Published online by Cambridge University Press: 04 March 2015
A large class of spatial models of elections converges upon a singleprediction: a candidate’s vote share increases in the congruence between herplatform and the median voter’s preferences. Though considerable empiricalresearch provides support for this prediction, these studies have notadequately identified the effects of platform positioning net of otherfactors. In this paper, we study the impact of challenger moderation on voteshares using data from 444 US House elections from 1996 to 2006 in whichsuccessive challengers competed against a common incumbent. Our findings arelargely null. We uncover no evidence that challengers increase their voteshares by adopting more moderate platform positions. This finding is robustacross a wide range of model specifications and subsets of districts.
B. Pablo Montagnes is an Assistant Professor in the Harris School ofPublic Policy, University of Chicago, 1130 E. 60th Street, Chicago, IL60637 (pmontagnes@uchicago.edu).Jon C. Rogowski is an Assistant Professor in the Department ofPolitical Science, Washington University in St. Louis, One BrookingsDrive, Campus Box 1063, St. Louis, MO 63130 (jrogowsk@artsci.wustl.edu). The authorsgratefully acknowledge Project Vote Smart for providing data used inthis paper, and Ethan Bueno de Mesquita, Will Howell, Eric Oliver,Drew Linzer, Boris Shor, Betsy Sinclair, two anonymous reviewers, andthe editor for helpful comments. Data used in this project wereobtained from Project Vote Smart, the Federal Election Commission, andthe Clerk of the US House of Representatives. Replication data can beaccessed through the Political Science Research and Methods Dataverse. To view supplementary material for thisarticle, please visit http://dx.doi.org/10.1017/psrm.2014.36