Why Pander to the Party: A Spatial Analysis of Senate Primary and General Elections
Author(s)
Ezrow, Lawrence
Abstract
Do candidates adopt very similar Left-Right policy positions, i.e. converge, under two-party plurality rule elections? If this is indeed the case, what are the logical consequences of primary contests on candidates’ policy positioning in general elections? In particular, do competitive primary elections lure eventual nominees towards the center of the distribution of their parties’ supporters, that is, towards more extreme policy positions than are held by the average voter in the general electorate, which in turn commits eventual nominees to take more extreme positions in the general elections? The paper addresses each of these questions, and arrives at three critical findings. First, a candidate increases her chances of victory if she is closer to her constituency’s mean citizen position than her competitor, thus supporting the Relative Convergence Hypothesis. The second conclusion confirms that divisive primaries do indeed hurt candidates in securing victory in their general elections. However, the third hypothesis only receives limited support, that asserts that divisive primaries harm candidates because they induce candidates to adopt more extreme positions. The Democratic candidates follow this tendency while the Republicans do not. This pattern reflects a temporal feature of the study that during this period “New Democrat” ideology, associated with policy centrism, was rising in influence in the Democratic Party. These findings have implications for our understanding of party strategies and political representation.
