Published online by Cambridge University Press: 08 May 2017
This article develops a novel explanation for the incumbency advantage based on incumbents’ ability to signal positions that are ideologically distinct from those of their parties. Using voter-level data from the Cooperative Congressional Election Study and controlling for unobserved district heterogeneity, the study finds that voters in US House elections primarily use information about the ideology of candidates’ parties to infer the location of challengers, while they instead rely on information about the individual candidates’ ideologies to place incumbents. In higher-profile Senate elections, the difference between challengers and incumbents is trivial. Decomposing the incumbency advantage into valence and signaling components, the study finds that the signaling mechanism explains 14 per cent of the incumbency advantage in House elections, but only 5 per cent of the advantage in Senate contests. It also finds that a 50 per cent increase in party polarization increases the incumbency advantage by 3 percentage points.
Assistant Professor, Department of Political Science, Emory University (email: firstname.lastname@example.org). I thank seminar audiences at Illinois Ohio State, Yale, the Harris School and Cyrus Aghamolla, Steve Callander, Kyle Dropp, Nick Eubank, Morris Fiorina, Alex Frankel, Justin Grimmer, Wesley Hartmann, Alex Hirsch, Keith Krehbiel, Neil Malhotra, Greg Martin, Eleanor Powell, Chris Stanton, Ken Shotts and Jonathan Wand for many helpful comments. I thank Gary Jacobson for generously providing his data on candidate experience in US House elections. An earlier version of this article circulated under the title ‘Candidate Positioning, Partisan Brands, and Election Outcomes’. Data replication sets are available at http://dataverse.harvard.edu/dataverse/BJPolS and online appendices are available at http://dx.doi.org/doi:10.1017/S0007123416000557.