SEMINAR CANCELLED Political (self-)selection and competition: Evidence from U.S. Congressional elections
I use data on U.S. Congressional primary and general elections for the years 1998-2014 to study the effect that political competition has on the entry and selection of high quality politicians, as measured by their previous legislative experience and how well they demographically represent their district. To identify causal effects, I rely on variation to competition caused by demographic changes resulting from decennial redistricting. Difference-in-difference estimates suggest that experienced candidates tend to avoid competition in primary elections in the electorally strong party. As opposed to this, experienced candidates and candidates with similar demographic characteristics as their constituents tend to run more frequently in primary elections in the weak party as competition increases. The estimates are mostly consistent with theoretical models that explain entry based on differing outside options. Overall there are no effects of competition on the quality and identity of the eventual winner of the general election.