Jeannie Suk Gersen is the John H. Watson, Jr. Professor of Law at Harvard Law School, teaching constitutional law, criminal law and procedure, family law, and the law of art, fashion, and the performing arts. In the free New Yorker article below (click to read), Gersen attended the one-day Supreme Court hearing in which Students for Fair Admission (SFFA) challenged Harvard and the University of North Carolina’s race-based admission practices, which, SFFA argued, discriminated against Asian-American applicants. Gersen’s been writing for the New Yorker for some time, and attended the hearings as a representative of the press.
As you may recall, the way Harvard kept down the number of Asian-American students (thereby giving minorities and whites preferential admission) was by application-readers in the admissions department giving Asians low “personality scores”, which reduced their chance of admission under the “holisitic” system. (Asian Americans also had lower athletic scores manifested as “extracurricular activities, but that wasn’t a bone of contention.)
Curiously, alums who (as part of the admissions process) actually interviewed Asian-American applicants did NOT give them lower “personality scores”, implying that there was some manipulation of scores by the admissions office itself. I believe this is the case, but two lower courts ruled that this did not constitute egregious race-based admission because any discrimination—and yes, the statistics show there was some—was based on “implicit biases. . . that could not be eliminated in a process that must rely on judgments about individuals”. In other words, evidence of anti-Asian American bias was evidence of race-based discrimination, but not evidence of intentional discrimination. (Didn’t the administrators receive “implicit bias” training?)
SFFA appealed all the way up to the Supreme Court, and it’s almost certain that the Court will rule for the plaintiffs, overturning the 1978 Bakke decision and affirmative action itself.
What’s interesting about Gersen’s report are two things: the lack of preparation for the defense on the “personality scores” issue, for which they had no good explanation, and Gersen’s take about how schools like Harvard will circumvent the Court’s likely decision to keep the minority enrollment from decreasing. Gersen, by the way, says she’s a supporter of affirmative action.
The cluelessness of the defense. Indented bits are quotes from Gersen’s piece. Bolding is mine
The strongest aspect of the discrimination claim against Harvard involves something called the personal rating. As early as 1969, the Crimson reported that the personal rating, assigned by admissions officers based on interviews, high-school officials’ reports, and essays, “has become by far the most important factor in Harvard’s admissions process,” because the increased academic strength of the applicant pool was making it harder to select students based on grades and test scores. It reported that, for the class of 1968, “there is just about no correlation between admission to Harvard and such factors as SAT scores, rank-in-class, and predicted rank list,” but “the correlation between admissions and the personal factor is better than 90 per cent.” The article quoted the dean of admissions saying, “We are justified and obligated to trust a hunch.”
When the Harvard case first went to trial, in 2018, S.F.F.A. alleged that Harvard uses the personal rating, in which admissions officers score applicants on qualities such as “integrity, helpfulness, courage, kindness,” even “effervescence,” to discriminate against Asian American applicants. Admissions records showed that, despite alumni interviewers, who met with applicants, having given Asian students scores that were as high or higher than those of white students, admissions officers, who normally did not meet with applicants, gave Asians the lowest personal ratings of any racial group. The trial, which I attended, focussed on these disconcerting questions: Did Asian students, who had higher academic and extracurricular ratings than white applicants, actually have worse personalities than all others? Or was the personal rating concealing an impermissible racial quota?
At Monday’s arguments, Justice Samuel Alito grilled Harvard about Asians’ low personal ratings. “It has to be one of two things. It has to be that they really do lack integrity, courage, kindness, and empathy to the same degree as students of other races, or there has to be something wrong with this personal score,” he said. “Why are they given a lower score than any other group?” The question was one that Harvard’s lawyer must have been preparing to answer for at least four years. And yet the seasoned Supreme Court advocate Seth Waxman, a former U.S. Solicitor General, seemed cornered and stuck. During several uncomfortable minutes, he at first tried to deflect the question; then, somewhere in the midst of multiple attempts by Alito to get him to answer, an assist from the Chief Justice, and Waxman’s telling assurance, “I’m not trying to filibuster you,” he managed to say that the personal ratings reflect “what teachers said, what guidance counselors said, what these students wrote” in essays. (Those inputs would have to be quite poor to offset the alumni interviewers’ high scores.) In other words, Asian applicants deserved the low personal ratings—or, perhaps, if there was any discrimination, it was by high-school officials, not Harvard.
