Harvard tries to reduce the number of Asian-American students by systematically downgrading the ratings of their personalities on applications

June 25, 2018 • 11:32 am

The U.S. Supreme court has ruled, and sustained, the use of race as an admissions criterion for colleges and universities as a tool for increasing ethnic diversity.  The stipulation, though, is that there cannot be racial quotas, and that there cannot be policies that “consciously aim at racial balancing.” This summary comes from today’s New York Times op-ed below (click on screenshot), and it confuses me.  There’s also a related article in last August’s New Yorker by Jeannie Suk Gersen, an Asian-American professor at Harvard’s Law School (click on screenshot). Both take up the issue of Harvard’s historical discrimination against Asian-American applicants, which of course is related to racial balancing.


I am in favor of affirmative action to rectify the historical discrimination against underrepresented and oppressed minorities, though the ultimate goal should be to eliminate affirmative action, accepting people solely on the basis of their achievements, interests, and other things that make for a good student body.  That, however, would require everyone to be given equal educational opportunity from the outset, so that nobody is disadvantaged by historical circumstances or poor life situations. And we’re ages away from that in America. And even then I’m not 100% sure that race or gender should be ignored, for even under equal opportunity there may be advantages of diversity that would outweigh a purely meritocratic approach.

What I don’t understand about the New York Times characterization of the law is that it seems to forbid “conscious aims at racial balancing”—yet that is exactly what colleges are doing. To try to keep student bodies diverse, Hispanic and African-American applicants are admitted with lower grade-point averages and test scores than are whites and Asians, with Asians facing the highest bar since they excel in qualifications on paper. Yang’s article gives the data:

The Asian-American population has more than doubled over the last 20 years, yet the Asian-American share in the student populations at Harvard has remained frozen. Harvard has maintained since the 1980s, when claims of anti-Asian discrimination in Ivy League admissions first surfaced, that there is no racial bias against Asian-Americans once you control the preferences offered to athletes and alumni.

The discovery process in this case has demonstrated that this claim is no longer supportable.

Mr. Arcidiacono found that an otherwise identical applicant bearing an Asian-American male identity with a 25 percent chance of admission would have a 32 percent chance of admission if he were white, a 77 percent chance of admission if he were Hispanic, and a 95 percent chance of admission if he were black. A report from Harvard’s own Office of Institutional Research found that even after alumni and athletic preferences were factored in, Asians would be accepted at a rate of 26 percent, versus the 19 percent at which they were actually accepted. That report, commissioned back in 2013, was summarily filed away, with no further investigation or action taken.

Were Asian-Americans admitted on the basis of academic achievement and extracurricular activities alone, regardless of ethnicity, Harvard would have a near-majority Asian student body (43%), something that Gersen says is undesirable. Rather, she says, “we should not want the composition of our elite universities to be wildly out of proportion to the racial composition of our country.” Readers can weigh in on that sentiment below, but in general I agree with it—diversity is a good thing. But not just diversity of race, but also diversity of thought, social class, ideology, and personality and interests.

But to keep the proportion of Asians down, since they are high achievers and also consciously engage in those crucial extracurricular activities that make one look “well rounded,” Harvard has engaged in an odious charade—one that they try to keep secret.

It can no longer be kept secret, though, since a lawsuit was filed against Harvard claiming that it discriminates against Americans of Asian descent. Harvard fought hard to avoid giving information on its admissions policies, but it finally had to as part of the discovery process. And what was revealed was a systematic downgrading of Asian’s personality scores compared to whites, blacks, and Hispanics. This of course plays into the stereotype of Asians as robotic grade-grinders who don’t have distinctive personalities. Yang says this:

. . . the Harvard admissions office consistently gave Asian-American applicants low personality ratings — the lowest assigned collectively to any racial group. She did not know that Harvard’s own Office of Institutional Research had found that if the university selected its students on academic criteria alone, the Asian share of the Harvard student body would leap from 19 percent to 43 percent. She did not know that though Asians were consistently the highest academically performing group among Harvard applicants, they earned admission at a rate lower than any other racial group between 2000 and 2019.

All she knew was what she had witnessed as an assistant principal and the single fact that she was shown by her deposers. But perhaps she intuited the rest.

