¹ See the discussion in Lijphart, Arend, Democracy in Plural Societies: A Comparative Exploration (New Haven and London: Yale University Press, 1977), pp. 75–83.Google Scholar In the Northern Irish case this view has been proposed by Aunger, Edmund A., In Search of Political Stability: A Comparative Study of New Brunswick and Northern Ireland (Montreal: McGill-Queen's University Press, 1981).Google Scholar

² For example, Dahl, Robert A., Dilemmas of Pluralist Democracy: Autonomy vs. Control (New Haven: Yale University Press, 1982), pp. 56–67.Google Scholar

³ Well-known proponents of this argument include Nordlinger, Erik A., Conflict Regulation in Divided Societies, Occasional Papers in International Affairs, no. 29 (Cambridge, Mass.: Harvard University Centre for International Affairs, 1972)Google Scholar; Sartori, Giovanni, ‘From the Sociology of Politics to Political Sociology’, in Lipset, Seymour M., ed., Politics and the Social Sciences (New York: Oxford University Press, 1969)Google Scholar, chap. 4; Przeworski, Adam, Capitalism and Social Democracy (Cambridge: Cambridge University Press, 1985)CrossRefGoogle Scholar; and Mair, Peter, ‘Explaining the Absence of Class Politics in Ireland’, in Goldthorpe, John and Whelan, Chris, eds, The Development of Industrial Society in Ireland (Oxford: Oxford University Press, 1992), pp. 383–401.Google Scholar

⁴ Horowitz, Donald L., Ethnic Groups in Conflict (Berkeley: University of California Press, 1985), p. 601.Google Scholar For example, Horowitz, referring to countries such as Nigeria, shows how ethnically-grounded political polarization can be reduced by providing electoral incentives for previously disparate groups to shift from their extreme positions. As he suggests, ‘the (institutional) rules can restructure the system so that the game itself changes’.

⁵ Roberts, Hugh, ‘Sound Stupidity: The British Party System and the Northern Ireland Question’, in McGarry, John and O'Leary, Brendan, eds, The Future of Northern Ireland (Oxford: Clarendon Press, 1990), chap. 4.Google Scholar

⁶ The current political system in Northern Ireland comprises five main parties, four of them ‘confessional’ parties, making the constitutional status of Northern Ireland their central political concern and appealing almost exclusively to either Catholics or Protestants, and one centre party. The unionist position is represented mainly by the Ulster Unionist Party (UUP) and the more hard-line Democratic Unionist Party (DUP). Nationalist concerns are represented mainly by the Social Democratic and Labour Party (SDLP) and the republican party, Sinn Féin (SF). The middle ground is held by the bi-confessional Alliance Party of Northern Ireland (APNI).

⁷ Roberts, , ‘Sound Stupidity’, p. 102.Google Scholar

⁸ For example, Beloff, Max and Peele, Gillian, The Government of the United Kingdom: Political Authority in a Changing Society (London: Weidenfeld and Nicolson, 1985), p. 211.Google Scholar

⁹ As evidence for the ‘cross-community potential for Labour politics’, integrationists point to the results of a recent CLRNI-sponsored survey in which, of the 1,100 Northern Irish people polled, more than half the Catholics and 28 per cent of Protestants were in favour of the Labour party fielding candidates in the region (see Mark Langhammer, CLRNI Spokesperson, quoted in ‘Majority wants Labour to fight in Northern Ireland: Poll’, Irish Times, 7 04 1993).Google Scholar

¹⁰ For example, O'Dowd, Liam, ‘Social Class’, in Stringer, and Robinson, , eds, Social Attitudes in Northern Ireland (Belfast: Blackstaff Press, 1991), chap. 5.Google Scholar

¹¹ For example, the 1974 Sunningdale experiment in devolution and power-sharing met with resounding failure. The outcome of the current ‘peace process’, the result, in part at least, of the joint governmental ‘Downing Street’ initiative, remains to be seen. Although the cease-fire announced in September 1994 appears to be holding, there are clear signs that the negotiations will stall many times in the face of the mutual suspicion and hostility between the involved parties.