Surely Waxman, who in general argued well, could have prepared better for Alito’s excellent question. But how could he have? The scores were significantly lower than those for other groups, but were lower only from Harvard officials who never met the applicants, not from those who personally interviewed the applicants. Ergo, blame it on the letters of recommendation. But that’s not a credible answer, nor is the claim that one’s “personality score” could be at all discerned from essays. All the evidence is that Harvard discriminated, supposedly out of “implicit bias.”
How colleges will circumvent the likely ruling.
But the “implicit bias” trope may, says Gersen, be the way that American universities get around the upcoming ruling. While they can’t use racial classification as a criterion for entry, they could use indications of it—not from “personality scores,” but aspects of a more “holistic admissions analysis” that would show racial minorities had compensating virtues, like overcoming difficulties.
It is conceivable that the Court could hold that the district court erred in finding that Harvard did not discriminate against Asians in assigning personal ratings, but such a ruling would not necessarily overrule cases allowing affirmative action; rather, it would mean that Harvard defied the Court’s precedents. It’s more likely that the Court will use this case to end or severely limit affirmative action, without disturbing the district court’s factual conclusion that Asians didn’t suffer intentional discrimination here. Such a decision would not make personal ratings go away, given that Harvard says they are supposed to be assigned without considering race. If the Court prohibits the use of race, so that race-neutral methods become the only permissible means to achieve diversity, schools will likely play with formulas to produce a diverse class in which Asian admissions don’t get unacceptably out of proportion.
It’s odd that Gersen, an Asian American herself, uses the words “unacceptably out of proportion”, which is invidious and implies a quota for Asians. What proportion of Asians, for example, is the upper bound on “acceptable”?
But what about test scores? Asian-Americans score higher than any other group. Here, from a 2018 story in the Harvard Crimson, is an 18-year series of SAT scores for admitted Harvard students. Asians are way up on top, followed by whites, and with non-Asian minorities together at the lower rank.
A Crimson analysis of the previously confidential dataset — which spans admissions cycles starting with the Class of 2000 and ends with the cycle for the Class of 2017 — revealed that Asian-Americans admitted to Harvard earned an average SAT score of 767 across all sections. Every section of the SAT has a maximum score of 800.
By comparison, white admits earned an average score of 745 across all sections, Hispanic-American admits earned an average of 718, Native-American and Native-Hawaiian admits an average of 712, and African-American admits an average of 704.
And here are Harvard’s own data on the ethnicity of students admitted in the class of 2026. Without the “personality score” adjustment, the proportion of Asian Americans would be substantially higher and that of everyone else lower (see below):
But the way around this is simply to devalue test scores (and grades), and raise the value of less tangible scores in a way that would benefit minorities. It’s a way to discern “other” characters that aren’t given a metric, like the notorious personality scores.
Gerson explains:
[Harvard] may reduce reliance on race-neutral factors in which Asians have done well, such as standardized tests, and increase reliance on race-neutral factors in which Asians have not done as well. The personal rating and similar mushy factors could become far more determinative, because they are places where admissions officers will continue to have great discretion to bump applicants up or down based on subjective assessments so long as they are not consciously using race. Prohibiting the explicit reliance on race may even push universities to fall back on the cover of implicit bias, which is not unlawful discrimination. This could leave Asian applicants worse off than they are now. If anything, the personal-ratings morass may suggest that what’s needed to check unconscious biases are more transparently and forthrightly race-conscious efforts, not less.
