¹ See Butler, David and Stokes, Donald, Political Change in Britain, 2nd edn. (London: Macmillan, 1974), pp. 193–208CrossRefGoogle Scholar, and Crewe, Ivor, Särlvik, Bo and Alt, James, ‘Partisan Dealignment in Britain 1964–74’, British Journal of Political Science, VII (1977), pp. 129–90CrossRefGoogle Scholar. I must thank these investigators for the use of their data which was made available through the ICPSR and the SSRC Survey Archive. This paper is restricted to England because the growth of nationalist voting, and national trends in support for the non-nationalist parties imply that a Britain-wide analysis of trends would simply mix and confuse a variety of divergent trends.

² Butler, and Stokes, , Political Change in Britain, pp. 68–71.Google Scholar

³ See, for example, the footnote on p. 175 of Butler, and Stokes, , Political Change in Britain.Google Scholar

⁴ Butler, and Stokes, , Political Change in Britain, p. 73.Google Scholar

⁵ See the footnote on p. 75 of the 1974 edition of Butler, and Stokes, , Political Change in Britain.Google Scholar

⁶ Rose, Richard, ‘Britain: Simple Abstractions and Complex Realities’, in Rose, Richard, ed., Electoral Behaviour: A Comparative Handbook (New York: Free Press, 1974), p. 501.Google Scholar

⁷ Butler, and Stokes, , Political Change in Britain, p. 73.Google Scholar

⁸ It is an interesting paradox that as more people have acquired occupations they have become less willing to admit thinking in occupational class terms.

⁹ Rose, , ‘Britain: Simple Abstractions and Complex Realities’, p. 510.Google Scholar

¹⁰ Butler, and Stokes, , Political Change in Britain, pp. 203–4Google Scholar, and Crewe, , Särlvik, and Alt, , ‘Partisan Dealignment’, p. 169.Google Scholar

¹¹ Alford, Robert, Party and Society (Chicago: Rand McNally, 1963)Google Scholar uses a unipolar Labour versus the rest approach. Butler, and Stokes, , Political Change in BritainGoogle Scholar, usually opt for a bipolar Labour-versus-Conservative analysis but when treating the religious alignment, p. 162, they use a unipolar Conservative-versus-the rest approach. Crewe, Särlvik and Alt return to Alford's definition: see ‘Partisan Dealignment’, p. 173.Google Scholar

¹² The classic statement of the consensual contact model is Berelson, Bernard, Lazarsfeld, Paul F. and McPhee, William N., Voting (Chicago: University of Chicago Press, 1954).Google Scholar

¹³ Przeworski, A. and Soares, G. A. D., ‘Theories in Search of a Curve: A Contextual Interpretation of the Left Vote’, American Political Science Review, LXV (1971), 51–68.CrossRefGoogle Scholar

¹⁴ Goodman, L. A., ‘Some Alternatives to Ecological Correlation’, American Journal of Sociology, LXIV (1959), 610–25.CrossRefGoogle Scholar

¹⁵ Rose, Richard, Governing without Consensus: An Irish Perspective (London: Faber, 1971), p. 235.Google Scholar

¹⁶ I must thank Ian McAllister for the use of his Northern Ireland data and his expert categorization of the parties. His own analysis of this data set appears in ‘Social Influences on Voters and Non-voters: A Note on Two Northern Ireland Elections’, Political Studies, XXXIV (1976), pp. 462–8.Google Scholar

¹⁷ A pencil and paper simulation exercise will prove this point. Take a society consisting of three equally-sized constituencies. Let the proportions middle class be 50, 33 and 20 per cent, respectively. Let the levels of Conservative support in both classes be 60 per cent in the first constituency, 50 per cent in the second, 40 per cent in the third. There is no individual level component of class polarization but a census or an error-free sample survey would show a 5 per cent class effect.

¹⁸ Using the Constituency Volume of the 1966 Census for 1964–66–70 and of the 1971 Census for 1974.

¹⁹ Taylor, A. J. P., English History 1914–45 (London: Oxford University Press, 1965), p. 129.Google Scholar

²⁰ For a technical description of ridge regression see Hoerl, A. E. and Kennard, R. W., ‘Ridge Regression – Biased Estimation for Nonorthogonal Problems’, Technometrics, XII (1970), 55–68.CrossRefGoogle Scholar

²¹ An introduction to trichotomous techniques is given in Miller, W. L., ‘Symmetric Representation of Political Trends in Three-party Systems with Some Properties, Extensions and Examples’, Quality and Quantity, XI (1977), 27–41Google Scholar. Trichotomous regression is described in Miller, W. L., ‘Trichotomous Regression Methods for Causal Modelling in Three-party Systems with Examples from Scotland, England and America’ (University of Strathclyde: CSPP paper no. 10, 1977).Google Scholar