The Justice Department didn’t need to be subtle about it. On Wednesday, it published findings showing that UC Berkeley Law spent the last two admissions cycles running an explicit racial spoils system — handing Black applicants 5.8 times the admission odds of comparable White applicants in 2025 alone, then dressing the operation in essay prompts about “diversity in our classrooms and community” so applicants would volunteer the demographic data the school was otherwise forbidden by the Supreme Court from demanding directly.
UC Berkeley Law calls this “merit-based opportunity.” The statistics the Justice Department pulled from its own admissions records tell a different story. Half of the Black students admitted in 2024 and 2025 had LSAT scores below those of 95% of admitted White students. More than a third sat below the 99th percentile of their White classmates. The median LSAT gap between Black admittees and their White or Asian counterparts ran roughly five points — a yawning chasm at the top of a test whose scoring band barely tolerates two.
This isn’t an accident. It’s a program.
Berkeley’s defense — that it “complies with all relevant laws” and admits students “based on their merit” — is the same hollow formulation we have now heard from Duke Law, from UC San Diego’s medical school, and from Yale’s medical program before it. The pattern is so uniform it has become a tell: when an institution has been caught engineering its class by race, it issues a statement that admits nothing and denounces nothing, and waits for the press cycle to move on.
The press cycle should not move on. Berkeley Law is the seventh professional school to be found against. UCLA David Geffen School of Medicine. Yale School of Medicine. UC Davis School of Medicine. UC San Diego School of Medicine. Duke University School of Law. George Washington University School of Medicine and Health Sciences. And now Berkeley Law. That is not a list of marginal actors. These are flagship programs at flagship institutions — the schools whose admissions practices set the model that two hundred lesser programs imitate. The seventh data point on the same curve.
But the cui bono question cuts in a direction the Justice Department’s press release did not anticipate. Berkeley kept the diversity-contribution prompt in place after the 2023 Supreme Court ruling in Students for Fair Admissions v. Harvard — Berkeley knew what it was doing. The DOJ knows what it found. The structural-interest question, cui bono in the analytical sense, asks whose interest the enforcement is built to serve.
The executive order that opened the compliance review is titled with characteristic plainness: “ending illegal discrimination and restoring merit-based opportunity.” It names its beneficiary in its own title. It treats race-conscious admissions as the obstacle to be cleared. It does not treat the upstream educational inequality that produces the credential gap between applicants the admissions office is selecting among as the obstacle. The application materials at Berkeley and elsewhere were the workarounds admissions offices built after the Supreme Court cut off the direct route. The workarounds are now the federal case. The upstream credential gap is not the federal case.
This is the donor-class cui bono. The anti-DEI enforcement campaign is built and bankrolled by an apparatus whose interest is not in evaluating the legality of any particular admissions prompt but in dismantling the diversity rationale at every professional school the executive branch can reach. The interest in Berkeley’s specific essay prompt is instrumentally valuable to that apparatus because the prompt is provable, documentable, and contested in a way that produces a clean Title VI case. Berkeley’s admissions dean could have written the same prompt twenty years ago and not attracted a federal case. The prompt became a federal case because the executive branch decided to make it one.
The relabeling is what the bad-faith catalog names: Berkeley’s deliberate outreach to applicants whose race correlates with educational disadvantage is recast as discrimination against applicants whose race correlates with educational advantage. Both characterizations are accurate at the level of formal description. The relabeling is what tells the rest of the story — whose disadvantage is treated as discrimination, whose disadvantage is treated as merit, and whose discrimination is treated as legitimate.
Berkeley, for its part, is using a vocabulary of merit the institution has never fully lived up to. The school admits students based on a bundle of criteria — LSAT, GPA, essays, recommendations, work experience, life experience — and the bundle has always included factors that correlate with class, with parental education, with the ability to afford LSAT preparation, with the ability to take time off work to study. The “merit” that an applicant from a household with two college-educated parents and $2,000 to spend on LSAT prep demonstrates is not the same merit an applicant from a household with neither demonstrates, even when both score identically on the test. Berkeley’s admissions workarounds, whatever their legal status, were trying to address that gap in the bundle. The legal question is whether the workarounds crossed the line the Supreme Court drew in 2023. The moral question is what an institution with Berkeley’s resources owes to applicants whose credentials are produced by an unequal educational pipeline.
Neither the Justice Department’s enforcement nor Berkeley’s response is asking the moral question. The department is enforcing a legal framework that ignores upstream inequality. Berkeley is defending itself within a legal framework that ignores upstream inequality. Both are operating as if the racial gap in credentials is a feature of the applicants and not a product of the system they came up through.
Assistant Attorney General Harmeet Dhillon, who runs the Civil Rights Division, called Berkeley’s conduct “a complete disregard for the rule of law.” That phrasing is correct but understated. What Berkeley Law has been doing is closer to contempt — a knowing, structured effort to obtain by indirection what the Court has forbidden by direct command. The Civil Rights Division is offering Berkeley Law a voluntary resolution agreement. That is the correct next step — and also the test. If Berkeley Law signs, abandons the essay prompts, retires the identity form, and submits to monitoring, the enforcement model works. If it refuses and waits for the political winds to change, the case becomes the vehicle for testing whether the Supreme Court’s 2023 ruling has any enforcement teeth at all. That is a fight worth having, and worth having at Berkeley.
Berkeley can comply. The country should hope it does, because the alternative is the slow-motion collapse of the proposition that an American law school, or medical school, is supposed to admit the strongest applicants it can find, full stop. But the country should also name what neither party is naming. The arc bends only when the upstream question gets named. It will not be named by the Justice Department under Dhillon’s successors. It will not be named by Berkeley’s general counsel. It will be named, when it gets named, by specific people — a Black applicant from a school district that has been defunded for two decades; a Latino applicant whose family could not afford LSAT preparation; an admissions officer who decides to ask the structural question in a forum the legal apparatus cannot reach.
The structural question is what an institution with Berkeley’s resources owes to applicants whose credentials are produced by an unequal educational pipeline. The work of pushing that question into the room — pushing past Berkeley’s workarounds and the Justice Department’s selective enforcement and the legal apparatus that lets both pretend the other is the whole story — is the work the next decade of admissions fights will turn on.