This offer apparently pits one minority group against other minority groups. But the SFFA has strategically combined two conceptually and legally distinct issues: a so-called Asian sanction and broader affirmative action. If Harvard does discriminate against Asian Americans, that wrong could in theory be righted without rejecting affirmative action.
The case is full of legal and statistical machinations that we and our students have faced in our Classes at Harvard and Stanford University, where we teach. It also provides an excellent example of two distinct theories of discrimination – disparate treatment and disparate impact – that underpin debates in education, criminal justice and beyond.
SFFA vs. Harvard by the numbers
Numbers are largely undisputed. Among applicants for Harvard’s Class of 2014 to Class of 2019, Asian American students had, on average, better academic and extracurricular records than white applicants. But while 80 out of 1,000 white applicants were admitted, only 59 out of 1,000 Asian American applicants were admitted, down 26 percent.
Part of that gap stems from higher admission rates for athletes and children of alumni — the so-called legacy candidates — who are disproportionately white. The rest is explained by Harvard’s consideration of factors such as geography, parental occupation and, more controversially, subjective ratings of “personal qualities” such as integrity and kindness, on which staff from Harvard rated Asian Americans lower than whites on average.
Among the most academically prepared applicants, Harvard gave 31% of white students a higher “personal” score, but only gave equally high marks to 23 percent of Asian American students. Harvard’s assertion that personal ratings are an unbiased measure of character requires accepting the idea that Asian American applicants have, on average, lower “personal qualities” than white applicants – which evokes corrosive stereotypes often applied to racial and ethnic minorities.
Disparate treatment claims
According to the SFFA’s claims, Harvard’s use of personal ratings would only be illegal if Harvard implemented them with a intention to admit fewer Asian Americans, a possibility the university vehemently denies. This type of discrimination — which jurists call “disparate treatment” — is what many people probably think of when they hear the word “prejudice,” with decisions designed to harm individuals based on their race, gender, or other legally protected traits.
In orchestra auditions, candidates often play behind a screen to limit disparate treatment in assessments. In our own work, to reduce the disparate treatment in prosecutors’ charging decisions, we have constructed a “blind load” algorithm which automatically hides race-related information from police reports. The tool is now used by district attorneys across California. In other research, we found that black drivers represent a smaller share of drivers stopped after sunset, when a “veil of darknessobscures the race of the drivers, suggesting patchy treatment in officer decisions.
Disparate impact: another form of discrimination
Disparate treatment, while important, is just one form of bias. Another legal approach examines what is called “disparate impact.” This type of discrimination occurs if a policy imposes disproportionate burdens on marginalized groups without valid justification. Unlike disparate treatment, disparate impact does not necessarily imply discriminatory intent. US law prohibits policies that have a discriminatory impact in certain important contexts, including employment, loans and housing. But even in areas not covered by disparate impact laws, this theory of discrimination offers a useful perspective for identifying policies that harm marginalized groups.
For example, in Nashville, the police department regularly stopped drivers for non-moving violations, such as broken taillights and tinted windows. We found that this policy disproportionately affected black residents, who were nearly 70% more likely to be arrested for no-move violations than white residents. Nashville police touted the practice as a crime-fighting measure, but our analysis showed it had little impact on burglaries, robberies and other serious crimes.
We have concluded that the department’s actions have imposed a disparate discriminatory impact on black residents — disproportionately penalizing them without any clear crime-fighting benefit — whether or not individual officers intended to discriminate. These results, along with a campaign supported by community groups, led Nashville to curb traffic stops nearly 80%, with little change in crime rates.
Disparate impact on college admissions
The Harvard case raises similar concerns about disparate impact. If the university’s stated preferences for legacy applicants and exceptional “character” do not further legitimate educational goals, the burden these criteria place on Asian American students could be found to be discriminatory, even if they do not. were not designed to exclude Asian Americans.
Claims of disparate impact in higher education are rare, and the SFFA lawsuit focuses exclusively on disparate treatment. But that can change. Under the Obama administration, the education and justice departments seemed willing to investigate claims of disparate impact in at least some educational settings. These efforts stalled under the Trump administration, but the Biden administration seems interested again.
As such practices risk legal action, universities that want fair admissions policies might consider expanding their definitions of discrimination to include disparate impact.
The future of affirmative action
The case raises legitimate questions about whether Asian Americans face barriers when applying to Harvard — and by extension, other universities with similar practices. If Harvard imposes an “Asian penalty” — intentionally or inadvertently — that could be rectified without dismantling affirmative action. For example, Harvard could restructure the way it uses personal scores or eliminate them altogether.
Corn SFFA and Harvard have repeatedly made the case an affirmative action referendum. The SFFA has urged conservative Supreme Court justices to end a policy it disapproves of. Harvard seems to be betting that the court wouldn’t reshape educational policy so drastically. The result could affect students for generations.
Sharad Goel (@5harad) is a professor of public policy at Harvard Kennedy School.
Julien Nyarko (@JulianNyarko) is an assistant professor of law at Stanford Law School.