An examination of congressional elections

This dissertation examines a number of issues that arise in U.S. Congressional elections. In the first chapter, I look at the source congressional campaign contributions. I show that the source of a campaign contribution is an im- portant factor in determining the effect the contribution has. Specifically I find that individual contributions have a greater impact on election results than corporate contributions. I show that the reason for this difference is that in addition to direct spending effects, individual contributions are also a reflection of a candidate’s quality and thus contain an indirect quality ef- fect. I then use 3SLS to obtain an estimate of all the effects involved with campaign contributions. The results present a better understanding of the true effects of money in an election. In the second chapter I look at the dynamics of party spending in congres- sional campaigns. I examine a model of political parties who must choose how to allocate campaign contributions over elections in several districts in order to maximize the chance that they will control the enacted legislature. The parties must compete in two areas, informed voters who only care about policy and uninformed voters who only care about campaign spending. I find that parties will choose to heavily concentrate their spending in whichever district is most competitive. In the third chapter I examine the causes of congressional gridlock. I examine a model of two parties that must negotiate in order to choose a policy to enact. The parties compete in elections in two bodies of a legislature and the winning candidates engage in a bargaining process to determine the implemented policy. I find that policy motivation and election uncertainty are the main drivers of gridlock. I also find that compromising can end up hurting a party and helping the non-compromising party. This leads to the possibility of an equilibrium with gridlock even though there are high costs to both parties associated with failing to compromise