulsions….”⁵⁹ But in 2005, McGarry and O’Leary were advising the Kurdistan National Assembly and arguing that “complex consociations” enable secession which is widely favoured among Iraqi Kurds.⁶⁰ Such shifts suggest the importance of context over Consociationalism’s tendency towards universal prescription.

The elitist and segregationist orientation of Classic Consociationalism is apparent in McGarry and O’Leary’s pessimism and advocacy of elitist and segregationist prescriptions. From 1989 to January 1993 they shifted from advocating partition (segregationist) to supporting joint authority (elitist), which involved joint British-Irish rule with little input from the people of Northern Ireland. This proposal had some support among Irish nationalists but was opposed by the overwhelming majority of unionists. A poll in The Belfast Telegraph in January 1986 suggested 16.9% of Catholics and just 1.6% of Protestants supported Joint Authority (7.9% overall) which was 6th out of 8 options. A poll published in Fortnight Magazine in April/May 1996 suggested 24% of Catholics favoured Joint Authority and Power Sharing as the 1st preference (ranked second after a united Ireland, 32%), whereas just 6% of Protestants supported the option (ranked 6th out of 8 options). In 1989 McGarry and O’Leary had argued that joint authority was unlikely to produce power sharing and more likely to create support for a unilateral declaration of independence by unionists.⁶¹ By 1995 they, again, acknowledged that imposing joint authority would be problematic.⁶²

In 1993 McGarry and O’Leary’s pessimistic essentialism led them to predict that the conflict was getting worse, ethnic antagonisms were “being reforged rather than resolved.”⁶³ A year after the IRA’s 1994 ceasefire and just three years before the Belfast Agreement were arguing that “The problem with Consociationalism is not its normative orientation, but rather that it has not worked – and the difficulties with Lijphart’s explanations of why it has failed.”⁶⁴

Consociationalism’s pessimistic essentialism could neither predict nor convincingly explain the peace process. McGarry and O’Leary predicted the demise of the hardline parties just as they were about to rise to dominance.⁶⁵ By 1998, they were arguing that “… As nationalist support grows through demographic change, hard line unionists will become a minority in the assembly – and that will require them to learn a new politics. …Irish governments present and future will have to prepare for the possibility of a federal Ireland in which there will be a very significant British minority.”⁶⁶

The IRA’s ceasefire in September 1994 and the publication of the Framework Documents in February 1995 turned Consociationalism’s pessimistic conservative realism into optimism. Lijphart and McGarry and O’Leary now argued that it was a British conversion to the principles of Consociationalism that had resulted in the peace process and the signing of the Belfast Agreement. Leading Consociationalists had claimed the integrationist power sharing deal in 1973 but they had a task to explain why British governments had not continued with this approach and only recently “converted” to consociationalism. Lijphart argued, The Framework Document 1995 “demonstrates that the need for power-sharing has been recognized even by the British, who are not avid power-sharers themselves.” The British government “is now proposing a completely consociational system….”⁶⁷ O’Leary argued that the Conservative Party under the leadership of Margaret Thatcher 1975–90 had rejected Consociationalism. Since then, it was only by a painfully slow policy of “ethno-national policy-learning” that Conservative governments resolved the “contradictions and inconsistencies” in British policy and by 1997 had relearned the Consociationalism of 1972–74.⁶⁸

British bipartisan policy had since at least 1972 recognised the need for integrationist power-sharing in Northern Ireland. The British state had experience of managing conflict in Empire and had used elitist forms of power-sharing in constructing constitutions for such countries as Nigeria, Malaysia and Cyprus even before Consociationalism was formulated. During the period since 1972 British policy towards Northern Ireland had fluctuated but been marked by a considerable degree of strategic continuity but with tactical adjustments. It is highly constrained and governments of either party – whether the Prime Minister or party is sympathetic to nationalism or unionism – have pursued a broadly similar constitutional approach promoting power-sharing with an Irish dimension.⁶⁹ Contradicting Lijphart and his own painfully slow ethno-national policy-learning thesis, O’Leary earlier argued that “British governments have been trying to develop a power-sharing or consociational settlement in Northern Ireland since 1972.”⁷⁰ Lijphart also, rather than stressing the British state’s recent conversion to consociationalism, has argued that the British have been pushing such a settlement since the late 1970s.⁷¹

