Left-right position matters, but does social class? Causal models of the 1992 British general election

British Journal of Political Science

Volumn 28, Issue 3, 1998, Pages 501-529

  Bartle, John ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0040248971     PISSN: 00071234     EISSN: None     Source Type: Journal    
DOI: 10.1017/S0007123498000222     Document Type: Article

References (105)

1. Patrick Dunleavy, 'Mass Political Behaviour: Is there More to Learn?' Political Studies, 28 (1990), 453-69.
2. J. Merrill Shanks, 'Unresolved Issues in Electoral Decisions: Alternative Perspectives on the Explanation of Individual Choice', in M. Kent Jennings and Thomas E. Mann, eds, Elections at Home and Abroad: Essays in Honor of Warren E. Miller (Ann Arbor: The University of Michigan Press, 1994), pp. 17-38.
3. Paul Whiteley, The Labour Party in Crisis (London: Methuen, 1983), p. 106.
4. J. Merrill Shanks and Warren E. Miller, 'Policy Direction and Performance Evaluations: Complementary Explanations of the Reagan Elections', British Journal of Political Science, 20 (1990), 143-235, p. 152.
5. cf. Douglas Rivers, 'Heterogeneity in Models of Electoral Choice', American Journal of Political Science, 35 (1991), 737-57; and Larry Bartels, 'Uninformed Votes: Information Effects in Presidential Elections', American Journal of Political Science, 40 (1996), 194-230.
6. cf. Douglas Rivers, 'Heterogeneity in Models of Electoral Choice', American Journal of Political Science, 35 (1991), 737-57; and Larry Bartels, 'Uninformed Votes: Information Effects in Presidential Elections', American Journal of Political Science, 40 (1996), 194-230.
7. Christopher H. Achen, 'Social Psychology, Demographic Variables, and Linear Regression: Breaking the Iron Triangle in Voting Research', Political Behaviour, 14 (1992), 195-211.
8. The term 'finer grain' of models is borrowed from Jon Elster, Nuts and Bolts for the Social Sciences (Cambridge: Cambridge University Press, 1989).
9. A notable exception is Mark Franklin, The Decline of Class Voting in Britain: Changes in the Basis of Electoral Choice (Oxford: Oxford University Press, 1985); cf. Anthony Heath, John Curtice, Roger Jowell, Geoff Evans, Julia Field and Sharon Witherspoon, Understanding Political Change: The British Voter 1964-1987 (Oxford: Pergamon Press, 1991), p. 51, fn. 16.
10. A notable exception is Mark Franklin, The Decline of Class Voting in Britain: Changes in the Basis of Electoral Choice (Oxford: Oxford University Press, 1985); cf. Anthony Heath, John Curtice, Roger Jowell, Geoff Evans, Julia Field and Sharon Witherspoon, Understanding Political Change: The British Voter 1964-1987 (Oxford: Pergamon Press, 1991), p. 51, fn. 16.
11. David Sanders, 'Economic Performance, Management Competence and the Next General Election', Political Studies, 44 (1996), 203-31.
12. Warren E. Miller and J. Merrill Shanks, 'Policy Directions and Presidential Leadership: Alternative Interpretations of the 1980 Presidential Election', British Journal of Political Science, 12 (1982), 299-356; Shanks and Miller, 'Policy Direction and Performance Evaluation'; J. Merrill Shanks and Warren E. Miller, 'Partisanship, Policy and Performance: The Reagan Legacy in the 1988 Election', British Journal of Political Science, 21 (1991), 129-97; and Warren E. Miller and J. Merrill Shanks, The New American Voter (Cambridge, Mass.: Harvard University Press. 1996).
13. Warren E. Miller and J. Merrill Shanks, 'Policy Directions and Presidential Leadership: Alternative Interpretations of the 1980 Presidential Election', British Journal of Political Science, 12 (1982), 299-356; Shanks and Miller, 'Policy Direction and Performance Evaluation'; J. Merrill Shanks and Warren E. Miller, 'Partisanship, Policy and Performance: The Reagan Legacy in the 1988 Election', British Journal of Political Science, 21 (1991), 129-97; and Warren E. Miller and J. Merrill Shanks, The New American Voter (Cambridge, Mass.: Harvard University Press. 1996).
