"The Waning of Racial Preferences at American Law Schools, 2021-2025," by My UCLA Colleague Rick Sander

From this new draft, cowritten by Prof. Sander and Henry Kim:

In June 2023, the Supreme Court broadly held that colleges and universities could not confer an admissions advantage to an applicant on the basis of the applicant's race. The ruling was fairly clear, but its likely effects were not. There were two major uncertainties. Would the institutions that had been using large racial preferences for decades actually change their practices in a substantive way? And if they did, would Black and Hispanic enrollments at elite colleges and professional schools crash?

The schools themselves were not forthcoming about their plans, aside from generally expressing a determination to preserve diversity. Higher education leaders had long maintained two conflicting precepts: first, that their admissions officers took account of race to only a modest degree, but second, that prohibiting these modest preferences would have a catastrophic effect on minority enrollments. They did not willingly share any data on actual admissions processes that would demonstrate whether either of these claims were true.

Journalists and education scholars thus watched closely when colleges and professional schools began to release enrollment data in the fall of 2024. The picture remained muddy: some schools, such as MIT and Amherst, reported large drops in undergraduate minority (especially in Black) enrollment; other schools, such as Yale and Princeton, reported either small drops or actual increases. Among professional schools, few reported large drops in minority enrollment. So, what was happening? Were elite colleges and universities generally ignoring the law? Had they found some way around the Court's decisions? No small part of the Trump Administration's offensive against many schools in 2025 seemed motivated by this perception. But pretty much everyone, including higher education leaders themselves, has had few real clues about how much actual practices were changing.

This article provides the first systematic analysis of how, in an important realm of higher education, the use of race in admissions has changed over the past few years. We focus on the two hundred accredited law schools in the United States, partly because the past patterns of racial preferences in law schools are particularly well understood. We make use of a new data source—a website on which applicants themselves report the outcomes of their law school applications—that allows us to draw credible conclusions about the operation of law school admissions in each of the past five admissions cycles. Over twenty thousand applicants have used the website "lsd.law" over the past five years to self-report their academic credentials, their race, and the outcomes of their law school applications.

This data is subject to a few caveats, and it is noisier—that is, it has more inaccuracies—than data directly reported by law schools, but we find that it nonetheless provides highly credible estimates of law school admissions practices. (See Section 2 for a more detailed assessment of the data.) In particular, we have enough observations to estimate accurately the average size of racial preferences across each of six "tiers" of law school, ranked by admissions selectivity, for each of the five admissions cycles from 2020-21 through 2024-25.

We find that racial preferences at elite law schools were much the same in 2021 and 2022 as they had been for decades—that is to say, very large and very widespread for Black applicants; about half as large, and less consistently applied, for Hispanic applicants. By the 2024-25 admissions cycle, however, the size of racial preferences for Black applicants had fallen by about half, and preferences for Hispanic students had similarly fallen by half in some tiers, and in others had fallen to levels too small to be detected in our data. Presumably because of the decline in preferences, longstanding credential disparities between Black students and their white and Asian classmates fell as well—indeed, perhaps fell even more sharply. But despite the diminished use of preferences (and possibly, because of the diminished use of preferences) Black and Hispanic applications to law schools rose in 2023-24 and 2024-25, and, minority enrollments in law schools as a whole rose in absolute numbers while falling only modestly in relative terms.

Six Key Findings

1. When we compare the admissions decisions of law schools in 2020-21 and 2021-22 with decisions in 2006-07 and 2007-08, we find virtually no change in the use of racial preferences at the top four law school tiers (i.e., the eighty most competitive law schools, which together account for about 60% of all law school seats). We do, however, find a that racial preferences shrank considerably over this period in the bottom two tiers. As we explain below, this is probably due to discussions within the American Bar Association ("ABA") during the 2010s about establishing minimum bar passage standards for accredited schools, and the eventual imposition of those standards in 2019. Importantly, the close similarity in the results of our models for 2006-08 (which use application data disclosed by law schools) and our models for 2020-22 (which use data from the lsd.law applicant website) gives us confidence that the website data is reliable enough to make valid inferences about changes in law school admission practices in recent years.
2. The use of racial preferences in the top four tiers fell significantly in the 2023-24 admissions cycle, and fell further in the 2024-25 cycle. Overall, Black preferences at the top four tiers were about half as large in 2024-25 as in the 2020-21 and 2021-22 cycles. Preferences for Hispanics fell as well, to a point where one cannot be statistically confident that they exist at all at many elite law schools. This strongly suggests that the Supreme Court's decision in the SFFA cases has had a powerful effect upon law school admissions, though it has caused very few elite law schools to abandon racial preferences for Blacks altogether.
3. If racial preferences for Black and Hispanic students fell, then we would expect that the "credential gap" between these students and their classmates to fall as well. This is exactly what we observe, at least for Black students; according to lsd.law information on enrollment, the median credential gap between Black versus white and Asian students fell from over ninety index points to less than forty index points over the five years we analyze. If the "mismatch hypothesis" is correct—and a great deal of evidence suggests that it is—it should be the case that Black performance in law schools is improving, and will continue to improve, in recent cohorts relative to earlier ones. Black GPAs should be rising relative to other students, and over the next few years Black graduation rates and (especially) bar passage rates should go up.
4. A key question in debates about affirmative action has been whether Black and Hispanic students would be discouraged from applying to colleges and professional programs if preferences were abolished. Many argued that minority applications would be severely "chilled." Others argued, with evidence, that the opposite effect (a "warming effect") was just as likely: minority students would prefer to attend schools where there was less or no stigma attached to their admission, and would more highly value degrees that prospective employers would not discount as "affirmative action degrees." Among law school applicants, something like the "warming effect" seems to have occurred: law school applications from both Blacks and Hispanics rose in 2023-24 and again in 2024-25, both in absolute terms and relative to applications from whites and Asian-Americans.
5. Despite the substantial decline in racial preferences, Black and Hispanic enrollment actually rose substantially, in absolute numbers, for the 2024 and 2025 entering classes. However, law school enrollments were growing generally in these two years, so Black enrollment fell in relative terms (i.e., as a percentage of all first-year law students) by about five percent from 2023 to 2025. Even so, given the large drop in preferences, a five percent decline in Blacks' relative share (from 8% to 7.6%) is more modest than most (but not all) observers expected. The decline was so modest because of three things: (1) applications from minorities rose; (2) applicants who were turned down by more elite schools were accepted by and matriculated at less-elite schools; (3) non-elite law schools had already largely phased out racial preferences by the time of the SFFA decision, so they continued to admit the sort of minority students they had admitted pre-SFFA, and also enrolled many additional minority students who had been rejected by more elite schools. As the tables in Section 6 show, by any measure the enrollment effects of reducing preferences have been mild at the great majority of law schools.
6. The decline in the use of racial preferences is not uniform across schools—indeed, there is substantial variation even with law school tiers in how much individual schools reduced preferences from 2020-22 to 2023-25. We provide some illustrative examples of this phenomenon; these give us additional confidence in lsd.law data, because the changes in preferences that we observe from that data correlate well with changes in minority enrollment we observe in the data reported by law schools.

The authors also had a Wall Street Journal op-ed on this earlier this week.