I’m very interested in confirmation bias, as regular readers know. A number of left-of-center bloggers have highlighted a new study by Giovanni Peri of the San Francisco Fed that highlights the beneficial impact of immigration on the economy’s productive capacity. Felix Salmon writes:
Anti-immigration forces are more likely to ignore it than attack it, I think, if they don’t like what it says. And George Borjas seems to have stopped blogging over a year ago, which is a shame, because he would be the perfect foil for Peri.
Well, I wouldn’t describe myself as belonging to the “anti-immigration forces.” But I will say that the study’s findings are hardly surprising. We’ve known for a pretty long time that immigration tends to increase wage dispersion by raising effective incomes at the top and depressing them at the bottom. And Peri’s findings regarding the impact on average income doesn’t tell us much about the distribution of gains. Felix excerpts the following from Peri:
Over the long run, a net inflow of immigrants equal to 1% of employment increases income per worker by 0.6% to 0.9%. This implies that total immigration to the United States from 1990 to 2007 was associated with a 6.6% to 9.9% increase in real income per worker. That equals an increase of about $5,100 in the yearly income of the average U.S. worker in constant 2005 dollars. Such a gain equals 20% to 25% of the total real increase in average yearly income per worker registered in the United States between 1990 and 2007. [Emphasis added.]
But how does immigration impact the median worker rather than the average worker? And how does it impact wages of workers at, for example, the 10th percentile?
In 2006, Peri co-authored a paper with Gianmarco I.P. Ottaviano of the Universita’ di Bologna that added an important wrinkle [PDF]:
Using our general equilibrium approach we estimate that physical capital adjsust promptly and fully to immigration (already within one year) and that immigrants are imperfect substitutes for US-born workers within the same education and experience group (because they choose different occupations and have different skills). These two facts, overlooked by the previous literature, imply a positive and significant effect of immigration on the average wage of U.S.- born workers, already in the short run. They also imply a small negative effect of immigration on wages of uneducated US born workers and a positive wage effect on all other US-born workers. Hence only a very small fraction of the increase in College/High School Dropout wage gap during the 1990-2004 period can be attributed to immigration.
A central question is how we weight the impact of immigration on “uneducated” US born workers. I tend to think the U.S. can accommodate a relatively large immigrant influx — I’d like to see an influx only slightly smaller than what we have now when we combine authorized and unauthorized immigrants, but with authorized and skilled immigrants much closer to 100 percent of the total than is presently the case. But that’s because I’m less concerned about wage dispersion than my left-of-center counterparts.
I wonder if my interlocutors would accept that we should only pay attention to the average effect of, say, changes to tax policy and ignore the impact on the median household. I doubt it.