14. Warren E. Miller and J. Merrill Shanks, 'Policy Directions and Presidential Leadership: Alternative Interpretations of the 1980 Presidential Election', British Journal of Political Science, 12 (1982), 299-356; Shanks and Miller, 'Policy Direction and Performance Evaluation'; J. Merrill Shanks and Warren E. Miller, 'Partisanship, Policy and Performance: The Reagan Legacy in the 1988 Election', British Journal of Political Science, 21 (1991), 129-97; and Warren E. Miller and J. Merrill Shanks, The New American Voter (Cambridge, Mass.: Harvard University Press. 1996).
15. Warren E. Miller and J. Merrill Shanks, 'Policy Directions and Presidential Leadership: Alternative Interpretations of the 1980 Presidential Election', British Journal of Political Science, 12 (1982), 299-356; Shanks and Miller, 'Policy Direction and Performance Evaluation'; J. Merrill Shanks and Warren E. Miller, 'Partisanship, Policy and Performance: The Reagan Legacy in the 1988 Election', British Journal of Political Science, 21 (1991), 129-97; and Warren E. Miller and J. Merrill Shanks, The New American Voter (Cambridge, Mass.: Harvard University Press. 1996).
16. Angus Campbell, Philip E. Converse, Warren E. Millerand Donald Stokes, The American Voter (New York: Wiley, 1960), pp. 24-7.
17. Shanks and Miller, 'Policy Direction and Performance Evaluation'.
18. The effect of race could, in theory, increase as controls are added for class. If this was the case, it could be taken as suggesting that 'perceptions of group interests or processes of group identification' directly affect voting behaviour (Heath et al., Understanding Political Change, p. 113). However, the models of left-right position examined in Section IV suggest that this does not happen here.
19. Ethnicity is not measured 'objectively' in the 1992 BES cross-section. It is a subjective sense of identity. The relevant item (Q. 915(a)) asks respondents 'To which of these groups do you belong?' and then shows them a card containing several ethnic groups.
20. Campbell et al., The American Voter, p. 24.
21. An alternative strategy would be to suggest that both predispositions have an effect and then attempt to model the interrelationship between them in non-recursive models. This strategy is rejected for reasons discussed below.
22. Shanks and Miller, 'Policy Direction and Performance Evaluation', pp. 173-4.
23. A point continually emphasized in Shanks and Miller. 'Policy Direction and Performance Evaluation'.
24. James Davis, The Logic of Causal Order, Sage University Paper Series on Quantitative Applications in the Social Sciences 007-055 (Beverley Hills, Calif.: Sage Publications, 1985) . In general 'block recursive models' produce rather 'conservative' estimates of the apparent total effect of variables. See Miller and Shanks, The New American Voter, chap. 8.
25. James Davis, The Logic of Causal Order, Sage University Paper Series on Quantitative Applications in the Social Sciences 007-055 (Beverley Hills, Calif.: Sage Publications, 1985) . In general 'block recursive models' produce rather 'conservative' estimates of the apparent total effect of variables. See Miller and Shanks, The New American Voter, chap. 8.
26. William D. Berry, Non-Recursive Causal Models, Sage University Paper Series on Quantitative Applications in the Social Sciences 007-037 (Beverley Hills, Calif.: Sage Publications, 1984), pp. 11-15; John E. Jackson, 'Issues. Party Choices and Presidential Votes', American Journal of Political Science, 19 (1975), 161-85.
27. William D. Berry, Non-Recursive Causal Models, Sage University Paper Series on Quantitative Applications in the Social Sciences 007-037 (Beverley Hills, Calif.: Sage Publications, 1984), pp. 11-15; John E. Jackson, 'Issues. Party Choices and Presidential Votes', American Journal of Political Science, 19 (1975), 161-85.
