Job Market Paper

"Failure to Launch: The impact of major party candidate dropout"

Abstract: Prominent partisan divides may inflate support for major party candidates, reinforcing their dominance. To test this, I exploit quasi-exogenous variation in the presence of one of two major political parties, arising from an unanticipated failure by the party to register candidates in some local government elections in New South Wales, Australia. I combine a difference-in-differences approach with detailed ballot-level ranking data. With this, I create a benchmark that separates out the mechanical substitution effect that arises from candidate absence even when voters rank candidates the same way. I find that the remaining major party's vote share falls significantly short of this benchmark, by 7.1 percentage points of enrolled voters. This shortfall is equivalent to 14.2 seats' worth of votes across all affected contests in which the party stood. Analysis of a corpus of candidate statements taken from online news articles is consistent with the framing of the contest changing in the absence of one major party—away from references to candidates' own parties, and towards communal moral language. These results suggest that major parties benefit from the partisan framing that their rivalry creates.