Affirmative Action Was Effective—and Unjust
What Sean Carroll Gets Wrong About Race-Conscious Admissions
Sean Carroll’s defense of affirmative action rests on two claims: that racial preferences worked, and that opposition to them is largely driven by racism, resentment, or bad faith.
The first claim is incomplete. The second evades the strongest objections.
Affirmative action worked in one clear sense: it increased the admission of Black and Hispanic students at selective universities. But selective admissions are zero-sum at the margin. A preference for one applicant lowers the probability of admission for another. The policy transferred opportunities toward favored racial groups and away from others, especially Asian applicants, many of whom had stronger conventional academic credentials.
That does not mean the beneficiaries were incapable of succeeding. Elite universities reject large numbers of applicants who would perform well. A university can lower the effective admissions threshold for one group while still admitting capable students and preserving strong graduation outcomes.
Carroll treats those outcomes as evidence that the policy succeeded. They establish that many beneficiaries gained from admission and that the institutions remained academically strong. They do not establish that the allocation was fair.
The standard institutional defense is that universities are constructing a class rather than rewarding applicants according to a single ranking. Demographic diversity may broaden the range of experiences, intuitions, and hypotheses available to a group facing open-ended problems. It may also improve discussion, collaboration, or institutional legitimacy.
Even granting those benefits, the moral question remains. Aggregate benefits do not automatically justify allocating opportunities by race. Defenders must show that the benefits are substantial enough to justify unequal treatment, that racial categories track the relevant differences with acceptable accuracy, and that less discriminatory alternatives would sacrifice comparable benefits.
Holistic admissions are not inherently objectionable. Universities may reasonably value leadership, adversity, unusual talent, geographic background, or achievement relative to opportunity. The problem begins when race itself changes the effective standard, or when vague holistic criteria provide cover for doing so.
Race is also a crude proxy for disadvantage. It can benefit affluent members of preferred groups while penalizing poor or otherwise disadvantaged Asians and whites. Socioeconomic measures are not perfect substitutes for race, because racial discrimination can persist independently of income. But universities can assess disadvantage more directly through family wealth, parental education, school quality, neighbourhood conditions, disability, family instability, refugee status, and documented experiences of discrimination. Those measures concern what happened to an applicant, not merely which ancestry category the applicant occupies.
Affirmative action could therefore be both effective and unjust.
Carroll acknowledges that race is an imperfect proxy. His defense is partly administrative. Identifying each person’s actual disadvantage is, he says, “completely unfeasible,” whereas group preferences are easy, simple, and bureaucratically undemanding.
Individualized assessment cannot perfectly reconstruct every applicant’s history of advantage and discrimination. Neither can race. The relevant comparison is not between a flawless individual process and an imperfect group proxy, but between two imperfect systems. Race is cheaper, but it is also coarser and assigns benefits and burdens partly by ancestry.
Administrative economy cannot by itself justify racial discrimination. A crude policy may be cheap and effective while remaining unjust.
Carroll’s treatment of dissent is worse. He treats opposition to affirmative action as part of a culture-war campaign ungrounded in science or research, attributes some of it to explicit racism, and characterizes another part as resentment toward “undeserving” people receiving opportunities that properly belong to someone else.
There are obvious non-bigoted objections.
Racial affirmative action judged applicants partly by ancestry. It imposed costs on people who did not cause the injustice being remedied. It used broad racial categories where more individualized evidence was available. It produced different effective admissions standards for different groups. It encouraged opaque admissions systems that concealed the magnitude and distribution of those preferences.
One can defend those choices. One cannot dismiss opposition to them as scientifically uninformed grievance.
Carroll is entitled to believe that the benefits justified the discrimination. He is not entitled to present that judgment as the uncontested conclusion of research.
The central question was never whether affirmative action helped anyone. It plainly did. The question was whether universities should allocate scarce opportunities by race.
Carroll assumes the answer, then assigns suspect motives to much of the opposition. That is political advocacy presented with scientific confidence.