I see your point, and thinking about it helped clarify my thinking, so first, thank you for that. I think another way of phrasing my opinion is that the US labor market is a large, reasonably efficient market; that it's reasonably well modeled by a simple supply and demand model; and that discussing it in other terms obscures the important issues. The notion of a shortage is fundamentally incompatible with a simple supply and demand model of a market. The market price in such a model is, by definition, the price at which supply and demand are balanced. So if you're talking about a shortage in the US labor market, you're either claiming that supply and demand is a poor model of the labor market, or you're talking nonsense. I think either is misleading or at best confusing. So to an extent I really am saying that shortages can't exist for anything you can buy on a large efficient market.

Of course not every market is a large efficient market. If you're talking about a price-regulated market, by all means, discuss shortages. Another good situation where discussing a shortage would be appropriate is markets that have very or perfectly inelastic supply over a relevant time period. Say, a vaccine that costs $1 but that takes 6 months to produce due to the biology involved, combined with a major epidemic in the relevant disease. It's fair to call that a shortage.

I think labour markets are some of the least efficient markets that exist. Without immigration it takes at least 4+ years for extra supply to become available based on increasing prices. And when you hire someone you really don't know what you're getting, particularly in software.

Constraining supply increases salaries precisely because it takes supply so long to incorporate information.

On top of that, supply and demand is a pretty leaky abstraction, I don't expect it to be very accurate in such a nonideal market.

Right, GP is leaving out A. necessary skills, and B. the size of the labor pool (multiplied by the labor force participation rate).

For any given skill there is a labor pool with some participation rate, of people who currently have that skill, such that it is accurate to say that the market has a shortage.

Answers to the shortage must solve for either the size of the labor pool, the rate of participation, or the amount of people who have that skill.

Those answers vacillate between importing labor to solve the problem now, or forcing the market to educate people by blocking imported labor.

The problem with importing labor is that it reduces market incentives to educate people. The problem with not importing people is that it means you'll have a labor shortage until the market adjusts and people re-educate.

Easy answers, on either side, are snake oil.

> I'm pretty sure the word shortage already implies that there is a shortage at the market price

"Shortage at the market price" seems to be a somewhat incoherent concept. The market price is the price point at which the supply and demand curves intersect. Other than a discontinuity in either supply or demand such that the curves pass without intersecting (which makes the market price ill-defined), demand equals supply at the market price.

I agree that it's poorly defined, but I would argue that a shortage at the market price implies that buyers would buy more at the market price if more of it was available.

We're pretty comfortable talking about shortages in consumer markets when we would buy things at the stated price, but can't because it is sold out (eg Pixel phone shortages), so I'm not sure why we can't say a similar thing about labour.

> I would argue that a shortage at the market price implies that buyers would buy more at the market price if more of it was available.

That can just be a current price below the equilibrium market price; I'd argue that a shortage is a specialized instance of that resulting from a transitory decrease from long-term supply (or increase from long-term demand) that is expected to revert to normal.

But if you need annual supplements of labor for the same field over a period of decades, that's not a labor shortage, its price being artificially suppressed (and inhibiting market signals that would develop domestic supply faster.)

> But if you need annual supplements of labor for the same field over a period of decades, that's not a labor shortage, its price being artificially suppressed

Theoretically, couldn't it also be that demand is consistently growing faster than anticipated (even after whatever corrections are made upon discovery that it previously grew faster than anticipated)?