¹ Carty, R. K., Party and Parish Pump: Electoral Politics in Ireland (Waterloo, Ontario: Wilfrid Laurier Press, 1981), p. 85.Google Scholar

² Chubb, Basil, The Government and Politics of Ireland, 2nd edn (London: Longman, 1982), p. 98, p. 104Google Scholar. The latter view is, however, subsequently implicitly qualified by the statement that ‘although it is not possible to attribute distinctive social bases to these two parties, perhaps there are differences that have their roots far back in history – in the attitudes of Irish people to national issues more than half a century ago that have been perpetuated’ (p. 108).

³ For an examination of the Irish party system from this perspective see Sinnott, Richard, ‘Interpretations of the Irish Party System’, European Journal of Political Research, XII (1984), 289–307.CrossRefGoogle Scholar

⁴ The stratified sample of seventy-five politicians included all current and former Ministers and Parliamentary Secretaries, all current frontbench spokesmen and fifteen prominent backbenchers identified by the political correspondents. A further eighteen randomly selected backbenchers were included to make up the total number. The interviews, which were of the open-ended conversational kind, were tape-recorded. Transcriptions of these recordings constitute the data base for this phase of the research. This approach to the study of elite political attitudes is discussed by Putnam (Putnam, Robert D., The Beliefs of Politicians: Ideology, Conflict and Democracy in Britain and Italy (New Haven and London: Yale University Press, 1973), pp. 1–11Google Scholar). It has been emphasized in the text that this research focuses on one of several sources of evidence relevant to the study of inter-party differences. For a study of the Irish party system that concentrates on the evidence of election manifestos, see Mair, Peter, ‘Issue-Dimensions and Party Strategies in the Irish Republic, 1948–81: The Evidence of Manifestos’, EUI Working Paper No. 41 (Florence: European University Institute, 1982).Google Scholar

⁵ TD: abbreviation for Teachta Dála (Dáil Deputy, literally messenger or delegate to the Dáil). For a glossary of Irish political terms see Chubb, , Government and Politics, pp. xvii–xviii.Google Scholar

⁶ The economic discussions were coded for topics or themes in a treatment somewhat similar to that developed by Robertson for the analysis of British election manifestos. Some of Robertson's topics were found to be directly applicable to the data, others required modification and it was also necessary to develop several additional constructs. A procedural difference from Robertson's approach is that, whereas he counted the number of occurrences of a particular topic, we coded the degree of importance or centrality of the theme in the discussion. See Robertson, David, A Theory of Party Competition (London: Wiley, 1976), pp. 76–9Google Scholar. In the case of the Northern Ireland discussions, it was necessary to construct an entirely new coding scheme. Both coding schemes are reproduced in the Appendix.

⁷ The use of discriminant analysis involves the deliberate decision to treat the data as if they were measured at the interval level. In adopting this course we are following the advice of Tufte, among others, who argues that ‘the distinction between interval and ordinal measurement is usually of little importance in data analysis. The wise assignment of numbers to ordered categories, coupled with the use of techniques that exploit the properties of numbers is generally preferable to working with ordered categories’ (Tufte, E. R., ‘Improving Data Analysis in Political Science’, World Politics, XXI (1969), 642Google Scholar). As to other assumptions about the data, Klecka points out that the technique is ‘rather robust’ and he argues that ‘in practical research situations, we must often be content with data that do not conform nicely to the assumptions of our statistical technique’ (Klecka, William R., Discriminant Analysis (London: Sage Publications, 1980), p. 63).CrossRefGoogle Scholar

⁸ This question is modelled on that used by Daalder and Rusk in their study of Dutch parliamentarians, with the substitution of individual politicians for the parties used as stimuli by Daalder and Rusk. See Daalder, Hans and Rusk, Jerrold G., ‘Perceptions of Party in the Dutch Parliament’ in Patterson, Samuel C. and Wahlke, John C., eds, Comparative Legislative Behavior: Frontiers of Research (New York: Wiley, 1972), pp. 143–98Google Scholar. The question was put to sixty-five of the seventy-five politicians interviewed. Omission of the question in ten of the interviews was due to time pressure on the respondents. Seven respondents declined to answer the question. This was a higher refusal rate than for any of the other questions, but given the potentially sensitive nature of the question, we can perhaps console ourselves that it was not even higher. Allowing for the one non-party deputy interviewed, the net outcome is that we have fifty-seven cases available for this analysis, the overall number consisting of thirty Fianna Fail respondents, nineteen Fine Gael respondents and eight Labour party respondents.

