Enough time has passed since the 2020 presidential election that we can now ask: What effect did COVID-19, arguably the biggest event of the year — of the century, even — have on the election outcome?
The answer to this question probably seems straightforward considering how abysmally Americans thought then-President Trump handled the pandemic. But the evidence we have points in many directions.
Partisans don’t just disagree, they hate one another | FiveThirtyEight
Let’s start with what history can tell us. That is, given what we know about elections held in the middle of a pandemic, what effect should we have expected the novel coronavirus to have had? If you’re scratching your head trying to think of a good comparison, that might be because we don’t really have one. The closest analogy to what we experienced in the U.S. in 2020 is the 1918-19 influenza pandemic, which also broke out during an election year and killed hundreds of thousands of Americans.1
The effect of the pandemic on the 1918 midterms has been studied, too. But political scientists Chris Achen and Larry Bartels found that it had no particular effect on the election outcome; the Democrats (in control of the White House at the time) did no worse in congressional elections in places where the disease hit hard than in places where it didn’t. A somewhat different approach by Leticia Arroyo Abad and Noel Maurer found only a very small effect on the congressional vote in 1918 and no subsequent effect on the 1920 election. Now, that doesn’t prove that a pandemic can’t affect an election. Maybe the fact that the 1918 election was a midterm election played a role here; that is, even if people did blame Woodrow Wilson’s presidency for the pandemic, they didn’t extend that to the rest of his party. And maybe the pandemic would have had a greater effect if the country hadn’t been engaged in World War I at the time. It’s also possible that many people didn’t yet think of the federal government as responsible for matters of public health.2
But returning to the present day, what do we know about the role COVID-19 played in the 2020 presidential election? One way to answer this question is to dig into state-level results and subtract Trump’s vote share in 2020 from his vote share in 2016, measuring how much his vote improved or declined across those two elections. What we find, however, is no statistically significant relationship. That is, Trump did no worse — and possibly slightly better — in states with higher COVID-19 mortality rates. The same is true if we compare the vote against per capita COVID-19 cases.
It turns out that economic growth, measured as the growth in per capita real disposable income from the first through third quarters of 2020, may explain some of what we’re seeing. That is, if we compare Trump’s vote share from 2016 to 2020 with the amount of economic recovery a state experienced, we find that Trump did much better in those states where the economy bounced back, even controlling for COVID-19 death rates. In other words, that $1,200 stimulus payment voters received back in the spring may have done a lot to help mitigate the political damage for Trump. In fact, had he and Congress been able to deliver some kind of additional economic relief prior to the election, that may even have saved his reelection bid.
Other researchers have also found this same pattern of Trump doing no worse, and possibly even better, at the county level in areas with higher COVID-19 mortalities. And, perhaps somewhat counterintuitively, I also found in my analysis that the economy appeared to be somewhat weaker in states where there were more COVID-19 cases, and somewhat stronger where there were more COVID-19 deaths. One possible explanation is that places with fewer health restrictions on businesses helped produce a stronger economy in those areas (helping Trump) even while spreading the disease, and in the end, the economy just had a greater effect on people’s votes. Researcher Solomon Messing discovered an added wrinkle in that more COVID-19 deaths seem to have hurt Trump in very white counties, while the same wasn’t true in counties where a large share of the population isn’t white.
To be clear, we still don’t have a great sense of why these patterns occurred, and none of this is to suggest that Trump did better in some areas because of the coronavirus. But, suffice it to say, this pattern is not the sort of thing many would expect given how poorly most Americans thought Trump handled the pandemic. What also makes it difficult to detect the effect of COVID-19 on the election? Like so many other issues in American politics, the pandemic was quickly interpreted through partisan lenses. The fact that the initial fallout in March didn’t give Trump much of a “rally-around-the-flag” effect, or a temporary boost in popularity given the crisis, is telling. But, then again, so is the fact that it didn’t seem to hurt him all that much either.
So, what can we ultimately say about the impact of COVID-19 on the 2020 election? Most likely, it worked against Trump. Had there been no pandemic, he may have still lost the popular vote, but considering how close the election was, he may have had a decent chance of winning the Electoral College. Yet the damage to his prospects was far from enormous, and that may have been mitigated somewhat by polarization. Indeed, a better response on Trump’s part that either helped reduce the spread of the disease or limit its economic impact could well have secured his reelection bid.
CORRECTION (Jan. 27, 2021, 7:25 p.m.): A previous version of this article misspelled Noel Maurer’s surname.