This is an article from the archives. Links and some facts and findings may be outdated.

... Or do they? Studies show otherwise


Professor Richard Sander’s recent Stanford Law Review article claims to empirically prove that because affirmative action places black law students in institutions where they cannot compete, they earn lower grades and have higher dropout and bar failure rates. Without affirmative action, blacks would be better off in institutions where they are not academically “mismatched” with their white classmates. How wrong he is. Let us count the ways:

(1) Sander admits that without affirmative action, black enrollment would fall to 1%-2% at elite law schools. (At UCLA, out of 1,000 students at the law school, 40 are black.) Yet in 2001-03, the black graduation rate for the top 25 law schools with affirmative action was almost 97%. How would resegregating elite schools improve black graduation rates?

(2) Sander recently claimed that his results had been replicated by Professors Ayres and Brooks at Yale Law School. Yet Ayres and Brooks’ forthcoming Stanford Law Review critique reports a “reverse mismatch effect,” meaning that in terms of bar passage rates, black law students who attend higher-status schools do better, not worse, than similarly credentialed black students attending lower prestige schools.

(3) Sander claims that ending affirmative action would likely increase the number of black lawyers by 8%. His predictions are impossible. Studies by other researchers applying Sander’s model to 2003 and 2004 data demonstrate projected declines of more than 20%. This stands to reason. The numbers of black lawyers jumped from 3,845 in 1970 to 39,000 in 2000 — a gain attributable in large part to affirmative action.

(4) One would expect that an article relying heavily on social science research methods would appear in a social science peer-reviewed journal instead of a law review journal that is reviewed only by law students. That one of the Stanford Law Review’s student editors is an “econometrician,” as Sander asserted in response to questions about the lack of scholarly vetting of his article, does not qualify as peer review.

(5) Respected social scientists have despaired at the research quality of Sander’s article. As noted Princeton demographer Marta Tienda explained, because Sander’s model has not tested his hypothesis, the article does not meet social science publication standards.

(6) Sander claims “academic mismatch” causes “devastating effects” on black bar passage rates, as illustrated by statistically significant correlations between incoming credentials, law school grades and bar passage. But evidence of correlation is not proof of causation.

(7) Statistical significance is particularly fraught in Sander’s work because the dataset on which he principally relies — the LSAC Bar Passage Study — has more than 21,000 subjects. As such, it is more likely that a host of factors — even trivial ones — will be statistically significant.

(8) Sander claims that racial disparities in performance are “simply a function of disparate entering credentials.” The Grutter v. Bollinger brief filed by UCLA law students of color identifies what Sander ignores: Institutional climate matters. Recent peer-reviewed research utilizing the LSAC data confirms that. Minorities, women and those over 30 earn lower law school grades than their entry numbers would predict, suggesting that institutional factors are affecting outcomes.

(9) Sander claims affirmative action produces “overwhelmingly negative effects” on black students’ grades, undermining their salary gains from attending higher-ranked schools. Part of the data on which he relies is not yet publicly available. So although he asserts that no one has yet contradicted his findings, in fact, some cannot be independently verified.

(10) One could go on. But space and common sense will not permit.

Media Contact