Disparities in Poll Closures in the Age of COVID-19: A Case Study of Wisconsin

Abstract

The COVID-19 pandemic of 2020 arose during the presidential primary process and spring elections for states across the U.S. One of the consequences of the pandemic on election administration was the closure of polling locations due to resource constraints and public health concerns. We aim to measure the extent to which disparities in electoral access and participation arose given polling place closures during the April 7, 2020, Wisconsin spring election. We run a spatial probit analysis on the probability of poll closures given population density and nonwhite populations of a given area using poll coordinate data provided by the Wisconsin Electoral Commission. We follow up to take advantage of the exogenous nature of poll closures to measure the impact of closure costs within the counties of Milwaukee and Dane via a multilevel logit analysis of turnout by mode. We find that the pandemic led to a series of poll closures that systematically impacted racial minorities and areas of high population density. Further, these poll closure costs as measured in increased distance to and size of polls are associated with reduced voter turnout, especially within Milwaukee County. We conclude that disparate access to polling places and costs of poll closures present challenges that must be overcome for purposes of election legitimacy and social cohesion.