¹ Niemi, Richard G., Whitten, Guy and Franklin, Mark N., ‘Constituency Characteristics, Individual Characteristics and Tactical Voting in the 1987 British General Election’, British Journal of Political Science, 22 (1992), 229–40, p. 229.CrossRefGoogle Scholar

² Heath, Anthony, Jowell, Roger, Curtice, John, Evans, Geoff, Field, Julia and Witherspoon, Sharon, Understanding Political Change: The British Voter 1964–1987 (Oxford: Pergamon Press, 1991), p. 54.Google Scholar

³ See Galbraith, John and Rae, Nicol, ‘A Test of the Importance of Tactical Voting: Great Britain, 1987’, British Journal of Political Science, 19 (1989), 126–36CrossRefGoogle Scholar; Johnston, R. J. and Pattie, C. J., ‘Tactical Voting in Britain in 1983 and 1987: An Alternative Approach’, British Journal of Political Science, 21 (1991), 95–128CrossRefGoogle Scholar; and Curtice, John and Steed, Michael, ‘Appendix 2’, in Butler, David and Kavanagh, Dennis, eds, The British General Election of 1987 (London: Macmillan, 1988).Google Scholar

⁴ Niemi, , Whitten, and Franklin, , ‘Constituency Characteristics’, pp. 230–1.Google Scholar

⁵ They also differ in that Heath et al. use 1983 constituency results rather than 1987 data used by Niemi, Whitten and Franklin. Given that at the time of voting the result of the 1987 election was both unknown and was presumably (according to Niemi, Whitten and Franklin's own argument) substantially influenced by tactical voting, it might have been better to use the 1983 election results, with adjustments for opinion poll estimates of trends, as a proxy for the informed tactical voters' decision-making context.

⁶ See, for example, Carmines, E. G. and Zeller, R. J., Reliability and Validity Assessment (Beverley Hills, Calif.: Sage, 1979)CrossRefGoogle Scholar; Bohrnstedt, G. W., ‘Measurement’, in Rossi, Peter, Wright, James and Anderson, Andy, eds, Handbook of Survey Research (New York: Academic Press, 1983)Google Scholar; Cronbach, L. J. and Meehl, P. E., ‘Construct Validity in Psychological Tests’, Psychological Bulletin, 52 (1955), 281–302CrossRefGoogle ScholarPubMed; Nunnally, J. C., Psychometric Theory (New York: McGraw-Hill, 1978).Google Scholar

⁷ Niemi, , Whitten, and Franklin, , ‘Constituency Characteristics’, pp. 231–2.Google Scholar

⁸ In their analysis Heath et al. find that among voters who have a clear party preference and whose party was in first or second place in the 1983 election, only 1.8 per cent report engaged in tactical voting. Whereas in more ‘appropriate’ circumstances – when their preferred party was in third place or lower and there was little difference between their first and second party preference – up to 35 per cent of voters did so. This suggests that Heath et al.'s measure of tactical voting and their analysis of the conditions under which such voting takes place have considerably better discriminating powers than the measures and conditions identified in Niemi, Whitten and Franklin's article. See Heath, et al. , Understanding Political Change, p. 57.Google Scholar

⁹ To a large degree we were able to replicate exactly Niemi, Whitten and Franklin's findings. However, there were certain differences between the figures presented by Niemi, Whitten and Franklin and our own estimates. In particular, we find that the tactical voting measures are slightly more strongly related to DCON than are the same measures calculated by Niemi, Whitten and Franklin. Why this is so is uncertain. However, it may be related to peculiarities in Niemi, Whitten and Franklin's estimates of the number of respondents with party preferences. Of the 3,295 respondents included in their Table 4, no less than eighty-nine refused to say which party they had voted for; two did not know which party they had voted for; another sixty-four said they had no preferred party; and one respondent, despite expressing a tactical motivation for her vote, was unable to recall the name of her preferred party. These missing data reduced the N to 3,142. How Niemi, Whitten and Franklin were able to obtain data on preferred party for all 3,295 respondents in their analysis is therefore somewhat puzzling.

¹⁰ The choice of ‘age at which respondents completed their full-time education’ as the measure of educational attainment is solely for comparison with Niemi, Whitten and Franklin's analysis. In Britain, if not in the United States, age at which full-time education was completed is a poor indicator of educational attainment as it is linked to changes in the minimum school leaving age and confuses policy changes with individual characteristics. Also, on p. 236 Niemi, Whitten and Franklin make the mistake of referring to university-educated respondents when discussing the figures in Table 4. However, even among respondents who completed their full-time education at age 17 or later, only 24 per cent attained degree level qualifications, and this includes those who obtained degrees from polytechnics and colleges of higher education. Even if we assume that a few respondents failed to complete courses and therefore did not receive degrees, it is clear that staying on beyond age 16 is not a suitable measure of higher educational experience.

¹¹ See Piazza, T., ‘The Analysis of Attitude Items’, American Journal of Sociology, 86 (1980), 584–603CrossRefGoogle Scholar, for a general analysis of such problems in multi-item indexes. For a recent exposition of the substantive inaccuracies resulting from inappropriate scaling procedures in the analysis of political attitudes in Britain, see Evans, G. A., ‘Is Britain a Class-Divided Society? A Re-analysis and Extension of Marshall et al.'s Study of Class Consciousness’, Sociology, 26 (1992), 233–58.CrossRefGoogle Scholar