Edit: In hindsight, this post and the next post were wrong. What you’re seeing below is mostly the last of the data reporting glitches occurring around the holidays. See Post #948 for a re-hash of this issue.
Original post follows.
Source: Calculated from NY Times Github COVID data repository. Data reported through 1/8/2021.
Recall that the previously-stated hallmarks of a holiday-driven surge would be an abrupt uptick in new COVID cases per day, occurring simultaneously across many or all states, arriving between twelve days to three weeks after the travel and socializing for that holiday began (Post #915). If all those pieces fall into place — abruptness, simultaneity, and timing — that’s about as good a case as you can make regarding causality — that the increase in cases is due to the prior holiday.
No such thing happened for Thanksgiving (Post #922). But it sure looks like we are now seeing the post-Christmas/New Year’s surge in new COVID-19 cases.
The circled area on the regional/national chart above satisfies all my criteria. You can see the simultaneous upswing in all six regions. The graph below illustrates the abruptness and timing criteria.
Above, I have sketched in, by eye, what I think a continuation of trend would have looked like (dotted line). And I’ve marked January 3, which is twelve days after the start of Christmas travel, based on air travel data (Post #939). I would characterize the steeply-sloped end of the blue line as an abrupt change, and it began to ramp up just about twelve days after the start of holiday travel.
(Separately, we’ve passed the point where this could even remotely be attributable to data reporting issues around the holidays (Post #941).
And so, any notion that we’re past the peak of the third wave was wrong (Post #930). That assumed no post-Christmas surge (because there was no post-Thanksgiving surge.) And that now appears totally incorrect.
How much higher, and how much longer, for this third U.S. COVID wave, is unknowable. If the “nearing herd immunity” hypothesis is right, there will be modest and short-lived uptick in the Midwest, owing to the high fraction of persons who have already been infected. Using a fairly conservative estimate of undiagnosed cases, somewhere around half the people in those states have already been infected (see Post #940). At some point, that has to start slowing the spread. Elsewhere, who knows.
And I would say that the biggest unknown is whether this will crest before the more-contagious British variant of COVID (Post #932) becomes prevalent enough to drive rates up again. In Britain, the lag between first detection and the spike in London cases was nearly three months. In the US? We have no way to know. As of today (1/9/2021), that new British strain has been identified in 8 states: CO, CA, FL, GA, NY, TX, CT, PA ).
The simple upshot of this is pretty straightforward. If states begin to roll back their re-opening plans, and impose additional restrictions, there is a concrete reason for that. So just deal with it. And, separately, however cautious you were in your own behavior, at the last peak of infection rates in your area, consider returning to that behavior now.
For my own family, shopping-as-entertainment is off the list of allowable activities for now. My wife and I separately came to that conclusion. No dropping by the grocery store to pick up a half-gallon of milk. No hitting the hardware store for a random item or two that might be needed. Batch your errands, get in, get your business done, and get out.
And wear the best mask you can get.
State-level detail follows.
Let me now just quickly graph the states separately, with minimal commentary by region.
- Northeast: Every state shows an upward trend. RI exceeded prior peak. The ludicrous peak in the NJ data reflect a reporting issue.
- South Atlantic: NC, SC, GA, FL show sharp upticks. MD-DC-VA show only modest increase in trend.
- South Central: Every state shows a sharp upward trend
- Midwest: Every state shows a mild upturn.
- Mountain: AZ, UT exceed or meet prior peaks. Other states show a less-sharp upturn.
- Pacific: Only Oregon shows no increase, on the mainland. Hawaii continues to do its own thing.
