Will 4% of Americans Who Get Coronavirus (10 Million People) Actually Die?

Sorry. I like this article. Kind of. But I question parts. Here’s my two cents along with quotes by medical experts saying 4-40 times less people may die per COVID-19 (Coronavirus) infection than stated by “Marketing Leader” Thomas Pueyo in his viral article.

David Shuey
10 min readMar 21, 2020


That’s 1 out of 30 citizens. But currently in South Korea they’ve stabilized at 1 in 540,000 citizens.

“For the back-of-the-envelope numbers: if ~75% of Americans get infected and 4% die, that’s 10 million deaths, or around 25 times the number of US deaths in World War II.” — Thomas Pueyo

First, I want to say: I’m all for using a “hammer” Tomas Pueyo talks about if necessary to stop the spread of this virus. i.e. extreme measures and social distancing. Most every leader is coming to this realization as they’re piecing together quickly moving epidemiological facts and projections. In fact, I’m writing this in social isolation in Chicago, Illinois, where the governor just declared a “stay at home” order. Most of you readers are doing the same.

This is verifiable and true. See news headlines today and Italy’s population.

Additionally, the healthcare sector should be given all the resources they’re asking for — test kits, ventilators, masks, swabs to administer tests, civil understanding (as surgeries are postponed, upsetting patients) — as they attempt to draw all their resources together for a worst-case scenario. It’s already happened in Italy, the hardest hit country, with hospitals at their breaking point and Italian military called in to impose a “coronavirus lockdown.”

But let’s be honest with the numbers. Especially when currently China’s only lost .0002% of their population to COVID-19. (The math: About 3,300 deaths on March 20 divided by 1.3 billion people.) Additionally and remarkably, China has stabilized new cases just like South Korea. In fact, South Korea’s also only lost .0002% of their population to COVID-19. Let’s hope it remains there.

In Italy, .01% of their population has died from COVID-19. (The math: 4,000 deaths on March 20 divided by 65 million residents.) Indeed, Italy is suffering the highest losses per capita of any country, with 1 person dying from Coronavirus out of every 15,000 citizens. They even list a case mortality rate of over 8% from the numbers I’m tabulating daily on a spreadsheet using WHO and CDC sources, a percentage also shared by mainstream sources.

But Pueyo’s conclusion that the U.S. will end up with a 4% case mortality rate mentioned in his Medium piece — a continuation of an explosively popular earlier piece with 40 million views — is way too high. Where does Thomas Pueyo get this figure? Does he factor in the fact these percentages mentioned now will certainly not be the final percentages, as we’re only testing a fraction of populaces?

He doesn’t really say clearly and provides no reference, as Daniel Sweeney an actual epidemiologist who also commented pointed out. It’s worth also saying that Pueyo doesn’t have health or science credentials. But hey, neither do I! So I’ll crunch some straight-forward math, too. Additionally, I’ll quote qualified experts with credentials who say the case mortality rate is far smaller than what Pueyo throws out with his gorgeous charts. I’m even tossing in some mediocre memes with 100% accurate facts.

Someday, I’ll learn to make some fancy graphs, as I do love this illustrative YouTube video based on Pueyo’s work watched by over 1 million viewers. Again, I think Pueyo has an important perspective to offer. And he does have some credentials: He’s a Silicon Valley “Marketing Leader” and “VP of Growth at Course Hero“ according to his LinkedIn bio. OK, this bit of information makes me a little skeptical.

Also note: Pueyo doesn’t quote any epidemiologists, but does credit a team of “normal citizens” that helped him, including “Dr. Carl Juneau (epidemiology)” who has a PhD in Public Health who runs a Dr. Muscle website specializing in exercise. It looks like Dr. Juneau first connected with Pueyo via a March 13 comment on Medium. He also has a solid website full of moving charts: https://dr-muscle.com/s/covid.html

Let’s Break Down the Numbers Further

One reason 4% is too high is simple: Many times more people have COVID-19 than will ever likely be tested. Some experts like Steve Goodman, a professor of epidemiology and population health at Stanford University, say this is as high as 16 times more in some areas of the U.S. Thus, the number who die is really the “canary in the coal mine” and later when the epidemic has run its course we’ll be able to figure out the final percentage. But for now, many are merely dividing number of confirmed cases by the number who have died. This, some experts argue, creates far too high a rate. For example, if South Korea had six times more cases than they tested their .6% case mortality rate would be .1%, as new deaths have slowed to a trickle.

And I’m just not saying that percentage out of thin air. An expert from Johns Hopkins University told it to neuroscientist Sam Harris.

This is verifiable and true. See WHO data and China’s 1.4 billion population.

Thus, I don’t understand why sensationalism is needed on such a widely respected and reported piece. Unless it’s for marketing purposes. I even posted on Facebook the Buzzfeed article about the author with an appropriately snarky title: “I’m Not An Epidemiologist But…”: The Rise Of The Coronavirus Influencers. (They note Pueyo is represented by a content marketing firm that promotes his work.) But I’ve seen these dramatic death count numbers said quite a few times for more than 10 days. I don’t question the modeling, per se, but I question how they stack up to past epidemics, and what we’re experiencing now.

We have to keep in mind that even the Spanish Flu killed less than 1% of Americans and likely just under 2% of the world’s population, and we had far less capacity to combat it. And even as Italy’s deaths far outstrip China’s, Italy would have to have at least 150 times more deaths than the current total of around 4,000 deaths (which took 1 month to occur) to lose 1 percent of their population. Again, I do understand that Italy’s health system is also completely overloaded and all countries want to avoid this nightmare.

This is verifiable and true. See WHO data and South Korea’s population.

