How COVID-19 death rates impacted 2020 presidential voting patterns

In the fall of 2020, some political analysts predicted voters mourning the loss of friends and loved ones to COVID-19 would have blamed then-President Donald Trump for mismanaging the pandemic and, in turn, be more likely to vote for Joe Biden than they were for Hillary Clinton. 

But that wasn’t the case, reports a new Northwestern Medicine study.

Counties across the United States with a high COVID death rate saw a smaller increase in Democratic voters in the 2020 presidential election compared to 2016, the study found. In contrast, counties with a lower COVID death rate saw a significantly larger increase in the Democratic vote margin from 2016. 

While the mean Democratic vote increased by 2.9% nationally from 2016 to 2020, there was only a 0.9% increase in the highest-mortality counties as compared to a 4.3% increase in the lowest-mortality counties, the study found.

Democrats have a challenge of how are they’re going to win over working-class voters, including those who have been hit hard economically and epidemiologically by the pandemic.”

“The COVID epidemic costs reflect the inequality in our society’s social and racial dimensions,” said senior and corresponding author Joe Feinglass, research professor of medicine (in the division of general internal medicine and geriatrics) at Northwestern University Feinberg School of Medicine. 

“Despite some predictions, Biden did not win because of a surge in turnout in the hardest-hit counties, many of which included large urban centers,” Feinglass said. “These counties actually had a significantly lower increase in overall turnout than safer, more affluent suburban counties.”

The study was published August 17 in the journal Public Health. 

While the study didn’t examine the reasons behind the voting patterns, Feinglass, also a longtime political activist who formerly taught history and political science at a local community college, offered possible explanations: 

High-mortality counties/smaller increase in Democratic voting

“Residents of some high-mortality counties could have been more fearful about in-person voting, especially in areas where lower in-person voting was not offset by mail-in voting — something typically utilized more by Democratic-leaning residents,”Feinglass said.

Additionally, some working-class voters, who tended to have greater concerns over pandemic-related job losses, disproportionately held “essential” or service-sector jobs with a high risk of exposure to COVID-19 and lived in higher-density housing, shifted toward Trump. These voters may have been reacting more negatively to stricter lockdowns and business closures imposed by or advocated for by Democratic policymakers, as that impacted their livelihood, Feinglass said. 

“Trump may have gained support among voters fearful of an economic collapse, especially in the absence of robust federal recovery spending in the months prior to the election,” Feinglass said. “It is also possible that residents of areas hit hardest by COVID-19 were simply exhausted by the economic and social disruption of the pandemic and may not have prioritized voting.”

Low-mortality counties/larger increase in Democratic voting

The 2020 U.S. election saw a large Democratic shift by more affluent, educated, suburban voters, ostensibly driven in part by the incumbent President Trump’s “anti-science” stance on the pandemic, Feinglass said. These voters also tended to hold “safer” jobs that permitted work from home, and therefore, were less likely to lose their jobs during the lockdown.  

“Democrats have a challenge of how are they’re going to win over working-class voters who have been hit hard economically and epidemiologically by the pandemic,” Feinglass said. “They’re gaining support from middle-class, educated voters but there’s a ceiling on that support.

“The income-inequality problem in our country has reached gargantuan levels and is roaring ahead every day. If Democrats are going to increase their vote share, they must begin to rebuild trust among working-class voters.” 

Study co-authors worked with COVID patients in the ICU 

The study’s co-authors were third-year medicine residents who were treating COVID patients in the intensive care unit at Northwestern Memorial Hospital and the Jesse Brown VA Hospital in Chicago during the pandemic. In the months leading up to the 2020 presidential election, they began collecting COVID, demographic and health status data at the county level, as they were interested in seeing how the pandemic would impact voting trends. 

“We anticipated the higher turnout across both parties, but did not expect such disproportionate suppression of Democratic votes in high-mortality counties,” said first author Dr. Aaron Parzuchowski, a third-year medical resident at Feinberg. “COVID greatly influenced this election and will likely have a lasting impact on American politics for years to come."

The authors collected presidential voting data from 2016 from the MIT Election Lab and 2020 voting data from Edison Research and the National Election Pool via the New York Times. The 2020 totals reflect 99.6% of the 2020 vote. 

Other Northwestern authors on the study, “County-level association of Covid-19 mortality with 2020 United States presidential voting,” include Andrew Peters, Cecelia Johnson-Sasso and Kelsey Rydland.