Published online by Cambridge University Press: 23 June 2016
A large class of theoretical models posits that voters choose candidates on the basis of issue congruence, but convincing empirical tests of this key claim remain elusive. The most persistent difficulty is obtaining comparable spatial estimates for winning and losing candidates, as well as voters. We address these issues using candidate surveys to characterize the electoral platforms for winners and losers, and large issue batteries in 2008 and 2010 to estimate voter preferences. Questions that were answered by both candidates and citizens allow us to jointly scale these estimates. We find robust evidence that vote choice in congressional elections is both strongly associated with spatial proximity and that individual-level and contextual variables commonly associated with congressional voting behavior condition the magnitude of its importance. Our results have important implications for theories of voter decision-making and electoral institutions.
Boris Shor, Assistant Professor, Department of Political Science, The University of Houston, 3551 Cullen Boulevard Room 447, Houston, TX 77204-3011 (firstname.lastname@example.org). Jon C. Rogowski, Assistant Professor, Department of Political Science, Washington University in St. Louis, One Brookings Drive, Campus Box 1063, St. Louis, MO 63130 (email@example.com). The candidate survey data in this paper relies on the hard work of Chad Levinson, while roll call data was assembled in a major collaboration with Nolan McCarty, and assisted by Steven Rogers and Michelle Anderson. The authors thank Josh Clinton and Keith Krehbiel for comments on previous versions of this manuscript, and Andrew Gelman, David Park, Gerald Wright, Rob Van Houweling, Will Howell, Stephen Jessee, Michael Bailey, and seminar participants at Princeton, Stanford, UNC, Rochester, and Chicago for helpful discussions about earlier ideas. The authors thank Project Vote Smart for making their data available to them. The authors thank Adam Bonica, Garry Hollibaugh, Lawrence Rothenberg, and Kristin Rulison for providing their candidate data to the authors for comparison. The authors thank Simon Jackman for his invaluable software and assistance. The authors thank anonymous reviewers and to the editors and staff at PSRM for their many helpful suggestions. Funding for the 2008 survey was generously provided by the University of Chicago Harris School. The authors welcome comments and questions. Any errors are the authors own. To view supplementary material for this article, please visit http://dx.doi.org/10.1017/psrm.2016.23