Here's a thread with some thoughts about the various forms of snark going round Twitter regarding expertise, and its absence, in discussions of COVID-19, epidemiology, economics, and so on, in case anyone is interested.

Fwiw: I’m a @BerkeleyLaw prof/former econ prof w/econ PhD & JD. My professional expertise: economics, statistical methods & their use, various areas of law (civil procedure, evidence, statutory interpretation, some securities law, law & econ).

I’ve seen lots of arguments of the form: “What do you [engineer/datasci/economist/lawprof/polisci] know? You’re not an epidemiologist?”

I’ve also seen plenty of: “I know math/how to read/I’ve taught cross-examination to doctors.”

So, who (do I think) is entitled to be taken seriously about technical aspects of the pandemic?

The answer isn’t just epidemiologists, bc we all have a stake, and because—and I say this based on having read/listened to epidemiologists—the technical issues aren’t so superspecial as to render everyone else incapable of understanding.

But—and this should but doesn’t seem to be obvious—some non-epidemiology people have failed to get informed enough about the substance to offer anything of use. Indeed, some such people have offered dangerous nonsense.

One way to know someone’s full of it is when he doesn’t engage with the underlying math or language of epidemiology. Someone who doesn’t mention/discuss R0, doesn’t engage with the reasons epidemics start off w/exponential growth, etc, just isn’t serious.

I say that as someone who had never heard of the concept of R0 before this pandemic. So feel free to disregard what I have to say, although I have tried to learn enough to say useful things, and not say really dumb/dangerous ones.

Things to discount: affiliations at fancy law schools or think tanks, and fancy-sounding titles like “Tisch Professor” or “Bedford Senior Fellow”.

Whatever value those credentials have generally, they don’t outweigh ignorance and bad reasoning.

Of course it’s not just @RichardAEpstein. Though he exemplifies the Dunning-Kruger-Epstein effect, he’s surely not the only one.

Again, pay attention to whether commentators are engaging with the language & concepts experts in the field use.

Having said all that, no one should refuse to listen to someone *merely* because that person's a lawprof/lawyer/economist/machinelearner/rando/etc.

There are many aspects of social behavior that matter a ton here, and people outside epidemiology have plenty of useful training in those areas.

Further, statistics is statistics, once you understand the phenomenon you’re modelling. (That latter part is critical, to be sure.)

I understand that this leaves people without any expertise/comfort in statistics/mathematically technical areas with little to go on, whereas “just listen to epidemiologists” might seem more manageable.

That’s ok, I guess, but the scope of this problem is too broad to limit the platform to only one area of study. You can get a lot wrong by just extrapolating math/stats relationships that work well in limited settings. Here, all aspects of society are affected.

Yes, we must—MUST—listen to & learn from epidemiologists.

But other expertise—e.g. economics, statistics, sociology, law—is valuable to evaluate plausibility & understand implications of epidemiological models.

Even if some commentators lack/misuse such expertise.

/fin