Even the SFFA suggested how this might work:
In the U.N.C. case, Justice Ketanji Brown Jackson worried that if “a university can take into account and value all of the other background and personal characteristics of other applicants, but they can’t value race,” that policy could actively disadvantage minorities who wish to convey the importance of race in their lives. But S.F.F.A. suggested that eliminating affirmative action does not mean disallowing applicants from writing about their racial backgrounds, or blinding admissions officers from knowing the race of an applicant. S.F.F.A.’s lawyer against U.N.C., Patrick Strawbridge, said that, though admissions officers could not credit an applicant’s race, they could credit an applicant’s cultural experience as an African immigrant. Kagan observed, “The race is part of the culture and the culture is part of the race, isn’t it? I mean, that’s slicing the baloney awfully thin.” Cameron Norris, S.F.F.A.’s lawyer against Harvard, said, “Culture, tradition, heritage are all not off-limits for students to talk about and for universities to consider.” He continued, “They can’t read that and say, ‘Oh, this person is Hispanic or Black or Asian, and, therefore, I’m going to credit that.’ They need to credit something unique and individual in what they actually wrote, not race itself.”
It seems that what S.F.F.A. is insisting on is a formal conceptual distinction—between crediting “race itself” and crediting individuals’ stories about their racial backgrounds—that makes little practical difference. If the Court issues a ruling that tracks with this idea, then, after affirmative action is gone, schools will not give any applicant a plus for “race itself,” but they will still consider race in the context of an applicant’s story.
. . . Justice Sonia Sotomayor made the point most plainly, saying that relying on race-neutral alternatives, including socioeconomic status, are really “all subterfuges to reaching some sort of diversity in race.” She echoed Justice Ruth Bader Ginsburg’s dissent in Gratz v. Bollinger, the 2003 case in which the Court held that a school could not automatically award extra points to underrepresented racial minorities. Ginsburg anticipated that what universities cannot do “in full candor,” they “may resort to camouflage” to accomplish: “For example, schools may encourage applicants to write of their cultural traditions in the essays they submit, or to indicate whether English is their second language. Seeking to improve their chances for admission, applicants may highlight the minority group associations to which they belong, or the Hispanic surnames of their mothers or grandparents.
As Gersen says, this is a distinction without a difference. And I’m not sure how colleges can evaluate intangible factors without giving them a ranking, but the Supreme Court may, in its ruling, even forestall this possibility by prohibiting “implicit bias” if there’s a statistical way to judge it (i.e., ratings of some sort). If those ratings happen to be correlated with race—as were the personality scores—then again we have unacceptable evidence for race-based admissions.
I’m still struggling with the issue of affirmative action, which I have favored, and Gersen’s essay didn’t help. One thing I know, though, is that the concept of an “unacceptably high” proportion of Asians (or of any race) is revolting. It’s a return to the old quota system, but with the quotas remaining implicit. One thing we know, though is that 43% of the student body being Asian is “unacceptably high”. On the other hand, schools lacking minorities don’t speak well of America. (Some like John McWhorter say that this is okay: those who required affirmative action for admission could simply go to colleges with less strict criteria for admission.)
This is from the Guardian (link above):
The lawsuit claims that, in 2013, Harvard killed an internal report about its admissions policies which acknowledged that it discriminates against prospective Asian American students.
The report found that Asian Americans would comprise 43% of admissions if only academic qualifications were considered and 26% when extracurricular activities and personal ratings were considered. Yet at the time the research was conducted, Asian Americans made up 19% of the share of admitted students. [JAC: see data above: the figure is now 27.9%.]
I have been favoring socioeconomic factors as things to consider during admissions. This would raise the proportion of minority students, I think. Sotomayor says this is just a “subterfuge”—a “back door” way of boosting minority attendance. But I think you can justify socioeconomic factors as being worthy of consideration on their own: as a way to raise the diversity not of ethnicity, but of social class and wealth. It’s a way of achieving class rather than racial equity, but they’re correlated.
Two things are certain. The court will rule, probably 6-3, for the plaintiffs, thereby killing affirmative action. But also certain is that affirmative action will rise from the dead as colleges figure out ways around the Court’s ruling so to achieve what they consider “acceptable” balance. After all, much of the administration of American colleges is involved in DEI efforts, and a reduction in the number of minorities could cost people their jobs. Given the elaborate DEI structure in many schools, downsizing it will be unacceptable.
“What proportion of Asians, for example, is the upper bound on “acceptable”?”
Can we agree that 100% every year is too high a bound? And that 0% is too low a bound?
Then there’s bound to be some range of acceptability.
We can start with the fact that 4.7 billion out of 8 billion humans are Asian, so 59% could be a target.