28. Berry, Non-Recursive Causal Models, p. 9.
29. Larry M. Bartels, 'Instrumental and Quasi-instrumental Variables', American Journal of Political Science, 35 (1991), 777-800.
30. Bartels, 'Instrumental and Quasi-instrumental Variables', p. 800.
31. Davis, 'The Logic of Causal Order', puts it quite forcefully: 'Although the methodological journals teem with crackpot proposals for determining causal order using only the numbers in the data set, most methodologists agree that causal order is a substantive or empirical problem to be solved by our knowledge about how the real world works, not by statistical gyrations', p. 11. See the exchanges in the following articles: David A. Freedman, 'Statistical Models and Shoe Leather Costs', Sociological Methodology, 21 (1991), 291-313; Richard A. Berk, 'Toward a Methodology for Mere Mortals', Sociological Methodology, 21 (1991), 315-24; Hubert M. Blalock Jr, 'Are There Really Any Constructive Alternatives to Causal Modelling?' Sociological Methodology, 21 (1991), 325-35.
32. Davis, 'The Logic of Causal Order', puts it quite forcefully: 'Although the methodological journals teem with crackpot proposals for determining causal order using only the numbers in the data set, most methodologists agree that causal order is a substantive or empirical problem to be solved by our knowledge about how the real world works, not by statistical gyrations', p. 11. See the exchanges in the following articles: David A. Freedman, 'Statistical Models and Shoe Leather Costs', Sociological Methodology, 21 (1991), 291-313; Richard A. Berk, 'Toward a Methodology for Mere Mortals', Sociological Methodology, 21 (1991), 315-24; Hubert M. Blalock Jr, 'Are There Really Any Constructive Alternatives to Causal Modelling?' Sociological Methodology, 21 (1991), 325-35.
33. Davis, 'The Logic of Causal Order', puts it quite forcefully: 'Although the methodological journals teem with crackpot proposals for determining causal order using only the numbers in the data set, most methodologists agree that causal order is a substantive or empirical problem to be solved by our knowledge about how the real world works, not by statistical gyrations', p. 11. See the exchanges in the following articles: David A. Freedman, 'Statistical Models and Shoe Leather Costs', Sociological Methodology, 21 (1991), 291-313; Richard A. Berk, 'Toward a Methodology for Mere Mortals', Sociological Methodology, 21 (1991), 315-24; Hubert M. Blalock Jr, 'Are There Really Any Constructive Alternatives to Causal Modelling?' Sociological Methodology, 21 (1991), 325-35.
34. Davis, 'The Logic of Causal Order', puts it quite forcefully: 'Although the methodological journals teem with crackpot proposals for determining causal order using only the numbers in the data set, most methodologists agree that causal order is a substantive or empirical problem to be solved by our knowledge about how the real world works, not by statistical gyrations', p. 11. See the exchanges in the following articles: David A. Freedman, 'Statistical Models and Shoe Leather Costs', Sociological Methodology, 21 (1991), 291-313; Richard A. Berk, 'Toward a Methodology for Mere Mortals', Sociological Methodology, 21 (1991), 315-24; Hubert M. Blalock Jr, 'Are There Really Any Constructive Alternatives to Causal Modelling?' Sociological Methodology, 21 (1991), 325-35.
35. Christopher J. Fleury and Michael S. Lewis-Beck, 'Anchoring the French Voter: Ideology Versus Party', Journal of Politics, 55 (1993), 1100-9;
36. Philip E. Converse and Roy Pierce, 'Comment on Fleury and Lewis-Beck, "Anchoring the French Voter: Ideology Versus Party" ', Journal of Politics, 55 (1993), 1110-17.
37. This article assumes that the effect of any given explanatory variable is both linear and additive. In practice, both assumptions may be violated. See Rivers. 'Heterogeneity in Models of Electoral Choice'.
