How Electoral Institutions Shape Citizen Participation and Legislative Behavior

Abstract

The electoral system is often treated as fixed, but throughout U.S. history significant changes in electoral institutions, or in political conditions dictated by electoral institutions, make it possible to identify more precisely the role that the electoral system plays in the democratic process. This dissertation examines three related questions, each focusing on an aspect of the influence of electoral rules on political behavior. How has the ability to directly elect representatives influenced other forms of citizen engagement with government? How has competitiveness influenced voter turnout? Finally, when separate elections lead to differences in partisan control over the branches of government, what is the effect on policymaking in Congress? The first chapter shows that petitioning campaigns have historically substituted for the communication and accountability obtained through direct elections. I estimate that rates of petitioning to the Senate declined by 30% when the passage of the 17th Amendment ended the practice of indirect election by state legislatures and replaced it with direct elections. The implication is that electoral reforms meant to improve representation may weaken other ties between citizens and lawmakers. The second chapter examines the relationship between electoral competition and turnout. Past research has found that citizens vote at higher rates in response to closer elections, either through instrumental voting at the individual level or through voter mobilization by elites. In contrast, this chapter demonstrates that citizens living in competitive congressional districts differ markedly from those in uncompetitive districts along a range of dimensions other than turnout. Using an individual panel based on voter files from all 50 states and exploiting variation in competitiveness induced by the 2012 redistricting cycle yields a precisely estimated null effect of competitiveness on turnout. The third chapter re-examines whether divided government reduces legislative productivity. After developing the most comprehensive database to date of significant acts of Congress---from 1789-2010---this chapter shows that unified control corresponds with one additional significant act passed per Congress in the 19th Century and four additional such acts in the 20th Century. However, party control of government cannot explain the broad historical trends in the rate at which Congress passes significant legislation.