¹² Budge, Ian and O'Leary, Cornelius, Belfast: Approach to Crisis-A Study of Belfast Politics, 1613–1970 (London: Macmillan, 1973), p. 365Google Scholar; Aunger, , In Search of Political Stability.Google Scholar

¹³ O'Leary, Brendan and McGarry, John, The Politics of Antagonism: Understanding Northern Ireland (London: Athlone Press, 1993), p. 298.Google Scholar

¹⁴ An exception was the performance of Conservative party candidate for North Down, Dr Laurence Kennedy, in the 1992 Westminster general election. However, this middle-class and largely Protestant district is not representative of Northern Ireland in general. Dr Kennedy left the Conservative party in July 1993 because its ‘understanding’ with the UUP, which, he believed, encouraged ‘the perpetuation of religiously divisive politics’. See Flackes, W. D. and Elliott, Sydney, Northern Ireland: A Political Directory, 1968–1993, 2nd edn (Belfast: Blackstaff Press, 1994), p. 200.Google Scholar

¹⁵ Curtice, John and Gallagher, Tony, ‘The Northern Ireland Dimension’, in Jowell, Roger, Witherspoon, Sharon and Brook, Lindsay, eds, British Social Attitudes: The Seventh Report (Aldershot: Gower, 1990), chap. 10.Google Scholar

¹⁶ O'Dowd, , ‘Social Class’, p. 47.Google Scholar

¹⁷ McNamara, Kevin, Stott, Roger and O'Brien, Bill, Oranges or Lemons? Should Labour Organise in Northern Ireland? (London: Kevin McNamara, 1993), p. 21.Google Scholar Integrationists such as Roberts blame the failure of the NILP not on the failure of class politics, but on the refusal of the British Labour party to formally recognize the NILP as an affiliate and thus give it the political clout to ride the sectarian challenges and assert its socialist credentials. However, others see the NILP's demise as evidence of the intractability of the ethnic base of politics.

¹⁸ We do not include the British Liberal Democratic party in our analyses since the integrationist argument centres around the Conservative and Labour parties (presumably because they see these as the anchors of class politics in Britain), and because the APNI is formally recognized as the Liberal Democrat proxy in Northern Ireland.

¹⁹ Religious affiliation was coded: Protestant/Catholic/No Religion. A negligible number of respondents expressed an alternative religious affiliation. Party identification was measured as a composite of responses to three questions: ‘Generally speaking, do you think of yourself as a supporter of any one political party?’, [if no] ‘Do you think of yourself as a little closer to one political party than to others?’ [if no] ‘If there was a general election tomorrow, which political party do you think you would be most likely to support?’ Those who had initially nominated a British party were asked: ‘If there was a general election tomorrow in which only Northern Irish parties were standing, which one do you think you would be most likely to support?’ Respondents were categorized as supporters of the British parties, or one of the five main Northern Irish parties – UUP, DUP, SDLP, Sinn Féin, or APNI. Those identifying with a range of smaller political parties were combined in an ‘Other party’ group. ‘Other’ parties included the Workers' Party, the Green Party, a number of small unionist parties, and a group campaigning for Labour representation in Northern Ireland.

²⁰ Compare our data, based on just those respondents who expressed a party identification (72.5 per cent of the sample), with the 1992 Westminster election results, where turnout was 69.7 per cent (Wilford, Rick, ‘The 1992 Westminster Election in Northern Ireland’, Irish Political Studies, 7 (1992), 105–11)CrossRefGoogle Scholar: 1989–91 party identifiers (%), UUP 41.3, DUP 13.4, SF 4.0, SDLP 21.4, APNI 12.7, Other 7.2; 1992 Westminster election (%), UUP 34.5, DUP 13.1, SF 10.0, SDLP 23.5, APNI 8.7, Other 10.2. It is likely that non-response may be more prevalent in those areas where the conflict is more marked and this may in part explain why there is an under-representation of Sinn Féin support. Underestimation of extremist views and support for extremist parties has also been reported in other studies. See, for example, Moxon-Browne, Edward, Nation, Class and Creed in Northern Ireland (Aldershot: Gower, 1983)Google Scholar; and Whyte, John, Interpreting Northern Ireland (Oxford: Oxford University Press, 1990).Google Scholar Another likelihood is that some (probably the more ambivalent) Sinn Féin supporters are in fact claiming to be SDLP supporters, perhaps because of social desirability pressures or because of the fear and suspicion that a situation of conflict engenders.

²¹ Given the associations between Labour and the nationalist SDLP, and given the strongly unionist stance of the DUP, this Labour/DUP link might seem surprising. However, it probably reflects the shared left-wing economic ideologies of the two parties and the more working-class nature of DUP support. See Evans, Geoffrey and Duffy, Mary, ‘Beyond the Sectarian Divide: The Social Bases and Political Consequences of Nationalist and Unionist Party Competition in Northern Ireland’ (British Journal of Political Science, forthcoming).Google Scholar

