Allyson Pellissier, California Institute of Technology

Online voter registration is one of the most recent efforts to stimulate turnout in American elections. Within the past decade, an increasing number of states have begun to allow their residents to register as voters electronically. Like other efforts to increase political participation, though, the actual impact on registration and turnout remains unclear. Although other voting liberalizations have received a fair amount of scrutiny, the peer-reviewed literature does not include a systematic exploration of how voters are responding to online registration. In this paper, I develop an individual-level model that point identifies an estimate for the impact of online registration on the likelihood of both registration and turnout. The results suggest that online registration may be one of the more successful implementations of convenience voting. Perhaps even more importantly, its effects seem to be concentrated most highly among young adults and those who have moved recently, two subgroups that are consistently underrepresented at the polls. I then use the individual-level model to predict changes in state-level aggregate turnout and identify states for which this registration alternative could have influenced the election. Although it is hard to know how the partisan distribution would shift, several states in both the 2008 and 2012 Presidential elections could have experienced different outcomes had they offered online registration. At both the individual and aggregate level, therefore, the introduction of online registration may have significant implications for American elections.