Electoral Effect of Protracted Presidential Primaries

Abstract

Divisive presidential primaries have become common in the United States following the McGovern—Fraser reforms. Both politicians and political scientists have argued that parties that experience divisive primaries may be disadvantaged in the general election, with several political scientists demonstrating a relationship between a nominee receiving a smaller share of the primary vote and a smaller share of the general election vote. However, demonstrating that divisive primaries cause worse general election results is difficult because candidate weakness or other factors may cause poor performance in both primary and general elections. I use a linear regression model to estimate the effect of divisive presidential primaries on general election outcomes. I compare near-border counties in states that did not experience contested primaries to near-border counties in states that did experience contested primaries. I find that there is generally a penalty for parties that experience divisive primaries, but the effect varies over election cycles.