Strategic Voting in the 2010 UK Election

This paper analyzes strategic voting in the 2010 UK General Election using survey data from the 2010 British Election Study (BES). The paper consists of two parts. In the first part, we use survey responses to generate respondent-level preference orderings among the three major parties. Using these preference orderings, we simulate the 2010 election as if it were a national election using four different electoral systems: pairwise comparisons (in search of a Condorcet winner), the Borda count, the alternative vote, and Coombs’ method. We found that the Lib-Dems were the Condorcet winner and, as the compromise party, won under every tested method except the alternative vote (which they advocated). In the second part, we empirically test a rational-choice model of strategic voting. The model predicts that voters should take into accounted the expected utility of their vote and vote for their second-most preferred candidate if their first choice is clearly in third place in terms of likelihood of winning. The 2010 election offers a compelling case study because of the strong performance of the Lib-Dems, who traditionally finished a distant third. Our statistical analysis of the BES survey data provides support for our theoretical model of strategic voting.