Thread by @Jabaluck

Good news everyone! I'm "a prominent economist".

But Twitter folks -- stop critiquing my calculation by pointing out that it makes assumptions and ignores factors.

https://twitter.com/BPredict/status/1442318150314467331

That is what calculations of this nature do. I thought it was important to get a rough order of magnitude sense of how many people need to wear a mask in public areas to save one life in the US. I did the most reasonable calculation I could think of given the data available.

I would be very pleased to see other calculations -- for example, embedding our results in a structural model which figured out the long-run impact on deaths given a race between vaccines and natural immunity would be extremely interesting.

But if you're just going to say, "You've made assumptions" as opposed to, "Here is a better calculation that makes more plausible assumptions", then you have not produced a better estimate of a number that a lot of people want to see.

This is a very general problem -- people look at quantitative models and dismiss the results because they don't like the assumptions when they have no quantitative model of their own.

Your intuitive guess is *also* based on assumptions that are equally questionable, just not ones that are clearly articulated.

If you don't like my estimate, give a better one, state your assumptions clearly, and then we can discuss which calculation makes more defensible assumptions.