Believe It or Not, Utah Is Showing Some Improvement

It can be hard to spot in the data, but Utah's COVID situation has been improving (modestly) for over a week. It seems plausible that this is due to the state blasting out all those emergency warnings about a month ago.

Anyway, the thing to look at is not the daily case totals. This is an unlogged growing series. It will set records all the time. It will also tend to show larger and larger daily increments (those aren't good for our healthcare system, but the outbreak can be improving in an important sense even if the continue to get bigger).

You also can't look at daily changes. There are weekly seasonals in the numbers (seasonals is the name for any repetitive calendar pattern in data, no matter what the length).

So a good, easy to explain to non-specialists, method is to divide the total cases on a day, by the total cases one week before. This produces a weekly growth rate, but not in (net) percentage terms, but rather as a (gross) ratio. 

Today that ratio is at 1.098. That means we have about 10% more total cases today than we did last Friday (206,165 vs . 187,775). That is not good. Our current wave began when that ratio cranked up from around 1.05, where it had been steady through late summer, to about 1.10. It did this over a span of 10 days right after Labor Day. The local news has attributed this to university students returning to school, and it's been apparent for sometime that casualness was biggest at BYU and UVU.

It cranked up further in mid-October, steadily rising through mid-November. That's what got the Governor so freaked out. It has been declining steadily since November 13th, when it peaked at 1.160. That' the improvement.

But do keep in mind that it has only dropped back into the elevated range that it was in from mid-September to mid-October. You can see this here:

The most recent wave is on the right. The wave in early summer is on the left. 

In this sense, that earlier wave was "more out of control". Having said that, it was starting from a much smaller base of cases: in the worst parts of that wave we were adding 200 cases per day. In the current wave we are getting 10 times that number.

Nonetheless we can see from the hump on the right that there is some room for optimism. But we can also see that we've just returned to the plateau from earlier in the autumn.

But again, having said that, what we continue to be watching (and experiencing) is a slow motion train wreck. That train is still accelerating, just not quite as fast.