Description:
How does competition affect the entry and selection of politicians? I use data on U.S. Congressional primary and general elections for the years 1998-2014 to study this question. I measure quality using previous legislative experience and the novel "identity match" quantifying how well candidates demographically represent their district. To identify causal effects, I rely on variation in competition caused by demographic changes resulting from decennial redistricting. Difference-in-difference estimates reveal differences between the electorally dominant and weak party. They show that experienced candidates avoid competition in primary elections in the strong party. As opposed to this, experienced candidates and candidates with a good identity match run relatively more frequently in primary elections in the weak party as competition increases. The effects of competition and entry overall cancel each other out so that there are no effects on the quality and identity match of the eventual winner of the general election.