After I published my article arguing that increased immigration even of low-skilled workers is good for the American economy, I meant to do a more technical followup but kept forgetting. So here goes. The concept developed in the column is of "complentary" factors of production. Increasing the supply of dishwashers and busboys increases employment opportunities for chefs, waiters, kitchen equipment manufacturers, and so forth all up and down the skill spectrum.
The research that really changed my thinking on this is ably covered in this great Heidi Shierholz did for EPI back in February 2010. Note that EPI is the premiere labor-liberal think tank in Washington and hardly a hotbed of apologism for the top one percent. The basic point here is that the old CW on low-skill immigration is that it raised real wages for high-skill workers but lowered them for low-skill workers. The key methodological advance comes from realizing that a very large share of low-skill workers in the United States are themselves immigrants. Since restricting low-skill immigration for the sake of low-skill immigrants is a little perverse, it's helpful to distinguish between the impact on immigrant workers and native-born workers.
Here's what they found:
One key finding here is that if you look at typical native-born working class Americans—folks with high school diplomas but no college degree—they win out thanks to immigration. And even if you restrict your attention to U.S.-born high school dropouts they win under most scenarios.
The losses from increased labor market competition are very real but they're concentrated among other immigrants. That's because it's all about complements. An increased supply of dishwashers and busboys increases the value of modestly educated people with complementary skills. To return to the restaurant, a waiter or a bartender needs to be able to speak English. In a world with no immigrants "can speak English" isn't much of a skill but when low-skill immigrants rush in suddenly it is. The people who lose out are the other workers who can't speak English, or who have specialized taco-making skills, or otherwise are extremely similar to new immigrants.
A lesson here, I think, is that we could probably make low-skill immigration even more beneficial to the United States by building a more diverse portfolio. Right now the stock of low-skill immigrants to the U.S. is heavily weighted to Mexico, since low-skill immigration is largely illegal and if you're going to sneak into the country Mexico is a good place to do it from. Actually allowing people to come would let us diversify and pick up more immigrants from Africa, Southeast Asia, etc., whose work might be complementary to the work of Latin American immigrants as well as to people born in the USA.