Experts Saying .1%-1% Case Mortality Rate

Source #1: On March 11, Sam Harris had infectious disease expert Amesh Adaljasay on his program. Adaljasay explained to Harris that because South Korea’s case mortality rate was .6% that this was the high end and therefore the low end could possibly be .1% (i.e. 1 out of 1000 people contracting the disease may die). This, of course, is 40 times smaller than Pueyo’s arbitrary number. Adaljasay isn’t a nobody. He is an MD working on pandemic policy at Johns Hopkins University Center for Health Security who’s been speaking almost constantly in major media about this outbreak. Adaljasay also warns that the case mortality rate could be as high as 1% in Vox on March 12 but doesn’t go anywhere near 4% while stating the same concerns as Pueyo:

I think the worst-case scenario is probably a 1 percent case fatality ratio with a lot of critically ill patients that need to be admitted to the ICU, and the majority of our ICU beds are already occupied. [Vox’s Julia Belluz explained why it’s so hard to pin down the death rate for Covid-19.] So the worst-case scenario is that we start running out of ICU beds and we start running short on mechanical ventilators and then we have to make hard decisions about how to optimize the scarce resources that we have.

I hope that we don’t get to that point, but we have to prepare for it.

Source #2: The New York Times posted it could be a .5% case fatality rate in a piece damning President Donald Trump’s response titled President Trump Is Unfit for This Crisis. Period.:

Because we’re testing only the sickest of the sick, the American fatality rate from the Coronavirus is roughly 4 percent. It’s a frightening and highly deceptive number, even higher than China’s. (Most experts predict it’s likely to wind up at 0.5 percent, which is five times more deadly than the typical flu, and it could be as high as 1 percent.)

Source #3: Physician Joel Zinberg said it could be close to .1% case fatality rate, too. He said so in this podcast:

Dr. Zinberg also wrote it could be .1–2% in a City Journal article:

The most recently reported case-fatality rate for laboratory-confirmed Covid-19, for hospitalized inpatients in China, was 1.4 percent; and since many more unreported, mild, or asymptomatic cases exist than the number of reported cases involving more severely ill or hospitalized patients, the fatality rate is likely much lower — perhaps as low as the 0.1–0.2 percent fatality rate of seasonal influenza. And unlike influenza, which has killed more than 125 children in the U.S. so far this season, Covid-19 appears to spare the pediatric population.

Yet Pueyo says the following in a bold pull out quote as to how the fatality rate could be close to 4%:

If 5% of your cases require intensive care and you can’t provide it, most of those people die. As simple as that.

“As simple as that.” That’s it. No real argument. Just an aside that the strain in the U.S. may be more severe than in China according to a CDC source. With all the math he renders, one would think he’d confront this “4%” figure head-on.

Then again, he’s just trying to get our attention. I’m hoping his intentions are to benefit society and not himself. Perhaps the ends justifies the means.

10 Million Americans Dead from COVID-19?

Also, 10 million dead is about 1/30th of our population, around 3%. Even the Spanish Flu with a 2.5% case mortality rate killed between 0.48 -0.64% of the population by infecting 28% of the population. And this was back in 1918! Do I need to get into detail how inadequately the world was prepared scientifically, civically, and economically to handle this? How is it Coronavirus is going to end up killing six times more Americans per capita a century later?

I do understand COVID-19 is more contagious than the common flu. I suppose I’m just curious how so many epidemiologists are saying a majority of humans in many countries could be infected:

66% of Germans are going to contract it, says Angela Merkel:

“When the virus is out there, and the population has no immunity and no vaccination or therapy exists, then a high percentage — experts say 60 to 70% of the population — will be infected, so long as this remains the case,” she said.

50% or more of Americans:

Last week, Republican members of Congress heard a sober warning in a closed-door briefing on Capitol Hill: There’s a good chance most people in the United States will eventually be exposed to the novel coronavirus, according to one former official.

So sure, hypothetically, if “75% of Americans get infected and 4% die” as Pueyo writes that is indeed “10 million deaths, or around 25 times the number of US deaths in World War II.” It sure smacks you in the head, like a catchy Mad Men ad copy. Yet couldn’t Pueyo have just gone with 50% contracting Coronavirus with a 1% case mortality rate and say 1.5 million die? That’s shocking enough. It’s possible 10 million is a nice round number and he worked backwards from there. Which isn’t exactly scientific, when you think about it.

This is my armchair opinion: Some people work with numbers and they go to the extremes. They paint doomsday scenarios because they see those possibilities. I’ll mention my bias: I tend to see how rarely those possibilities come to fruition. I downplay things. This is because I often see people — and the media — exaggerating or manipulating data to argue bias in a system that is inherently broken. I don’t see it as thus.

I also remember how SARS only killed 774 worldwide and H1N1 (Swine Flu) killed 12,469 in the United States.

I’m currently running the odds of how this virus compares to the flu, which apparently is the most ridiculous and unscientific thing you can do.

Well, we’ll see if we surpass the flu. I still have my doubts. And if we don’t, many will certainly say it took a historical groundswell of behavioral changes to stave off mass death. We won’t have the counterfactual. Still, if we don’t even come close, will anyone say, “Well, maybe we could have approached this differently?” I doubt it.

The question I’m still asking for the past 10 days: “Why don’t we commit to the same actions for the flu? It kills 500,000 worldwide every year.” I know the answer: This is different and unknowable and we don’t have a vaccine.

I suppose it’s better to be safe than sorry as the economy suffers potentially its most colossal slowdown in history. Together we can prevent the worst outcomes. But let’s stick to the facts along the way.



David Shuey

Writer. Researcher. Designer. Human seeking better outcomes for all. Empiricism, relevant facts, and logical arguments > simple narratives.