The Cato Institute’s Alex Nowrasteh posted a very interesting piece that reproduces and expands the findings in my Mariel paper. Although the title, “The Mariel Boatlift Raised the Wages of Low-Skilled Miamians,” is very misleading (making it a good example of Cato publicists gone wild), I actually liked the essay and recommend it to anyone interested in the subject.
My paper showed that the wage of high-school dropouts fell significantly after Mariel, but recovered by 1990. Here’s the graph that goes with that conclusion (where the shaded area indicates the margin of error):
In early drafts of my Mariel paper, as well as in We Wanted Workers (Figure 7.5 on p. 148), I documented that the wage drop was not experienced by high school graduates, the group of workers on the next rung up the skill distribution. In contrast to the dropouts, high school graduates actually saw their wages rise. Here’s the graph from my book:
A few months ago, Joan Monras and I followed up on this insight to document a pattern common to many refugee supply shocks: refugees have harmful wage/employment effects on the workers that they most resemble, but beneficial effects on the workers that are different.
Nowrasteh’s conclusion that “low-skilled Miamians” gained comes about because he pools the two groups of workers (high school dropouts and high school graduates) into a large “low-skilled” workforce, and he shows that the wage of the average worker in this group increased as a result of Mariel. He concludes: “The Marielitos redistributed wages from dropouts to workers with only a high school degree with a net positive effect on all low-skill workers.”
The exercise illustrates two very important points that Nowrasteh does not emphasize. First, it shows just how easy it is to hide the adverse wage impact of immigration by redefining skill groups. This is a trick that, unfortunately, is used much too often to “prove” that immigration is good for everyone. As I wrote in We Wanted Workers (p. 196): “The more one aggregates skill groups, the more likely one hides away the specific group of affected workers–making it harder to document whether immigration made anyone worse off. The more laser-focused the group of native workers examined, the easier it is to detect that immigration affected the targeted group.”
Second, Nowrasteh (perhaps unwittingly) blows up the cornerstone underlying the Card-Peri argument that immigration has not made low-skill Americans worse off. That cornerstone is the assumption that high school dropouts and high school graduates are productive clones (or “perfect substitutes”). That assumption is what gives the researcher “permission” to pool those two groups. Because there are many fewer dropouts, the wage trend will essentially reflect whatever happened to the wage of high school graduates.
Nowrasteh’s documentation that Mariel had very different effects on high school dropouts and high school graduates flatly contradicts the assumption that the two groups are productive clones. If the two groups were clones, they should have reacted in exactly the same way to Mariel. But they didn’t. Instead, they are complements, implying that the two groups should be studied separately. Those who buy into the Card-Peri argument need to go back to the drawing board if they want to salvage the conclusion that immigration didn’t really harm the least skilled Americans.