²² Social class was measured using the schema devised by Goldthorpe and colleagues (see Erikson, Robert and Goldthorpe, John H., The Constant Flux: A Study of Class Mobility in Industrial Societies (Oxford: Clarendon Press, 1992)).Google Scholar The schema was condensed to provide five classes: (1) the salariat (professionals and managers); (2) routine non-manual workers; (3) petty bourgeoisie (the self-employed); (4) foremen and technicians; (5) the working class (skilled, semi-skilled and unskilled manual workers). Since foremen and technicians tend to be similar to the working class in occupational characteristics and political attitudes, in the multivariate analyses these groups were combined to avoid problems with small cell sizes. See Evans, Geoffrey, ‘Testing the Validity of the Goldthorpe Class Schema’, European Sociological Review, 8 (1992), 211–32.CrossRefGoogle Scholar

²³ Evans, Geoffrey, Heath, Anthony and Payne, Clive, ‘Modelling Trends in the Class/Party Relationship, 1964–1987’, Electoral Studies, 10 (1991), 99–117.CrossRefGoogle Scholar

²⁴ A logit model is used because of the dichotomous dependent variable. With the Conservative party taking the value 0 and Labour the value 1, the model uses the maximum likelihood method to predict the probability of being a Labour supporter, given the explanatory variables. See Aldrich, John H. and Nelson, Forrest D., Linear Probability, Logit and Probit Models (California: Sage Publications, 1984).CrossRefGoogle Scholar

²⁵ The chi-square for the first model is the difference between the – 2 log likelihood for a base model with only a constant term and the model with religion included as a predictor. In Models 2 to 5, the nested model chi-square is the difference between the – 2 log likelihood following the fitting of the previous model and the – 2 log likelihood when the new variable has been added.

²⁶ Nationalist/unionist ideology was measured by the question: ‘Generally speaking, d o you consider yourself as unionist, nationalist or neither?’

²⁷ To measure left-right ideology, we used a scale comprising five agree-disagree items tapping issues relating to the redistribution of wealth and the economic gap between rich and poor: ‘Government should redistribute income from the better-off to those who are worse off’; ‘Big business benefits owners at the expense of workers’; ‘Ordinary people do not get their fair share of the nation's wealth’; ‘There is one law for the rich and one law for the poor’; ‘Management will always try to get the better of employees if it gets the chance’. Responses to each item were measured along 5-point Likert scales, where ‘strongly agree’ was given the value 1 and ‘strongly disagree’ the value 5, with the usual graduations in between and a centre point of 3. A summated scale score ranging between 5 and 25 was created by adding the scores for individual items. Scores were arranged so that higher values represented increasingly right-wing attitudes. The scale has a good level of internal consistency (Cronbach's alpha = 0.81) and has been shown to have acceptable predictive validity and over-time stability. See Evans, Geoffrey, Heath, Anthony and Lalljee, Mansur, ‘Measuring Left-Right and Libertarian-Authoritarian Values in the British Electorate’, British Journal of Sociology, 46 (1995, forthcoming).Google Scholar

²⁸ A measure of secularization is given by over-time increases in the proportions declaring no religious affiliation. Rose's Loyalty Study, carried out in 1968, reported less than 0.5 per cent of those surveyed as having no religious identity. See Rose, Richard, Governing without Consensus: An Irish Perspective (London: Faber, 1971).Google Scholar In the 1989 Northern Irish Social Attitudes survey, this had risen to almost 12 per cent.

²⁹ O'Leary, and McGarry, , The Politics of Antagonism, p. 298.Google Scholar

³⁰ However, this does not mean, especially in the current situation where UUP support for the government is often crucial, that the Conservative party has no price to pay for this support. It must measure its actions so as not to alienate or risk losing the confidence of the unionists.

³¹ McNamara, , Stott, and O'Brien, , Oranges or Lemons?, p. 9.Google Scholar

³² Rose, , Governing without Consensus, p. 218.Google Scholar

³³ Downs, Anthony, An Economic Theory of Democracy (New York: Harper and Row, 1957).Google Scholar

³⁴ See, for example, Barry, Brian, Democracy and Power: Essays in Political Theory, 1 (Oxford: Clarendon Press, 1991), p. 416.Google Scholar

³⁵ Horowitz has recently argued, for example, that without the existence of various ‘moderating’ factors, some ethnic cleavages may be difficult to alter by any means. See Horowitz, Donald L., ‘Democratic Transition and Ethnic Conflict’Google Scholar (Plenary address, Seventeenth Annual Conference of the International Society of Political Psychology, Santiago de Compostela, July 1994).

³⁶ McAllister, Ian and Rose, Richard, ‘Can Political Conflict Be Resolved by Social Change? Northern Ireland as a Test Case’, Journal of Conflict Resolution, 27 (1983), 533–57.CrossRefGoogle Scholar

³⁷ O'Leary, Brendan, ‘The Limits to Coercive Consociationalism in Northern Ireland’, Political Studies, 32 (1989), 562–88.CrossRefGoogle Scholar