Harvard Clarifies Race Data Reporting Practices Following Confusion

Harvard College released clarifications to its racial breakdowns for the Class of 2028 after a Crimson report that found inconsistencies between the school’s posted comparisons with the Class of 2027 and data the school shared last year.

The school quietly announced via a Friday note on its website that the data — which showed a modest dip in the share of Black students in the freshman class — did not account for international students, a distinction the College had not initially reported.

College officials previously said they changed their methodology by reporting the percentage of students who chose to report their race on their application. Until this year, the College used to present percentages out of the full matriculated class, including international students and students who did not report their race.

This new clarification suggests that — relative to the full freshman class — the true share of students from each demographic group may be lower than what was implied by the initial release.

The College’s decision to post clarifications to the data for its first class after affirmative action is an implicit admission that the numbers, as initially presented, were hard to fully parse.

But while the clarifications answer some questions, they fail to fully address why Harvard moved to change its years-old methodology for reporting race data, especially during a period of intense scrutiny following a Supreme Court loss.

College spokesperson Jonathan Palumbo declined to provide details of that decision, though he said the College would report its race data in a consistent manner going forward.

The College’s clarification about its data source also raised a new question about how the school chose to construct the comparisons presented in its announcement.

While 4 percent of the Class of 2027 chose not to identify their race at the time they applied, per Palumbo and the Wednesday release, the data submitted by Harvard to the Common Data Set only lacks race data for about 2 percent of the same class.

Palumbo declined to comment on the inconsistency between the two numbers. Though he wrote in a statement that the data used in the release was obtained using students’ responses on their application forms, he declined to say whether the data in the Common Data Set came from a different source — and, if that was the case, why the College elected to use a seemingly less complete data source.

—Staff writer Elyse C. Goncalves can be reached at elyse.goncalves@thecrimson.com. Follow her on X @e1ysegoncalves or on Threads @elyse.goncalves.

—Staff writer Neil H. Shah can be reached at neil.shah@thecrimson.com. Follow him on X @neilhshah15.