Voting Plans and Mask-Wearing Practices During the 2020 U.S. National Election

Background: While U.S. continues to face increasing rates of COVID-19, there is concern that voting behavior during the 2020 U.S. Presidential election may contribute to additional outbreaks and infections among communities. The purpose of the current study was to assess the impact of the spread of the COVID-19 infection and mask-wearing behavior on voting behaviors and political affiliation during the 2020 national election. Methods: During a two-week period from September into October 2020, YouGov, in association with Saint Louis University, conducted an online cross-sectional survey consisting of participants that were likely to vote in Missouri in the upcoming national election (n=931). The sample was stratified into two groups: those that reported always wearing a mask or face-covering in public spaces and those that did not. Individual socio-demographics and environmental factors were compared between these groups to identify significant differences according to mask-wearing patterns. Additionally, two adjusted multivariate models were constructed to determine probability of (1) reporting always wearing a mask and (2) planning on voting in-person on election day. Indicators in each model included reported political party affiliation, and urbanicity, presence of mask mandate, and recent COVID-19 rate, respective of reported Missouri county of residence. Results: The sample consisted of 931 participants across Missouri that were likely to vote during the 2020 Presidential election. Among this sample, 38.5% resided in counties with a mask mandate at the time of the survey. Individuals who resided in either suburban or urban counties were twice as likely to report always wearing a mask compared to rural residents. In addition, while individuals from counties with a mask mandate were over twice as likely to report always wearing a mask, county COVID-19 infection rates were not found to be a significant predictor of mask-wearing. Republicans and Independents were significantly less likely to report always wearing a mask. Compared to Democrats, Republicans were 4 times, and Independents were 2 times, more likely to vote in person on election day compared to Democratic party members. These results were significant even when adjusting urbanicity, residing in a county with a mask mandate, and county COVID-19 case-rate. While urbanicity and COVID-19 infection rate were determined to not add significantly to model performance, those that lived within a county with a mask mandate were nearly 50% less likely to vote in person on election date. Discussion: Overall, this study identified significant relationships that are likely to contribute to the spread of COVID-19. Individuals who identified as Republican and Independent party members were more likely to vote in-person on election day and less likely to always wear a mask in public spaces. The interaction between political party affiliation and mask wearing highlights the concerning dichotomy within political discourse and highlights an opportunity to develop novel interventions that reduce the current political division that exists within the U.S.