No, the CBO report doesn’t mean immigration brings down wages

June 19, 2013

Tuesday's CBO score of the Gang of Eight immigration reform bill was pretty much a slam dunk for the bill's supporters. As Ezra put it, "This isn’t just a good CBO report. It’s a wildly good CBO report." But that doesn't mean there aren't holes that the bill's opponents can try to poke in that picture.

Sen. Jeff Sessions (R-AL), for instance, sent out a news release noting that the report also found that Americans' average wages would decline in the first ten years of the bill's enactment. "As Dr. [George] Borjas has illustrated and as CBO confirms, this increased GDP will be at the expense of poor and working-class Americans," the statement said. "The benefit will go to the business owners while the wages of U.S. workers — which should be growing — will instead decline. That is indisputable, and while CBO did not sort out the effects among U.S. residents, the impact will be harshest for today’s low-income Americans."

With due respect to Sessions, let's dispute that "indisputable" conclusion. First - what does the CBO report actually say about wages? The headline conclusion is that wages would be 0.1 percent lower in 2023 but 0.5 percent higher in 2033 than under current law:

But as Ezra notes, that 0.1 percent decline includes the fact that the average now includes new immigrants, who'll be disproportionately low-skilled and push down the average wage. The CBO stresses that the wage numbers "do not necessarily imply that current U.S. residents would be worse off," though it "has not analyzed the full economic effects of the legislation separately for the incomes of people who would be U.S. residents under current law."

The 0.1 percent is a tiny decline in average wages and after adjusting for the effect that adding new workers to the average has, it'll be tinier still. But the effects aren't equally distributed. While by 2033, all income groups will see wages rise due to the legislation, in 2023 the middle three quintiles (for reference, that meant those making between $20,262 and $101,582 in 2011) would gain relative to those on the top and bottom.

That middle group would see 0.5 percent higher wages relative to the bottom and top quintiles, which would see 0.3 percent lower wages relative to the middle. That still has to be adjusted for the effect of adding new workers to the average, but it means that some groups would still likely see lower average wages even if you exclude new immigrants from the picture.

So, the bill hurts poor and rich workers' incomes, then? Nope. Even if you take new immigrants out of the picture, you still have to disaggregate the effect of immigration on the wages of native-born workers as opposed to existing immigrants. After all, new immigrants, and particularly low-skilled or agricultural immigrants, are likely to enter the same industries as existing immigrants, and so one might expect existing immigrants to compete with them more, and for those existing immigrants to thus bear the brunt of any decline in wages.

That's exactly what research by Gianmarco Ottaviano and Giovanni Peri, economists at Bocconi University and UC-Davis, respectively, finds. They estimate that between 1990 and 2006, immigration to the United States reduced average wages of existing immigrants by about 6 percent, but it actually modestly increased average wages of native-born workers by about 0.6 percent. The Economic Policy Institute's Heidi Shierholz, analyzing immigration between 1994 and 2007, found the same, concluding, "any negative effects of new immigration over this period were felt largely by the workers who are the most substitutable for new immigrants—that is, earlier immigrants."

That holds up even for native-born workers with no high school diploma, whose wages changed by -0.1 to 0.6 percent under Ottaviano and Peri's preferred specification. And considering that immigration has been found to spur more people to graduate from high school precisely because it increases competition for low-skilled jobs, as Michael Clemens explained here, even a finding that the least-educated native-born workers were hurt wouldn't be a slam-dunk for the anti-immigration case.

To be sure, this is an area of academic dispute, with Borjas, who Sessions cites, arguing that the least-educated workers suffer due to immigration. But even Borjas finds that U.S.-born workers benefit overall, as this chart from Michael Greenstone and Adam Looney at the Hamilton Project shows:

And lest you think this is just a dry technical dispute, some of these academic papers are being taken seriously by members of the Obama administration's economic team as they consider the effects of immigration reform.

Continue reading 10 minutes left
Show Comments



Most Read Business