As you no doubt have already heard, on Friday the Heritage Foundation accepted the resignation of one Jason Richwine, who in 2009 had completed a Harvard dissertation in which he probed the nexus between immigration and IQ.
The decision revealed a shocking unwillingness on the part of Heritage to stand up to bullying and protect the academic freedom of its researchers. Perhaps the only good thing to emerge from all this has been the wide-scale distribution of the dissertation itself, a worthy if highly debatable document. It’s a pity that none of Richwine’s detractors seem to have seriously engaged the paper, because an actual discussion of the ideas therein would be fruitful.
Conservative think tanks emerged as a parallel institution — they were intended to provide a safe haven for right-leaning academics in light of the fact that academia itself was hostile to politically incorrect thought. In this context Richwine’s dismissal seems like a scene out of Bizarro World: The dissertation earned its author a doctorate, and it bore the signatures of Harvard professors Christopher Jencks, a social-policy researcher who once sat on the board of The American Prospect; George Borjas, a labor economist; and Richard J. Zeckhauser, a political economist. And yet years after its publication it caused the resignation of its author from a conservative think tank.
Of course, regardless of its founding purpose, Heritage is a private organization and it can employ whomever it pleases. And as an ideological think tank, as opposed to a university, it arguably has less of an obligation to continue employing researchers whose past work it finds harmful to its cause. But it asks to be taken seriously in the public-policy arena as an “educational institution” — a designation that ought to imply at least a healthy respect for controversial arguments so long as they are competently executed. And Richwine’s dissertation is most certainly competently executed.
The dissertation makes two significant contributions to the immigration debate. One, it offers a detailed analysis of Hispanic assimilation as measured by IQ scores. And two, it offers a serious argument for using IQ testing in immigration policy. In addition, the dissertation takes a stance on the fraught question of whether racial IQ gaps are genetic in origin; on this Richwine’s position, that they are, is highly contestable, but it is also plausible given the current state of scientific knowledge.
This is not the place to lay out a full defense of intelligence testing — this article from 1995 remains a good starting point — but a few things are beyond dispute. One, regardless of what intelligence is, psychometricians have developed useful tests based on the idea of measuring it; a person’s IQ correlates strongly with everything from his grades in high school to his likelihood of going on welfare. Two, some groups of people perform better on these tests than others; these gaps can’t be explained by bias in the tests, they correspond to gaps in real-world outcomes, and one such gap is between Hispanics and non-Hispanic whites. And three, regardless of why these group differences exist, they do not always disappear — and sometimes they barely budge — when social conditions improve for underperforming groups.
It may be unseemly to ask what might happen with Hispanic IQ as Hispanics move from immigration to assimilation. But given the above facts, and given the reality of large-scale Hispanic immigration to the U.S., there is no denying that the answer is relevant to the future of this country.
Using the National Longitudinal Survey of Youth — a massive government study whose participants take an intelligence test — Richwine compares the IQ scores of Hispanic immigrants with those of later-generation Hispanic Americans. He finds some improvement between immigrants and the second generation, but no additional improvement in the third. This finding is remarkably similar to that of Edward E. Telles and Vilma Ortiz, who studied educational achievement instead of IQ. Richwine also notes other widely known trends; for example, while Hispanic immigrants themselves commit few crimes, subsequent generations are more likely than the native population to wind up behind bars.
In short, though we are constantly assured that the current crop of immigrants will integrate into American society just as easily as previous waves did, much of the actual data on Hispanic assimilation provide reason for concern. The addition of IQ scores to this body of evidence is helpful.
Richwine also suggests using IQ tests in the immigration process. I was initially skeptical: We know we want immigrants with skills, immigrants with education, etc. — why not just base our policy on these factors directly, rather than using IQ as a proxy? With so many powerful groups clamoring for massive low-skill immigration, a policy favoring immigrants with high levels of skill and education will be difficult enough.