Thread by @shashj

My new piece: how heavy are Russian casualties in Ukraine, and how do we know? It depends greatly on the assumptions one makes about wounded-to-killed ratios in war, which in turn hinges on military medicine and the evolution of tactics over the invasion. economist.com/europe/2022/07…

In the first & second world wars, roughly three soldiers were wounded for every one killed. That soared over the 20th century, as @tanishafaza's research shows. But the Russian ratio is lower: three to one, according to @CIA director Bill Burns last week. economist.com/europe/2022/07…

Other estimates posit a higher wounded to killed ratio. It's probably risen as artillery, which wounds through shrapnel, has come to play a dominant role. But high ratios that mean even modest numbers of deaths can equate to enormous casualties overall. economist.com/europe/2022/07…

For more on wounded to killed ratios, see the work of the @dupuyinstitute, which draws on historical data. dupuyinstitute.org/blog/2016/10/2…
dupuyinstitute.org/blog/2022/06/0…

.@tanishafazal's excellent paper on the topic is below. She points out: "because of improvements in military medicine, the same conflict that produced 1,200 fatalities in 1860 is likely to have produced 800 fatalities in 1980". But relatively more wounded.
belfercenter.org/sites/default/…

.@KofmanMichael's thread from May is here. He notes: "The problem with higher range KIA counts is that you quickly develop input/output problems. There are more casualties than could feasibly be involved in the fight...certain numbers become implausible"

And now we have another US casualty estimate, which reinforces my point that Bill Burns’ public figure of 15,000 Russians killed & 60,000 casualties overall was chosen from the *lower* end of the US spectrum.

I’m going to go mad tracking these casualty estimates. Because a few weeks after the CIA director said 60,000 Russian casualties, DoD says 70-80,000. Different people plucking figures from across the estimated interval?