The Role of Traits in Voter Choice: Evidence from U.S. Blanket Primary Reforms

• In Event: Emerging Scholars in Political Economy

Abstract

In 2008, Washington adopted a top-two primary system to replace the blanket primary. Blanket primary reforms in Washington and other states have changed political competition, allowing the possibility that one of the major parties won’t be represented in all competitive races on the general election ballot. When co-partisans face off in the general election, how do voters respond? Do out-party voters behave differently than same-party voters? And what candidate characteristics matter the most when both candidates are from the same party?

Answering these questions poses three methodological challenges. First, observing aggregate voting outcomes but not individual choice creates an ecological inference problem. Second, parties behave strategically and may alter their support for candidates in districts that may be prone to co-partisan run-offs. Potential candidates observe the change in party strategy and may reconsider their choice to emerge as a candidate. Finally, I need to be able to discern whether changes to turnout were a consequence of election reform or because of some other unobserved or unmodeled process.

To address these methodological concerns, I use a difference-in-differences research design to identify causal effects on turnout from Washington’s adoption of the top-two primary. In order to better estimate counterfactuals to get at causal effects, I then use a structural model to estimate a system of demand for candidate traits. Results show how the importance of traits vary considerably among subsets of the electorate and provide a rubric for identifying the determinants of choice in other U.S. elections.

Author

• T. Ryan Johnson, Washington University in St. Louis