The first thing President Trump said at the second presidential debate was false.

“As you know,” he said in response to a question about his handling of the coronavirus epidemic, “2.2 million people modeled out were expected to die."

Trump was referring to a study published by Imperial College London in March.

“In total,” it concluded, “in an unmitigated epidemic, we would predict approximately 510,000 deaths in [Great Britain] and 2.2 million in the U.S.," even excluding the number of deaths that would result from hospitals being filled with coronavirus patients.

The key word there is “unmitigated.” That's what the death toll could have been by midsummer if the country were to do literally nothing: keeping everything open, yes, but also not even isolating sick people. Even had the federal government done nothing, states would nonetheless have acted, as some did in advance of the White House's eventual embrace of shutdown measures. In other words, it was not the case that 2.2 million deaths was the baseline of what should have been expected.

It's obvious why it's useful for Trump to cite that number, of course: the bigger the worst-case outcome, the better the actual outcome looks. By the White House's own measure, though, the actual outcome has been bleak.

When Trump's coronavirus task force first called for closing parts of the economy to contain the virus in March, it produced a graph using that figure as the upper limit of what could have happened. A mitigated pandemic, on the other hand, would mean that only 100,000 to 240,000 deaths would occur.

As of writing, at least 222,000 people have died of the virus. The key phrase here is “at least,” but we'll come back to that.

Shortly after Trump's 2.2-million claim, former vice president Joe Biden used the confirmed death toll to criticize how the administration had handled the pandemic.

“Two hundred and twenty thousand Americans dead,” Biden said. “If you hear nothing else I say tonight, hear this: anyone who's responsible for not taking control, in fact, [saying] I take no responsibility initially — anyone who's responsible for that many deaths should not remain as President of the United States of America."

This, too, is misleading. Trump can't be considered accountable for 220,000 American deaths from the coronavirus. At least: not yet.

Assessing the number of people who might have died had the federal government acted differently is tricky for three reasons. First, the actual number of deaths so far is a bit murky. Second, the number of deaths the country might have seen involves a fair amount of speculation. And, third, people are still dying at the rate of 1,000 a day, meaning that we're nowhere near knowing what the final toll from the virus will be.

How many have died?

As noted above, there are at least 222,000 confirmed deaths to date. Many of those came at the outset of the pandemic, when undetected infections spread from person-to-person before containment measures were implemented.

Because of how the virus works — infections are identified a week or two before patients succumb — surges in new cases have preceded surges in deaths. You can see that in the recent data: cases began to increase at the end of last month; deaths began to increase over the past week.

What isn't captured is any unconfirmed deaths, deaths which followed undiagnosed infections or which preceded America's slow rollout of testing. To estimate that total, we look at excess deaths, how many people died in 2020 compared to preceding years. Deaths follow seasonal patterns, and exceptional mortality events therefore stand out statistically.

A report from the Centers for Disease Control and Prevention released earlier this week estimates that “an estimated 299,028 excess deaths have occurred in the United States from late January through October 3, 2020, with two thirds of these attributed to COVID-19,” the disease caused by the virus.

That's roughly in line with the confirmed total through the same date, but this estimate, too, has limitations. More recent deaths might not yet have been counted, for one, and because “deaths from other causes might represent misclassified COVID-19 — related deaths,” as the report notes. Nonetheless, it seems fair to assume that the number of deaths is around the 222,000 total.

How many might have died?

This is a trickier question to answer.

One way to look at it is to compare the death toll in the United States with outcomes in other, similar countries. That's the approach taken by researchers with Columbia University's National Center for Disaster Preparedness. This week, they released a report estimating that between 130,000 and 210,000 deaths from covid-19 didn't need to occur, had the country's response been as effective as those of other nations.

In South Korea, where the first infection was detected on the same day as the United States, there have been only 0.85 deaths per 100,000 residents. That's largely a function of the robust containment measures the country undertook right off the bat, unrolling a broad testing regimen that allowed it to halt the progression of the virus. Had the United States been able to do something similar and achieved a 0.85 per 100,000 death rate, only about 2,800 people would have died here.

Even if we'd matched Germany, where the death toll was 11.72 per 100,000 residents, we'd still have seen fewer than 40,000 deaths. Instead, the number of deaths in the United States is the highest in the world, and among the highest when controlling for population.

What might the United States have done instead? Delays in implementing a national containment strategy might have led to 36,000 additional deaths, according to another team at Columbia. Then, of course, there are personal mitigation strategies, like encouraging the wearing of face masks and avoiding crowded indoor spaces.

“If we just wore these masks,” Biden said at the debate, holding one up, “the president's own advisers have told him we could save 100,000 lives."

Which is true — and effects what comes next.

How many more people will die?

Many assessments of how the country has fared center on the extent of the pandemic to date. But we're no better able to fully evaluate our success or failure now than we are to know who is winning a horse race halfway around the track.

An estimate of future coronavirus deaths in the United States compiled by the Institute for Health Metrics and Evaluation at the University of Washington puts the death toll here at nearly 400,000 by early February of next year.

There's some statistical uncertainty to that estimate, indicated by the shaded area. One takeaway, though: by mid-November, the country will have surpassed 240,000 deaths, the upper-bound on what the White House in March identified as a successful pandemic response.

There's also uncertainty because we don't know how effective Americans will be at limiting the virus's spread. If everyone wears masks all the time, the estimate drops to about 322,000 deaths by Feb. 1. If we ease containment measures, the toll rises to 485,000.

A separate study from the IHME team published in the journal Nature Medicine had a higher estimate of the toll over the next few months.

“Projections of current non-pharmaceutical intervention strategies by state — with social distancing mandates reinstated when a threshold of 8 deaths per million population is exceeded (reference scenario) — suggest that, cumulatively, 511,373 (469,578—578,347) lives could be lost to COVID-19 across the United States by 28 February 2021. We find that achieving universal mask use (95% mask use in public) could be sufficient to ameliorate the worst effects of epidemic resurgences in many states. Universal mask use could save an additional 129,574 (85,284—170,867) lives from September 22, 2020 through the end of February 2021, or an additional 95,814 (60,731—133,077) lives assuming a lesser adoption of mask-wearing (85%), when compared to the reference scenario.”

That's a lot of numbers, but the central points are twofold. First, that the country could see more than 500,000 deaths by the end of February without therapeutic drugs being widely available. Second, that near-universal mask-wearing could shave 130,000 deaths off that figure. Put another way, nearly half of the expected toll could be prevented if Americans consistently wore face masks.

These are estimates, but they are obviously ones in our power to affect. It's probably more in Trump's power to affect, given that his general indifference to encouraging mask-wearing is almost certainly a significant driver of the partisan divide in doing so. A Yahoo!-YouGov poll earlier this month found a wide gap in self-reported mask-wearing by party.

Put all of this together and we arrive at a fair assessment of the situation that runs as follows:

  • At least 200,000 people have died of covid-19.
  • Tens of thousands of those deaths, at least, were preventable.
  • Tens of thousands of more deaths will occur, many of them also preventable.

Biden’s assertion that Trump bears the blame for 220,000 American deaths was overstated. But it’s still demonstrably more accurate than Trump’s shrugging claim that 2.2 million people might have died on his watch and his past claims that the government’s response was an unmitigated success.