Shifty Characters: Ideological Shifting and Electoral Outcomes in U.S. Senate Elections

Abstract

Very little research has investigated how a two-stage electoral process (a primary election to nominate the party candidate, and a general election where the parties' candidates face off) affects candidate behavior. Here I argue that candidates are attracted to the median voter position of the electorate in which they are running. And that differences between the ideological positions of the primary median voter and the general election median voter means that candidates have incentive to shift their ideological positions to align with the relevant median voter. I test for the presence of candidate ideological shifting in U.S. Senate elections. I measure candidate ideological shifting by recovering two ideological positions for each candidate (one during the primary campaign and another during the general election campaign) utilizing donations from political action committees. Using these measures I find that candidates do engage in ideological shifting. Candidates align closely with the primary median voter during the primary campaign, and then shift toward the middle to moderate their ideological position to align more closely with the general election median voter. My results also indicate that ideological movement has electoral implications. Candidates get punished in the general election for their primary ideological positions. If candidates were extreme during the primary campaign they received fewer general election votes than candidates whose primary ideologies were moderate.