Two Percent of Americans Are Currently Covid-Infected

The question isn’t how many people have been infected since the beginning of the pandemic. The question is: how many are currently infected by the virus, currently carrying active virus inside their bodies?

The Assumptions

On average, a person infected with covid remains infected for 2 weeks. So the total number of infected people on any given day is the sum of those who have been newly infected each day for the prior 14 days.

The daily covid death count is the most accurate indicator of the spread of the virus. However, deaths lag behind infections by about three weeks; i.e., the average person who dies from covid was infected 3 weeks prior to dying.

Somewhere around 0.65 percent of those who’ve been infected with covid eventually die from the disease. So, divide the most recent 14-day death count by .0065 to estimate the number of people infected 21 days ago.

The daily case count in the US underestimates the daily infections. However, over the past month the change in daily case counts has fairly accurately mapped onto the change in estimated daily infections.

The Math

Over the past two weeks, November 6-20, 18,100 Americans died of covid. So, as of three weeks ago, there would have been 18,100/.0065 = 2.8 million Americans actively infected by the virus.

Over the past two weeks (11/6 – 11/20), the 14-day total new case count was 2,212K. Three weeks prior (10/15 – 10/29) the 14-day total new case count was 992K. So, the current case count is 2212/992 = 2.23 times what it was two weeks ago. Assuming proportionality of changes in case counts to changes in new infections, then the total number of infected people today is 2.23 times the number of people who were infected three weeks ago.

2.8 million infected 3 weeks ago x 2.23 = 6.24 million Americans are currently infected by covid.

The total US population is 328 million. 6.24/328 = 0.019. So, about two percent of the American population is currently infected by live covid virus.

6 thoughts on “Two Percent of Americans Are Currently Covid-Infected

  1. I had recently remembered when you found that sign that said CAUTION and came home after your run and told Anne “I just threw caution to the wind”. And then Anne called her mother and she said “Tell him I said HA HA HA”. I imagine her as somewhat like a character in Lonesome Dove, and she shot up some varmints too. Groundhogs, I think.

    I go out every day, usually for at least a mile, and often two. Not that things aren’t getting bad again, but I’m used to it. At first, there was such panic you were supposed to use hand-sanitizer before you touched a bank door and after you went out. I was wearing a good mask early on and still have two KN95’s, but I don’t think about it that much now that I was finally able to remember to put on my mask every time I go out. I forgot it maybe 9 times before I made a mental exercise for it. I do the same finally for my reading glasses, because not being able to find your glasses is a special form of hell.

    My god, you 2 were writing before Thanksgiving. I had Duck Rillettes, Stilton Bleu and an Apple-Orange Cake with Grand Marnier Butter Cream Frosting. I remember your elaborate cooking. Do you still do it? I hope so. It’s important to be pompous in the middle of a crisis…


  2. Right, I wrote this post on 21 November — time to update it. I’ll look at the current numbers and trends, mostly to see whether the predictive assumptions and algorithms from this post are validated by updated data. I need precisely 3 weeks’ worth of new data to do that, which will be tomorrow.

    Nearly all of the posts in this series have been motivated by the same question: how many people have been infected? There are no direct measures — dx-positive case counts are contingent on test availability and criteria for selecting people to be tested. So there’s a lot of estimation and uncertainty, repeatedly iterating between hypotheses and equations and data. It’s a kind of ficticity in the sense that the knowledge isn’t read directly off the world, but is rather a constructed abstract alternate reality. The challenge is isn’t to make the abstraction represent the actual world point by point, but rather to have the interrelationships of objects and forces comprising the model correspond with the empirical residue left behind by virus-to-human transmission occurring the actual world.

    I’ve written only one extended post about covid preventive behaviors, but I posted that one on Carl Dyke’s blog because it fit more with that blog’s agenda than with this one’s. Here’s a calc about behavior that’s more up my alley:

    Let’s say that 2 percent of the US population currently has covid and is contagious. That means that, for any random person you or I might encounter in the world, there’s a 2 percent chance that that person is currently infected. Suppose I’m in an indoor setting with 10 unmasked people: what’s the likelihood that at least one of them is currently infected/contagious? It’s 1 – (1 – .0210) = 18 percent. Suppose I’m in such a 10-person setting once a day for a week: now the likelihood is 1 – (1 – .0270) = 76 percent that at least one of those people is currently infected/contagious. How much risk am I willing to take?

    I get out a couple of times a day walking and/or running, maintaining the 6-foot social distancing rule on sparsely trafficked routes rather than masking up. Your Thanksgiving feast sounds marvelous. I'd put up a photo of the chicken liver and pork terrine I made the other day, but I don't know how to do that in a blog comment. It was delicious featured on the plate in a charcuterie dinner, along with a bottle of old French wine.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s