1. Virus Data
A reader who is a data scientist has been playing around with the numbers to see if there are any interesting trends that aren’t being reported widely. And last night he sent me this graph:Let’s talk about what we’re looking at here.
On the x-axis we have tests performed per capita. On the y-axis we have positive cases per capita. And he’s plotted the trends for states that have reported numbers of tests performed. If a state isn’t here, it’s because they’re doing negligible testing.
If things are going well, then you’ll be in the bottom-right quadrant of the graph: You’re doing a lot of testing and getting very few positives. That means you’ve got the contagion under control.
If things are going terribly, you’ll be in the top-left quadrant: You aren’t doing enough testing to know the real state of play, but clearly a lot of people are sick. You just can’t see the iceberg.
If you’re in the bottom-left, it means you don’t really yet have a handle on where things are.
And if you’re in the top-right, it means things are bad.
So look at the graph again: Obviously, New York and New Jersey are the big outliers. But also note that the Northeast region in general: Connecticut, Vermont, Massachusetts. Hot spots radiate outward.
The thing that jumped out at me here was Louisiana. So we’ve had a hotspot in the Pacific Northwest which seems to be coming under control. We’ve got the Northeast out of control. What we did not have yet was hard evidence of something very bad going on in the South.
Louisiana’s numbers are that signal and the spot that we should be watching closely over the next two weeks. Because if Louisiana turns into a hotspot, then you run the real risk of blooms in Florida and Texas.
And as a supporting chart, take a look at this ranking of the states by tests performed as a total of population:I’ll get to why this chart is important in a minute.
So let’s turn to this Axios piece. They’ve put together an interactive chart that’s so good I won’t put a picture of it here because I want you to click on the link and go see it for real.
This data set looks at metro areas (not counting NYC) by total number of cases and percentage increase in the last four days. And sure enough, what you see here is what our last graphic suggested: Real trouble in New Orleans, Boston, Philadelphia, Miami, Detroit, Dallas, and Houston. All of them have seen over 100 percent increases in just the last four days.
There have been people who have tried to argue that COVID-19 is really a New York City problem. It’s true that NYC is the epicenter in America. But that’s because it’s the densest, biggest city. The outbreak is so widespread that it was already in all 50 states. And while it got to NYC early, because it’s such a big travel hub, it also migrated out to other cities, which are simply a week or so behind NYC.
Hopefully the fact that no other metro area is as dense as NYC will prevent the spread from being as catastrophic. But you have to understand that this wave hasn’t yet crested. It’s still gathering strength.
If you want to get really worried, look at the rate of increase in the heartland: In the last four days, St. Louis has had a 200 percent increase in total number of confirmed cases. Indianapolis has had a 230 percent increase.
Now, jump back up to that blue chart ranking states by total testing percentages. What you see is that Indiana and Missouri are near the very bottom of the scale—they’ve performer fewer tests per capital than 80 percent of the states. And yet their big cities are showing some of the highest growth rates.
What does that tell you?
(Many thanks to reader T.D. for putting together those charts.)