I've seen the argument made that adjusting for population size doesn't make sense in cross-country comparisons. I'm not an epidemeologist but I understand the math. The argument is until we get near saturation, spread from one person happens to a discrete 1/n

number of people, so population size is irrelevant in comparing spread. That makes sense to me. But it also seems like it's only right if you can truly view each thing that you're comparing as a monolith (i.e. "USA" is one thing, "Italy" is one thing). 2/n

But that's not right, I don't think. The US has more different discrete regions than most other countries. And as a result, it means that different parts of the country will at any given time be at very different places on their "Curve." If you add all the curves 3/n

together, the combined curve would have a very different shape than any individual curve. The US curve looked steep when NYC was steep and now as NYC levels, it will keep looking steep because of other places climbing the curve. 4/n

So what I'm saying is that neither an absolute comparison of US deaths to Country X deaths OR a per capita comparison makes sense to me. Outbreaks are regional. Compare NYC to London. Compare NYS to Lombardy. There's no way to Compare the USA to, say Belgium, as a whole. 5/5

PS: Relatedly, it's just wrong to say the US outbreak is "better" or "worse" than any other country. It's hundreds of separate outbreaks, some horrific and others seemingly better-controlled. We need a nat'l response but that doesn't mean we should view it as 1 nat'l outbreak.

Even when we talk about the bad outbreak in "Italy" we mostly mean northern Italy. It would be great to be able to compare northern Italy and southern Italy, separately, to other outbreaks. Some places show this data. But thinking of countries as monoliths leads to crude measure.