We’re now a week our from Thanksgiving. Here’s what the graph looks like.
As expected, we’ve seen a big dip in cases and deaths over the last week. Remember that dip is a dip in reporting, not in actual cases or deaths. The virus doesn’t take time off for long holiday weekends, but the people who record and report the numbers do. Also as expected, the numbers have already started to rebound. That rebound will accelerate over the next few days as the Thanksgiving numbers drop out of the 7 day average and reporting catches up. Tonight’s numbers will show an especially large jump as they will replace Thanksgiving day in the 7 day average. Thanksgiving showed about 80,000 cases and 800 deaths fewer than we would have expected if it had been a regular Thursday.
The net effect of all this is that over the next week the moving average will be artificially high. This is because it will include not only the actual new cases and deaths from each day, but also the reporting of cases and deaths that really happened over Thanksgiving. This is exactly what we saw over the Labor Day weekend. Here is the same graph zoomed in to the period since August.
I’ve inartfully circled the Labor Day pattern. Expect to see something like this, but bigger, in the two weeks after Thanksgiving. We won’t have a complete picture for another week.
With that said, it’s clear that cases and deaths continue to rise across most of the country. The upper plains and Midwest that had been the hardest hit a week ago are showing signs of improvement, but the states around them are doing worse. Overall, the effect is a bit like a ripple on a pond, expanding outward from roughly South Dakota. That’s good news for states that were hit hard in October and November, but it means that the worst is yet to come for highly populated states at the coasts. To their credit, many of these states recognize the danger and are taking steps to minimize the harm. We can debate the degree to which those steps will work, but at least they aren’t adopting the “Let my people die” approach of South Dakota.
Yesterday saw reports that more than 100,000 people are currently hospitalized with COVID. This is worrisome as stressed hospital systems will likely lead to a higher case mortality rate. It’s difficult to provide top notch care when your staff is worn out and you are overcapacity on patients.
Another danger is the mental gymnastics some people will do to refuse to believe there is significant danger. One of my son’s friends insists that California has had fewer than 80 deaths over the last week, despite the CDC, California, and every other tracker clearly reporting more than 400. Instead, this friend latched on to a poorly labeled statistic on the CDC website, did his own arithmetic, and came to the lower number from which he will not be swayed. Motivated reasoning is a real problem in the best of times, in these times it can be fatal.
I’m not going to do any deeper analysis until the numbers stabilize. Vaccines are coming, but they won’t be here for most of us until at least spring. In the meantime, please do your best to stay safe and help others do the same.