COVID-19 likely to weigh on U.S. election turnout, outcomes

The outbreak of the COVID-19 pandemic has collided with the 2020 presidential election cycle in the U.S., a mix that has the potential to change voting preferences and outcomes.

In a paper soon to appear in COVID Economics, Dan Honig, assistant professor of international development at the Johns Hopkins School of Advanced International Studies, and James Bisbee, a postdoctoral fellow at Princeton's Niehaus Center for Globalization and Governance, examine how anxiety from COVID-19 will affect U.S. voters in the months ahead.

Image caption: Dan Honig, assistant professor of international development at the Johns Hopkins School of Advanced Studies

The Hub caught up with Honig to discuss the impacts on turnout, voter behavior, and the logistics of holding elections during a pandemic.

Your work focuses on the effects this pandemic has had on voter behavior. How has the COVID-19 pandemic affected the dynamics of the Democratic presidential primary race?

When you compare U.S. counties that had reported a confirmed case of COVID-19 prior to voting to similar counties where COVID-19 appeared only after votes were cast, there was a noted "flight to safety" of voters toward Joe Biden and away from Bernie Sanders.

I think of the "flight to safety" in the financial sense, as being typified by a desire to de-risk. When there's anxiety and uncertainty in a financial market, what we want to do is reduce our exposure to risk. So a "flight to safety" moves toward safer assets, often things like gold or U.S. Treasury bonds. In the political context, when we face an uncertain future, we are going to gravitate toward a political option that feels less risky. In other words, when we feel anxiety or fear we gravitate toward the safety of the known which, in this case, is Joe Biden.

The pandemic has affected the primary race in other ways as well. Some states decided to postpone their primaries or shift to vote-by-mail. Others are battling in court to determine whether they can change their primary date. Should policymakers weigh the impact of COVID-related anxiety in determining when and how to hold these elections?

Yes, I think that they probably should. First, vote by mail is a great response in part because some of the fear from voters is fear of catching the disease at the voting booth, which will certainly decrease voter turnout. This is conjecture, but COVID-related anxiety could dissuade voters differentially—higher risk groups are likely to be less willing to show up at the polls, which would affect the composition of the electorate and the eventual winner of the election.

The impact of this anxiety on the electorate is just one factor policymakers should keep in mind while ensuring that the country can maintain free and fair elections in 2020.

Obviously many other factors shaped the dynamics of this race in the months before the pandemic arrived in American communities, and a number of states cast their ballots before widespread concern over COVID-19 became a top-of-mind issue for voters. How decisive was the "flight to safety" factor in influencing the outcome of this race?

When you compare counties with COVID-19 to those that had not yet recorded a case, we noted a non-trivial "flight to safety" toward Joe Biden, accounting for a 7% slippage in voter share for Bernie Sanders in 2020 compared to 2016. This occurred in counties where there were cases of COVID-19 reported in the local media market. This is a big effect, but given the size of Biden's delegate lead by the time COVID-19 fears took hold, it likely was not the decisive factor—I don't believe our results support the conclusion that without COVID-19 Sanders would have secured the nomination.

That said, anxiety can affect other elections, too—and if a 7 percentage point effect is typical, that's an effect large enough to potentially shift outcomes in close elections. In the 2018 U.S. House elections, for example, about 65 seats were won with a margin of victory less than 7 percentage points.

Johns Hopkins responds to COVID-19

Coverage of how the COVID-19 pandemic is affecting operations at JHU and how Hopkins experts and scientists are responding to the outbreak

How might anxiety related to COVID-19 shape the U.S. general election?

Who would benefit from a "flight to safety" between Donald Trump and Joe Biden? It's not clear in which direction the flight to safety would run. A well-crafted voter survey could give insight into what voters determine as the "safer" candidate in this matchup.

Donald Trump, as president, represents the status quo and, in many circumstances, the incumbent is the "safe" pick. "Don't change horses in midstream," as Lincoln put it when running for reelection in the midst of the Civil War. Trump is far from a traditional candidate and president, however. It is possible that voters might see Joe Biden as a more traditional, mainstream candidate, and thus he would benefit from a "flight to safety" effect.

It's also worth noting that while I am not aware of evidence of any actor—foreign or domestic—seeking to increase anxiety from the pandemic in order to shape this election, COVID-19 anxiety could potentially be exploited by those seeking to influence the general election results.

What are the broader political implications of COVID-19 in the U.S.?

First, it is possible we're underestimating the effect of COVID-19 on voting (or, conversely, overestimating it). Consider the effect that COVID-19 might have in the House of Representatives, or in the state legislatures, where just a small shift in voting share can change the outcome of races. I think the anxiety related to COVID-19 has the ability to shape voting in many other elections and politics in ways that aren't currently well understood.

Thus far we, as a society, have focused a lot on how COVID-19 is changing our social lives, as well as its serious consequences in terms of life, health, happiness, and the economy. But I think the political implications of the pandemic and the fear it's inducing are also potentially large and have thus far gone unexamined. As the pandemic extends from a month-long, to a quarter-long, to potentially a longer-term shift in American life, it's worth thinking through the longer-term implications of the disease, not the least of which are the political impacts.