Regulars on this site have probably already read Ron Unz’s provocative analysis of  admissions policies at Harvard, Yale, and Princeton. If you haven’t, you should. Drawing on an apparently unprecedented array of data, Unz argues that these policies amount to a de facto quota for students of Asian origin. At the same time, they tend to inflate the number of less accomplished Jewish students.

Unz’s piece is welcome for a number of reasons. First, it’s another blow to the myth of meritocracy, which I’ve discussed in a review of Chris Hayes’s Twilight of the Elites. Americans like to believe that scarce goods like Harvard degrees (and the glamorous careers to which they can lead) are generally earned by the most capable and hard-working among us. Unz contributes to the mounting pile of evidence that’s not simply true. Second, Unz effectively reframes the debate about  affirmative action. As Dan McCarthy points out, the problem in elite universities is not controversial but relatively public preferences for blacks and Hispanics. It’s that “[a]lmost everyone is getting affirmative action or other unearned advantages of one kind of another, but only some minorities have been marked out as, in effect, charity cases.” Finally, Unz deserves credit for breaking the taboo surrounding Jews’ vastly disproportionate presence in the upper reaches of American higher education. If we’re going to talk about what fairness requires the student bodies of these institutions to look like–and the Supreme Court has just engaged in such a discussion–this fact cannot be left off the table.

Nevertheless, I find several of Unz’s conclusions unconvincing. I’ll leave entirely aside statistical issues on which I am unqualified to pronounce (I scored only 580 on the math SAT, and was not what Unz calls a “quality” applicant). And I’ll merely echo Tyler Cowen’s discomfort with Unz’s methods of counting members of various ethnic and religious groups, which include inferring their backgrounds from their names. The problem is not only that these methods resemble the Jew-hunting of classical anti-Semitism. It’s that they ignores the complexity of origins and identity in a society characterized by intermarriage among religious and ethnic groups.

Instead of dwelling on these concerns, however, I want to challenge Unz’s finding of a “collapse in Jewish academic achievement” in the last decade or so. This surprising conclusion is based on three trends: the declining number of (apparently) Jewish winners in math competitions including the U.S. Math Olympiad and Putnam Exam; the declining number of (apparently) Jewish winners in various science competitions; and the declining number of (apparently) Jewish National Merit semifinalists.

Of these measures of academic achievement, only the last involves any major verbal component, which is nevertheless weighted against math scores on a 2:1 basis. Achievement in “productive” verbal pursuits, namely writing, and substantive knowledge of history, foreign languages, and literature, are not measured at all. As such, Unz presents strong evidence of a collapse of (apparently) Jewish accomplishment in achievement in math and science. He presents only weak evidence of a “collapse” of Jewish accomplishment in other academic areas.

This is important because Unz argues that Jews are disproportionately represented at elite universities not only relative to the population as a whole, but also relative to the national cohort of high-achieving students. The cohort of high achievers as Unz defines it, however, is significantly biased toward students who excel in mathematics, hard sciences, and engineering.

These subjects are only part of the picture. A more plausible account of academic accomplishment would consider the humanities and liberal arts. The standards for achievement in these areas are more subjective than success on math exams. But Unz could have considered the results of national competitions in more humanistic pursuits, such as debate, student journalism, and so on. He could also have looked at achievement on AP and SATII exams, which are offered in a number of humanities subjects and languages.

The narrowness of Unz’s definition of academic achievement is connected to a broader defect of the piece. In addition to neglecting students’ ability in the liberal arts, Unz does not consider the liberal arts’ contribution to the university as whole. Unz’s model of meritocracy is Caltech. Not coincidentally, Caltech is an engineering school, which has only a vestigial presence in the humanities and liberal arts. Caltech is a wonderful institution. But would Harvard be more “meritocratic” if its student body and course offerings were more like Caltech’s? Would it be a better university? I doubt it, and not only for reasons of self-esteem.

The differing missions of tech schools and the Ivy League universities could also help clarify one of the underlying uncertainties in Unz’s analysis. Since he doesn’t have access to data about who applies to various universities, he assumes that their numbers are roughly proportionate to the national cohort of high-achievers. This leads to the conclusion that Jews are overrepresented.

But what if Jews are disproportionately likely to apply to Ivy League universities? In other words, what if their representation on those campuses is reflective of their representation in the application pool? On the other hand, qualified Asian students may be more likely to apply to technical schools than to Yale. That could explain their relative underrepresentation at the latter.

Unz dismisses this possibility, asking: “Why would high-ability non-Jews be 600 percent or 800 percent more likely to apply to Caltech and MIT than to those other elite schools, which tend to have a far higher national profile?” The answer may be precisely that  Caltech and MIT  focus on math and science, in which Asian students achieve the very success that Unz documents.

This kind of self-selection could also be self-reinforcing. More high-achieving Asians might apply to Caltech or MIT to pursue their interests and abilities–or to satisfy their parents. As a result, more Asians would be admitted, giving these schools a reputation for friendliness to Asian students, encouraging yet more applications.

A complicating factor in college admissions is that applications are driven as much by reputation and rumor as by rational evaluation of students’ chances for admission and success. That’s one reason colleges are fanatical about producing brochures including photographs of students from every possible ethnic group: they’re trying desperately to counteract the false assumption, for example, that there are “no blacks” at Yale.

I don’t know if any of the possibilities I’ve mentioned are true. In combination with the narrowness of Unz’s definition of academic achievement, however, they suggest that there’s a lot more evidence to be sifted and thinking to be done before we can hope to understand why some students rather than others are invited through the gates of Harvard Yard.

We also should not assume that current trends will continue forever. Based on anecdotal evidence, it appears that many Jewish students have shifted their interest from quantitative pursuits, in which they excelled in the recent past, to humanities subjects, which were once dominated by WASPs. If they are really the “new Jews,” Asians may do the same over time. That could create room for students of different backgrounds to make their mark in math and science.

Yet Unz is not entirely at fault for the thinness of his evidence and the ambiguity of his conclusions. The bulk of the blame goes to the gatekeepers themselves, who wield enormous influence but resist almost every attempt to open their decisions to public scrutiny.