I’ve been emailing with Ari Schulman, the editor of the New Atlantis, and he made a comment that I thought was worth (with his permission) quoting:

We just do not know enough about Covid to make reliable general conclusions from a single country. And for any general conclusion we could reach, we can come up with many counterexamples. The Times had a great writeup of this yesterday. If you want to say that Covid spares warm-weather regions, that does hold up as a broad trend — but there are plenty of counterexamples. If you want to say that countries that don’t do official lockdowns turn out fine, there are indeed plenty of examples besides Sweden. But then there are also Italy, Spain, Iran, New York. If you want to say that these variations must be explained by different strains, so that anywhere that hasn’t already been badly hit won’t be — well, that makes a certain amount of sense (it seems hard to imagine Montana taking off at this point). But it also gets into epicyclic territory until it’s backed by actual genetic analysis, and there are counterexamples too (Russia, Turkey, the resurgence in Singapore).

It’s a choose-your-own-adventure game. The number of unknowns about Covid, and the number of parameters across which it varies markedly, are astonishing. The main difference I see between analyses right now is whether they view this situation as basically clearcut — here, look at the back of my envelope, I’ll explain it to you — or as deeply murky, a terrible conundrum.