Today there’s a big jump in the seven-day moving average of U.S. new COVID-19 cases per 100,000 per day. But I don’t think it signals any fundamental change. I think it’s mostly an artifact of this year’s holiday data reporting.
Details follow.
As of today, it’s still a U.S. winter wave in just two regions: Midwest and Northeast. The rest of the country has new case rates that are steady — more or less what they were at the start of the U.S. winter wave on 10/25/2021.
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 12/3/2021, from https://github.com/nytimes/covid-19-data.” The NY Times U.S. tracking page may be found at https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html.
That said, there was a big jump in the seven-day average of new cases today. And, not coincidentally, today is the day that the literal Thanksgiving day cases fall out of our seven-day moving average.
My best guess is that the true (long-term) trend is flat, and that today’s uptick is a consequence of even poorer data reporting this year, compared to last year.
Above, we saw an even greater dip in cases around Thanksgiving this year than we did last year. That’s indicated by “Lower” above. But that was probably due to worse reporting, not to an actual decline in cases.
So the greater volume of “missing” cases, from the “lower” period, simply got reported a few days later. And now, as we drop those under-reported days from the 2021 data, we’re left with those extra cases still in the seven-day data window. Leading to the second I’ve labeled “higher” on the graph above.
Every holiday, some cases get reported late. They aren’t reported on the holiday day itself, but get reported a few days later. I think that this year, that effect is larger, owing to generally worse data reporting (e.g., more public health offices closed, for longer periods, over the holiday).
The upshot is that I think this bump-up in new case rates is temporary. That’s a little bit of reading the tea leaves, because I can’t (easily) directly measure the day-by-day under-reporting. It’s just a bit too much of a coincidence for me to believe that we got a big, true change in trend — on the exact same day that we’d expect to see any artifacts of changes in data reporting appearing.
We’ll know one way or the other by the middle of next week.
