It’s the academic dust-up that will never end. All the way back in September of 2015, Jason Richwine wrote in this space about a then-new study by Harvard’s George Borjas casting doubt on the common wisdom about the “Mariel boatlift.” In 1980, a lot of low-skilled immigrants flooded into Miami from Cuba, and a famous 1990 study claimed that native workers weren’t harmed. Borjas contended that, in fact, wages for non-Hispanic high-school dropouts in Miami plummeted.
Borjas’s study included a lot of technical detail, but here is a simple chart that nicely illustrates the finding. It compares the wages of working-age, non-Hispanic men without high-school degrees in Miami with their peers everywhere else in the U.S.:
Naturally, the study generated a lot of highly complicated pushback; I detailed some of it in my profile of Borjas earlier this year for The American Conservative. But a new paper alleges that Borjas is wrong in a simple and straightforward way.
Right around 1980, the year of the boatlift, the Census started making a concerted effort to include more low-income black males in its surveys, on the belief that this demographic had previously been underrepresented. Low-skilled blacks also arrived in Miami around this time from Haiti. Nationwide this didn’t make an enormous difference, but Miami in particular saw a higher concentration of low-skilled blacks in its surveys, for reasons having nothing to do with the Mariel boatlift. Since black dropouts make less than white dropouts, this demographic shift by itself drove down the average wage. Indeed, in the authors’ telling, the shift can account for the entirety of Borjas’s result.
I’m a journalism major, but I do have a basic familiarity with the Census survey Borjas and his critics are working with. (Last year, Borjas featured on his blog my explanation of how to recreate the chart above using nothing but an online tool and a basic spreadsheet program like Excel.) I can confirm some of the basic facts of the critique. The sample of low-skill Miami men Borjas looked at did become much more heavily black right around 1980. And even among men without high-school degrees, blacks do in fact make considerably less than whites do.
The obvious way to fix the problem would be to just analyze whites and blacks separately, but that isn’t advisable in this case, because the sample of low-skilled Miami men is incredibly small even with all races included. (The chart above, for example, uses three-year “moving averages” to triple the amount of data that goes into each point. If you want to, you can see my hackish and inconclusive attempt to make similar charts for whites and blacks separately here.)
I checked with Borjas, as well as with Joan Monras, who coauthored another paper with Borjas that also touched on Mariel, to see if they had any response. Unfortunately neither has had time to read the paper carefully and replicate it, but Borjas raised a number of potential objections in an email:
Yes, it’s true that the proportion of blacks went up. The question is why? Could it be that it had something to do with Mariel—that somehow many of the white low-skill workers left Miami because they (perhaps unlike blacks) had the opportunity to do so? To the extent that this type of mobility was driven by Mariel, it’d be wrong to focus on this issue. And it’s certainly plausible that Mariel had an impact on low-skill mobility rates in the 1980s. Moreover, the sampling issue seems to be less relevant (though again I haven’t checked the data) in the ORG [another Census survey].
Second, I would read this paper with the utmost skepticism—similar to the skepticism that one uses when one reads drug research funded by pharmaceutical companies. The paper was paid for by Silicon Valley open-border plutocrats, and I’m pretty sure they wouldn’t buy or commission research that didn’t fit their priors.
You can see the financial disclosures of the Center for Global Development, which published the study, here. They do receive a lot of Gates Foundation cash and tend to criticize immigration restrictions.
Regardless, the new paper strikes me as a very serious criticism. I look forward to seeing more technical responses to it.