38. Respondents are asked to describe the sort of jobs that their fathers had when they were around 14 years of age and they were then assigned a 'Goldthorpe class'. This article does not explore the possibilities of measurement error that may result. Any error will weaken the strength of the relationship between father's social class and left-right position or voting.
39. Campbell et al., The American Voter, conceded that the funnel of causality 'like all physical analogies for complex and intangible processes ... becomes misleading if pressed too far' (p. 24).
40. Of the people who reported being a 'member' of the Church of England in 1987, 91.2 per cent were still 'members' in 1992. The figure for Catholics was 83.4 per cent. Similarly, 96.4 per cent of homeowners in 1987 were still homeowners in 1992, while 70.75 per cent of council tenants in 1987 were still tenants in 1992. Evidence on regional mobility is contained in Dorren McMahon, Anthony Heath, Martin Harrop and John Curtice, 'The Electoral Consequences of North-South Migration', British Journal of Political Science, 22 (1993), 419-43.
41. The stability of social class depends on the schema used. When the five-fold Goldthorpe schema is used 1987-92 BES panel data suggests that 79.5 per cent of the 1987 salariat were still in that group in 1992. Equivalent figures for the petty bourgeoisie are 68.6 per cent, for foremen and technicians 41.8 per cent, for routine non-manual workers 62.2 per cent and for the working class 78 per cent. These figures may reflect some combination of social mobility and measurement error. The stability figures for the eleven-class schema are: high-grade salariat 60.1 per cent, low-grade salariat 53.0 per cent, routine office workers 55.3 per cent, sales personnel 45.0 per cent, small petty bourgeoisie with employees 59.1 per cent, small petty bourgeoisie without employees 43 per cent, farmers 57.2 per cent, foremen and technicians 41.8 per cent, skilled manual workers 47.8 per cent, semi-skilled manual workers 64.4 per cent and agricultural employees 46.0 per cent. Most of the movements in this eleven-class schema are between contiguous classes (such as high- and low-grade salariat).
42. My thanks to Anthony Heath for this point. It could equally be argued that it is actually income which determines housing. Unfortunately, it was not possible to include income as an explanatory variable, because there was a great deal of missing data.
43. Anthony Heath, Roger Jowell and John Curtice, How Britain Votes (Oxford: Pergamon Press, 1985): 'we suspect that there is now a greater choice in housing market than in the labour market' (p. 49: see also p. 54, fn. 4).
44. The last variable (whether the respondent has privatized shares) raises particular problems about reciprocal relationships between social characteristics and values. However, in this model, ownership of privatized shares is interpreted as an indicator of wealth and therefore prior to left-right values
45. This statement must be qualified. Voters are quite stable in their identity as middle-class and working-class but are less consistent about expressing such an identity without prompting. Given the apparent importance of this distinction in models of left-right choice (see Section IV below), the class identity variables are treated as highly proximate to voters' left-right positions.
46. David Butler and Donald Stokes, Political Change in Britain: The Evolution of Electoral Preference (London: Macmillan, 1974), p. 93.
47. Campbell et al., The American Voter, chap. 7.
48. Campbell et al., The American Voter, p. 135.
49. Campbell et al., The American Voter.
50. Campbell et al., The American Voter, p. 141.
51. Achen, 'Social Psychology, Demographic Variables and Linear Regression', p. 196; cf. Donald Philip Green and Bradley Palmquist, 'Of Artifacts and Partisan Instability', American Journal of Political Science, 34 (1990), 872-902, Donald Philip Green and Bradley Palmquist, 'How Stable is Party Identification?' Political Behaviour, 16 (1994), 437-66; and R. J. Johnston and C. J. Pattie, 'The Strength of Party Identification Among the British Electorate: An Exploration', Electoral Studies, 15 (1996), 295-309.
52. Achen, 'Social Psychology, Demographic Variables and Linear Regression', p. 196; cf. Donald Philip Green and Bradley Palmquist, 'Of Artifacts and Partisan Instability', American Journal of Political Science, 34 (1990), 872-902, Donald Philip Green and Bradley Palmquist, 'How Stable is Party Identification?' Political Behaviour, 16 (1994), 437-66; and R. J. Johnston and C. J. Pattie, 'The Strength of Party Identification Among the British Electorate: An Exploration', Electoral Studies, 15 (1996), 295-309.
