Did anti-vaxxer deaths from COVID swing elections?

Reading Time: 6 minutes

[Previous: The 2022 midterms: Not a red wave but a blue wall]

The 2022 midterm elections were unexpectedly good for progressives.

The predicted red wave petered out into a red trickle. Democrats lost the House, barely, but gained a Senate seat and netted two governorships. They also flipped the Minnesota Senate, the Pennsylvania House of Representatives, and both chambers of the Michigan legislature. This was the first time since 1934—under FDR!—that the president’s party didn’t lose any Senate seats or state legislative chambers in the midterm.

What can explain this precedent-shattering result?

Previously, I discussed three factors that loomed large: Americans are fed up with Trumpism; the loss of abortion rights motivated strong pro-choice turnout; and young, secular voters showed up in record numbers.

But there’s one other reason, ugly though it is, that shouldn’t be ignored. Thousands of Republican voters didn’t show up in 2022… because they died of COVID.

The COVID partisan death gap

Is it possible that enough conservative voters died to alter the outcome of the election? Let’s look at the numbers.

Depending on whose estimate you accept, the U.S. has suffered around 1,100,000 deaths from COVID. The U.S. has 330 million people, meaning that COVID deaths subtracted about 0.33% of the population.

At the human level, this is an enormous toll of pain, suffering and grief. The deaths of 58,000 Americans in the Vietnam War left scars on our society for decades. COVID has killed almost twenty times as many. But on the level of impersonal numbers, one question is whether it’s enough to sway elections.

If deaths were evenly split between Republicans and Democrats, there’d be no electoral impact. However, we have grounds for believing this isn’t the case. There are stark demographic differences between the average members of each party, and some of these differences influence who gets sick and who dies.

Republican voters are more likely to be elderly, more likely to be men, more likely to live in rural areas with lower hospital capacity, and importantly, more likely to reject social distancing, masks and vaccines. All of these factors make them more susceptible to the virus. (COVID’s higher fatality rate among the elderly is well-known; it’s also more deadly to men, for reasons that are less clear.)

Counterbalancing this, COVID hit earliest and hardest in densely populated cities, which lean more Democratic. Also, Democrats are more likely to be Black or from other marginalized groups which, for historical reasons, are more likely to have comorbidities and more likely to lack access to health care.

Which of these factors matters most? The upshot is that death rates were roughly the same between Republicans and Democrats—until vaccines became available. After that, a partisan gap opened up, and it’s been widening ever since.

How wide a gap is it? A recent paper has some numbers:

Average excess death rates in Florida and Ohio were 76% higher among Republicans than Democrats from March 2020 to December 2021, according to a working paper released last month by the National Bureau of Economic Research.

… [The] paper found that the partisan gap in the deaths widened from April to December 2021, after all adults became eligible for Covid vaccines. Excess death rates in Florida and Ohio were 153% higher among Republicans than Democrats during that time, the paper showed.

“Covid death rates are higher among Republicans than Democrats, mounting evidence shows.” Aria Bendix, NBC News, 6 October 2022.

The 76% figure implies 1.76 Republican deaths for every Democratic death. If this estimate holds for months since December 2021, and if you assume that Florida and Ohio’s death rates are representative of the nation as a whole, then that yields roughly 700,000 deaths among Republican voters, compared to 400,000 deaths among Democratic voters.

(Does the existence of independents who don’t belong to a party complicate this result? Probably not, because the vast majority of independent voters lean strongly toward one party.)

These are ballpark estimates. They’re not, and can’t pretend to be, precise numbers. The most obvious complication is that it’s not only deaths that matter. An unknown number of people survived COVID but incurred serious disability, and some of them may not have been healthy enough to go out and vote. However, this gives us an idea of the magnitude of the effect we could expect to find.

Again, although these hundreds of thousands of deaths are an enormous tragedy on the human scale, they yield smaller numbers when divided among the populations of the fifty states. These aren’t the kinds of numbers that usually decide an election.

Of course, the key word is “usually”. In our hyper-polarized era, many elections are extremely close. A few voters not showing up in a critical race could make a big difference.

For example, the 2016 presidential election was decided by as few as 107,000 votes in three states—just 0.09% of all votes cast. A COVID-scale catastrophe could very well change the outcome of a future presidential election that’s this close.

