Diverse districts and the electoral connection: Electoral coalitions and legislators' pursuit of policy preferences

This dissertation asks under what conditions legislators are able to pursue their policy preferences when representing their constituency. The question is motivated by an empirical and theoretical puzzle---a persistent gender gap in Congressional voting is unexplained by current theories of legislative behavior. I propose a theory of politically exploitable heterogeneity, which argues that legislators want to pursue their preferences in office but will only do so when they are able to build an electoral coalition consistent with their policy preferences. Legislators will have the opportunity to build such coalitions only in districts with politically exploitable heterogeneity, defined as districts where constituents disagree both on their issue positions and priorities. To test my theory, I use a unique dataset of constituent preferences measured with individual contributions to political action committees. I find, contrary to expectations, that in districts where constituents disagree on both position and priority, legislators tend to converge in their voting.