¹ By the ITN/Harris Exit Poll; cited in Fishman, Nina and Shaw, Andrew, ‘TV87: The Campaign to Make Tactical Voting Make Votes Count’, in Crewe, Ivor and Harrop, Martin, eds, Political Communications: The General Election Campaign of 1987 (Cambridge: Cambridge University Press, 1989), p. 299.Google Scholar

² 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), 95–128CrossRefGoogle Scholar; Galbraith, John W. and Rae, Nicol C., ‘A Test of the Importance of Tactical Voting: Great Britain, 1987’, British Journal of Political Science, 19 (1989), 126–36CrossRefGoogle Scholar; Curtice, John and Steed, Michael, ‘Appendix 2’Google Scholar, in Butler, and Kavanagh, , The British General Election of 1987.Google Scholar

³ Heath, Anthony, Curtice, John, Jowell, Roger et al. , Understanding Political Change: The British Voter 1964–1987 (Oxford: Pergamon Press, 1991), p. 54.Google Scholar

⁴ Heath, et al. , Understanding Political Change, p. 60.Google Scholar

⁵ Catt, Helena, ‘Tactical Voting in Britain’, Parliamentary Affairs, 42 (1989), p. 551.Google Scholar

⁶ These are fourteen individuals who claimed to have voted for one of the three reasons indicated in the text but who in response to a follow-up question about their preferred party named the party they said they voted for.

⁷ All of our analysis is based on voters only. Eighty-six per cent of the respondents claimed to have voted; the actual turnout rate was 75.3 per cent.

⁸ Heath et al. give the figure 6.0 per cent in their text (p. 53) and cite a figure of 6.3 per cent (note 2) as the basis for their later analyses.

⁹ Question 15B, codes 187, 287–8, 387, 487–8, 587 and 687–8. We did not differentiate between the first to eleventh reasons cited.

¹⁰ It is well known from studies in psychology that individuals are not able fully to explain their own behaviour or articulate their own thinking. An example from election studies in which people do not explain themselves fully is the failure of many American respondents to use liberal/conservative terminology in response to open-ended questions about the parties and candidates when it is clear from later questions that they recognize and understand such ideological descriptions and can correctly apply them to the parties. See Converse, Philip E., ‘The Nature of Belief Systems in Mass Publics’, in Apter, David E., ed., Ideology and Discontent (New York: Free Press, 1964).Google Scholar

¹¹ In contrast, Heath, et al. (p. 54)Google Scholar argue that responses to the second question merely reinforce the results from the first item by showing that a similar proportion cite tactical motivations even when allowed to explain their vote in their own words.

¹² ‘You say you have voted (Conservative, Labour, Liberal/SDP/Alliance) in a general election since 1964. Why do you not support them nowadays?’

¹³ Another important question is the effect of tactical voting on election outcomes. Although we have no independent estimate, we see no reason to question Curtice and Steed's estimate of perhaps one Labour and seven Alliance and Nationalist victories being partly attributable to tactical voting. We would, however, point out that this is no small thing in proportion to the total number of Alliance and nationalist victories in 1987. See Curtice, and Steed, , ‘Appendix 2’, p. 340.Google Scholar

¹⁴ Tactical voting is usually thought to be limited to individuals whose most preferred candidate is in third place or lower. In developing expectations relating to constituency characteristics, we shall assume that to be the case.

¹⁵ Ideally one would use individual pre-election expectations or poll results in individual con stituencies. Not having such expectations or results, we shall follow the lead of Cain in ‘Strategic Voting in Britain’, and of Black, Jerome, ‘The Multicandidate Calculus of Voting: Application to Canadian Federal Elections’, American Journal of Political Science, 22 (1978), 609–38CrossRefGoogle Scholar. These authors use actual election results as a proxy. Cain's (p. 647) perspective on this point is worth quoting: ‘proxies for [the expected closeness of the race] are constructed from the actual constituency results on the assumption that they capture with some random measurement error people's expec tations of the eventual outcome’. The alternative is to use constituency outcomes from the most recent past election. In this case that would mean 1983; not only would people's recollections have dimmed, but the objective situation – even the number of candidates itself – would have changed in some constituencies. In any event, we tried that variant and the results were little different from those shown below in that DCON is strongly related to tactical voting.

¹⁶ Technically, of course, the incentive to vote tactically depends on the utility differentials (relative strengths of preference) for the various candidates. Utilities are sometimes estimated directly from thermometer ratings, which are not available in the 1987 British study; see Cain, , ‘Strategic Voting in Britain’Google Scholar; Abramson, Paul R., Aldrich, John, Paolino, Phil and Rohde, David W., ‘“Sophisticated” Voting in the 1988 Presidential Primaries’, American Political Science Review, 86 (1992), forthcomingCrossRefGoogle Scholar. Later we shall use strength of party identification as a partial indicator of strength of preferences.

¹⁷ Heath, et al. , Understanding Political Change, p. 57Google Scholar, also make the point that tactical voting varies according to circumstances, and they show that 35 per cent of one group of voters claim to have voted tactically. However, their primary emphasis is on the low level of tactical voting, even under the most compelling circumstances.

¹⁸ Riker, William H. and Ordeshook, Peter C., ‘A Theory of the Calculus of Voting’, American Political Science Review, 62 (1968), 25–42.CrossRefGoogle Scholar

¹⁹ Felsenthal and Brichta found that university-educated respondents in Israel claimed to have voted tactically (in some past election) much more frequently than those with only a high school education. See Felsenthal, Dan S. and Brichta, Avraham, ‘Sincere and Strategic Voters: An Israeli Study’, Political Behavior, 1 (1985), 311–24.CrossRefGoogle Scholar

²⁰ Liking or disliking the parties is given in variables 13A–F.

²¹ If we define the dependent variable by responses to the first three items in Table 1, the coefficients differ little; tactical voting is almost equally frequent under the circumstances described below.

²² Mean values are: DCON = 2.3; COMP = 8.7; strength of partisanship = 2.1; education = 2.6; recalled winner = 0.8; negative towards winning party = 2.7.

²³ Lanoue and Bowler included negative feelings about the Prime Minister, Margaret Thatcher, and the Labour leader, Neil Kinnock, as well as a measure of negative feelings for the party winning the constituency. Their results suggest that it may have been anti-Thatcher feelings that prompted tactical votes. See Lanoue, David J. and Bowler, Shaun, ‘The Sources of Tactical Voting in British Parliamentary Elections, 1983–1987’ (unpublished paper, University of California, Riverside).Google Scholar

²⁴ We say that pressure groups ‘apparently’ got the message across because we lack over-time data; we cannot confidently say that tactical voting increased over prior years. Note also that we take no issue with the argument of Heath, et al. , Understanding Political Change, p. 54Google Scholar, that tactical voting was not all directed against one party.

²⁵ It may be, for example, that some respondents refer to tactical behaviour when they engage in what most political scientists call bandwagon voting.

²⁶ In addition to studies cited earlier, see Fisher, S. L., ‘The Wasted Vote Thesis: West German Evidence’, Comparative Politics, 5 (1973), 293–9CrossRefGoogle Scholar; Bartels, Larry, Presidential Primaries and the Dynamics of Public Choice (Princeton, NJ: Princeton University Press, 1988)Google Scholar; Biais, André, Johnston, Richard, Brady, Henry E. and Crête, Jean, ‘The Dynamics of Horse Race Expectations in the 1988 Canadian Election’ (paper presented at the Annual Meeting of the Canadian Political Science Association, Victoria, BC, 1990)Google Scholar. Tactical voting was also found in a small experimental study of approval voting; see Niemi, Richard G. and Bartels, Larry, ‘The Responsiveness of Approval Voting to Political Circumstances’, PS, 17 (1984), 571–7.CrossRefGoogle Scholar