53. Achen, 'Social Psychology, Demographic Variables and Linear Regression', p. 196; cf. Donald Philip Green and Bradley Palmquist, 'Of Artifacts and Partisan Instability', American Journal of Political Science, 34 (1990), 872-902, Donald Philip Green and Bradley Palmquist, 'How Stable is Party Identification?' Political Behaviour, 16 (1994), 437-66; and R. J. Johnston and C. J. Pattie, 'The Strength of Party Identification Among the British Electorate: An Exploration', Electoral Studies, 15 (1996), 295-309.
54. Achen, 'Social Psychology, Demographic Variables and Linear Regression', p. 196; cf. Donald Philip Green and Bradley Palmquist, 'Of Artifacts and Partisan Instability', American Journal of Political Science, 34 (1990), 872-902, Donald Philip Green and Bradley Palmquist, 'How Stable is Party Identification?' Political Behaviour, 16 (1994), 437-66; and R. J. Johnston and C. J. Pattie, 'The Strength of Party Identification Among the British Electorate: An Exploration', Electoral Studies, 15 (1996), 295-309.
55. Butler and Stokes, Political Change in Britain, p. 69. The measurement of class in terms of the manual/non-manual dichotomy made sense because they found overwhelming evidence that most people thought of class in these terms.
56. Party identification is measured by responses to the question 'Generally speaking do you think of yourself as Conservative, Labour, Liberal Democrat or what?' Those voters who decline an identity, but who feel themselves 'closer to' a party are not classified as identifiers. This practice is in accordance with the observations made in Warren E. Miller, 'Party Identification, Realignment and Party Voting: Back to Basics', American Political Science Review, 85 (1991), 557-68. See Charles H. Franklin, 'Measurement and the Dynamics of Party Identification', Political Behaviour, 14 (1992), 297-309, for a critique of Miller's work.
57. Party identification is measured by responses to the question 'Generally speaking do you think of yourself as Conservative, Labour, Liberal Democrat or what?' Those voters who decline an identity, but who feel themselves 'closer to' a party are not classified as identifiers. This practice is in accordance with the observations made in Warren E. Miller, 'Party Identification, Realignment and Party Voting: Back to Basics', American Political Science Review, 85 (1991), 557-68. See Charles H. Franklin, 'Measurement and the Dynamics of Party Identification', Political Behaviour, 14 (1992), 297-309, for a critique of Miller's work.
58. The full list of variables appears in Appendix 1 (b) below.
59. The general strategy being pursued in my research is to model voting behaviour as a series of dichotomous choices: Conservative or Labour, Conservative or Liberal and Labour or Liberal. Because the dependent variable is dichotomous, it is preferable to use a logistic regression. However, the ordinary least square (OLS) results are more easily interpretable and are reported here. Logistic regressions confirm the patterns outline above, with few variables being found to be significant in the OLS and not the logit. On the question of model choice, see Guy D. Whitten and Harvey D. Palmer, 'Heightening Comparativists' Concern for Model Choice: Voting Behaviour in Great Britain and the Netherlands', American Journal of Political Science, 40 (1996), 231-60 and R. Michael Alvarez and Jonathan Nagler, 'Economics, Issues and the Perot Candidacy: Voter Choice in the 1992 Presidential Elections', American Journal of Political Science, 39 (1995), 714-44.
