Primary Concerns: The Lack of Forward-Looking Strategic Voting in Primary Elections

A few districts and states with competitive general elections now decide control ofcongressional chambers. In these closely contested generals, the identities of parties’nominees can win or lose the election. As such, we must think about elections in thesedistricts as a single, two-stage game beginning with the primary. Yet we do not knowwhether primary voters think about primary elections as a first stage that affects thesecond stage. Do primary voters vote strategically to give their party the best chanceof winning the general election? I use six experiments to investigate the willingnessof primary voters in both parties to vote for the most “electable” candidate. I findthat, though primary voters increase support for candidates they are told have the bestchance of winning the general election, they do not vote strategically. Instead, the entirepositive effect of treatment is derived from voters that have their preconceptions aboutelectability confirmed. When treatment confirms a voter’s perceptions about primarycandidates’ general election chances, treatment increases support for the more electablecandidate enormously. When treatment presents counter-attitudinal information aboutelectability, primary voters are less likely to support the more electable candidate.That primary voters do not vote strategically runs counter to work on strategic votingin general elections, and suggests novel limits on party elites’ ability to control theirchances in general elections and the influence of partisan concerns on voter behavior.Future research will explore why primary voters do not switch support to the moreelectable candidate: are primary voters spurred to change their preferred candidate’schances or are they motivated reasoners who reject counter-attitudinal information?