The European Issue and Party Choice at British General Elections, 1974–2005

Abstract

Political scientists have long debated whether “issues” influence vote choices. It is now generally agreed, however, that policy preferences and evaluations of party performance have some limited impact on vote decisions. It has also been shown that the influence of issues varies both between individuals and between elections. This article adds to our understanding of these issues by examining how the European issue influenced individual vote decisions at each general election since 1974. We use Manifesto Research Group (MRG) data to show that the effect of Europe on individual vote decisions increases as the parties became more polarized and decreases as they converge on the European issue.

Acknowledgements

The authors would particularly like to thank Judith Bara for providing us with some of the data used in this study. Also, thanks to Kyriaki Nanou, Tom Quinn and the three anonymous reviewers for their helpful comments and suggestions. Ben Clements would like to acknowledge the assistance of the ESRC through Award No.: PTA‐030‐2004‐00228.

Notes

1. Butler and Stokes (1974: 290) combined the first two conditions: “A first condition is a straightforward extension of the need for an individual to perceive an issue and to form some attitude towards it if it is to influence his behaviour. For an issue to have much impact in the whole of the electorate the bond of issue to self must be formed in the minds of a substantial body of electors.”

2. It should finally be noted that Campbell et al. (1960) and Butler and Stokes (1974) also noted that if the issue was to affect election outcomes one further following condition had to be met: the distribution of preferences should be skewed in one “direction” so that they led many voters to switch from one party to another.

3. Nevertheless, psychological studies imply that people are influenced by subconscious factors of which they are entirely unaware (see Nisbett & Wilson, 1977).

4. Bold in the original BES question.

5. Wlezien (2005) demonstrates that the related “most important problem” item asked by Gallup appears to influence what is thought to be a problem rather than important. Nevertheless, many scholars use responses to this question as an indicator of importance (see Baumgartner & Jones, 1993; Jones & Baumgartner, 2005; Jennings & John, 2008).

6. We should note, however, that if policy preferences have some prior effect on partisanship then controlling for the latter will produce biased estimates of the impact of the former.

7. There is no measure of tactical voting on the BES until 1987. It is omitted for this reason.

8. For sake of completeness it is worth noting that the correlation between the measure of Conservative and Labour policy distance and MORI salience is 0.36 (N = 8, ns) and that between Conservative and Liberal policy distance and Mori salience is 0.38 (N = 8, ns).