Hostname: page-component-586b7cd67f-dlnhk Total loading time: 0 Render date: 2024-11-23T20:46:41.347Z Has data issue: false hasContentIssue false
  • English
  • Français

Reconstructing Past Partisanship in Britain

Published online by Cambridge University Press:  27 January 2009

Richard S. Katz ,
Richard G. Niemi  and
David Newman
Get access Rights & Permissions [Opens in a new window]

Extract

Most of the variables in survey-based studies of electoral behaviour are measured only at second hand. Although sex or race can be coded from direct observation, and whether an individual voted can often be ascertained from documentary evidence, researchers must usually rely on respondents' accuracy and truthfulness in reporting attitudes and in recalling past behaviour. Both of these may involve significant biases. Research has shown that many events are underreported by survey respondents, although clearly desirable activities like voting may be overreported. Other work has suggested the likelihood that ‘non-attitudes’ will be reported by many respondents. When these two are compounded, so that respondents are asked to recall previous attitudes or dispositions, the danger of misrepresentation is multiplied.

Type
Notes and Comments
Information
British Journal of Political Science , Volume 10 , Issue 4 , October 1980 , pp. 505 - 515
Copyright
Copyright © Cambridge University Press 1980

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 Clausen, Aage, ‘Response Validity: Vote Report’, Public Opinion Quarterly, xxxii (1969), 588–606Google Scholar; Traugott, Michael W. and Katosh, John P., ‘Response Validity in Surveys of Voting behavior’, Public Opinion Quarterly, XLIII (1979), 359–77.CrossRefGoogle Scholar

2 Converse, Philip E., ‘The Nature of Belief Systems in Mass Publics’, in Apter, David, ed., ideology and Discontent (New York: Free Press, 1964).Google Scholar

3 Campbell, Angus, Converse, Philip E., Miller, Warren E. and Stokes, Donald E., The American Voter (New York: Wiley, 1960)Google Scholar; Butler, David and Stokes, Donald E., Political Change in Britain (London: Macmillan, 1974)CrossRefGoogle Scholar; Andersen, Kristi, The Creation of a Democratic Majority (Chicago: University of Chicago Press, 1979).Google Scholar

4 Niemi, Richard G., Katz, Richard S. and Newman, David, ‘Reconstructing Part Partisanship: the Failure of the Party Identification Recall Questions’, American Journal of Political Science, xxiv (1980), 633–5.CrossRefGoogle Scholar

5 These data were supplied by the Inter-University Consortium for Political and Social Research. Neither the Consortium nor the original investigators bear any responsibility for the analysis or interpretations presented here.

6 Throughout the paper we use the 718 respondents interviewed in all four waves of the panel – 1963, 1964, 1966 and 1970. (This is about 39 per cent of the original sample after an allowance is made for deaths.) The sample is weighted by sex, level of political knowledge, and number of political attitudes in order to correct for differential survival rates among categories of these variables. The weighted N used here, before the elimination of Independents for the reasons noted below, is 936·1.

7 For example, Butler, and Stokes, , Political Change in BritainGoogle Scholar; Crewe, Ivor, Särlvik, Bo, and Alt, James, ‘Partisan Dealignment in Britain 1964–1974’, British Journal of Political Science, vii (1977), 129–90.CrossRefGoogle Scholar

8 Partisans were asked, for example, ‘Was there ever a time when you thought of yourself as Labour or Liberal rather than Conservative?’. Leaners were asked ‘Was there ever a time when you felt closer to Labour or the Liberals rather than to the Conservatives?’. Independents were asked if they had ever felt closer to one of the parties, but, as noted above, they have been eliminated from this analysis.

9 Considerable care had to be exercised for those who changed during the panel period. For example, if a respondent who reported in 1970 that she had changed in, say, 1967, her 1966 report could be different from, and yet be consistent with her 1970 report. A minor problem was the handling of ‘don't know’ responses and responses referring only to a personal event (e.g., ‘when I got married’). These were automatically considered consistent with ranges of years reported in other waves.

10 An additional cause of inconsistency, and of the inaccurate recollections discussed below, would be recording, transcription, and punching errors. An indication that these sources account for only a minor part of the error found here can be obtained from Schreiber, E. M., ‘Dirty Data in Britain and the USA: the Reliability of “Invariant” Characteristics Reported in Surveys’, Public Opinion Quarterly, IXL (1976), 493–506.Google Scholar Schreiber reports three-wave reliabilities of over 98 per cent for race and sex in Butler and Stokes's data. Question unreliability may account for an additional portion of the error, but it is difficult to estimate reliability figures. See Niemi, Katz and Newman, ‘Reconstructing Past Partisanship: The Failure of the Party Identification Recall Questions’.

11 Butler, and Stokes, , Political Change in Britain, p. 60.Google Scholar

12 We constructed tables exactly like that of Butler and Stokes, but using an interval of seven rather than thirteen years. One table was based on recalled changes in partisanship while the other was based on actual changes. The latter table showed smaller differences in the percentage of strong partisans between those with long-term rather than short-term party ties for two of the three age groups and over all. For the 40–59 age group, the difference was slightly larger based on actual rather than recalled change. Additionally, both of our tables show a strong relationship between age and strength of partisanship with duration controlled, flatly contradicting Butler and Stokes's results. Similar results are obtained with American data, thus calling into question the analogous table in Campbell, et al. , The American Voter, p. 164.Google Scholar

13 In fact, to check adequately the accuracy of recalled parental partisanship when the current respondents were young, we would need interviews conducted with parents at that time since the parents' recall cannot be relied upon.

14 ‘When you were young’ being a rather ambiguously specified time, some of the inconsistency in recall of parental partisanship may be the result of real change in the parent's preference. This underlines one danger of using filial recall data to reconstruct past parental partisanship. A second possibility is that respondents were inspired by the 1963 questions to ask their parents about their previous preferences. While this would account for the decline in the numbers reporting their parents to have had no party preference, this contamination of the sample is of little use to those working with single-wave studies; nor can it overcome failure of the parent to recall correctly. Furthermore, it does not account for the pattern of bias reported below.

15 Dreyer, Edward C., ‘Change and Stability in Party Identifications’, Journal of Politics, xxxv (1973). 713–22.Google Scholar

16 Reports of the contemporary political attitudes of parents and others are also biased towards respondents' own preferences. See Niemi, Richard G., How Family Members Perceive Each Other (New Haven, Conn.: Yale University Press, 1974), Chaps. 3, 4, 7.Google Scholar

17 Himmelweit, Hilde T., Biberian, Marianne Jaeger and Stockdale, Janet, ‘Memory for Past Vote: Implications of a Study of Bias in Recall’, British Journal of Political Science, viii (1978), 365–75.CrossRefGoogle Scholar