No, I can’t agree. Why would 100% be too high a bound? You either believe in meritocracy or you don’t. And if you seek to redefine meritocracy because the current definition admits too many Asian students, I call foul.
If a college has to admit some of the less meritorious in order to fill out the football team or keep alumni donations humming, that’s reality. But it’s a need to get those ringers in, not to keep Asians out.
The argument against 100% asian actually seems much stronger to me than the argument against 0% black — which (crudely stated) is the goal of the ordinary kind of affirmative action.
By not helping a low-performing 10% group, your society may have 10% relatively poor people who feel left-out, and can riot etc. But ceding elite institutions entirely to a visible high-performing minority risks something much darker. The left-out 80% may eventually come to view this as… rule by foreigners, something like colonialism?
We tend to think always of european examples, but this pattern of minority domination is in fact quite common worldwide, and quite volatile. I highly recommend Amy Chua’s book about this:
https://en.wikipedia.org/wiki/World_on_Fire_(book)
I increasingly think that pure meritocracy is in fact one of those “nice idea, wrong species” ideas. At least for humans living in multi-ethnic societies.
Maybe, instead of diversity-sensitivity training, people need meritocracy training.
It’s worth pointing out that there is zero evidence that undergoing “implicit bias training” has any effect on people’s real-world decision making. (Despite the billions of dollars spent on such training.)
But anyhow, that’s irrelevant since Harvard’s discrimination against Asians (and in favour of some other racial groups) is entirely conscious and deliberate.
Yes, I know this; my comment was sarcastic.
The ethnicities in the table in the post total to 59.4%. Is there some double counting or do white students at Harvard represent just 40.6% of all admissions? I assumed that affirmative action sought assist underrepresented minorities – perhaps to match their proportion in the US national demographic, but they seem to be massively overrepresented here. Or have I misunderstood something?
That is sooo last-year. Anti-racism now means that nothing less than over-representation is acceptable. Have you noticed (as just one example) the fraction of black actors in adverts compared to the national demographic?
Indeed – and here in the UK, were the demographics are very different. No wonder people in opinion polls dramatically overestimate minority populations.
A couple of comments: First, these cases are coming from highly selective institutions (for out of state students, UNC rivals Harvard in terms of the competitiveness of admission), but the impact of the Court’s decision will apply across the board, even to the less selective institutions that the vast majority of college students attend. Second, let’s face it – college admissions, especially at the elite level – are subjective and or discriminatory in lots of ways (favoring, among others, legacies, children of large donors and athletes). So why not include race/ethnicity in the mix? In my idea world, all of these preferences would go away, but realistically it ain’t gonna happen.
So how about the following for the elites – use a set of defensible standards (e. g. grades, test scores, class rank) to identify the top 20% of applicants and then determine admits by lottery. And place the onus on the university to recruit a diverse applicant pool that will be judged on uniform standards. Might require some tweaking down the road, but perhaps it would level the playing field in a nondiscriminatory fashion.
Bruce, you write “these cases are coming from highly selective institutions … but the impact of the Court’s decision will apply across the board, even to the less selective institutions that the vast majority of college students attend.
I don’t know. Susan Dynarski, in 2018 she was a professor of education, public policy and economics at the University of Michigan (since July 2021 she is professor of education at the Harvard Graduate School of Education). She wrote then in an article about affirmative action in the New York Times:
“it’s important to remember that affirmative action is not an issue that directly affects most college students, because the majority attend schools that are not at all selective”
Ungated article: https://archive.ph/4e3tG
According to the data above, are not people of pallor significantly under-represented, and should this not be considered “problematic” by the tenets implicit within “woke” dogma?
I don’t like to repeat this again and again, but affirmative action based solely on socioeconomic factors would be devastating for Blacks. Poor Whites (and Asians even more so) still outscore wealthy Blacks on the SAT. There is not much overlap. And this is the reason why universities chose not to use parental income as a proxy for race, as was was indeed mentioned during the hearings.
Isn’t comparing poor whites to rich blacks based solely on economic factors and no socio factors? Presumably there are reasons for a large difference in scores other than race membership? So what are these factors? Will merely admitting such scorers to college offset these factors or will these students continue to undeperform?
