The COVID-19 pandemic of 2020 severely impaired election administration capacity and services across the states. Among the most significant effects of the pandemic proved to be the number and quality of in-person voting sites, which fell susceptible to unplanned closures or limited resources to process voting. We explore the impact of the COVID-19 incurred polling closures on voter participation as enabled through GIS integrated records on polling places and geocoded voter coordinates. We seek to answer first, who suffered the brunt of the polling place closures, and second, how did these closures impact participation. Through an analysis of 2020 primary participation in the state of Wisconsin and several other large counties in other states with polling place address records, we find substantially higher poll closure rates associated with areas with higher population densities and nonwhite populations. Additionally, we find poll closures substantially increased the travel costs to vote, which in turn is strongly associated with decreased voter turnout. We therefore offer strong evidence of the costs incurred on electoral participation due to the pandemic, as mediated by degree of preparedness to open polling places. Further, our results offer support to the necessity of maintaining a well-developed system of geographic information systems records in order to identify high at risk areas and take preventative measures to mitigate the impact of the pandemic or similar emergencies on election administration.

Presenters:

• John Curiel, Research Scientist, MIT Election Data & Science Lab
• Charles Stewart III, Founding Director, MIT Election Data & Science Lab
• Jesse Clark, Ph.D. Candidate, Massachusetts Institute of Technology