60. The general strategy being pursued in my research is to model voting behaviour as a series of dichotomous choices: Conservative or Labour, Conservative or Liberal and Labour or Liberal. Because the dependent variable is dichotomous, it is preferable to use a logistic regression. However, the ordinary least square (OLS) results are more easily interpretable and are reported here. Logistic regressions confirm the patterns outline above, with few variables being found to be significant in the OLS and not the logit. On the question of model choice, see Guy D. Whitten and Harvey D. Palmer, 'Heightening Comparativists' Concern for Model Choice: Voting Behaviour in Great Britain and the Netherlands', American Journal of Political Science, 40 (1996), 231-60 and R. Michael Alvarez and Jonathan Nagler, 'Economics, Issues and the Perot Candidacy: Voter Choice in the 1992 Presidential Elections', American Journal of Political Science, 39 (1995), 714-44.
61. Party identification also reduces the apparent effects of policy preferences and evaluations by factors between 2 and 8, compared with models which control for left-right position. Those analysts who believe that much voting is of the habitual or expressive variety will not be overly concerned about these effects. Those who believe that voting represents a conscious choice between the parties would suggest that this indicates that party identification is quasi-tautological. Such analysts could also point out that the effect of party identification in Britain (controlling for a similar set of social characteristics) is considerably larger in Britain than the United States (0.88 compared with 0.70). (See Miller and Shanks, The New American Voter, Table 11.1, p. 286.) This again adds to the suspicion of tautology.
62. Harold D. Clarke, Marianne Stewart and Paul Whiteley, 'Can the Tories Win Again?' (paper presented at the Specialist Group Conference on Elections, Public Opinion and Parties, September 1995, London).
63. Clarke et al., 'Can the Tories Win Again?' p. 9. Similar points are made in Herbert F. Weisberg and Charles E. Smith Jr, The Influence of the Economy on Party Identification in the Reagan Years', Journal of Politics, 53 (1991), 1077-92.
64. Clarke et al., 'Can the Tories Win Again?' p. 9. Similar points are made in Herbert F. Weisberg and Charles E. Smith Jr, The Influence of the Economy on Party Identification in the Reagan Years', Journal of Politics, 53 (1991), 1077-92.
65. See Morris P. Fiorina, Retrospective Voting in American National Elections (New Haven, Conn.: Yale University Press, 1981).
66. But see Anthony Heath and Roy Pierce, 'It Was Party Identification All Along: Question Order Effects on Party Identification in Britain', Electoral Studies, 11 (1992), 93-105.
67. Stanley Feldman, 'Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values', American Journal of Political Science, 32 (1988), 410-40; Anthony Heath, Geoffrey Evans and Jean Martin, 'The Measurement of Core Beliefs and Values: The Development of Socialist/Laissez Faire and Libertarian/Authoritarian Scales', British Journal of Political Science, 24 (1994), 115-32; Geoffrey Evans, Anthony Heath and Mansur Lalljee, 'Measuring Left-Right and Libertarian-Authoritarian Values in the British Electorate', British Journal of Sociology, 47 (1996), 93-112; Harvey D Palmer, 'Effects of Authoritarian and Libertarian Values on Conservative and Labour Party Support in Great Britain', European Journal of Political Research, 27 (1995), 273-92; Jan W. van Deth and Elinor Scarbrough, eds, The Impact of Values (Oxford: Oxford University Press, 1995)
68. Stanley Feldman, 'Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values', American Journal of Political Science, 32 (1988), 410-40; Anthony Heath, Geoffrey Evans and Jean Martin, 'The Measurement of Core Beliefs and Values: The Development of Socialist/Laissez Faire and Libertarian/Authoritarian Scales', British Journal of Political Science, 24 (1994), 115-32; Geoffrey Evans, Anthony Heath and Mansur Lalljee, 'Measuring Left-Right and Libertarian-Authoritarian Values in the British Electorate', British Journal of Sociology, 47 (1996), 93-112; Harvey D Palmer, 'Effects of Authoritarian and Libertarian Values on Conservative and Labour Party Support in Great Britain', European Journal of Political Research, 27 (1995), 273-92; Jan W. van Deth and Elinor Scarbrough, eds, The Impact of Values (Oxford: Oxford University Press, 1995)