“Socio-economic” could also include things like “first person in the family to go to a college/university”, since education and occupation together with income determine SES. But if there was any silver bullet hidden there, it would have long been found. AA was instituted over 50 years ago. Back then, great expectations might have been reasonable, but no longer.
It seems to me too charitable to call anti-Asian bias “implicit bias.” It appears quite explicit, much as Jews were kept out of the Ivy League during much of the 20th century.
Indeed, removing race from consideration will not change outcomes much, since there remains a long list of correlated attributes that will still be fair game and that admissions departments will use to achieve their aims.
As comment #2 points out, the Harvard claim that “implicit” bias explains the working of a deliberate policy is bullshit. It follows that Harvard’s motto, Veritas is also bullshit. As a matter of fact, we all know that many academic admissions offices have been turning handsprings of deceptiveness on behalf of “affirmative action” for years, the very name of the policy being an example of disingenuous language.
Permit me to suggest that these departures from veritas might have contributed to the other departures in academia that we call “woke”. Perhaps the normality of this kind of bs supplies the example, the template, for the intellectually dishonest verbiage about “disparate impact”, “white empiricism”, the “colonial” bases of Genetics or Math or Chemistry, the “harm” of reporting on facts, the “spectrum” of sexes, etc. etc. etc. etc. .
Jay Caspian Kang, The New Yorker staff writer (formerly New York Times):
https://archive.ph/dEJTx
Not to mention that practically no one at Harvard is stupid or dramatically unecucated – there may be plenty of foolish people, but by design, pretty much everyone there should have an IQ well beyond 100, and no one dropped out of high school. Whereas leaders have to take into account that a massive number of citizens cannot read and write properly, are basically innumerate, and cannot find China on a map. So there are natural limits to the concept of “representation”.
Bit off topic, but, speaking of Justice Alito and SCOTUS, I see that the Rev. Rob Schenck — a former far-right-wing evangelical anti-abortion activist — has claimed that Alito divulged the outcome of the controversial 5-4 decision in the 2014 Hobby Lobby case weeks before the formal opinion was released. There seems to have developed an all-too-chummy relationship between the Court’s most conservative members and certain right-wing machers by virtue of the latter’s copious contributions the Supreme Court’s Historical Society.
In this regard, there’s still been no announcement regarding the results of Chief Justice Roberts’s investigation into who leaked Alito’s draft opinion in the Dobbs case. Last I heard, the circumstantial evidence pointed in the direction to the conservative wing — done to pin down Brett Kavanaugh as the fifth vote needed to provide a majority for Alito’s opinion, for fear that Kavanaugh would succumb to John Roberts’s blandishments that he tergiversate to join Roberts in charting a middle course by upholding Mississippi’s 15-week abortion ban without overruling Roe v. Wade outright.
Clarence Thomas is clearly the Court’s craziest justice, but Alito is as responsible as anyone for SCOTUS’s declining legitimacy in the eyes of the public. There is something essentially humorless and unlikeable about the guy. It came out during Alito’s contentious 2006 confirmation hearings, and it’s stuck to him like a second skin since. He’s basically Nino Scalia minus the wit and charisma.
Interesting article; persistent challenges to creating systems that are both fair and inclusive. Perhaps these are fundamentally incompatible values…
Long ago (mid 1970s) as a junior officer in the USAF, I lodged a complaint against HQ AF for their policy of including photographs in promotion folders. In my opinion, the inclusion of a photograph gave the promotion board members an opportunity to exercise their personal prejudices about a whole range of characteristics that were not supposed to be factors in determining an officer’s suitability for promotion. These factors included race, personal attractiveness, weight, hair style and (most importantly) the presence of facial hair (i.e., mustaches or sideburns). My grievance was declined, and the IG explained that promotion board members had almost unanimously agreed that having photographs enhanced their confidence in the selections they made. This is not surprising: the opportunity to exercise prejudices (especially implicit ones) is deeply satisfying and reassuring.
This practice continued well into the 21st century, but recently a former student sent me a brief article indicating that the Department of Defense had removed photographs from promotion folders.
