2 Comments
User's avatar
⭠ Return to thread
Be's avatar

This is a good point but it’s not adverse selection, it is simply that the current marginal high skill immigrant is net good whereas if we expanded immigration a lot then the counterfactual marginal immigrant would be substantially different and probably substantially worse. It is not that the worst people choose to immigrate (at least legally) but that actually the best do currently but that would change if we moved the margin. The implication being that we should cautiously permit more high skill immigration but stop before the point where we are legally allowing the sorts that are currently asylum seekers (which obviously should be halted).

Expand full comment
Tanmay's avatar

I think that part of the explanation for why the counterfactual marginal immigrant with expanded immigration would be substantially worse *is* adverse selection. When you reduce the degree to which you vet offers, you should expect the (marginal) proportion of instances of adverse selection to increase, since vetting is the one lever you have available to pull to decrease adverse selection.

Put another way: I think an essential lesson of adverse selection is "the person who is putting an offer on the book (the market maker) is at a disadvantage because the people looking at the book (the market takers) can see the offer and decide (with the information that the offer is on the book) whether to take it or not.” This is why you should always expect market makers to show spreads; it’s also the advantage of being “in position” in poker. But if you’re the taker, you give up this “taker’s advantage” if your strategy for taking is "lift every offer on the book" (much like how you don’t have an advantage “in position” if your strategy is to call everything). If that is your strategy (and if it is well-known that that is your strategy), then you're the one who is extremely vulnerable to adverse selection, because your strategy is equivalent to putting a permanent bid of +infinity on the book!

(This is adapted from an email I wrote to Prof. Caplan expanding slightly on my post above.)

Expand full comment