Evolution Understands Selection Bias
Juice or Peanut Butter?
Alice is a chimpanzee. She likes peanut butter as well as juice popsicles. When given the choice between the two, though, she'll usually choose peanut butter (good choice).
Now, unfortunately for her, you're not here just to feed her. You're here to study the endowment effect.
So, you give Alice a juice popsicle. She's happy. She likes juice popsicles.
But, just as she finishes the first, delightful suck, you reappear, standing there with a tube of peanut butter. You make her the offer: "give me back the popsicle, and you can have this peanut butter!"
She looks you in the eyes for a few seconds, half the popsicle hanging out of her mouth... and proceeds to continue slurping on her popsicle.
She prefers peanut butter, yet she chose to keep the popsicle? How could she?
It does indeed seem like she's made an irrational decision, choosing to forgo a trade that would make her happier. Okay, you might have some misgivings about the validity of this experimental setup, but this effect has been seen not just in chimps, but in us, perfectly rational, perfectly optimizing humans. Therefore, this is seen as a counterexample to homo economicus.
(This description is based loosely on Endowment Effects in Chimpanzees - ScienceDirect)
Mr. Powell
Now, let's consider another scenario:
Case 1: I tell you that you can, on even odds, bet that the next inflation report will, or will not, see an increase of 3 percentage points.
Well... you should probably pile as much money as you have on the side that, no, the inflation report will not increase by 3%.
Case 2: Jerome Powell approaches you with the following offer: "I bet you $10,000 (on even odds) that the CPI next report will see an increase of 3%."
Do you take it?
On the one hand, it's the same bet that you were willing to make before. But on the other hand, this is Jerome Powell. Chances are, he would not be willing to make this bet with you if he didn't have some information (or he's drunk).
What happened here? Well, sort of the same phenomenon! When I was presented with two options: (A) bet for a 3% increase in CPI, (B) bet against it, you showed a preference for (B). But when Mr. Powell offered you (B), you refused it. The very fact that Mr. Powell was willing to do this trade with you gave you information that, hmm, maybe I'm better off giving him the finger, thanking him politely for waging our battle against inflation, then running away.
Any connection?
It'd be cool if you could say something along the lines of "evolution has selected for selection bias." That is, too many apes in the past were screwed over by their mischievous counterparty, trading their good banana for a taxidermied banana peel.
However, my guess is that the more likely explanation is a sort of "distributional" version of selection bias. That is, maybe the exchange isn't adversarial: Alice knows that this popsicle tastes good, while there's still some uncertainty as to how good the peanut butter is, even if it's expectation is higher.