What about 2022? Were there any close races where COVID deaths could conceivably have dictated the result?

Race #1: Arizona attorney general

As of 2022, Arizona had 32,000 total COVID deaths. About 60% of Arizonans are registered voters, so 19,000 of those deaths should be from this group on average.

Of those 19,000, if we assume the Florida and Ohio patterns hold up, this implies about 7,000 were Democratic voters and 12,000 were Republican voters. That’s a statewide partisan gap of 5,000 votes.

That sounds like a tiny margin, and it is. Except… in 2022, the Arizona attorney general’s race was won by the Democrat, Kris Hayes, by a mere 280 votes.

This is a plausible case where Republican voters dying of COVID changed the outcome. And if the 2024 election comes down to swing states like Arizona, it could matter very much which party holds the attorney general’s seat in case of a court challenge.

As far as I’m aware, there were no other statewide elections that were this close. Below the statewide level, things get more dicey, because there isn’t a reliable count of COVID deaths at this level of granularity. However, if we make a few additional assumptions, there were several local races decided by statistically interesting margins.

Race #2: Pennsylvania House District 151

In 2022, the Democrats flipped the Pennsylvania State House by winning 102 seats to the Republicans’ 101. One of those races, District 151, was won by a Democrat, Melissa Cerrato, by a mere 63 votes.

Pennsylvania has had 48,000 COVID deaths out of 13 million people, a statewide death rate of 0.37%. District 151 has 62,000 residents, so if we apply the statewide death rate, there should have been around 230 deaths in this district.

With a statewide 66% voter registration rate, we should expect 150 of those deaths to be among registered voters. Applying the 75% excess figure yields a result of 55 Democratic deaths to 95 Republican deaths, a 40-vote partisan gap.

That number is smaller than the winning margin, but this is only a rough estimate. If this district had a COVID death rate even slightly in excess of the statewide average, that may well have made the difference in this race, and therefore for control of the state House of Representatives.

Race #3: Michigan Senate District 12

In 2022, Democrats flipped the Michigan Senate, winning 20 seats to the Republicans’ 18. One of those races, District 12, was won by a Democrat, Kevin Hertel, by 404 votes.

Michigan has had 40,500 COVID deaths out of a population of 10,100,000, a statewide death rate of 0.40%. District 12 has 275,000 residents, so if we apply the statewide death rate, we get 1100 deaths in this district.

Michigan has a high rate of voter registration, nearly 80%. We should expect 880 of those deaths to be among registered voters. Applying the 75% excess figure yields a result of 320 Democratic deaths to 560 Republican deaths, a 240-vote partisan gap.

That number is smaller than the winning margin, but this is only a rough estimate. If this district had a COVID death rate in excess of the statewide average, that may well have made the difference in this race, and therefore for control of the state Senate.

Races #4 and #5: Massachusetts House, 1st Middlesex & 2nd Essex Districts

The final two cases are particularly interesting. In 2022, the 2nd Essex District of the Massachusetts state House of Representatives was won by a Democrat, Kristin Kassner… by one vote.

Another state district, the 1st Middlesex, was also won by a Democrat, Margaret Scarsdale, by seven votes.

Because Democrats already had a supermajority in the Massachusetts statehouse, the results of these elections won’t decide control of the chamber. However, these are two clear cases where even a tiny excess of COVID deaths among conservative voters could have decided the results.

What the future holds

The biggest question is what the future holds. SARS-CoV-2 could evolve into a less lethal form, becoming like other coronaviruses which only cause common colds. In that case, COVID death rates would decline over time and the electoral impact would fade away.

But there’s no law of nature that says this has to happen. COVID could keep circulating, and deaths could keep mounting. Unvaccinated people lean Republican by a 3-to-1 margin, and they die at up to fifteen times the rate of vaccinated people.

For that matter, vaccination in general is becoming a political football. There’s now a partisan divide in flu vaccination, which wasn’t the case before 2020. Measles, chickenpox and even polio are also making a comeback among unvaccinated communities.

It’s not a reach to speculate that, if Republicans become entrenched as the anti-vaccine party, diseases of all kinds will continue to exact a disproportionate toll on them. This means that the COVID-induced partisan gap, which was relatively small in 2022, will continue to grow and will make a bigger and more significant difference in future elections.