Letter on ABA Proposal for "Diversifying" Law Schools

Signed by Richard Peltz-Steele (Massachusetts), Richard Sander (UCLA), Robert Steinbuch (Arkansas-Little Rock), and me.

I thought I'd pass this along; you can also see the PDF:

Last month, the ABA's Council of the Section of Legal Education and Admissions to the Bar approved for Notice and Comment proposed revisions to Standards 205, 206, 303, 507, and 508 of the ABA Standards and Rules of Procedure for Approval of Law Schools. The revisions to Rule 206 would significantly alter the responsibilities of law schools to achieve "diverse" and "equitable" environments. In response to your solicitation of comments, we offer the following:

[1.] The proposed Rule 206(b) provides that "a law school shall take effective actions that, in their totality, demonstrate progress in diversifying the student body, faculty, and staff…." There appear to be no exceptions, indicating that the language requires that all law schools must demonstrate progress.

"Diversifying," to judge from the annotations to the rule, means "adding people of color" (not "minorities," which the annotations say is an outdated term). Yet "progress" is nowhere defined; indeed, there is not even a hint of what it means to fully satisfy this standard. According to the ABA's own website, which reports the proportion of first-year law students in 2020-21 who are "minorities" (we assume this means "people of color"), the makeup of the 197 ABA-accredited law schools ranges from 8% "minority" to 100% "minority." According to the ABA data, minorities make up more than 90% of students at four schools, and more than half the students at 24 schools. Presumably, these schools are also mandated to achieve greater diversity; does that mean they must find ways to enroll more whites?

If there is an implicit goal, is it the same nationwide, or does it depend on the demographics of a school's region? Any useful effort to create usable guidance to law schools must, at a minimum, address these and other similar questions. The standard, as written, is so vague that it will give enormous discretion to ABA accreditation committees to exert arbitrary control over important and sensitive policy issues.

The proposal fails to account for the fact that among the current population of law school applicants, there are very large disparities in credentials that correlate with race. For example, among all students taking the LSAT, there is about a 1.0 standard deviation gap between the mean score of white takers and the mean score of black takers. The white-black gap in college grades is smaller but still very large (about .8 standard deviations). It is difficult to argue that either of these credentials is discriminatory, since they are predictive of law school grades and subsequent bar performance, and their predictions are as valid for blacks as for whites. Indeed, to the extent there is a debate over the relationship between black credentials and black law school performance, it is whether LSAT scores and college grades overpredict law school performance.

The large credential gap means, of course, that law schools have resorted to large racial preferences as the main method of increasing the numbers of enrolled blacks, Hispanics, and American Indians. The best data we have on this come from admission records released in 2007-08 by 41 public law schools in the U.S., which in the aggregate show that roughly 60% of blacks entering these law schools had academic credentials that were at least a standard deviation below those of their median classmate. (This was also true for about 30% of Hispanic first-years, compared to about 6% of Asian-American students and 4% of whites.) A major failing of the proposed Rule, therefore, is that since it provides no guidance on how the existing pool of law school applicants can be meaningfully expanded, it necessarily implies that greater "diversity" should be achieved by using even more aggressive racial preferences.

[2.] The proposed Interpretation 206-2 asserts that "the enrollment of a diverse student body has been proven to improve the quality of the educational environment for all students" but cites no evidence to this effect. So far as we are aware, no one has even attempted to study, in a scientifically credible way, the effect of diversity on legal education quality or outcomes.

Careful studies have been done at the undergraduate level, but these studies come to very different conclusions. Importantly, the leading studies that find positive educational benefits from diversity (notably, those by Patricia Gurin and her colleagues) do not take into account how those benefits are affected when schools use large racial preferences to achieve diversity (as nearly all law schools do). The research that does take large preferences into account (such as the work of Arcidiacono et al. at Duke, or the work of Carrell et al. at the Air Force Academy) finds that large preferences can directly undermine the goals of a diverse environment and increase racial segregation and isolation. There is also, of course, the very real danger that if race correlates very highly with class performance — an outcome difficult if not impossible to avoid if large racial preferences are used — then the single-minded pursuit of diversity will create, rather than erode, racial stereotypes.

[3.] The proposed Rule and accompanying interpretations conspicuously ignore the likelihood of "mismatch" — that is, the potential harmful effects of very large preferences upon the intended beneficiaries (in terms of law school grades, bar passage, and long-term outcomes). In 2007, the U.S. Commission on Civil Rights issued a lengthy report on law school mismatch, finding grave cause for concern and urging further investigation, but the ABA has never taken up this question. This inaction persists despite the fact that the Journal of Legal Education recently accepted for publication a new empirical study showing compelling evidence that law school mismatch has large, negative effects upon bar passage. [Richard Sander and Robert Steinbuch, Mismatch and Bar Passage: A School-Specific Analysis (2020).]

There is heavy attrition of students admitted with large preferences, first in terms of graduation from law school and second in terms of passing state bar exams, and this is at least arguably the major reason the legal profession remains as predominantly white as it still is. The committee's proposal not only ignores this fundamental problem, but creates pressure on schools to worsen it.

[4.] Finally, the proposed Interpretation 206-1 states that "The requirement of a constitutional provision or statute that purports to prohibit consideration of race, color, ethnicity … in admissions or employment decisions is not a justification for a school's non-compliance with Standard 206 …. [Such a school must] demonstrate the effective actions and progress required by Standard 206 by means other than those prohibited by the applicable constitutional or statutory provisions." Setting aside the problem noted earlier — that "effective actions and progress" are nowhere defined — the predominant method that schools have used to increase the number of enrolling members of underrepresented racial groups is the use of ever-larger admissions preferences.

Case law in the states that have prohibited the use of race-based preferences makes clear — not surprisingly — that such preferences do, in fact, violate the law. In the absence of any explanation or documentation of other, proven methods by which schools can make "progress," the proposed standard places these schools in an impossible bind — violate the law and the civil rights of applicants, or risk losing accreditation. Putting schools in this impossible bind would be an abuse of the ABA's professional responsibility as an accreditor.