Waste Not, Vote Not (for Third Parties)?

Abstract

I develop and demonstrate evidence for a theory of strategic, state-based voting in American presidential elections. Such voting occurs when individuals anticipate the behavior of their neighbors and cater their voting decisions accordingly. I turn to an infrequently-studied phenomenon to uncover evidence for this theory: support for minor-party presidential candidates. The Electoral College in the United States determines the victors of presidential elections. Since most states allocate their electors on a winner-takes-all basis, and because voting outcomes in some states are more unpredictable than in others, presidential candidates tend to cluster their campaign resources in competitive or “battleground” states. Voters who reside in these states live in different campaign environments than voters who reside in uncompetitive, “blackout” states. I suggest that voters in competitive states anticipate the decisions of their fellow residents and perceive that their votes have higher pivot values than votes in other states. The perceived value of votes is expected to vary across states due to differences in electoral competitiveness and the winner-takes-all nature of awarding Electoral College votes. As a result, I propose that voters in competitive states are more likely to vote strategically than voters in other states. To test my hypotheses, I turn to the election returns of third-party candidates in two recent presidential elections. I find that support for minor parties is positively associated with wider two-party margins of victory, after adjusting for ballot access and campaign resources.