^t This is an updated version of a paper originally delivered at the annual conference of the Association for the Study of Nationalities at Columbia University, New York, 5 April 2001. The session was sponsored by the European Centre for Minority Issues. My thanks go to Margaret Moore and Patti Lenard for their helpful comments and to the Social Sciences and Humanities Research Council of Canada (SSHRCC) and the Carnegie Corporation of New York for funding my research. ' The most comprehensive account of consociational theory is contained in Arend Lijphart, Democracy in Plural Societies: A Comparative Exploration (New Haven, 1977); but see also id., The Policies of Accommodation: Pluralism and Democracy in the Netherlands ( Berkeley, 1968 and id., "Consociational Democracy", 21(2) World Politics (1969). 207-25. For an extensive summary of the main criticisms of consociational theory, see id., Power-Sharing in South Africa (Berkeley, 1985). ² Brendan O'Leary, "The Nature of the Agreement", 22(4) Fordham Journal of International Law ( 1999 ), 1628-67.

³ Lijphart lists nine factors that make the establishment of a consociational democracy more likely: no majority segment; segments of equal size; small number of segments; small population size; external threats; overarching loyalties; socio-economic equality; geographical concentration of segments; tradi- tion of accommodation. Lijphart, Power-Sharing ..., 120. ⁴ Lijphart, Democracy in Plural Societies.

⁷ Donald Horowitz, "The Northern Ireland Agreement: Clear. Consociational and Risky", in John McGarry (ed.), Northern Ireland and the Divided World (Oxford, 2001), 89-108, at 89. ⁸ Horowitz argues that Northern Ireland's Agreement was the result of exceptional conditions. In short. Northern Ireland's parties may have managed to settle on a consociational package, but we should not expect other warring factions to do the same.

⁹ See footnote 1. endogenous factors were also important, particularly demographic change. The Protestant and union- ist share of Northern Ireland's population is in decline, and sits currently around 55 per cent. There is a possibility, and an even stronger perception, that there will be a Catholic and nationalist majority at some point in the foreseeable future. This has undercut unionists' enthusiasm for majoritarian democracy and increased their support for power-sharing. For a more extensive treatment of the exogenous and endogenous factors behind the Northern Ireland Agreement, see John McGarry, "Political Settlements in Northern Ireland and South Africa", 46(5) Political Studies (1998), 853-70. " A 1994 republican document on the peace strategy, TUAS (reputedly an acronym for either Totally Unarmed Strategy or Tactical Unarmed Struggle), was explicit about the importance of the American role, noting that "there is potentially a very powerful Irish-American lobby not in hock to any particular party in Britain or Ireland" and that "Clinton is perhaps the first US President in decades

to be influenced by such a lobby". Cited in Roger MacGinty, "American Influences on the Northern Ireland Peace Process", 43 Journal of Conflict Studies ( 1997), 31-50, at 34. ¹² For a more extensive analysis of the exogenous factors that led to Northern Ireland's Good Friday Agreement, see Brendan O'Leary and John McGarry, Explaining Northern Ireland (Oxford, 1995); and John McGarry, "Globalization, European Integration and the Northern Ireland ConBict", in Michael Keating and John McGarry (eds.), Minority Nationalism and the Changing International Order (Oxford, 2001 ), 295-324. " Personal communication from Professor Andrew Wilson, who interviewed Trimble on this matter. '^a Margaret Thatcher had several reasons for signing the Anglo-Irish Agreement, but an important one was pressure from the United States. From the early 1980s, leading US politicians had applied pressure on Britain to cooperate more closely with the Republic of Ireland, and US President Ronald Reagan, whom Thatcher respected, put his personal clout behind this message. American pressure, then, pre- pared the groundwork for a power-sharing settlement even before Clinton assumed power. ¹⁵ Indeed, the option of integration, which had been abandoned since 1979, was reconsidered during the Major years.

'⁶ I do not think that exogenous forces can promote stable consociational settlements when endogenous forces are strongly unfavourable, but I also do not subscribe to the view that outsiders can make no appreciable difference. For an interesting essay on how external governments can use food aid to promote power-sharing, see Ed Luttwak, "Aid is a Weapon. Let's Use it", Glo6e and Mail (Toronto, 22 November 2001), A25. " This need for interstate institutions is a feature of the Northern Ireland conflict that Lijphart overlooks: in an otherwise masterly dissection of the Northern Ireland conflict in a 1975 article, in which he lists the obstacles to power-sharing, he omits the fact that nationalists have always opposed any institu- tional arrangement restricted to Northern Ireland. See Lijphart, "Review Article: The Northern Ireland Problem: Cases, Theories and Solutions", 5 British Journal of Political Science (1975), 83-106. '⁸ The Agreement committed both parts of Ireland to a further six functional areas of cooperation - including some aspects of transport, agriculture, education, health, the environment and tourism. It

