Association for Public Policy Analysis & Management

This paper sheds light on inconsistencies in past research regarding the relationship between state-level voter turnout and registration levels, and state voting and registration laws. Specifically, it examines the influence of early and absentee voting provisions, voter identification requirements, election-day registration, and online registration, while closely examining the importance of key details, such as the length of early voting periods, the strictness of voter identification laws, and how long ago the policies were enacted. It evaluates which policy features are correlated with state-level turnout, and tests which policy changes between 2012 and 2016 had a causal effect on turnout, using difference-in-differences techniques. Importantly, all estimated models control for other state-level changes that might separately explain swings in registration and turnout rates, such as the closeness of the national election at the state level, the closeness of concurrent senate and governor elections, and the composition of the state’s population during each period. To address possible policy endogeneity, I assess the extent to which registration and turnout levels respond similarly in red states, blue states, and swing states.