The Hidden Acceptance Rates: US Universities That Stopped Publishing Their Numbers

Five of the thirteen we track: Stanford from 2018, Cornell from 2020, Princeton and Penn from 2022, and Harvard most recently. UChicago is a sixth case, having never routinely announced its figures, and it is the only one that discloses nothing about its early rounds either.

The stated reasons are about the harm the number does. Princeton’s admissions dean wrote that highlighting the statistic instils anxiety and fear, and that neither applicants nor the university benefit from admissions being reduced to a single figure in a headline. Stanford’s provost pointed to coverage that exists mainly to rank universities by scarcity. Sceptics note that the same silence is equally compatible with wanting application volumes to keep climbing.

From the filings the universities are obliged to make. Every institution taking part in US federal student aid must report admissions data to the Department of Education, and most also publish a Common Data Set on their own website. Harvard is the clearest illustration: it declined to announce a rate, then published a Common Data Set on a Harvard domain giving 47,893 applicants and 2,003 admitted, which is 4.2%.

A standardised template, developed jointly by universities and publishers, that reports admissions and enrolment figures using agreed definitions. It is voluntary but widely followed, and it is more detailed than federal reporting because it separates early decision and early action figures, which federal filings do not. Most universities post it on their institutional research pages.

More reliable, if anything. They come from the university’s own reporting using standard definitions, rather than from a press release that can choose its framing and its denominator. The trade-off is timing rather than accuracy.

Between 9 and 15 months. Federal admissions data is collected over the winter and released the following autumn, and Common Data Sets typically appear in the summer after the cycle closes. This means the current year’s rate is genuinely unknowable until roughly a year later, for anyone, including the press.

Not necessarily, and the filings show movement in both directions. Since going quiet, Stanford and Cornell have become more selective, Princeton is slightly less selective than its pandemic-era low, Penn is broadly stable, and Harvard has become less selective, with applications down 12% over two cycles. Silence is not evidence of anything in particular.

Probably not, or at least not as a direct cause. Stanford stopped in 2018 and Cornell in 2020, both well before the 2023 decision, and none of the universities has given this as a reason. It is a tidy explanation that the timeline does not support.

Each time a new filing appears, which in practice means autumn for federal data and summer for Common Data Sets. Every row is flagged with the entry year it describes, so you can see the vintage of each figure rather than having to assume.