also provided for a British-Irish Intergovernmental Conference to promote bilateral cooperation between the Irish and British Governments on all matters of mutual interest within their jurisdiction, and for a British-Irish Council, a forum to bring together not just the representatives of the British and Irish Governments but also those within the UK's various devolved institutions (Scotland, Wales, the Isle of Man, Jersey, and Guernsey). '⁹ Michael Keating and John McGarry, "Introduction: Minority Nationalism in the Changing International Order", in Michael Keating and John McGarry (eds.), Minority Nationalism and the Changing International Order (Oxford, 2001), 1-15, at 10. the traditional argument against permitting such interstate institutional linkages is that they would legitimize and strengthen irredentism, both among the minority and its ethnic kin in the neighbouring state. This is similar to the argument that is used to deny autonomy to minorities, i.e., it will facilitate secession. However, interstate institutions, like autonomous institutions, can help to reconcile minorit- ies to the state in which they live. To allay fears about strengthening irredentism, interstate institutional linkages can be accompanied by explicit (treaty-embedded) acknowledgements that the territorial integrity of the states concerned is respected.

²¹ For details on how the Agreement dealt with these various issues, see O'Leary, "The Nature of Agreement", 22(4) Fordham Journal of International Law (1999) 1628-67. ²² Post-Agreement survey data from Evans and O'Leary show that 86 per cent of Catholics and 62 per cent of Protestants are in favour of power-sharing. See Geoffrey Evans and Brendan O'Leary, "Northern Irish Voters and the British-Irish Agreement: Foundations of a Stable Consociational Settlement?" 71 Political Quarterly (2000), 78-101. ²³ Kirsten Schulze, "Taking the Gun out of Politics: Conflict Transformation in Northern Ireland and Lebanon", in John McGarry (ed.), Northern Ireland and the Divided World: Post-Agreement Northern Ireland in Comparative Perspective (Oxford, 2001), 253-75.

24 see the speeches by Alliance Party representatives during the Northern Ireland Assembly debates of 2 and 5 November 2001, in the Northern Ireland Assembly Official Report (Hansard), at http://www.ni-assembly.gov.uk/hansard.htm. ^zs Kevin Rooney, "Institutionalising Division", Fortnight (Belfast. June 1998), 21-2. Rooney worries that by establishing institutions that "celebrate difference", the Agreement has "put an end to the prospects for overcoming these divisions". ^zs Robert McCartney. "Devolution is a Sham", Obseruer (20 February 2000). see at http:// www.guardian.co.uk/Devolutionj Story/0,2763,190960,OO.html. ²⁷ Spokespersons for Democratic Left and the Alliance Party have criticized the Agreement as, respect- ively, a pact "between the two dominant sectarian and tribal blocs" and as emphasizing "two commu- cities" rather than all the people who share "common values and principles". See, respectively, Paddy- Joe McLean, "Five Reasons why Socialists Should Say Yes to the Deal", Irish News (9 May 1998); and Philip McGarry, "Why the Agreement May Fail", Belfast Telegraph (16 September 1999). ²⁸ R_upert Taylor, "Consociation or Social Transformation", in John McGarry (ed.). Northern Ireland and the Divided World (Oxford, 2001 ), 36-52. at 47.

^Z9 The unionist intellectual and politician, Robert McCartney, ironically mimics anti-Agreement repub- licans in claiming that the power-sharing institutions are "impermanent", "dysfunctional" and "unworkable", and that it is only a matter of time before this "macabre parody of real democracy" is brought to a halt by its "inherent defects and weaknesses", see McCartney, "Devolution ...". denims Kennedy, "Evidence is Growing that Agreement Did not Work", Irish Times (16 February 2000); Patrick Roche, "A Stormont without Policy", Belfast Telegraph (30 March 2000). ³' Donald Horowitz, "The Northern Ireland ...", 92-3. Opponents of the grand coalition model claim to have been vindicated by the problems that have beset the power-sharing institutions since 1998. For an example, see R. Wilford and R. Wilson, "A 'Bare Knuckle Ride': Northern Ireland", in R. Hazell (ed.), The State and the Nations: The First Year of Devolution in the United Kingdom (Thorverton, UK, 2001), 79-116. Wilford and Wilson are good at describing the problems, but not at offering viable alternatives. ³² Some liberal democrats are unable to appreciate the difference between ethnic cleansing and apartheid, on the one hand, and power-sharing between groups on the other. Thus when Brendan O'Leary and I expressed our support for consociationalism in Explaining Northern Ireland, one critic wrote that these views could be seen as "condoning ... ethnic cleansing". See Paul Dixon, "The Politics of Antagonism: Explaining McGarry and O'Leary", 11 Irish Political Studies (1996), 139. See also our reply, John McGarry and Brendan O'Leary, "Proving our Points on Northern Ireland", 11 Irish Political Studies (1996), 142-54. For a perspective that appears to equate the defence of the rights of multicultural minorities with the policies of apartheid in South Africa, see Laurence Piper, "Whose Culture? Whose Rights? A Critique of Will Kymlicka's Multicultural Citizenship", paper presented at the International Political Science Association (IPSA) XVIII World Congress, Quebec City, 1-5 August 2000.

