The Supreme Court did not tell universities to stop caring about diversity. It told them they could no longer say so directly — and certainly could not act on it directly. What has followed is an institutional performance of pursuing equity through language that has been legally sanitized of its original intent.
The enrollment data makes the cost of that performance visible. At Harvard, Black student enrollment dropped from 18% to 14% in the Class of 2028. At Boston University, it fell from 9% to 3%. MIT reported underrepresented minority enrollment at roughly 16%, down from approximately 25% in prior years. These are not rounding errors. They are structural outcomes — the predictable result of removing a mechanism that was explicitly designed to counteract the compounding disadvantages that standardized metrics have never been built to capture.
What universities have done in response is recalibrate without naming what they are recalibrating toward. Socioeconomic indicators, first-generation status, zip code, and targeted community outreach have emerged as the new architecture of diversity pursuit. Harvard’s recent class profile includes over 20% first-generation students, with more than half of enrolled students receiving need-based aid. MIT has explicitly pointed to race-neutral alternatives as part of its evolving admissions strategy. These are genuine commitments. They are also incomplete ones, because socioeconomic disadvantage and racial disadvantage are related but not identical. A first-generation white student from rural Appalachia and a first-generation Black student from Roxbury may share a financial profile while navigating entirely different structural realities. Proxies can approximate. They cannot substitute.
This is the central problem with the post-affirmative action moment: institutions are being asked to produce equitable outcomes through tools that were not designed for that purpose, while being prohibited from using the one tool that was. The result is a system that still values diversity in principle, but is structurally constrained in how directly it can name race as a variable — even when race is precisely what the data shows has been affected. The language has shifted from acknowledgment to euphemism, from remedy to approximation.
That shift has consequences beyond enrollment numbers. When an institution cannot name what it is addressing, it struggles to build accountability around whether it is actually addressing it. Targets become vague. Progress becomes harder to measure. And the communities most directly affected — Black and Brown students, their families, the K-12 pipelines that feed these institutions — are left navigating a system that signals commitment while obscuring the metrics by which that commitment could be evaluated.
Boston’s elite campuses are among the best-resourced institutions in the world. If they cannot close the gap under the current legal framework, the gap will remain — described in softer language, but no smaller in practice. The question is no longer whether diversity is valued. It is whether the tools now permitted are capable of producing it, or whether the language of equity has been preserved while the infrastructure of equity has been quietly dismantled.