clock menu more-arrow no yes mobile

Filed under:

The Asian penalty in college admissions is still here — even without affirmative action

Legacy and geographic preferences will continue to favor white applicants, a new study finds.

People walk past a red brick building.
Pedestrians pass a dormitory in Harvard Yard at Harvard University.
Craig F. Walker/The Boston Globe via Getty Images
Fabiola Cineas covers race and policy as a reporter for Vox. Before that, she was an editor and writer at Philadelphia magazine, where she covered business, tech, and the local economy.

The Supreme Court’s decision to strike down affirmative action in college admissions — effectively banning the consideration of an applicant’s race — will not end advantages awarded to white students during admissions. A new study backs up this idea with ample evidence.

Researchers analyzed almost 700,000 college applications from white and Asian students and found that admissions at selective colleges rewarded privileged applicants who are disproportionately white. In other words, the report reinforces the idea that there is a potential open bias against Asian American applicants. Because of this alleged Asian penalty, it’s likely that Asian American applicants will continue to be admitted to selective schools at lower rates than similarly qualified white applicants — even with affirmative action gone.

The report identifies two key factors that cause the alleged penalty: the favorable treatment given to children of alumni, who are usually white, and admissions patterns based on geography.

Since the affirmative action decision, the practice of legacy admissions has come under fire. The Department of Education is now investigating Harvard’s preferential treatment of legacy applicants. The report highlights just how jarring the legacy boost is. On average, legacy applicants are two to three times as likely to be admitted to a selective college than non-legacy applicants with comparable academic credentials.

I talked to the report’s lead author, Josh Grossman, a PhD candidate in computational social science at Stanford University, about the report’s findings and why he and his co-authors say it’s past time to rethink legacy admission. Grossman walks me through the report’s methodology and what sets it apart from previous similar studies. One key detail is that the researchers analyzed outcomes separately for East Asian students, Southeast Asian students, and South Asian students, recognizing that there were major differences between the groups.

Ultimately, the report suggests, the gaps between white and Asian applicants will not change with the elimination of affirmative action. “The raw numbers of Asian American and white students will probably go up,” Grossman said, “but those differences in admission rates are still going to be there.”

Our conversation has been edited for length and clarity.


Fabiola Cineas

Are Asian American students admitted to selective colleges and universities at lower rates than white students with similar academic qualifications?

Josh Grossman

Yes. Our study, and many prior studies, have shown that this is the case. If you just condition [admissions] on academics — compare students who have similar test scores and other academic qualifications — what is found time and time again at these selective colleges is that Asian American students are admitted at lower rates just based on academics. I do want to emphasize that our study focused on academic qualifications, though there are other qualifications taken into account.

Fabiola Cineas

You looked at close to 700,000 anonymized applications from white and Asian applicants to a group of selective colleges for five years, starting in the 2015-2016 school year. But you didn’t have the full application materials. What elements of students’ applications were you able to analyze and which elements were you unable to analyze in this study?

Josh Grossman

What we did have was quite substantial. Our data include students’ high school GPAs, standardized test scores, extracurricular participation, leadership record, their AP classes, and what their courses look like. We knew where their parents went to college. That’s important because while legacy status is somewhat controversial, it can be a boost in admissions for students who had parents who went to the same school.

We didn’t have access to their specific transcripts so we couldn’t see progression through high school. We know that’s important in admissions. We didn’t have access to their intended majors and we know that’s another thing that’s taken into account during admissions. We didn’t observe whether an applicant was an athletic recruit or not, but we tried to infer it based on how they prioritize their reporting of sports on their application and when they were admitted. Since admissions works differently for recruited athletes, we excluded them from our main analysis.

We also didn’t have access to student essays. And one thing to note is that these other factors themselves, that we didn’t have access to, can potentially encode their own biases. We saw in the Harvard case that the personal ratings appeared to have some degree of bias against Asian American applicants. So that’s something else to think about.

Fabiola Cineas

You also didn’t have actual admissions outcomes. So out of the applications you analyzed, you didn’t know where the students were admitted or denied admission. To circumvent that, you created a proxy based on enrollment choices. What does that mean and how did it allow you all to draw conclusions?

Josh Grossman

Particular schools have a measurement called the yield rate, where if a student is admitted to a particular school it’s the proportion of admitted students who actually enroll. There are yield rates as high as like 80 percent for schools like Harvard, the University of Chicago, MIT, and Stanford. That’s a pretty high yield rate but keep in mind, there’s still that 20 percent who aren’t committed. They might be going to one of the other schools I just mentioned. Yield rates drop off pretty quickly after that. In our study, all of the schools we included had relatively high yield rates. If we were to just look at enrollment for a specific school, it wouldn’t be telling us the full picture about admissions because there are students who didn’t enroll who were admitted. So that’s why it’s sort of challenging to look at an individual school.

Instead of thinking about admission to a specific school, we decided to think about admission to a specific group of schools. And if you think about this group of schools, we can make this assumption that if you’re admitted to at least one of these schools, we’re going to assume that you are going to matriculate at one of these schools. And if you do that, then a record of enrollment at one of these schools can be viewed as a record of admission to one of these schools and that anyone who didn’t enroll in one of these schools was not admitted.