Consociational Idealist Optimism: The Belfast Agreement 1998

The outline of the Belfast Agreement 1998 was widely anticipated because previous landmark documents – the Downing Street Declaration 1993, The Framework Documents 1995, the Heads of Agreement 1998 – had sketched out the ground on which compromise between the political parties would be possible. The Belfast Agreement was, therefore, shaped by what was perceived to be “politically possible” rather than the adoption of a constitutional blueprint, whether Consociational or any other. What was remarkable about the Belfast Agreement was that it won the support of the leadership of both Sinn Féin and the Ulster Unionist Party, when the deal fell so far short of their previous, publicly stated positions. This was facilitated by the “creative ambiguity” of the Agreement which allowed the deal to be sold to unionists as strengthening the Union and to nationalists as a step down the road to a united Ireland.⁷²

Consociationalists do not provide any evidence that the key actors who negotiated the peace process or the Belfast Agreement were influenced by Consociationalism, although there may have been some minor interest.⁷³ The memoirs and diaries of all leading actors do not mention Consociationalism but rather the pragmatism of the negotiators.⁷⁴ When there is no evidence of Consociational influence the champions of the theory claim that Consociational agreements are the result of a “natural” but unconscious creative political response by politicians.⁷⁵

Consociationalists and their Cosmopolitan critics both argue that the Belfast Agreement is Consociational. Cosmopolitan critics had argued that the Belfast Agreement is Consociational because it is a “sectarian carve up” that reinforces the dominant, antagonistic identities and this underpins the power of the most hardline political parties, the DUP and Sinn Féin. They would prefer a more ambitious reformist and integrationist, or intercultural, approach. Robin Wilson does, however, recognise that the deal “was contradictory in its Consociationalist versus integrative aspects.”⁷⁶

Consociationalism is the (elitist and segregationist) theory that informs the implementation of their prescriptions. Although McGarry and O’Leary argue that the Belfast Agreement is “unarguably Consociational” they have also stated that it was more complex than Consociationalism. If the Agreement had been limited to Consociational institutions “there would have been no settlement.”⁷⁷ Their later embrace of “hardline” powersharing in 2007 represents a rejection of the intent of the Belfast Agreement.

British and Irish party and governmental actors (the SDLP, APNI and UUP) in 1998, were explicitly integrationist rejecting Classic Consociational theory.⁷⁸ The intention of the two governments and the moderate political parties in Northern Ireland was, most likely, to create an historic, integrationist accommodation based on and bolstering the moderate political parties while including but marginalising the hardline parties and preventing a return to violence. Communalist aspects of the Belfast Agreement were designed to ensure powersharing and had a strong reformist agenda. The triumph of the hardline DUP and SF in 2007 was unforeseen and a “plan B” after the collapse in electoral support for the moderates in 2003.⁷⁹ Proposals for a Civic Forum, elections and a referendum on the Agreement resembled more the democratic enthusiasm of the Cosmopolitan approach. Integrationism is explicit in the Agreement:

… An essential aspect of the reconciliation process is the promotion of a culture of tolerance at every level of society, including initiatives to facilitate and encourage integrated education and mixed housing.⁸⁰

The joint declaration (April 2003) by the British and Irish governments reiterated their integrationist approach. They recognised “the importance of building trust and improving community relations, tackling sectarianism and addressing segregation, including initiatives to facilitate and encourage integrated education and mixed housing.”⁸¹ This integrationist approach has even survived SF/DUP power sharing. Together: Building a United Community (2013) produced by the First and Deputy First Minister is explicitly integrationist and seeks to tackle segregation.

Comparing the Belfast Agreement against Lijphart’s relatively coherent “Classical Consociationalism” suggests that the Agreement is not Consociational whether using a full or highly restricted interpretation of the theory (see Table 3). Not only was Consociationalism’s pessimistic realist analysis of the conflict inaccurate, its preference for “segmental autonomy” and rule by elite cartel were explicitly rejected. Reducing Consociationalism to its 4 institutional prescriptions means that it bears some resemblance to the Belfast Agreement but, to reiterate, the intention of the two governments and most of the parties was to promote integration. Although Classical Consociationalism’s primary instrument is supposed to be a consensual or cooperative grand coalition, the Executive has been designed to divide or split power and is adversarial, with ministers using their veto power to frustrate rivals and at points trying to bring down the agreement.

Table 3. Comparing and contrasting Consociational theory (Lijphart 1977) to the Northern Ireland peace process.