⁹ Duverger, Maurice, Political Parties: Their Organization and Activity in the Modern State, trans. Barbara, and North, Robert, 2nd English edn (London: Methuen, 1959), pp. 215–16Google Scholar; Downs, Anthony, An Economic Theory of Democracy (New York: Harper and Row, 1957), pp. 115–16.Google Scholar

¹⁰ Robertson, David, A Theory of Party Competition (London: Wiley, 1976), p. 59.Google Scholar

¹¹ Robertson, , A Theory of Party Competition, p. 148.Google Scholar

¹² Sartori, Giovanni, Parties and Party Systems: A Framework for Analysis, Volume I (Cambridge: Cambridge University Press, 1976), p. 335.Google Scholar

¹³ Sartori, , Parties and Party Systems, p. 337.Google Scholar

¹⁴ Sartori, , Parties and Party Systems, pp. 336–7Google Scholar; in this context it may be worth noting Daalder's comment on a subsequent treatment of this issue by Sartori. According to Daalder, Sartori and his co-author, Sani, ‘seek to reduce a multidimensional world of party identifications to a unidimensional universe of electoral competition’ (Daalder, Hans, ‘The Comparative Study of European Parties and Party Systems: An Overview’, in Daalder, Hans and Mair, Peter, eds, Western European Party Systems (London: Sage Publications, 1983), p. 15).Google Scholar

¹⁵ Stokes, Donald, ‘Spatial Models of Party Competition’ in Campbell, Angus et al. , eds, Elections and the Political Order (New York: Wiley, 1966), p. 176.Google Scholar

¹⁶ Budge, Ian and Farlie, Dennis, Voting and Party Competition: A Theoretical Critique and Synthesis Applied to Surveys from Ten Democracies (London: Wiley, 1977), pp. 3–5.Google Scholar

¹⁷ Budge, and Farlie, , Voting and Party Competition, p. 5.Google Scholar

¹⁸ For an overview of the proposed theoretical synthesis see Budge, and Farlie, , Voting and Party Competition, pp. 497–500Google Scholar. The data analysis (of voting surveys from ten countries) employs a quasi-Bayesian statistical technique to place voters in a ‘likelihood ratio space’. Although the input into this sort of analysis is data on the distribution of party supporters over demographic and attitudinal variables, the space itself is ‘party defined’ in the sense that, in the case of the British party system, for example, the ends of the line or dimension are ‘defined by complete certainty of being Labour or the converse, complete certainty of being Conservative’ (Budge, and Farlie, , Voting and Party Competition, p. 20).Google Scholar

¹⁹ Budge, and Farlie, , Voting and Party Competition, p. 4.Google Scholar

²⁰ Budge, and Farlie, , Voting and Party Competition, p. 228.Google Scholar

²¹ Budge, and Farlie, , Voting and Party Competition, p. 228.Google Scholar

²² Budge, Ian, Crewe, Ivor and Farlie, Dennis, Party Identification and Beyond (London: Wiley, 1976), p. 16.Google Scholar

²³ Klecka, , Discriminant Analysis, p. 7.Google Scholar

²⁴ Kruskal, Joseph B. and Wish, Myron, Multidimensional Scaling (London: Sage Publications, 1978), p. 7.CrossRefGoogle Scholar