³³ The fact that 50 per cent of either bloc can prevent the election of the First Minister and Deputy Minister, as happened on 2 November 2001, is a weakness in the Agreement's decision-making rules. I will address this later. ³⁴ For an analysis of the Agreements, see O'Leary, "The Nature ...", and the series of articles by McCrudden, McGarry and O'Leary in the Sunday Business Post (Dublin, 1998). Also see Brendan O'Leary, The British-Irish Agreement: The Second Peace by Ordeal (provisional title) (Oxford, forthcoming).

³⁵ Arend Lijphart, "Self-Determination versus Pre-Determination of Ethnic Minorities in Power-Sharing Systems", in Will Kymlicka (ed.), The Rights oJMinority Cultures (Oxford, 1995), 275-87. '⁶ There is an argument for making the executive even more inclusive by extending its size. A larger executive, constituted by the d'Hondt mechanism, would give a seat to the Alliance Party and might in future give seats to other small parties. Alternatively, the executive could be constituted by the Sainte-Lague mechanism, which is more advantageous for small parties than d'Hondt. For an explana- tion of the difference between d'Hondt and Sainte-Lague, see McGarry and O'Leary, Explaining .., 373-5.

37 it does not follow from my argument that the inclusion of radicals in government in 1974 would have improved the stability of the power-sharing government. In 1974. these radicals were virulently opposed to power-sharing and would have refused to take any positions offered to them. However, when radicals are prepared to participate in power-sharing institutions, and are not bent on their destruction, it makes sense to include them. I give reasons for this later. Anti-consociationalists tend to see Sinn Fein's rise as evidenee of increasing extremism. and sometimes attribute it to the 'unworkable' nature of Northern Ireland's consociational institutions. See Wilford and Wilson, "A 'Bare Knuckle Ride ...�. However, it makes more sense, given Sinn Fein's clear movement from physical force republicanism to constitutional politics, to see its electoral growth as

a result of its increasing moderation. See Paul Mitchell, Brendan O'Leary and Geoffrey Evans, "Northern Ireland: Flanking Extremists Bite the Moderates and Emerge in their Clothes", Parliamentary Affairs (2002, forthcoming). Other factors are also responsible, such as the party's articulate and capable leadership, and the growing Catholic share of the population combined with the tendency of young Catholics to vote Sinn Fein. '⁹ Donald Horowitz, "Making Moderation Pay: The Comparative Politics of Ethnic Conflict Management", in J.P. Montville (ed.), Confiict and Peacemaking in Multiethnic Societies (Lexington, 1989), 451-75. '° Trimble and Durkan were elected on 5 November 2001, after members of the Alliance Party, who had originally designated as 'Others', temporarily redesignated as 'Unionists'.

⁴¹ At the time this article is being submitted (November 2001), a review of the Agreement's rules is about to begin. For an analysis of the rules, and our preferred changes, see John McGarry and Brendan O'Leary, "Reviewing the Rules" (unpublished paper, available from the authors), 2001. There are alternatives to using d'Hondt as the default rule for electing the First and Deputy First Minister team. One is to use the Agreement's weighted majority rule: 60 per cent of the Assembly plus at least 40 per cent of both the nationalist and unionist blocs. Another is to rely on a simple weighted majority of 60 per cent - a threshold that is high enough to ensure that support from both communities is required, but low enough to prevent vetoes by rejectionists. O'Leary and I prefer d'Hondt as a default rule because we believe these other defaults would result in moderate parties colluding to deprive Sinn Fein or the DUP of one of the top positions (should either come to command a majority in its bloc). It is our view, as I have expressed in my arguments against Horowitz in this paper, that consociational coalitions will be stronger if they include radical parties and give them incentives to compromise. It would also be useful if the Northern Ireland Act (1998) was revised to remove the requirement that the resignation of one of the co-premiers automatically triggers the resignation of the other. This provision is not in the Agreement itself. The ability of one co-premier to bring the other down was used by both David Trimble and Seamus Mallon as a destabilizing bargaining chip during the 1999-2001 period. A resigning co-premier should be replaced in the first instance by someone from his or her party. If this is not an option, the vacant position should be filled by the d'Hondt procedure with the proviso that it could not be occupied by someone from the other co-premier's bloc.