Consociationalism	Northern Ireland peace process
Description or analysis of conflict
Pessimistic primordialism or ethnonationalism (the conflict is getting worse)  Negotiations consensual	Optimism of peace process and agreement, radical shift in ideologies (SF, DUP) Negotiations adversarial Based on moderate parties Integrationist Political skills
Favourable conditions	Not present in Northern Ireland
Prescription
“Segmental Autonomy” or Segregation	Explicit integrationism of Agreement. Social inclusion; community relations; equality; education; housing; human rights.
Sceptical of democracy  Negotiated by elite cartel, top down	Political process of negotiation Top down AND bottom up Participatory democracy: Civic Forum and referendum 1998
Consensual elite cartel, moderate  Power sharing	Adversarial but “moderate” (1998), then “hardline” (2007) Power splitting (dividing), some sharing and use of veto against rivals
Four institutional prescriptions
i) Grand coalition  Power sharing  Consensual  Voluntary	i) Power sharing, splitting or veto to obstruct ministerial rival Opposition of SDLP/APNI/UUP 2016–17 Adversarial, very limited cooperation D’Hondt
ii) Proportional representation (List PR)	ii) PR STV. No proportionality in state resources, limited recruitment quotas for PSNI
iii) Mutual veto	iii) Partial veto. This leads to deadlock in the Executive because of lack of consensus. External British veto used to suspend devolution and sovereignty
iv) Autonomy	iv) “Two traditions” but also integrative bias of BELFAST AGREEMENT
Extra consociational and institutional aspects of the Belfast Agreement
Security: silence	Security: the most important issues of the peace process were non-constitutional: decommissioning, prisoners, policing, demilitarisation, “On the Runs.” Policing: there was a rejection of Consociationalism’s federated/segregated police structure.
External dimension: silence	External dimension: British-Irish, All Ireland, US role, international mediation
Politics: top down	Politics: lacks a “realistic” understanding of dynamic politics of peace making, top down and bottom up, “constructive ambiguity” and pragmatic realism
Materialism and reformism: skeptical	Extensive reformist aspirations

Classic Consociationalism’s technocratic focus on constitutional architecture makes it ill-equipped to account for the reality that many of the most toxic and important issues in the peace process related to security. The internal focus of Classic Consociationalism also made it blind to the international British-Irish (even US and European) dimension of the Belfast Agreement because it focusses on powersharing within a state rather than across them. Most importantly, Consociationalism fails to properly analyse the politics of the peace process. The Belfast Agreement was a milestone in the ongoing, dynamic, and pragmatic negotiations of the peace process, with important issues kicked into the future with an awareness that flexibility and creativity – rather than an intransigent constitutionalism – lay at the heart of a successful process.

Consociationalists and the Peace Process: Chasing the Case

Leading Consociationalists, McGarry and O’Leary, took a “Constitutional Traditionalist” approach to the peace process by claiming that they had the “correct” Consociational interpretation of the Belfast Agreement. This interpretation – following Lijphart’s Classic Consociationalism – required the uncompromising implementation of the Agreement to reinforce communalism and oppose integration. Consociationalists advocated segregated communal police forces for different areas of Northern Ireland, but these were rejected.⁸² They were, therefore, relaxed about the levels of segregation in Northern Ireland and the lack of grassroots reconciliation because they believed such divisions would naturally “biodegrade.” By 2019, “de-pillarisation” had not taken place and this, in spite of consociationalism’s segregationist preferences, appeared to be a cause for concern.⁸³

Consociationalism’s description of benign political elites negotiating and imposing a consensual (at the elite level) agreement assumes away the reality of political conflict at every level. The peace process was elite led but it could not be imposed in the way advocated by Consociationalists because agreement among political elites was very difficult and there was an awareness that their parties and voters could not be forced into an agreement. McGarry and O’Leary acknowledged the difficulties of political agreement by advocated joint authority and after the Belfast Agreement 1998 electoral opinion continued to polarise resulting in the triumph of the most hardline parties.

McGarry and O’Leary’s Consociationalism was combined with an intransigent “Constitutional Traditionalism.” This correlated with a neo-nationalist, and then increasingly neo-republican position which threatened to drive pro-Agreement unionism out of the process. Ian Lustick, discussing the Middle East peace process, describes the way that insincere supporters of peace processes attempt to undermine them by calling for an intransigent implementation of any agreement which undermines the flexibility necessary to sustain an ongoing negotiating process (Lustick 1997a).

