Immigration, Jobs, and Wages

Like most economists, I favor increased legal immigration to the US. I have to say, though, to my fellow immigration advocates: let’s get our story straight!

Immigration advocates typically make two assertions:

1) Immigrants do jobs that Americans aren’t willing to do, and
2) Immigration has little effect on wages

These statements cannot both be true. Consider what would happen if we got rid of all the immigrants. Somebody would have to do the jobs that the immigrants had been doing. Since, according to statement 1, Americans aren’t willing to do those jobs, we would have to raise the wages dramatically to get Americans to do them (which contradicts statement 2).

Theoretically, the way to get out of this contradiction is to assert that statement 1 is so thoroughly true that there are really no Americans who compete with immigrants in the labor market. If that’s the case, then, even though immigration affects the wages associated with jobs, it doesn’t necessarily affect the wages received by individuals. If we got rid of the immigrants, all the jobs with newly increased wages would be filled by people from other (presumably initially higher paying) jobs, so no individual would necessarily receive a large wage increase.

Logically, that works, but the premise is thoroughly implausible. I can believe that people who compete with immigrants are a small fraction of the population, but I cannot believe that they are an empty (or virtually empty) set.

Of course, you could say, since they’re a small fraction of the population, we don’t care about them. That, first of all, is not a very nice thing to say. But second of all, you have to recognize that, the smaller the fraction, the more dramatic is the harm done to them by immigration. If these immigrant-competers are actually a large fraction of the population, then statement 1 is false, so statement 2 can be true, and we don’t have to worry about them. On the other hand, if they’re a very tiny fraction of the population, then statement 1 is extremely true, which means that statement 2 is extremely false.

If I had to guess (given what I’ve heard about the research), I would say the truth is probably somewhere between statements 1 and 2. That is, there is a small but non-negligible fraction of the native population that is willing to do the jobs that immigrants do (or jobs that functionally substitute for those jobs), and the wages of this fraction are affected significantly but not overwhelmingly by immigration. I have always gone along with the argument that, since immigration helps the immigrants a lot more than it hurts these Americans, it’s still a good idea. Some people say I should be more concerned with my fellow citizens than with foreigners, and this leads to all sorts of contentious discussions about nationalism and the nature of democracy. Perhaps I’ll go into that stuff in another post.

Monetary Policy, Not Immigration Policy

The following quote from Paul Craig Roberts (taken from a piece quoted in a comment to this Economist’s View post) typifies a common point of view that I think has got things completely backwards:

“Job growth over the last five years is the weakest on record. The US economy came up more than 7 million jobs short of keeping up with population growth. That's one good reason for controlling immigration. An economy that cannot keep up with population growth should not be boosting population with heavy rates of legal and illegal immigration.”

First of all, many economists believe that the failure of job growth is a supply problem rather than a demand problem. Certainly part of the reason that job growth has not kept up with population growth is that the fraction of the population in their prime working years has declined. And part of the reason is that young people are choosing to delay working in order to become better educated. And part of the reason may be that people (now including men as well as women) are beginning to prefer caring for their own children to working. To the extent that it is a supply problem, the slow job growth could be taken as an argument for increasing rather than decreasing immigration.

Personally, I don’t agree that the main problem is on the supply side. But many at the Fed do think so, and that’s just where the problem is. As long as the Fed believes the labor market is approaching a capacity constraint, reducing immigration can only make things worse.

What would happen if we were to reduce immigration? Some of the businesses that now hire immigrants would have to hire Americans. In itself, that might be a good thing, but the story doesn’t end there. As businesses hire more Americans, the unemployment rate would fall, and the Fed would pursue its tightening policy farther than it currently intends. The Fed’s tightening would result in a weaker economy, which would cause other businesses to lay off Americans. The net effect on employment of Americans would be close to zero (and the direction of the net effect is unclear).

If you want to argue that immigration reduces the real wages of Americans, there is certainly an argument to be made, but I’m not going to get into that here. To argue that immigration is bad because the economy already isn’t creating enough jobs is to misunderstand macroeconomic policy.