The Impact of Party Leaders in Britain: Strong Assumptions, Weak Evidence

There has been a widespread belief in Britain in recent years (which appears to be corroborated by survey evidence) that leaders? personalities strongly influence the way that people vote and therefore the results of elections, although in a parliamentary system such as Britain?s, this is less obviously correct than in presidential systems such as the United States, France and Russia. However, British political science has usually reached negative conclusions on this belief, pointing out that popular leaders have often lost elections, and that only two of the eleven elections fought since 1964 (those of 1964 and 1997) stand out as elections in which leaders? personalities may have been decisive. This chapter looks at the 1997 election, for which three separate studies belonging to the British Election Study are available: the cross?sectional study, XBES; the campaign study, BECS; and the British Election Panel Study, BEPS. It first discusses the effect of leaders? personalities by constructing a series of vote models, and then looks at the relationship between leaders? personalities and the vote. Each of the three British Election Studies are then analysed; only the BEPS study (which has the least comprehensive data set) indicates that leaders? personalities had anything other than a marginal effect on the election outcome.