By contrast, the peace process had been successfully pursued by a pragmatic realist approach involving political skills including flexibility, choreography and creative ambiguity which more reflected a “New Constitutionalism.”⁸⁴ Political actors both attempted to gain political advantage but also recognised and often attempted to accommodate the needs of rival political actors. Seamus Mallon, a key nationalist negotiator, suggested there was a morality to a pragmatic peace and an immorality to continued violence:

… We had gone beyond the position of who was right and who was wrong. We had got to the point where it doesn’t matter who’s right. If you can’t move, then it’s politically wrong, no matter how valid it is…⁸⁵

Behind the scenes, but sometimes also front stage, nationalists and unionists and even more hardline republicans and loyalists recognised the importance of attempting to meet the political needs of rival actors by compromise and flexible politics. Arguably, the inflexible implementation of McGarry and O’Leary’s prescriptions threatened to undermine pro-Agreement unionism and, therefore, the peace process. Their research on public opinion suggested that it was pro-Agreement unionism that was most politically vulnerable in the wake of the Belfast Agreement, but they did not draw the logical conclusion from this analysis that pro-Agreement unionism needed to be supported through concessions.⁸⁶

By February 2000 the leader of pro-Agreement unionism, David Trimble, was in political crisis having entered devolution in December 1999 on the understanding that there would be IRA decommissioning of weapons. When the IRA didn’t decommission, the British government suspended devolution to protect Trimble and pro-Agreement unionism. This initiative was opposed by McGarry and O’Leary even though it was widely expected that pro-Agreement unionism would be almost fatally undermined if suspension did not take place.⁸⁷ Suspension did preserve Trimble’s position so that he could participate in the restoration of devolution later that year.

In Summer 2001, O’Leary called for fresh assembly elections even though this threatened the prospects of both moderate unionists and nationalists. Former Taoiseach (Prime Minister), John Bruton, was critical of O’Leary for calling for these elections and being too tolerant of the IRA’s failure to disarm. Bruton criticised O’Leary’s “unrealistic view” arguing that his proposals would result in the triumph of the hard-line parties and deadlock the peace process.⁸⁸

The Belfast Agreement failed in the sense that it was intended to produce powersharing based on the “moderate” parties (SDLP, APNI, UUP) in Northern Ireland.⁸⁹ By 2003 the SDLP and the UUP had been overtaken by hardline rivals, Sinn Féin and the DUP. Against expectations, these two parties reached the St Andrews Agreement 2006 and restored powersharing in 2007. Consociationalists now shifted from justifying the triumph of the moderate parties to supporting the victory of the hardline parties for bringing greater stability. Sinn Féin and the Democratic Unionist Party were described as “tribune” parties, “The ethnic tribune party can be simultaneously pragmatic over resources and intransigent about identity” thereby preventing “ethnic outbidding.”⁹⁰ By 2015 McGarry and O’Leary regarded the conflict as “mostly having been resolved, they are ‘more interested in places like Syria, and Yemen, and Iraq and Cyprus in particular’: where they can apply their consociational framework,” although apparently without success.⁹¹

Continuing segregation and the lack of underlying reconciliation were not significant because McGarry and O’Leary assumed that Consociations could be left to naturally “biodegrade.” This assumption contrasted with their earlier position, which was more in keeping with Classic Consociational theory, that “Consociationalists also realize that communal or ethnic divisions are resilient rather than rapidly biodegradeable, and that they must be recognized rather than wished away.”⁹² In “Consociational” Belgium, Lebanon, Cyprus and Bosnia communal divisions did not seem to have biodegraded in the way that Consociationalists predicted for Northern Ireland.

From Consociationalism to Irish Unity

Ironically, as old and “New Generation” Consociationalists claimed Northern Ireland as a great success for their theory, McGarry and O’Leary dropped Consociationalism and the peace process and advocated for an integrated united Ireland. The “New Generation” claimed a “new wave” of Consociational settlements and emphasised the importance of Northern Ireland claiming that it was “arguably the most stable ‘new-wave’ consociation.” There had been a “transformation” since 2007, it was a “watershed moment in a peace process historically characterised by intransigence.” A “stark change” in party political behaviour by Sinn Féin and the DUP took “advantage of the institutional incentives to move toward more moderate positions.” Northern Ireland, she claimed, “has seen the most success under consociational rules.” These changes represented an “important rebuttal” to Consociationalism’s critics and demonstrated that consociationalism can be designed “to include incentives for moderation.”⁹³

