Post #1261: COVID-19 trend to 9/21/2021

Posted on September 22, 2021


The U.S. now stands at 41.8 new COVID-19 cases per 100K per day.  Cases are down 13% in the last seven days, and are down 19% since the 9/1/2021 peak of the Delta wave.

Data source for this and other graphs of new case counts:  Calculated from The New York Times. (2021). Coronavirus (Covid-19) Data in the United States. Retrieved 9/22/2021, from”  The NY Times U.S. tracking page may be found at

Looking ahead

Alaska remains the only state with more than 100 new cases per 100K per day.

The center of case growth continues to shift northward.  To me, that northward shift is  expected, and looks like the prelude to the start of the U.S. winter wave.

Pct change in new COVID-19 cases, week ending 9/21/2021

Map courtesy of  Scaling on this map differs from prior growth maps because some states are outside the bounds of the scale on prior maps.

History doesn't repeat itself, but it rhymes.  Below are the first (blue) and second (red) years of the pandemic, starting April 1 2020 and 2021, respectively.  In green, I've sketched in a plausible path for daily new cases.

My best guess is that at some point -- probably mid-October -- the declining Delta wave will be overtaken by a rising 2021 winter wave.

When (if) that happens, the pandemic will probably become a race between herd immunity and hospital capacity.  Will the hospitals fill up before enough people have gotten some immunity, either via vaccination or via recovery from infection?

My guess -- presented in my last post on the COVID-19 trend -- is that if we have a winter wave that is as bad as last year's winter wave, many states will reach herd immunity.  That was based on my crude count-em'-up method of adding vaccinations to (gu)estimated total infections, less assumed overlap of the two.  If you count 'em up under my assumptions, a lot of states should end up in the neighborhood of herd immunity with a repeat of last year's winter wave.

I just had an astute reader offer his own estimate of that, using a statistical measure rather than my accounting-based approach.  He notes a strong negative correlation, at present, between state-level new case growth and state-level COVID-19 immunity (calculated as the sum of estimated infections plus all vaccinations, less overlap).  The higher the immunity level, the slower the current new case growth.  Most states are in the mid-70-percents on that immunity measure right now.  Based on that observed relationship, he estimates that new case growth will fall to zero as states pass 91% immunity.  That is, they'll reach herd immunity when the sum of vaccinations plus infections, less overlap, exceeds 91% of the population.

If you step back from it, his conclusion based on that statistical analysis, is more-or-less the same as mine, based on my "herd immunity" table:  A repeat of the winter wave should push a lot of states into herd immunity status.

A couple of lessons from Virginia

Finally, it's worth taking a moment to re-work the Virginia data on infections by vaccination status.  Because I think they tell a fairly important story about which segment of the population is going to be responsible for straining hospital capacity in the winter wave.

Here's the chart I showed yesterday:

Source:  Virginia Department of Health COVID-19 dashboard.

Now let me re-do that, not in terms of the infection rate, but in terms of the fraction of all new COVID-19 cases.  Per the Virginia Department of Health, the population of Virginia currently breaks out as follows:

  • 58.8% of the population is fully vaccinated.
  •   9.2% of the population is partially vaccinated
  • 32% of the population is unvaccinated/

(N.B., A large chunk of the unvaccinated are children for whom no vaccine has yet been approved.  If we restrict this to the adult population, more than 70% of Virginians are fully vaccinated.)

Assuming I can do the math right, then the share of all new COVID-19 cases in Virginia is as follows:

If the hospitals start to overflow this winter, it's clear which segment of the population will be to blame for it.

On the brighter side, I still see no evidence of any explosion of cases from K-12 schooling in Virginia.  Below is the picture as of yesterday.

Source:  Calculated from Virginia Department of Health case counts by age and health district, crosswalked to school districts grouped by school start date.

Whatever cases are being generated by K-12 Virginia public schools, their impact is too small to be seen in terms of the share of all new cases attributable to the under-20 population.  That share has held steady at 25% since mid-August and it appears completely unrelated to the timing of the opening of K-12 schools.