There has been a ton of bad data analysis around this issue.  An especially egregious example is the pdf file at the end of this article.  People are coming up with all sorts of creative ways to avoid the fact that New York, New Jersey, Connecticut, and Michigan all did terrible jobs managing this disease.  I offer a simple, easy-to-understand measure of COVID19 state performance.

(click for larger image)

First we need an easy-to-use measure that is a well-defined event on a specific date.  I propose death from COVID19.  This is easy to understand and a good binary choice: either someone is alive or they are dead. (I’m excluding zombies, vampires, and others who most likely are not susceptible to this disease.) And the date of death is a matter of public record.  The only remaining issue is reporting error. All we can say is that reports of COVID deaths are probably no less accurate than other measures.  Naturally we want COVID deaths per 109,000 population.

There are many sources for this data.  I used the Washington Post for two reasons.  First, it was among the first that popped up in my search. Second, very few people believe that WaPo has a conservative bias.

I sorted this table from highest to lowest deaths per 100,000 population.  Of the first 20, 14 are solid blue states,  Nine of the top ten are controlled by the Democrat party. There is some question about Maryland because they have a Republican governor. In the Excel workbook and the table below Maryland is coded red.  You may do your own research by downloading my Excel workbook

(click for larger image)

Here’s the example of a truly lame attempt at justifying the actions taken by governors of several blue states.