ABSTRACT
A central debate in our understanding of primaries is simple: does primary type matter? Most of the discussion around this debate has centered on the question of whether or not different primary types lead to different kinds of candidates, with some evidence that there is a correlation between primary type and the kinds of representatives that emerge (Gerber and Morton 1998; Kanthak and Morton 2001) but a good deal more suggesting that primary type matters little in the subsequent behavior of legislators (Bullock and Clinton 2011; Hirano, Snyder, Ansolabere and Hansen 2010; McCarty, Poole and Rosenthal 2006; McGhee 2010; McGhee, Masket, Shor, Rogers and McCarty 2014; Pearson and Lawless 2008; Rogowski and Langella 2015). 1 Yet this creates something of a puzzle, because the evidence is clear that primary voters are more extreme than general election voters (Brady, Han, and Pope 2007; Carey and Polga-Hecimovich 2006; Jacobson 2012; Kaufmann, Gimpel and Hoffman 2003; Sinclair 2006). 2 Why don’t more ideologically extreme voters nominate similarly extreme candidates in a manner consistent with the Downsian (Downs 1957) spatial logic that undergirds much of the literature on legislative behavior? We attempt to answer that question by taking a step back in the electoral process to before the election, at the stage of candidate emergence. Voters obviously can select only from those candidates who have chosen to run. In order, for example, for open primaries to result in more moderate legislators, open primaries must attract more moderate candidates. Otherwise, these relatively more moderate voters have no relatively more moderate candidate to select. In the current project, then, we explore the question of how ambitious candidates respond to primary type to determine whether or not those candidate emergence patterns are compatible with the patterns we would need to see for primary type to have an effect on candidate selection and subsequent legislative behavior.
