After the Supreme Court ended race-conscious college admissions in 2023, the 2024-25 academic year was seen as a kind of test: What effect would the decision have on freshman classes?

At the start of the school year, the Upshot asked selective colleges for the racial and ethnic composition of their incoming classes. We obtained this data for 66 colleges, allowing us to put together the most detailed look yet at how the makeup of these colleges changed after the end of affirmative action.

While it’s still early to draw definitive conclusions — it will be years before we understand the full impact of the ban — here are three things we learned.

1. Black and Hispanic enrollment declined on average

One of the arguments for affirmative action was that it helped compensate for disadvantages faced by Black and Hispanic students, and that without it, their enrollment would fall. So did that happen?

Black and Hispanic students have historically been underrepresented at selective colleges compared with their share of high school graduates. Over the past decade and a half, their numbers have gradually risen at selective colleges.

That changed in 2024.

Of the 66 colleges for which we have data, 59 reported it in a standard format that let us make comparisons with past public records. That format narrowly defines Black students as U.S. residents who are Black but not Hispanic or multiracial — which means it’s an undercount of all students who identify as Black, even as it allows for a straightforward comparison with past data.

By that definition, the average share of incoming Black students at those colleges dropped by about one percentage point — from about 7 percent to 6 percent.

The share of incoming Hispanic students at these colleges also fell by nearly one percentage point — from about 14 percent to 13 percent.

Together, these changes represent the largest annual drop in the average share of Black or Hispanic students across these colleges since 2010.

Outcomes varied across individual colleges. Some experienced large drops in their shares of Black or Hispanic students, while others experienced modest changes or even increases (we have more details on this variation below).

But the overall trend was downward: In 2024, 17 of these 59 colleges experienced their largest drops in Black enrollment in 14 years, while only three had their largest gains. Similarly, 15 colleges had their largest drops in Hispanic enrollment in over a decade, while four had their largest gains.

2. The data didn’t show a comparable increase in Asian and white enrollment

Over the 14 years for which we have data, the share of white students has decreased at selective colleges — in part reflecting the changing demographics of America’s youth.

Asian students are overrepresented at selective colleges compared with the makeup of high school graduates. Their share of college enrollment has continued to rise in the past 14 years.

Critics of affirmative action argued that it suppressed the share of Asian and white students at top colleges. There was a broad expectation that banning race-conscious admissions would drive up Asian and white enrollment.

We didn’t see sizable changes.

Across the 59 selective colleges we could compare with historical data, the average share of Asian students was essentially unchanged.

And the average share of white students increased by under one percentage point.

There were outliers, of course: The share of incoming domestic Asian students at Johns Hopkins rose by 18 percentage points, and the share of incoming domestic white students at Middlebury rose by 10 percentage points.

The findings by race and ethnicity so far raise a question: College enrollment is zero sum, so if the Black and Hispanic share went down, why didn’t the white and Asian share rise similarly?

Our third finding helps explain this discrepancy.

3. Many more students did not disclose their race in 2024

In 2024, the share of students at these 59 colleges who did not disclose their race or ethnicity increased from about 2 percent to 4 percent. That share had been generally declining since 2010.

This means we don’t definitively know how much of the changes that we see in 2024 are caused by genuine shifts in racial makeup, and how much are because students didn’t report their race.

Others have tried to answer this question. A 2020 study by the economist Zachary Bleemer found that after California banned affirmative action in public universities in 1998, more than twice as many applicants to the University of California system left out their race or ethnicity on their applications the following year.

By using their name, high school and neighborhood to infer their race, he estimated that the vast majority of the students who left out their race were white or Asian.

“Is this true again today? I don’t know for sure, but I think it’s a good guess,” Professor Bleemer said.

If most of the students who left out their race were white or Asian — perhaps out of wariness of the admissions process amid heightened media coverage of affirmative action — the charts above wouldn’t fully reflect the rise in their share. But it’s also possible that the lessons from California don’t apply here.

“It’s important to be careful not to speculate too much about who the unknowns are,” said James Murphy, who directs postsecondary policy at Education Reform Now and has been tracking the effects of the ban. “This is a very unusual year.”

By contrast, in 2014 there was a dip in the number of students who didn’t report a race. Those students would have applied after the Supreme Court decided a high-profile case about race-based college admissions, essentially allowing the practice to continue. The reasons for that dip are ultimately mysterious as well.

The share of multiracial students at these colleges declined in 2024 by about half a percentage point. Previously, multiracial students made up an increasing share of college enrollment.

Because the available data doesn’t get any more granular, this decline could reflect any of several things: a smaller number of multiracial Black or Hispanic students enrolling, a tendency for students to leave out their race altogether, or fewer multiracial white students listing a second race, for example. In time, we may understand these shifts better as researchers pore through additional data.

But the overall picture resists simple generalizations.

A wide variation in outcomes

The charts above show the average enrollment. But as mentioned, there’s a wide spread in the outcomes at individual colleges.

In addition to the 59 colleges that reported their enrollment figures in the standard format used by the government, seven (Barnard, Brown, Columbia, Duke, Harvard, M.I.T. and Yale) reported their enrollment figures using a different method. For completeness, we included those colleges in the charts below.

You can select a circle to see more about that college.

Some caveats

Although this data offers a detailed look at shifts in enrollment, there are several caveats.

For one thing, the data for 2024 is preliminary — we collected it before colleges were required to submit these figures to the government. And as we noted, many more students didn’t disclose their race in 2024.

Here are a few more notes of caution:

1. There were problems signing up for financial aid in 2024.

A new application system for a federal program that tens of millions rely on for financial aid faced major problems in 2024, which may have altered the pool of college applicants.

2. We have only one year of post-affirmative-action data.

A single year of data isn’t a lot. Colleges are still making sense of the Supreme Court’s decision, and exploring ways to create a diverse college class without affirmative action. It will be a few years before we gain a more comprehensive picture.

3. Enrollment isn’t admissions.

To enroll in a college, three things have to happen: A student has to apply, be admitted and decide to attend. We have only a picture of college enrollment — which combines all three steps.

So we can’t yet disentangle how much of these shifts are because of changes in admissions (which the Supreme Court’s decision directly affects), versus changes in who decided to apply beforehand or who decided to accept afterward.

For now, with these caveats in mind, you can go through the enrollment data for the colleges that we analyzed below.

Look up a college

For the 59 colleges that reported enrollment using the federal categories for race, you can look up how their racial makeup has changed over time below: