Hostname: page-component-586b7cd67f-r5fsc Total loading time: 0 Render date: 2024-12-03T19:20:47.620Z Has data issue: false hasContentIssue false
  • English
  • Français

People Who Live in Glass Houses: A Response to Evans and Heath's Critique of our Note on Tactical Voting

Published online by Cambridge University Press:  27 January 2009

Richard Niemi ,
Guy Whitten  and
Mark Franklin
Get access Rights & Permissions [Opens in a new window]

Extract

In a recent Research Note, Evans and Heath cast doubt on the validity of our measure of tactical voting. We and our critics agree on a set of ‘Main reason’ tactical voters – those who said they voted tactically in response to the question, ‘What comes closest to the main reason you voted the way you did?’ Unlike our critics, we reasoned that many of those who cast a tactical vote would not be identified so simply. Some respondents may have been reluctant to say they had voted for other than the best party. Others may have convinced themselves that the ‘best’ party was one that had a chance of winning and therefore answered in such a way as to place themselves squarely among non-tactical voters even though in other tactical circumstances they would have voted differently. Consequently, we sought to identify tactical behaviour through additional questions. Our final measure of tactical voting can be regarded as containing two components: the ‘main reason’ voters, who expressly told interviewers they voted for one party while ‘really preferring’ some other party, and ‘Other reason’ voters, whom we identified as tactical by their responses to additional questions. It is the ‘Other reason’ voters who are at issue.

Type
Notes and Comments
Information
British Journal of Political Science , Volume 23 , Issue 4 , October 1993 , pp. 549 - 553
Copyright
Copyright © Cambridge University Press 1993

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Evans, Geoffrey and Heath, Anthony, ‘A Tactical Error in the Analysis of Tactical Voting: A Response to Niemi, Whitten and Franklin’, British Journal of Political Science, 23 (1993), 131–7CrossRefGoogle Scholar; Niemi, Richard G., Whitten, Guy and Franklin, Mark, ‘Constituency Characteristics, Individual Characteristics and Tactical Voting in the 1987 British General Election’, British Journal of Political Science, 22 (1992), 229–54.CrossRefGoogle Scholar

2 Distance from contention is the difference between the percentage vote for the respondent's most preferred candidate in his or her constituency and the percentage vote for the lower of the two top contenders. The distance from contention measure scored zero if the respondent's most preferred candidate was in first or second place. Evans and Heath criticize our use of 1987 rather than 1983 voting statistics for the purpose of defining this variable. This is an important point, and we dealt with it fully in fn. 15 of our original presentation.

3 Evans, and Heath, , ‘A Tactical Error’, pp. 131–2Google Scholar, also argue that some of the responses we categorized as indicative of tactical thinking are only implicitly so at best. We agree that the responses are not as clear-cut as one would like, but we think it is at least as misleading simply to ignore the respondents' suggestions of tactical motivations.

4 That most of the ‘Other reason’ voters are 0 on distance from contention is obvious from Table 2 in our original article. For each row, one can calculate the number of respondents represented by the difference between the second column (‘Main reason’ tactical voters) and the third column (‘Any reason’ tactical voters). These are the ‘Other reason’ voters. That is, 10.3 per cent of the 2,668 respondents scored 0 equals 275 respondents; other rows yield 22, 13, 13, 9 and 7 respondents, for a total of 339 ‘Other reason’ voters, with 275/339 = 81 per cent in row 0.

5 It is difficult, even with direct questions, to identify voters' true preferences. Many people, for example, probably do not fully evaluate an alternative they consider non-feasible, even if their issue preferences, social characteristics or some initial, limited knowledge suggest that they would prefer that alternative. None the less, we think it likely that some voters who did not say outright that they had voted tactically did, in some sense, prefer another party.

6 Answers to questions 15B1 to 15B11 were coded as follows: pro-Conservatives are 286 and 288; anti-Conservatives are 287 and 789; pro-Labour are 486 and 488; anti-Labour are 487 and 489; pro-Alliance are 686 and 688; anti-Alliance are 687 and 689. Answers to the question series 17 and 18 take into consideration which party the respondent did not vote for. With regard to ‘Why not Conservatives?’ pro-Conservative answers are 286, 386, 586, 387, 487, 587, 687, 288, 489 and 689; anti-Conservative answers are 287 and 289. With regard to ‘Why not Labour?’, pro-Labour answers are 186, 486, 586, 187, 587, 488, 289 and 689; anti-Labour answers are 489. With regard to ‘Why not Alliance?’ pro-Alliance answers are 186, 386, 686, 187, 287, 387, 487, 688, 289,489; anti-Alliance answers are 687 and 689.

7 In a few cases, respondents offered negative mentions of B but no positive mentions of A, having voted for C. We did not impute a preference on the basis of these exclusively negative comments.

8 We continue to use the party voted for in the case of the respondents for whom we could not reasonably impute a preference. We also correct a small error that Evans and Heath (fn. 9) spotted. We inadvertently coded eighty-nine respondents as preferring a party in first or second place when in fact they had refused to indicate the party they voted for; likewise for two others who could not remember the party they voted for. Eliminating these respondents from Table 2 does not affect our overall estimate of the percentage of voters who behaved tactically. In the present Note we have also gone along with Evans and Heath and deleted sixty-four respondents who at one point said there was no other party they would have preferred to vote for.

9 Using Table 2 in this Note and applying the same sort of calculations as in fn. 4 above yields 201 ‘Main reason’ voters, with 101/201 = 50.2 per cent in row 0.