Ultimately, race itself, is a social construct based on appearance and as long as photographs are included in admission (or promotion) folders, individual raters are likely to be influenced by their own perceptions as well as their prejudices. As long as applicants meet minimal criteria that indicate they can benefit from and contribute to higher learning, the idea of a blind lottery has many advantages for creating a system that is both fair and inclusive.
I don’t want to underplay the value of a university education, both socially and culturally (if not politically). But to believe that these four years of late adolescence are going to be major factors in exposure to “diversity” and its benefits is a bit naive. The major contributor to an individual’s socialization and education, aside from the family, will be what happens AFTER graduation, when Real Life begins, not the protected period cushioning students from the shock of leaving home and living with other people. The same goes for any level of education. Early childhood education is important for the basic equipment of living in society, but the values of the family are really the ones that will form the child’s mind and behavior. The same is true at any stage of education. The university is not going to heal pre existing problems or prejudices. Most important is this overlooked fact: race is relatively trivial (aside from family or religious traditions) in the broad scheme of things because, putting aside foreign students, all the students are Americans, share its culture, participate in its politics and institutions, adopt (or sometimes challenge) the same values and manifestations of THE SAME CULTURE.
The fact that some students have foreign born parents is irrelevant and no more
important than the fact that older people like me also had foreign born parents but grew up as Americans, not half-Hungarian or with Asian grandparents. Every student today is part of the same country and culture and society. Race is irrelevant to their
achievement or place in society. How can anyone dismiss EIGHTEEN YEARS of living and schooling in the US and still think of someone of a different race as NOT American? If some students are disadvantaged is has nothing to do with their race.
It has to do with their family and upbringing and early life. If we decide poverty imposed on American blacks, or intellectual shortcomings needs correction, it has nothing to do with race nor does it require affirmative action. It requires improvement in the conditions under which the student grew up, economic and family and values.
This reminds me of claims that a person’s “race” is not a biological reality but just a social construct. However, studies referenced on this site show that genetic analysis can quite accurately predict one’s self-identified race / ethnicity / ancestry.
Since one’s self-identified race also correlates with various measures of past academic success and predicted future success, there must be something real here also. It’s clearly not “THE SAME CULTURE”. Today there are several obvious painful divisions in the “same country” of America; race certainly appears to be one of these divisions.
Nothing?? The disadvantages of being disrespected, feared, hated, and even killed on account of one’s apparent race are surely something.
And when Chris Rock confronts a white audience with the fact that not one of them would choose to exchange their life for his, even though he is rich and famous, there is something inescapably real going on.
“And when Chris Rock confronts a white audience with the fact that not one of them would choose to exchange their life for his, even though he is rich and famous, there is something inescapably real going on.”
I think that a person with good mental health wouldn’t choose to exchange his life for anyone else’s. Even in extremely misogynist cultures, very few women would want to exchange their lives for men’s. But to that particular question – quite a few people, e.g. Rachel Dolezal, would agree to the exchange.
All selection involves discrimination. Rather than Gov banning various flavors of discrimination and action, let the river run. Prohibit Government from banning discrimination. Let a university (or a factory, or a church, or a golf country club) select by whatever criteria. Let consumers judge based on that. Do you really want to engage a doctor with a degree from a medical school that selects on any criteria other than excellence in skill?
Given that subterfuge takes more work than honest affirmative action, the workload (and thus workforce) of DEI will probably increase.
>From the New Yorker article:
>>In the U.N.C. case, Justice Ketanji Brown Jackson worried that if “a university can take into account and value all of the other background and personal characteristics of other applicants, but they can’t value race,” that policy could actively disadvantage minorities who wish to convey the importance of race in their lives. — emphasis added to reinforce that the direct quote ends where it does.
Did a Justice of the Supreme Court of the United States really say what the New Yorker attributes to her in the indirect quote that follows?
Richard Sander had interesting things to say about this decision on this podcast:
https://www.manifold1.com/s2/23
He sees 3 possible outcomes. Besides the extremes of “Harvard wins” and “all racial preferences are unconstitutional”, the 3rd option is to declare that Harvard’s preferences broke civil rights law but not the constitution.
That (if I understood right) kicks the ball back to congress, who by ordinary process can change the law about exactly what kind of racial preferences are allowed in future. Which sounds rather like the Dobbs case. Maybe the most interesting outcome.