¹ See, for example, Curtice, John and Steed, Michael, ‘Electoral Choice and the Production of Government: The Changing Operation of the Electoral System in the United Kingdom since 1955’, British Journal of Political Science, 12 (1982), 249–98CrossRefGoogle Scholar; Heath, Anthony, Jowell, Roger and Curtice, John, How Britain Votes (Oxford: Pergamon, 1985)Google Scholar; Johnston, R. J., Pattie, C. J. and Allsopp, J. G., A Nation Dividing? The Electoral Map of Great Britain 1979–87 (Harlow, Essex: Longman, 1988)Google Scholar; Denver, David and Halfacree, Keith, ‘Inter-Constituency Migration and General Election Results in Britain’ (North West Regional Research Laboratory Working Paper, University of Lancaster, 1991).Google Scholar

² See Heath, Anthony et al. , Understanding Political Change: The British Voter 1964–87 (Oxford: Pergamon, 1991), chap. 7.Google Scholar

³ Campbell, Angus, Converse, Philip E., Miller, Warren E. and Stokes, Donald, The American Voter (New York: Wiley, 1960), chap. 16, pp. 448–9.Google Scholar

⁴ Converse, Philip E., ‘On the Possibility of a Major Realignment in the South’, in Campbell, Angus, Converse, Philip E., Miller, Warren E. and Stokes, Donald, Elections and the Political Order (New York: Wiley 1966).Google Scholar

⁵ Converse, , ‘On the Possibility of a Major Realignment in the South’, p. 227.Google Scholar

⁶ Brown, Thad A., ‘On Contextual Change and Partisan Attributes’, British Journal of Political Science, 11 (1981), 427–47CrossRefGoogle Scholar; see also Brown, Thad A., Migration and Politics (Chapel Hill: University of North Carolina Press, 1988).Google Scholar

⁷ For further details, see Heath, et al. , Understanding Political Change, pp. 230–3.Google Scholar

⁸ A separate document giving further details of Super Profiles is available from Stan Openshaw, Department of Geography, University of Newcastle upon Tyne.

⁹ In the analysis of intra-generational migration we exclude respondents who were not eligible to vote in 1979, and we thus avoid the overlap that would otherwise occur between inter- and intra-generational migration.

¹⁰ For some evidence counter to the usual assumption that age reduces political learning see Black, Jerome, Niemi, Richard and Powell, G. Bingham, ‘Age, Resistance and Political Learning in a New Environment: The Case of Canadian Immigrants’, Comparative Politics, 20 (1987), 73–84.CrossRefGoogle Scholar

¹¹ The 1964 and 1966 election studies distinguished these six regions only. We can thus compareour results over time only if we use the collapsed version. In 1987 the six regions were constructed from the eleven regions of the standard official classification as follows: the North, North-West and Yorkshire and Humberside were merged to form the North region; West Midlands, East Midlants and East Anglia were merged to form the Midlands; and London and the South-East were merged to form the South-East. The South-West, Wales and Scotland were left as they are in the official classification.

¹² There were also 170 respondents who were migrants to Britain from overseas, including Ireland. These have been excluded from all the analyses reported in this article.

¹³ There is to our knowledge only one other study of inter-generational regional migration, namely Coté, Guy, ‘Moving On: Area, Migration and Socio-Economic Attainment in Sociological Perspective’ (doctoral dissertation University of Oxford, 1983)Google Scholar. Coté's study is based on the 1972 Oxford mobility study and gives broadly comparable results to ours.

¹⁴ Classes have been defined according to the procedures described in Heath, et al. , Understanding Political Change, p. 66.Google Scholar The four educational levels which we distinguish here are: (1) Degree, (2) ‘A’ level (including higher school certificate), (3) ‘O’ level or CSE (including school certificate), (4) None of the above. We exclude the various technical qualifications which our respondents had received, as these are often obtained in the course of an individual's working career and cannot therefore be regarded as antecedents.

¹⁵ Details of the multivariate analysis are available on request.

¹⁶ Whether one includes the Midlands with the north or south is a matter of some dispute. We here follow the practice of Curtice and Steed, ‘Electoral Choice and the Production of Government’, who used the differential patterns of regional swing since 1955 as their criterion.

¹⁷ Fienberg, S. E., The Analysis of Cross-Classified Categorical Data, 2nd edn (Boston, Mass: MIT Press, 1987).Google Scholar For an introduction to logit models see Aldrich, J. H. and Nelson, F. D., Linear Probability, Logit and Probit Models (Beverly Hills, Calif.: Sage, 1984).CrossRefGoogle Scholar We chose a logit model rather than the more familiar linear regression since we have a dichotomous dependent variable.

¹⁸ If we exponentiate the parameters we obtain the fitted odds ratios. Thus the parameter of 0.86 for migration tells us that, controlling for the other variables in the model, the predicted odds that a migrant will not vote Labour are 2.36 times the predicted odds that a non-migrant will not vote Labour. That is to say, the predicted migrant:nonmigrant odds ratio is 2.36.