²⁵ Budge, and Farlie, , Voting and Party Competition, p. 52.Google Scholar

²⁶ Herbert, E. G., Weisberg, F. and Rusk, Jerrold G., ‘Dimensions of Candidate Evaluation’, American Political Science Review, LXIV (1970), 1167–85.Google Scholar

²⁷ Daalder, and Rusk, , ‘Perceptions of Party in the Dutch Parliament’, pp. 143–98.Google Scholar

²⁸ Kruskal, and Wish, , Multidimensional Scaling, pp. 48–9.Google Scholar

²⁹ Rabinowitz, George B., ‘An Introduction to Nonmetric Multidimensional Scaling’, American Journal of Political Science, XIX (1975), 369.Google Scholar

³⁰ I am grateful to Charles L. Jones for suggesting this strategy of examining individual stress values and removing highly exceptional cases. The first case (stress value 0·691) was a Fine Gael respondent whose preference order was as follows: Cosgrave (FG), FitzGerald (FG), Corish (Lab), Tully (Lab), Keating (Lab), Ryan (FG), Haughey (FF), Burke (FG), Lynch (FF), O'Brien (Lab), O'Malley (FF), Colley (FF). There are several exceptional features here, the most obvious of which are the low placement of Ryan and especially of Burke, given that the first choice was Cosgrave and, secondly, the preference for Haughey over the other Fianna Fáil politicians and indeed over Burke of Fine Gael. The much more common Fine Gael preference vis-à-vis the Fianna Fáil stimuli was Lynch or indeed Colley or O'Malley over Haughey. The second exceptional case (stress value of 0·551) was a Fianna Fáil respondent with the following ranking: Haughey (FF). FitzGerald (FG). O'Malley (FF). Colley (FF). Cosgrave (FG). Lynch (FF). Coristi (Lab). Tully (Lab). Keating (Lab). Burke (FG). Ryan (FG). O'Brien (Lab). Here the unusual features are the placement of FitzGerald in second position after Haughey and the placement of another but very different Fine Gael politician (Cosgrave) before the Fianna Fáil leader Lynch. The third exceptional case (stress value 0·541) was a Labour deputy whose preferences were as follows: Keating (Lab). FitzGerald (FG). Haughey (FF). Corish (Lab). Tully (Lab). Ryan (FG). O'Malley (FF). Lynch (FF), Cosgrave (FG). Cruise O'Brien (Lab). Colley (FF). Burke (FG). Here the exceptional feature is the spanning of all three parties with the first three preferences and the unusual juxtaposition in second and third place of FitzGerald and Haughey. It should be emphasized that the argument is not that any of the three cases is meaningless but that, whether assessed in terms of their poor fit in the two dimensional solution or in terms of their individual ranking, they are clearly exceptional.

³¹ Kruskal, and Wish, , Multidimensional Scaling, p. 45.Google Scholar

³² See especially O'Brien, Conor Cruise, States of Ireland (London: Hutchinson, 1972).Google Scholar

³³ Sartori, , Parties and Party Systems, p. 338.Google Scholar

³⁴ See Manning, Maurice, ‘The Political Parties’, in Penniman, Howard, ed., Ireland at the Polls: The Dail Elections of 1977 (Washington, D.C.: American Enterprise Institute, 1978), pp. 88–99.Google Scholar

³⁵ For details of these variables see the coding schemes in the Appendix.

³⁶ This variable was coded on a five-point scale with ends defined as an unqualified internal solution on the one hand and an unqualified united Ireland solution on the other, with degrees of emphasis on these objectives in between. See Appendix for details.

³⁷ In regard to this variable in particular it should be noted that a low score simply means that an individual did not mention such a policy proposal, not that he explicitly opposed it or proposed its opposite.

³⁸ See Sinnott, , ‘Interpretations of the Irish Party System’, pp. 301–5.Google Scholar

³⁹ As indicated in footnote 6, in constructing this code use was made where possible of the coding scheme developed by Robertson for the analysis of British election manifestos (Robertson, , Theory of Party Competition, pp. 72–5).Google Scholar