The UK’s decision to leave the European Union in 2016 has put Irish unity back on the political agenda. Brendan O’Leary argued in Making Sense of a United Ireland: Should It Happen? How Might It Happen? (2022) for a referendum and an integrated united Ireland.⁹⁴ He argues that his preference would be for the continuation of the Belfast Agreement with a simple change of sovereignty from the UK to the Republic of Ireland. But he argues that this is unacceptable to public opinion in the Republic of Ireland and so an integrated united Ireland should be the goal.⁹⁵ In the past Lijphart and McGarry and O’Leary – with their elitist prescriptions – were unconcerned about the lack of public (and particularly unionist) support for their solutions whether repartition or joint authority. They were also appeared to be unconcerned about the impact of declining public support on unionism in sustaining the peace process. When “nationalists” were a minority, he argued for minority guarantees in the Belfast Agreement, now they might be a majority he does not support these same guarantees for unionists. O’Leary’s position on Irish unity – like his position on Iraq – suggests a lack of commitment to the principles of “Consociationalism” when these clash with his political commitments to the Kurds or Irish nationalism and then republicanism.

Leading nationalists have defended the powersharing principles of the Belfast Agreement. Seamus Mallon (SDLP) and Taoiseach Bertie Ahern (1997–2008), who were intimately involved in negotiating the peace process as well as Taoiseach Leo Varadkar (2017–20, 2022–), have proposed parallel consent in a referendum in Northern Ireland. Mallon argued that a 50%+1 vote would “not give us the kind of agreed and peaceful Ireland we seek.”⁹⁶ There is both a principled argument, that parallel consent is in the spirit of the Belfast Agreement, and a pragmatic one, that unless there is strong unionist consent a unification process, or even the prospect of it, is reckless and could well reignite violence. An argument for parallel consent would have to come from nationalists because the consent of a Northern Irish majority for constitutional change has long been a key principle of British policy and is in the Belfast Agreement. To be questioning this now when a nationalist majority could emerge is, understandably, likely to be seen as provocative by nationalists, republicans and dissident republicans. O’Leary has opposed this powersharing principle and described these interventions as “unhelpful.”⁹⁷

Conclusion

Consociationalism has been constantly defined and redefined so that it now has a Wonderland definition in which Consociationalism means whatever Consociationalists currently choose it to mean. Consociationalism has been interpreted both to reinforce sectarian authoritarianism and to oppose it; to be synonymous with power-sharing and with no power-sharing at all; to recognise alternative approaches to conflict management and the incorporation or denial of their existence. Consociationalism is not successful (5 cases) and spectacularly successful (48 cases). This confusion means that although Classic Consociationalists supported sectarian authoritarianism to manage conflict, the New Generation of Consociationalists argue that consociationalism somehow opposes sectarian authoritarianism.

Consociational theory has changed over time first, to deflect criticisms of its essentialism and prescription for segregation. Second, to more credibly claim apparently successful cases of conflict management and distance itself from failure. Consequently, there are myriad versions of Consociationalism which leads to contradiction and incoherence. While Classic Consociationalism does have some degree of coherence, subsequent revisions and complexities have made it difficult to discern what is the theory’s analysis of conflict and prescriptions for conflict management.

Why has a theory that is so ambiguous, contradictory, and incoherent been so successful? Paradoxically, although the Wonderland definition of Consociationalism has led to confusion about the theory (not least among Consociationalists themselves) it has been a highly successful rhetorical strategy in further building the Consociational paradigm within academia.⁹⁸ Consociationalism’s power within the academy has combined with claims to objectivity, universal application and success to influence the management of conflict in the “real” world and deflect attention from the political agendas of its advocates.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Matthijs Bogaards, Ludger Helms, and Arend Lijphart, “The Importance of Consociationalism for Twenty-First Century Politics and Political Science,” Swiss Political Science Review 25, no. 4 (2019): 4–5, 6; Mahmoud Farag, Hae Ran Jung, Isabella C. Montini, Juliette Bourdeau de Fontenay, and Satveer Ladhar, “What Do We Know about Power Sharing after 50 Years?” Government and Opposition, (2022), online.

