A Statistical and Geostatistical Survey of the Relationship Between COVID-19 and By-Mail Balloting in the 2020 North Carolina General Election

Keywords: By-Mail Voting, COVID-19 Mapping, Cluster Analysis, Election Mapping, Geographic Information Systems
Session Type: Virtual Paper Abstract
Day: Saturday
Session Start / End Time: 2/26/2022 02:00 PM (Eastern Time (US & Canada)) - 2/26/2022 03:20 PM (Eastern Time (US & Canada))
Room: Virtual 30

Abstract

Adding to the already polarizing 2020 general election was the COVID-19 pandemic. One way in which this pandemic greatly impacted the election was through an increased participation in by-mail, or mail-in ballots. The state of North Carolina experienced a 316% increase in by-mail votes between 2016 and 2020, when approximately 977,186 by-mail votes were cast. It is no surprise this increase is due to the COVID-19 pandemic; however, these by-mail voting patterns are spatial in nature and vary across the state. This research measures to what degree COVID-19 rates impacted by-mails voting rates. Using GIS data developed from robust tabular files provided by the North Carolina State Board of Elections, by-mail votes were calculated and mapped at zip code scale and compared to COVID-19 rates taken at different dates. By-mail rates taken from final absentee tallies for the highest and lowest COVID-19 zip codes saw no significant differences across multiple dates (9/30/2020 and 10/31/2020) when COVID-19 data were collected. COVID-19 hot spots (high COVID-19 rates surrounded by other high COVID-19 rates) were extracted using geostatistical techniques and compared to COVID-19 cold spots (low COVID-19 rates surrounded by other low COVID-19 rates). It was found the lowest by-mail rates actually occurred in these COVID-19 hotspots across both dates, as well a metric that expressed percent change in COVID-19 rates in the month before the 2020 election.