Above is a diagram I drew back around the end of April 2020.

I thought it would be worth doing a one-point-in time update of that same sketch, with current data. And then maybe review where things stand. Quick answer is, probably about half-way there, toward herd immunity.

I can’t update that prior diagram with a lot of precision. But I can do some estimate of it. So that’s what I line out, in this post.

There are four pools of people now assumed to have some immunity to COVID-19.

As if 1/21/2021:

- 024.1M Diagnosed with COVID, reference.
- 096.4M Had COVID, no formal diagnosis, estimated from CDC studies (per Post #940).
- 012.1M Partially vaccinated (1 dose, calculated), ~54% immunity, reference
- 002.2M Fully vaccinated (2 doses), ~90% immunity, reference.

Now for the uncertainties.

The numbers for Pools 1, 3, and 4 are known.

** The number for Pool 2 — by far the largest — is an estimate,** with a lot of uncertainty around it (see Post #933, Post #940). What you see there is my best guess, based on an estimate of five actual COVID cases for every diagnosed case.

**The overlap between Pools 1+2 and Pools 3+4 is unknown,** and will probably never be known. As noted in a recent post on vaccines (Post #959), the CDC recommends that you get vaccinated even if you’ve had COVID. And so, some fraction of Pools 1+2 will be double-counted in Pools 3+4. But we have no idea what fraction that is.

The best I can do is estimate that the overlap occurs at random, that is, in proportion to the two populations. As if you flipped one coin to become part of Pools 1+2, and then flipped a separate count to become part of Pools 3+4. And when I de-duplicate the counts, I’ll take the count out of the vaccinated pool, because I want to count 100% of those overlapping individuals as being immune to COVID-19.

And so, how many people in the US are immune to COVID now? Here’s my calculation.

And so, best guess, 38% of the U.S. population is currently immune to COVID-19. Or, in round numbers, half of what at one time was considered the herd immunity target, before they started moving the goalposts (Post #928). If you don’t want to de-duplicate for potential overlap, then count 39% as immune.

**Except for showing you how things have changed, almost nothing about this graph is right or certain.** The scale at the bottom is wrong (we’ve taken longer to get here), the shape of the line is wrong (COVID has progressed in waves, not smoothly). The assumption that 70% will give us herd immunity is just that — an assumption. And the actual you-are-here point depends critically on the ratio of all COVID cases to diagnosed COVID cases.

Other than that, this is a fine summary of where things currently stand.

I’ll just finish off with the graph of recent US trend. New cases per day continues to fall, and that’s because it’s now falling in almost every U.S. state. The rate is now back down to were it was in early December 2020.

Source: NY Times Github COVID data repository, data reported through 1/20/2021.