69. Stanley Feldman, 'Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values', American Journal of Political Science, 32 (1988), 410-40; Anthony Heath, Geoffrey Evans and Jean Martin, 'The Measurement of Core Beliefs and Values: The Development of Socialist/Laissez Faire and Libertarian/Authoritarian Scales', British Journal of Political Science, 24 (1994), 115-32; Geoffrey Evans, Anthony Heath and Mansur Lalljee, 'Measuring Left-Right and Libertarian-Authoritarian Values in the British Electorate', British Journal of Sociology, 47 (1996), 93-112; Harvey D Palmer, 'Effects of Authoritarian and Libertarian Values on Conservative and Labour Party Support in Great Britain', European Journal of Political Research, 27 (1995), 273-92; Jan W. van Deth and Elinor Scarbrough, eds, The Impact of Values (Oxford: Oxford University Press, 1995)
70. Stanley Feldman, 'Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values', American Journal of Political Science, 32 (1988), 410-40; Anthony Heath, Geoffrey Evans and Jean Martin, 'The Measurement of Core Beliefs and Values: The Development of Socialist/Laissez Faire and Libertarian/Authoritarian Scales', British Journal of Political Science, 24 (1994), 115-32; Geoffrey Evans, Anthony Heath and Mansur Lalljee, 'Measuring Left-Right and Libertarian-Authoritarian Values in the British Electorate', British Journal of Sociology, 47 (1996), 93-112; Harvey D Palmer, 'Effects of Authoritarian and Libertarian Values on Conservative and Labour Party Support in Great Britain', European Journal of Political Research, 27 (1995), 273-92; Jan W. van Deth and Elinor Scarbrough, eds, The Impact of Values (Oxford: Oxford University Press, 1995)
71. Stanley Feldman, 'Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values', American Journal of Political Science, 32 (1988), 410-40; Anthony Heath, Geoffrey Evans and Jean Martin, 'The Measurement of Core Beliefs and Values: The Development of Socialist/Laissez Faire and Libertarian/Authoritarian Scales', British Journal of Political Science, 24 (1994), 115-32; Geoffrey Evans, Anthony Heath and Mansur Lalljee, 'Measuring Left-Right and Libertarian-Authoritarian Values in the British Electorate', British Journal of Sociology, 47 (1996), 93-112; Harvey D Palmer, 'Effects of Authoritarian and Libertarian Values on Conservative and Labour Party Support in Great Britain', European Journal of Political Research, 27 (1995), 273-92; Jan W. van Deth and Elinor Scarbrough, eds, The Impact of Values (Oxford: Oxford University Press, 1995)
72. Anthony Downs, An Economic Theory of Democracy (New York: Harper & Row, 1957), p. 96
73. Ian Budge, 'A New Spatial Theory of Party Competition: Uncertainty, Ideology and Policy Equilibria Viewed Comparatively and Temporally', British Journal of Political Science, 24 (1994), 443-67.
74. Philip Converse, 'The Nature of Belief Systems in Mass Publics', in D. Apter, ed., Ideology and Discontent (New York: The Free Press, 1964).
75. Elinor Scarbrough, Ideology and Voting Behaviour: An Exploratory Study (Oxford: Clarendon Press, 1984).
76. The importance of ideology or ideological position will depend upon the voter's ability to relate their general beliefs to specific choices or preferences. For an important statement of this principle, see John R. Zaller, The Nature and Origins of Mass Opinion (Cambridge: Cambridge University Press. 1992).
77. Anthony Heath and Sarah MacDonald, 'The Demise of Party Identification Theory?' In David Denver and Gordon Hands, eds, Issues and Controversies in Voting Behaviour (London: Harvester Wheatsheaf, 1991), p. 230-8.
78. Christopher H. Achen, 'Mass Political Attitudes and Survey Response', American Political Science Review; 12 (1975), 195-211.
79. This definition is Lipset's. It is quoted in D. Fuchs and H. D. Klingemann, 'The Left-Right Schema', in M. K. Jennings et al., eds, Continuities in Political Action: A Longitudinal Study of Political Orientations in Three Western Democracies (New York: Walter de Gruyter, 1989), pp. 203-34, at p. 224.