Fabiola Cineas

What did you all determine when it came to preferential treatment for children of alumni?

Josh Grossman

We estimated that legacy applicants were, on average, two to three times as likely to be admitted to a selective college than non-legacy applicants with similar academic credentials. And legacy applicants are much more likely to be white.

Fabiola Cineas

Something that sets your study apart is that it breaks down the broad “Asian American” category into smaller groups. What did you discover along those lines and why was it important to analyze the data for these smaller groups?

Josh Grossman

A lot of past studies have really treated Asian Americans as sort of this monolithic group. And it’s not necessarily because they wanted to do that. A lot of the time, it just has to do with data limitations. In this case, we had access to that data. And I think it’s important to recognize that there’s heterogeneity between these groups.

For example, East Asian and Southeast Asian migration to the United States started picking up in the 1970s, whereas it started picking up in the 1990s for South Asian students. Because of that, East Asian and Southeast Asians on average have an extra year for perhaps their parents to attend a US-based university, which can provide social capital and in the case of legacy admission, might even provide an added boost of getting it. Asians make up a large part of the world’s population, so it’s odd that we lump everybody together.

One thing we hope is that this paper will lead education researchers, when they’re able, to start analyzing these groups separately. We found that there’s a huge difference in the estimated likelihood of admission across these groups.

Probably the most salient comparison in the paper when it comes to legacy status is white students are six times more likely to have legacy status than South Asian students, when you’re only looking at these high-scoring applicants. But if you look at white students versus East Asians and Southeast Asians, it’s closer to three to four times as likely. So it’s still more likely but the difference isn’t as stark. That’s one important difference, that South Asians among these three groups are the least likely to have legacy.

Fabiola Cineas

You all also found how legacy admissions impact other groups of students.

Josh Grossman

Yes. We found that high-achieving white students were about twice as likely as high-achieving Black and Hispanic students to have legacy status. Legacy admissions also limit the number of low-income students at these top colleges. We found that students who received an application fee waiver were one-eighth as likely to be legacy students than those who didn’t.

Fabiola Cineas

Something that isn’t as widely discussed that the paper covers is geography, and how it affects an applicant’s admissions chances. What did you learn about the role geography plays when it comes to admissions for Asian American students?

Josh Grossman

Legacy is just one piece of the puzzle, and the geography component is also pretty interesting. I’m certainly not an admissions counselor, but I think it’s pretty open knowledge that admissions offices often work on a regional basis. And while they don’t have a fixed quota for each area, because that would be illegal, there’s roughly the same proportion of students coming from each part of the country each year. For this reason, you often see, at least in our case, that admission rates for states like California are lower because so many more students are applying and maybe they want to make sure — similar to the popular vote with the Electoral College — that there’s sufficient representation from other parts of the country. This preference came out of a nefarious place when they were trying to limit the number of Jewish applicants earlier in the 20th century.

But now, it’s morphed into something that’s acceptable, though I think it’s still nefarious. What ends up happening is if you reduce the admission rate for students from California in this way, and a lot of Asian students live in California — the second highest in our data set concentration of Asian American applicants, among all states, and DC — that’s going to have a disparate impact on Asian American applicants. They’re going to be admitted at lower rates.

Geography explains part of the gap we see. If you just compare applicants from California who are white and Asian, the gap shrinks a bit. Does that mean it is a justifiable shrinking? That’s the ultimate question for universities to decide.

Fabiola Cineas

Which states stand out in the data as being particularly tricky for Asian American applicants?

Josh Grossman

The states that really stand out are New York, with a pretty high admissions rate and a lower proportion of Asian Americans, and on the other side there’s California, with a lower admissions rate and a pretty high proportion of Asian Americans. That’s driving a lot of that pattern. But there are other states like Texas, Washington state, and Georgia that are all clustered near California in the figure we created. There are other northeast regions up there with New York.

Fabiola Cineas

Why don’t we talk about geography in the admissions conversation as much?

Josh Grossman

In the case, lawyers on both sides steered the conversation toward affirmative action. Affirmative action, at least how they were talking about it in the case, didn’t have a lot to do with geography. They weren’t openly taking it into consideration. Because of that, the big question in recent conversations has been affirmative action.

Affirmative action and the alleged Asian penalty, can be really thought of as orthogonal issues. Before affirmative action was eliminated, you could reduce barriers; you could reduce things that benefit white students over Asian American students like legacy, some subset of athletic recruitment, some of this geography stuff, while still retaining a preference for groups that are typically underrepresented in higher education. The plaintiffs in this case sort of attached the two issues and merged them.

Fabiola Cineas

Ultimately, what does your research tell us about how the decision to ban affirmative action impacts Asian American applicants?

Josh Grossman

After you eliminate affirmative action, the differences in admission rates between similarly qualified Asian American students and white students aren’t going to go away. The raw numbers of Asian Americans and white students will probably go up, depending on how things sort of shake down, but those differences in admission rates are still going to be there.

Sign up for the newsletter Today, Explained

Understand the world with a daily explainer plus the most compelling stories of the day.