2 Ian Lustick, “Lijphart, Lakatos, and Consociationalism,” World Politics, 50, no. 1 (1997).

3 Allison McCulloch, Power-Sharing and Political Stability in Deeply Divided Societies (London: Routledge, 2014), 64; Bogaards et al., “Importance of Consociationalism,” 5; Farag et al., “What Do We Know,” 6.

4 Brendan O’Leary, Making Sense of a United Ireland: Should It Happen? How Might It Happen? (Dublin: Sandycove, 2022).

5 Arend Lijphart, Democracy in Plural Societies (New Haven: Yale, 1977).

6 Arend Lijphart, “Consociational Democracy,” World Politics 21, no. 2 (1969); Lijphart, Democracy; Paul Dixon, “Paths to Peace in Northern Ireland (I): Civil Society and Consociational Approaches,” Democratization 4, no.2 (1997); Paul Dixon, “Paths to Peace in Northern Ireland (II): Peace Process 1973-74, 1994-96,” Democratization 4, no.3 (1997).

7 Lustick, “Lijphart,” 110.

8 Arend Lijphart, “Cultural Diversity and Theories of Political Integration,” Canadian Journal of Political Science 4, no. 1 (1971): 11; Lijphart, “Consociational Democracy,” 219; Lijphart, Democracy, 140.

9 Lijphart, Democracy, 42, 48, 140–41.

10 Ibid., 25.

11 Bogaards et al, “Importance of Consociationalism,” 404, 405.

12 John McGarry and Brendan O’Leary, Explaining Northern Ireland (Oxford: Blackwell, 1995), 210, 307.

13 John McGarry, “Explaining Ethnonationalism: The Flaws in Western Thinking,” Nationalism and Ethnic Politics 1, no. 4 (1995): 133–34; Arend Lijphart, “Review Article: The Northern Ireland Problem; Cases, Theories, and Solutions,” British Journal of Political Science, 5, no. 1 (1975): 186.

14 Lijphart, Democracy, 54.

15 Matthijs Bogaards, “The Favourable Factors for Consociational Democracy: A Review,” European Journal for Political Research, 33, no. 4 (1998): 478.

16 Rogers Brubaker, Ethnicity without Groups (Cambridge: Harvard University Press, 2004).

17 Bogaards, “Favourable Factors,” 490.

18 Lijphart, Democracy, 53.

19 Matthijs Bogaards, “Interview Making a Difference: An Interview with Arend Lijphart,” Zeitschrift für Vergleichende Politikwissenschaft, 9, no. 1-2 (2015).

20 Lijphart, Democracy, 1, 21, 5; Arend Lijphart, “The Wave of Power-Sharing Democracy,” in The Architecture of Democracy, ed. A. Reynolds (Oxford: Oxford University Press, 2002), 41.

21 Lijphart, Democracy, 25.

22 Lijphart, Democracy, 228; John McGarry and Brendan O’Leary, “Power Shared after the Deaths of Thousands,” in Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict, ed. Rupert Taylor (London: Routledge, 2009), 68.

23 Colin Hay, Political Analysis (Basingstoke: Palgrave, 2002).

24 Marysia Zalewski, “Gender Ghosts in McGarry and O’Leary and Representations of the Conflict in Northern Ireland,” Political Studies, 53 (2005): 213.

25 Arend Lijphart in Bogaards, “Interview,” 92; 2015: 92; Brenda O’Leary, A Treatise on Northern Ireland, Volume 3 (Oxford: Oxford University Press, 2019), 5.

26 Lijphart, Democracy, 48.

27 McGarry and O’Leary, Explaining, 339–40.

28 Lijphart, Democracy; McGarry and O’Leary, Explaining.

29 M. Hewstone, et al. “Intergroup Contact in a Divided Society: Challenging Segregation in Northern Ireland,” in The Social Psychology of Inclusion and Exclusion, ed. D. Abrams et al. (Philadelphia: Psychology Press, 2005).

30 Lustick, “Lijphart,” 111–12; Matthijs Bogaards, “The Uneasy Relationship between Empirical and Normative Types in Consociational Theory,” Journal of Theoretical Politics 12, no. 4 (2000): 402.

31 Bogaards, “Uneasy Relationship,” 408.

32 Bogaards, “Interview,” 92.

33 Ibid., 92–93.

34 Lustick, “Lijphart”; Dixon, “Paths to Peace 1”.

35 Rupert Taylor, “A Consociational path to peace in Northern Ireland and South Africa?” in New Perspectives on the Northern Ireland Conflict, ed. A. Guelke (Aldershot: Avebury, 1994), 166.

