A Power Struggle: Do Contested Primaries Between Moderates and Extremists Harm the Party?

As intra-party disagreements rise, primary challenges between candidates with ideological differences occur. A deeper understanding of how these primary challenges affect party success will assist potential candidates, and the parties themselves, in making safer and more helpful decisions in their pursuit to change platforms and policies. This thesis attempts to study the effect of contested primary elections between ideologically distanced candidates on their party’s general election success. Merging three existing datasets, I use a custom dataset with candidate-level data from 1980-2016. I employ several fixed-effect multivariate regression models measuring the effect of ideological difference in a primary race (utilizing CFscores as an ideological marker) and the closeness of the primary race itself on the general election vote percentage of the respective political party. I am unable to find a relationship between these variables measured in an interaction, nor am I able to find a relationship between ideological distance in a primary and any effect on general election vote percentage. I do measure a positive correlation between the closeness of a primary race, and the general election vote percentage for the respective primary winner; the less competitive the primary election is, the better the candidate performs in the general election. I conclude that the data utilized are robust in their ideological mapping, however, current ideological scores (CFscore and DW-NOMINATE) are insufficient in measuring extremism as a variable. Further research within this theoretical framework ought to attempt a measurement of political extremism, rather than ideological differences measured in relation to their respective party members