Life Expectancy and Voting Patterns in the 2020 U.S. Presidential Election

Introduction In the 2016 U.S. Presidential election, voters in communities with recent stagnation or decline in life expectancy were more likely to vote for the Republican candidate than in prior Presidential elections. We aimed to assess the association between change in life expectancy and voting patterns in the 2020 Presidential election. Methods With data on county-level life expectancy from the Institute for Health Metrics and Evaluation and voting data from GitHub, we used weighted multivariable linear regression to estimate the association between the change in life expectancy from 1980 to 2014 and the proportion of votes for the Republican candidate in the 2020 Presidential election. Results Among 3,110 U.S counties and Washington, D.C., change in life expectancy at the county level was negatively associated with Republican share of the vote in the 2020 Presidential election (parameter estimate −7.2, 95% confidence interval, −7.8 to −6.6). With the inclusion of state, sociodemographic, and economic variables in the model, the association was attenuated (parameter estimate −0.8; 95% CI, −1.5 to −0.2). Conclusion Counties with a less positive trajectory in life expectancy were more likely to vote for Republican candidates in the 2020 U.S. Presidential election, but the association was mediated by demographic, social and economic factors.