36 Brendan O’Leary, “Foreword” to M. Kerr, Imposing Power Sharing (Dublin: Irish Academic Press, 2005).

37 Michael Kerr, Imposing Power Sharing (Dublin: Irish Academic Press, 2005), 40.

38 O’Leary, “Foreword,” xviii.

39 Jonathan Steele, Defeat: Why They Lost Iraq (London: I B Taurus, 2007).

40 Susan Halpern, “The Disorderly Universe of Consociational Democracy,” West European Politics 9, no. 2 (1986).

41 Brendan O’Leary, “Debating Consociational Politics: Normative and Explanatory Arguments,” in From Power-Sharing to Democracy, ed. S. Noel (London: McGill-Queens University Press, 2005), 34.

42 Lijphart, Democracy, 30, 31, 33, 36; Paul Dixon, “Consociationalism and the Northern Ireland Peace Process: The Glass Half Full or Half Empty?” Nationalism and Ethnic Politics 3, no.3 (1997).

43 Lijphart, Democracy, 31; O’Leary, Treatise, 2 states that “powersharing” is not a synonym for Consociation.

44 Rogers Brubaker, “Myths and Misconceptions in the Study of Nationalism,” in The State of the Nation, ed. J. A. Hall (Cambridge: Cambridge University Press, 1998).

45 McCulloch, Power-Sharing, 5, 18; John Nagle, “Between Conflict and Peace: An Analysis of the Complex Consequences of the Good Friday Agreement,” Parliamentary Affairs, 71 (2018).

46 Bogaards et al., “Importance of Consociationalism,” 7.

47 Toby Dodge, Iraq – From War to a New Authoritarianism (London: Routledge, 2012).

48 O’Leary, Debating Consociational Politics, 14, 12–15.

49 John McGarry and Brendan O’Leary, “Consociational Theory, Northern Ireland’s Conflict, and its Agreement. Part 1: What Consociationalists Can Learn from Northern Ireland,” Government and Opposition, 42, no. 1 (2006): 63, 62. John McGarry, Brendan O’Leary, and R. Simeon, “Integration or Accommodation? The Enduring Debate in Conflict Regulation,” in Constitutional Design for Divided Societies: Integration or Accommodation, ed. S. Choudhry (Oxford: Oxford University Press, 2008), 58.

50 Brendan O’Leary, How to Get out of Iraq with Integrity (Philadelphia: University of Pennsylvania Press, 2009), 81, 19.

51 O’Leary, Making Sense, 4.

52 O’Leary, Iraq integrity.

53 Lijphart, Democracy, 136; McGarry and O’Leary, Explaining, 321.

54 Lijphart, “Northern Ireland,” 99.

55 Lijphart, Democracy, 137, 141, 134–41; Brendan O’Leary, “The Limits to Coercive Consociationalism in Northern Ireland,” Political Studies, 37 (1989).

56 Arend Lijphart, “The Power-Sharing Approach,” in Conflict and Peacemaking in Multiethnic Societies, ed. J. V. Montville (New York: Lexington Books, 1991), 496.

57 McCulloch, Power-Sharing, 64; Nagle, “Conflict and Peace,” 399. Allison McCulloch, “Consociational Settlements in Deeply Divided Societies: The Liberal-Corporate Distinction,” Democratization 21, no. 3 (2014): 507. For a critique of “New Generation” Consociationalists see Paul Dixon, “The Politics of Conflict: A Constructivist Critique of Consociational and Civil Society Theories,” Nations and Nationalism 17, no. 4 (2011).

58 O’Leary, “Limits,” 587–88.

59 John McGarry and Brendan O’Leary, Consociational Engagements (Oxford: Oxford University Press, 2004), 43–44.

60 Brendan O’Leary, “Debating Consociational Politics: Normative and Explanatory Arguments” in From Power-Sharing to Democracy, ed. S. Noel (London: McGill-Queens University Press, 2005), 25–26.

61 O’Leary, “Limits,” 586.

62 McGarry and O’Leary, Explaining, 371–72.

63 Brendan O’Leary and John McGarry, The Politics of Antagonism (London: Athlone, 1993), 325.

64 McGarry and O’Leary, Explaining, 338.

65 Ibid., 405.

66 The Guardian, April 7, 1998.

67 Arend Lijphart, “The Framework Documents on Northern Ireland and the Theory of Power Sharing,” Government and Opposition, 31, no. 3 (1996): 267–68, 273.

