Does Election Day Registration Make a Difference? Evidence from Illinois

Abstract

While many scholars have investigated whether election day registration (EDR) boosts turnout and who is influenced by it, empirical evidence is still mixed. Utilizing within-state registration rule difference in Illinois, this paper examines the effects of EDR on turnout rate, registration rate, and partisan composition of voters. In 2016 elections, counties in Illinois (i) with a population over 100,000 or (ii) using electronic poll books needed to allow their citizens to register to vote on election day at every precinct, while other counties only had to adopt the rule at one precinct per county. Analyses using randomization inference approach to regression discontinuity design reveal that EDR exerted small or no effects on turnout rate and registration rate (0-3%, statistically insignificant) but it seems to have somewhat increased the Democratic vote share (6-12%, statistically significant in most of the cases).