80. Shanks and Miller, 'Partisanship, Policy and Performance', p. 154.
81. David Weakliem and Anthony Heath, 'Rational Choice and Class Voting', Rationality and Society, 6 (1994), 243-70. These authors use 1987 cross-sectional data to suggest that, since some of the effects of class persist after controlling for attitudes, there is some form of 'social determinism' at work (p. 266). The evidence presented here suggests that the direct effects of class might disappear once enough controls are added for political attitudes.
82. If the models failed to control for any stable predisposition (party identification or left-right position), a statistically significant direct effect for the working-class variable would also remain.
83. Heath et al., Understanding Political Change, 'social class (and class related variables) dominates analysis of the left-right dimension', p. 174.
84. Heath et al., How Britain Votes, p. 14.
85. Heath et al., How Britain Votes, p. 14.
86. See David Rose, A Report on Phase One of the ESRC Review of OPCS Social Classifications (Swindon: Economic and Social Research Council, 1995).
87. Heath et al., How Britain Votes, chap. 2.
88. Heath et al., How Britain Votes, p. 15.
89. See Giovanni Sartori, 'The Sociology of Parties: A Critical Review', in Peter Mair, ed., The West European Party System (Oxford: Oxford University Press, 1990), pp. 150-82.
90. Where class is measured by the eleven-fold schema and entered in the models as a series of ten dummy variables, the models suggest that there are statistically significant differences between the groups that make up the five classes (for example, between the high and low grade salariat). Details of such analyses are available from the author upon request.
91. The notable exception are farmers, who do belong to a powerful trade union.
92. The average scores for each of the classes can be calculated by simply adding the coefficient (which may be negative) to the constant from the first regression. Thus, the average score for the petty bourgeoisie is 10.80 + 1.06 = 11.86.
93. There is a great deal of missing data on father's class. Such cases are omitted from the statistical analyses.
94. The effect of being a Roman Catholic is not diminished if controls are added for Irish origins.
95. There are reasons to be suspicious about the effect of the Midlands regional variable in the models. This variable includes the East and West Midlands and East Anglia. Some 62 per cent of the BES sample in the West Midlands reported having voted Conservative, compared with the 46.6 per cent actually recorded. See David Butler and Dennis Kavanagh, The British General Election of 1992 (London: Macmillan, 1992), p. 268.
96. Campbell et al., The American Voter, were alert to this problem: 'Measurement close to the behaviour runs the risk of including values that are determined by any event we are trying to predict, that is, the vote decision' (p. 35).
97. Heath et al., How Britain Votes, p. 49; cf. Dunleavy, 'Mass Political Behaviour', p. 460.
98. Heath et al., How Britain Votes, p. 49; cf. Dunleavy, 'Mass Political Behaviour', p. 460.
99. Of course, it might be that those who had previously joined and subsequently left a trade union originally joined because of their (then existing) 'left-wing' views.
100. The BES asks respondents, 'Do you ever think of yourself as belonging to any particular class? (If yes) Which class is that?'; those who report anything other than middle or working class or don't know are then prompted, 'Most people say they belong to the middle class or the working class. If you had to make a choice, would you call yourself ... middle class or working class?' (Questions 55(a) and 55(b), 1992 cross-sectional study).
101. Although Heath et al., Understanding Political Change, p. 174, argue that 'class (and class related variables)' dominate analysis of left-right positions, it is unclear what the authors mean by both 'dominate' and 'class related variables'.
102. Butler and Stokes, Political Change in Britain, were similarly unclear as to the nature of the class-party relationship, saying that 'too little attention has been paid to the beliefs that link class to party in the voters mind', p. 67.
103. Shanks, 'Unresolved Issues', p. 19, makes the point forcefully.
104. Heath et al., Understanding Political Change, p. 96.
105. Anthony Heath, 'Comment on Dennis Kavanagh's "How We Vote Now" ', Electoral Studies, 5 (1986), 29-30.