The Partisan and Racial Impact of Changing Voting Hours: A Voter-Level Analysis

• In Event: Constraints on Election Administration

Abstract

The contemporary Republican Party is programmatically committed to restricting access to the ballot box, an effort that has continued following the 2020 election. The best social science research in this area has focused on voter ID laws, but has found no evidence that voter ID laws reduce voter turnout. Relatedly, evidence that the availability of early voting affects voter turnout is mixed. Why, then, is so much effort expended on enacting restrictive voting laws across all levels of government? We offer the first evidence using fine-grained, individual-level data that Republican-led voter suppression, in the form of reductions in early voting hours, reduces turnout, with the most pronounced effects among Democrats in general and Black Democrats in particular. Focusing on the 2016 presidential election in North Carolina, we leverage county-level variation in changes to early voting hours to generate causal estimates of the effect of reducing or increasing voting hours on affected populations. Drawing from commercial voter file data that provides detailed demographic information on registered voters, we are able to conduct a systematic and comprehensive search for the best model that breaks confounding, matching similar individuals across counties while imposing minimal modeling assumptions. We find that reducing early voting hours caused a decrease in turnout for Black Democrats, with no effect for whites, while increasing early voting hours increased turnout among whites, but white Republicans in particular.

Authors

• Marc Thomas Ratkovic, Princeton University

• Eric Manning, Princeton University

• Will Lowe, Hertie School