68 Brendan O’Leary, “The Conservative Stewardship of Northern Ireland, 1979–97: Sound-Bottomed Contradictions or Slow Learning,” Political Studies, 45 (1997).

69 Michael J. Cunningham, British Government Policy in Northern Ireland, 1969–2000 (Manchester: Manchester University Press, 2001); Paul Dixon Dixon, “British Policy toward Northern Ireland 1969–2000: Continuity, Tactical Adjustment and Consistent ‘Inconsistencies’,” British Journal of Politics and International Relations, 3, no.3 (2001).

70 McGarry and O’Leary, Explaining, 321.

71 Arend Lijphart, “The Wave of Power-Sharing Democracy,” in The Architecture of Democracy, ed. A. Reynolds (Oxford: Oxford University Press, 2002), 43.

72 Mo Mowlam, Momentum (London: Hodder and Stoughton, 2002), 231; Jonathan Powell, Great Hatred, Little Room: Making Peace in Northern Ireland (London: Bodley Head, 2008); Tony Blair, A Journey (London: Hutchinson, 2010), 190.

73 Graham Spencer, The British and Peace in Northern Ireland (Cambridge: Cambridge University Press, 2015), 160, 298.

74 Robin Wilson, The Northern Ireland Experience of Conflict and Agreement: A Model for Export? (Manchester: Manchester University Press, 2010), 85, 149–50.

75 O’Leary, Treatise, 178; Arend Lijphart, Thinking about Democracy (London: Routledge, 2007), 269, 278.

76 Wilson, Northern Ireland Experience, 157.

77 McGarry and O’Leary, Consociational, 348.

78 Paul Dixon, “Why the Good Friday Agreement in Northern Ireland Is Not Consociational,” Political Quarterly 76, no. 3 (2005).

79 Mowlam, Momentum, 362–63.

80 The Belfast [Good Friday] Agreement (London: HMSO, 1998), 18, paragraph 13.

81 Joint Declaration by the British and Irish Governments (2003), 8, para. 27.

82 McGarry and O’Leary, Consociational, 403.

83 O’Leary, Treatise, 328.

84 Vivien Hart, “Constitution-Making and the Transformation of Conflict,” Peace and Change, 26 (2001); Paul Dixon, Northern Ireland: The Politics of War and Peace (Basingstoke: Palgrave, 2008), Chapters 9 and 10.

85 Seamus Mallon, Brook Lapping Interviews, Liddell Hart Military Archive, London.

86 Geoffrey Evans and Brendan O’Leary, “Northern Irish Voters and the British-Irish Agreement: Foundations of a Stable Consociational Settlement?” Political Quarterly, 71, no. 1 (2000), 98–99.

87 The Guardian, February 2, 2000.

88 The Guardian, July 13, 2001; Irish Times, October 8, 2001.

89 Paul Dixon, Performing the Northern Ireland Peace Process: In Defense of Politics (Cham: Palgrave, 2018), for an analysis of the peace process and for a more empathetic view of David Trimble’s dilemma see Chapter 8.

90 Paul Mitchell, Geoffrey Evans, and Brendan O’Leary “Extremist Outbidding in Ethnic Party Systems Is Not Inevitable: Tribune Parties in Northern Ireland,” Political Studies, 57, no. 2 (2009), 403, 416.

91 Corey Jentry, “The Trouble with Studying the Troubles: How and Why an Epistemic Community Emerges,” PhD, LSE, 166; on Consociational complacency see also John McGarry, “Conclusion: What Explains the Performance of Power-Sharing Settlements?” in Power-Sharing: Empirical and Normative Challenges, ed. Allison McCulloch and John McGarry (London: Routledge, 2017), 269.

92 McGarry and O’Leary, Explaining, 338.

93 McCulloch, Power-Sharing, 20, 6, 75, 76, 77, 87; Nagle, “Between Conflict.”

94 O’Leary, Treatise, xvi; McGarry, “Conclusion,” 289 did not argue for Irish unity but “something similar” to the Belfast Agreement.

95 O’Leary, Making Sense.

96 O’Leary, Making Sense? 91; BBC News, May 24, 2019.

97 O’Leary, Making Sense? 91, 342 fn.11.

98 Lustick, “Lijphart.”