¹⁹ There is a great deal of research on patterns of intra-generational regional migration. See, for example, Ogilvy, A. A., ‘Population Migration Between the Regions of Great Britain, 1971–9’, Regional Studies, 16 (1982), 65–73CrossRefGoogle ScholarPubMed; Fielding, A. J., ‘Inter-Regional Migration and Social Change: A Study of South East England Based upon Data from the Longitudinal Study’, Transactions of the Institute of British Geographers, NS 14 (1989), 24–36CrossRefGoogle Scholar; Stillwell, John, Boden, Peter and Rees, Philip, ‘Trends in Internal Net Migration in the UK: 1975–1986’, Area, 22 (1990), 57–65Google Scholar; and the annual reports published in Population Trends. These sources generally show slightly higher levels of internal regional migration than does our survey. The difference can probably be explained, firstly, by the fact that these sources report annual rates of movement (some annual movements cancelling out over our eight-year period) and, secondly, by our use of a smaller number of regions than the eleven official ones.

²⁰ See van der Eijk, C. and Niemoller, B., ‘Recall Accuracy and its Determinants’, Acta Politica, 14 (1979), 289–342.Google Scholar

²¹ For an account of the index of concentration see Haberman, S. J., ‘Analysis of Dispersion of Multinomial Responses’, Journal of the American Statistical Association, 77 (1982), 568–80.CrossRefGoogle Scholar

²² These figures are calculated from the parameter estimates in Table 9. That is to say, they are the fitted probabilities obtained from our model.

²³ On social and geographical mobility, see Bell, Colin, Middle Class Families: Social and Geographical Mobility (London: Routledge and Kegan Paul, 1969)Google Scholar; Lansing, John B. and Mueller, Eve, ‘The Geographic Mobility of Labor’ (Ann Arbor Survey Research Centre, 1967)Google Scholar; Fielding, A. J., ‘A Search for the ‘Missing Link’ between Social and Geographical Mobility’, Revue de Geographie de Lyon, 65 (1990), 24–36CrossRefGoogle Scholar; Savage, Mike, ‘The Missing Link: The Relationship between Social and Geographical Mobility’, British Journal of Sociology, 39 (1988), 554–77CrossRefGoogle Scholar; Cote, , ‘Moving On’.Google Scholar

²⁴ See also Kelley, Jonathan and McAllister, Ian, ‘Social Context and Electoral Behavior in Britain’, American Journal of Political Science, 29 (1985), 564–86.CrossRefGoogle Scholar

²⁵ Hughes, G. A. and McCormick, B., ‘Migration Intentions in the UK. Which Households Want to Migrate and Which Succeed?’, Economic Journal, Supplement, 95 (1985), 113–23.CrossRefGoogle Scholar

²⁶ One further control which we have included is age. Older people tend to be more Conservative and the migrants in our sample are also somewhat older (not because people migrate when they are old but because the old have had more opportunity to migrate). When age is added to the analysis, it has the expected significant relationship with vote, but again fails to reduce the migration parameter to non-significant levels.

²⁷ A further refinement is to include an interaction term in order to take account of possible asymmetrical effects between upward and downward mobility. See Clifford, P. and Heath, A. F., ‘The Political Consequences of Social Mobility’, Journal of the Royal Statistical Society, series A, forthcoming.Google Scholar

²⁸ See Brown, , Migration and Politics, pp. 88–104.Google Scholar

²⁹ In defining local environments for this purpose we use a ten-fold classification in place of the two-fold one used before.

³⁰ See Galbraith, J. W. and Rae, N. C., ‘A Test of the Importance of Tactical Voting: Great Britain 1987’, British Journal of Political Science, 19 (1989), 126–136CrossRefGoogle Scholar; Heath, et al. , Understanding Political Change, chap. 4Google Scholar; Johnston, R. J. and Pattie, C. J., ‘Tactical Voting in Great Britain in 1983 and 1987: An Alternative Approach’, British Journal of Political Science, 21 (1991), 95–108.CrossRefGoogle Scholar

³¹ Denver, and Halfacree, , ‘Inter-Constituency Migration and General Election Results in Britain’, p. 12.Google Scholar

³² We reached this conclusion by the following method. First we calculated the distribution of the vote in 1979 (using the recall measure provided by our respondents in 1987) in the north and south respectively. This gave figures of 37.8 per cent Conservative in the north and 54.3 per cent Conservative in the south and an index of dissimilarity of 16.5 points. We next made a hypothetical calculation of what the distribution of the vote would have been if migrants were not subject to political conversion. To do this we simply deducted the 1979 votes of the north-south (intra-generational) migrants from the northern total and added them to the southern total (and vice versa in the case of the south-north migrants). This gave us hypothetical figures of 36.6 per cent Conservative in the north and 54.4 per cent Conservative in the south, an index of dissimilarity of 17.8. We have to use recall measures of 1979 vote for this exercise since our migrants' 1987 votes will have been subject to political conversion.

³³ Our estimate of a twelve-point change is based on the 1979 and 1987 British Election Surveys, not on recall measures. In 1979 the index of dissimilarity between north and south was 13.3 and in 1987 it was 25.7.

³⁴ See, for example, Kinder, Donald R. and Kiewiet, D. Roderick, ‘Sociotropic Politics: The American Case’, British Journal of Political Science, 11 (1981), 129–61.CrossRefGoogle Scholar For a full discussion of the north-south divide itself see Johnston, , Pattie, and Allsopp, , A Nation Dividing?Google Scholar