A Crisis Of Epistemic Overconfidence

Alchmemy illustration in the Italian book Della tramutatione metallica sogni tre (1572?) by Gio. Battista Nazari], via Wikimedia Commons

The world is complicated, so sometimes we don't know what's up. When you don't know what is going on, you can still have opinions about the best thing to do. But what you can't do is present those opinions with confidence, like Oh that, we have that all sorted out. And yet -- it seems like that's exactly what people do. It's infuriating.

Often I feel like when you look into why people told you to believe something, it was like some version of "sounds reasonable," "makes sense to me," "sounds legit."

For example, the more we learn about nutrition, the more it seems that whole idea of eat-less-fat be-less-fat doesn't really add up. Just like the whole eating-cholesterol causes high cholesterol didn't add up. Just like the whole massive doses of vitamin C or E or whatever didn't add up.

And then you find out that the reason people believed them had to do with things like "fat has more calories per unit weight than carbohydrates," or "some vitamins are good so more is probably great" combined with some vaguely suggestive empirical studies.

Sounds reasonable. Also sounds absurdly oversimplified. And turns out, it was. Yet things like this are often presented as well established and they've had huge effects on our lives.

There are so many things like this in medicine. When I was a kid, everyone said that even a low fever should be treated with aspirin, since it seemed reasonable that a fever would be bad for you. Everyone also said to treat swelling with ice, since it seemed reasonable that swelling was bad and you should try to counteract that. But now we know: fevers and swelling are both ways the body helps fight whatever is wrong with it. Which, once you think about it, also seems reasonable.

Of course, the same thing happens in economics. In this Times op-ed defense of economics being a science, Raj Chetty says the fact that there are constant disagreements, epistemic uncertainty, and extremely limited opportunities for controlled studies doesn't mean economics isn't a science, since "big picture" medicine has the same problem, and it is certainly a science.

It seems to me maybe if you have enough big picture problems it'd be more accurate to say "could be a science" than "is a science," but leaving that aside: the reason the science issue seems important to people is mostly because of the truth, objectivity, and certainty aspects that come in its wake. People want to know: does this activity present results that tell us what is going on?


And the answer there seems to be mixed at best.

It's fine that economics and big picture medicine are difficult, and that answers there "remain elusive," as Chetty puts it. But if you don't know, act like you don't know. Don't go throwing your weight around, shaming the butter-loving, the aspirin-reluctant, the anti-capitalist. 

They might not know what is going on, but face it, you don't know what is going on either.