We typically think the demand curve for labor is downward sloping, right? Empirically, the modal estimate of the elasticity of labor demand is something like -0.4. (So a 10% increase in wages means a 4% drop in the number of workers employers want to hire.)
So in light of the fact that we know the demand curve for labor slopes downward, are you guys puzzled by the following considerations?
CONSIDERATION #1: The number of employed workers is much higher today than in 1800. And yet, real wages haven’t plummeted.
CONSIDERATION #2: If the U.S. were to make Puerto Rico the 51st state, the official US population would rise, and the number of employed US workers would rise. And yet, I’m guessing wages wouldn’t noticeably fall.
Does anyone think either of the above is hard to reconcile with downward sloping labor demand curves? I wouldn’t. And then, by the same token, I don’t see why Bryan Caplan thinks this is weird:
While labor demand elasticity is pretty clearly negative, virtually all estimates have an absolute value less than 2. Yet estimated effects of immigration on native wages are tiny…
How are both these results possible? The easy answer is that “wage elasticity of labor demand” and “wage elasticity of immigration” are conceptually distinct. Quite true, but they’re also conceptually related. Indeed, unless labor supply is fairly elastic, a low wage elasticity of labor demand seems to imply a high wage elasticity of immigration.
Bryan then goes on to list some possible reconciliations, but I don’t see him listing what is (to me) the most obvious direct explanation: that when you bring in more people, you are simultaneously boosting consumer demand. (At least two people in the comments at Bryan’s post said the same thing.)
If we are holding “other things” (including total population) constant, and we want to induce employers in a certain firm or sector to hire more workers, then you would need the price of the workers (i.e. wages) to go down. But that is totally distinct (right?) from increasing the population–whether through births or immigration–and then asking what needs to happen to wages, in order for employers to expand hiring.
Looked at differently, suppose a bunch of new people came into an economy with disposable income (perhaps they were landlords for tenants in another country), and none of the immigrants wanted jobs. But, they started renting apartments, buying food, going to the movies, etc. What would happen to the wages of domestic workers? Wouldn’t they go up, at least in many sectors?
So now if instead we say that the immigrants are earning an income from working domestically (in their new country), but are renting apartments, buying food, going to the movies, etc., then wouldn’t their spending still (by itself) cause domestic wages to increase, and this effect would be counterbalanced by their offering of new labor supply?
To be sure, I’m not saying the two things would automatically cancel out (though with a simple enough model, they would), I’m just saying that this would be the default starting point, I would think. It would only be where you started introducing heterogeneity in the original domestic population (land owners, skilled workers, unskilled workers, etc.) that I think you would start to get pronounced impacts on certain groups from a large influx of immigrants.