Dismissing concerns that states are reopening too soon, President Donald Trump incorrectly said that a newly revised model projecting 134,000 COVID-19 deaths by August “assumes no mitigation.” In fact, the model assumes states will keep their existing social distancing measures in place, unless suspensions have already been announced.
Trump’s error came when he was asked about a model from the University of Washington’s Institute for Health Metrics and Evaluation. The group’s modeling efforts previously figured prominently in presentations from the White House’s coronavirus task force, and until a few days ago, had been projecting around 72,000 deaths in the U.S. from the coronavirus by August. According to Johns Hopkins University’s COVID-19 tracker, there were already more than 70,000 deaths as of May 5.
Reporter, May 5: Mr. President, on the IHME model, which is now showing 134,000 deaths by August, doubling its previous prediction, are you concerned that that’s happening because some of these states are relaxing guidelines too early?
Trump: No. No, I’m not, because that — that assumes no mitigation.
Reporter: Assumes less?
Trump: And we’re going to have mitigation. No, we’re letting people out. But the fact that we’re letting people go and go to their jobs — they have to do it.
When asked later if he believed the death toll numbers, Trump again suggested that the higher estimate from the model was because it didn’t incorporate any mitigation efforts.
“Well, that’s with no mitigation. We’re doing mitigation. We have a lot of mitigation,” he said. “The fact that they’re out — they’re mitigating. They’re social distancing. They all know that. They’re washing their hands a lot. But we have to get our country open. We have to open our country.”
The president then proceeded to draw attention to a different death estimate, made by Imperial College London early on in the pandemic, that unlike the IHME model, was done assuming no mitigation.
“I’ve seen models that are very inaccurate,” Trump said. “But, you know, one model that’s very important is that if we did this a different way, we would have lost more — much more than 2 million people. And we did it the right way. We did everything right. But now it’s time to go back to work.”
In a March 16 report, researchers with Imperial College London projected as many as 2.2 million deaths in the U.S. from the coronavirus in “the (unlikely) absence of any control measures or spontaneous changes in individual behaviour.”
It’s fair for Trump to note that some models haven’t performed well, and to be skeptical of them in general.
Many experts have been especially critical of IHME’s methods, as they essentially take previously reported COVID-19 death data from around the world and try to extrapolate those patterns and apply them to other outbreaks, without accounting for differences in how those outbreaks are being handled. That’s in contrast to epidemiological modeling, which actually tries to use what’s known about the virus and how it spreads to model disease transmission in a population.
But Trump is wrong to say IHME’s model doesn’t include any mitigation. As Amelia Apfel, a media relations officer for IHME, told us, “[O]ur model has always assumed mitigation.”
The new IHME model, which was released on May 4 and now includes some elements of disease transmission modeling, incorporates a variety of government-mandated social distancing policies.
This is most obvious when viewing IHME’s projections for individual states. The top of each page shows which policies the model is assuming are in effect.
“As of May 4, our forecasts now account for locations that have lifted social distancing measures and assumes those measures will remain lifted through August 4,” the website says. “For locations that have not lifted social distancing, the model assumes measures will stay in place through August 4.”
The state of New York, for example, has not suspended any of its mandates, so the model’s projections are based on the idea that schools and businesses remain shuttered, and gathering restrictions and stay-at-home orders are in place. In Colorado, some social distancing measures still apply, but the model is factoring in that the stay-at-home order was lifted and some businesses were allowed to reopen.
IHME’s website further explains that it only classifies measures as being in place if they pertain to the entire population within a given state. So, if a city had ordered schools to be closed, but that order didn’t pertain statewide, the measure wouldn’t be part of the model.
Apfel said that the team incorporated information about states’ social distancing policies, including any that had ended or were scheduled to end, as of May 3. “[A]s we update going forward,” she said, “we’ll be rolling in changes to state social distancing policies that we’re able to verify.”
The model, then, reflects some relaxation of policies as states begin to reopen, but is still trying to capture the mitigation efforts that remain. This is one of the key reasons why the IHME numbers increased — to 134,475, with a range between 95,092 and 242,890 — relative to its earlier estimates.
“The revised projections reflect rising mobility in most US states as well as the easing of social distancing measures expected in 31 states by May 11, indicating that growing contacts among people will promote transmission of the coronavirus,” explains a press release for the new model.
It remains to be seen whether the new IHME estimate is anywhere close to being accurate, but the president was incorrect to say the model “assumes no mitigation.”
FactCheck.org does not accept advertising. We rely on grants and individual donations from people like you. Please consider a donation. Credit card donations may be made through our “Donate” page. If you prefer to give by check, send to: FactCheck.org, Annenberg Public Policy Center, 202 S. 36th St., Philadelphia, PA 19104.