Earlier this month, we learned that a review of more than 160,000 individual student files contained in six years of Harvard’s admissions data found that Asians outperformed all other racial groups on every measure of academic achievement: grades, SAT scores and the most AP exams passed. They had more extracurricular activities than their white counterparts. They were rated by interviewers who had met them as virtually on par with their white counterparts in their personal qualities. Yet Harvard admissions officers, many of whom had never met these applicants, scored them collectively as the worst of all groups in the one area — personality — that was subjective enough to be readily manipulable toThe report by the plaintiff’s expert witness, the Duke University economist Peter Arcidiacono, revealed that Harvard evaluated applicants on the extent to which they possessed the following traits: likability, helpfulness, courage, kindness, positive personality, people like to be around them, the person is widely respected. Asian-Americans, who had the highest scores in both the academic and extracurricular ratings, lagged far behind all other racial groups in the degree to which they received high ratings on the personality score.

“Asian-American applicants receive a 2 or better on the personal score more than 20% of the time only in the top academic index decile. By contrast, white applicants receive a 2 or better on the personal score more than 20% of the time in the top six deciles,” wrote Mr. Arcidiacono. “Hispanics receive such personal scores more than 20% of the time in the top seven deciles, and African Americans receive such scores more than 20% of the time in the top eight deciles.”

This sounds like a conscious decision by Harvard as a way to keep the number of Asian-American students down. Note that some of the students didn’t even get to show their personalities, via a personal interview, with admissions officers. It’s no surprise, then, that Harvard fought tooth and nail to avoid disclosing it. In fact, as Yang notes, the exact same tactic was used in the past to keep the number of Jewish students down:

Harvard has been here before. “To prevent a dangerous increase in the proportion of Jews, I know at present only one way, which is at the same time straightforward and effective,” wrote A. Lawrence Lowell, Harvard’s president in the 1920s, “and that is a selection by a personal estimate of character on the part of the Admission authorities, based upon the probable value to the candidate, to the College and to the community of his admissions.” The opacity of its admissions procedure could veil what Lowell’s written correspondence would later disclose to be a fully intended policy of discrimination.

I’ll add one other thing: Harvard gives preferential admission to “legacies” (relatives of those who attended previously, and of donors) and to rich people, preferences that can outweigh test scores. This is done so Harvard can keep building up its multibillion-dollar endowment. I object to that vehemently, but that’s the way they roll.

This raises a number of questions—not just about Asians, but about racial balancing in general. As I said, I am in favor of affirmative action, but we should be more honest about it: not using “personality denigration” as a way to effect it.  But are quotas the answer? I don’t know many people who are comfortable with explicit quotas.

So here are the questions, and readers can weigh in:

1.)  Do you favor affirmative action by ethnicity so that college student bodies mirror to some extent the composition of the country as a whole? Or do you think admission should be genetics-blind, based on criteria like scores, grades, and other achievements and activities?

2.) If you favor affirmative action to increase diversity of gender or ethnicity, do you favor increasing diversity of other traits, like social class, ideology, politics, and so on?

3.) If you favor affirmative action, should the goals be in the form of quotas?

4.)  Do you agree with Dr. Gersen that we should strive to keep Asian students from dominating the composition of the student body?

5.) If your answer to #4 is “yes”, and Asians score the highest in all criteria save personality, is it ethical to downgrade Asians on personality scores to keep their numbers down? (I myself can’t see this as ethical at all, nor do I believe that Asians are uniform, robotic, and without distinctive personalities. That simply  hasn’t been my experience with Asian students here.) Do we then practice anti-affirmative action with Asians?

There are other questions as well, but I’ll stop here. All I know is that the situation is a mess, and difficult to tackle. The only way it would be easy is if you believe in a pure meritocracy—in fact, one in which ethnicity is not even specified in the application. Once you start bringing in diversity as a desirable situation, then you run into trouble, especially in view of the Supreme Court’s ambiguous stand. Those of us who think diversity of multiple traits is a good thing to have in college—after all, what good is a college where you don’t confront different people with different views?—will have to face up to this mess.

97 thoughts on “Harvard tries to reduce the number of Asian-American students by systematically downgrading the ratings of their personalities on applications

  1. The present rule seems to be Harvard can’t discriminate unless they do it in a way approved by the court. Makes sense to me. Not good sense of course. More like say one thing and do another.
    I say eliminate athletic and alumni preferences. That would be a start. Then go from there.

  2. I favor a leg up for students from disadvantaged backgrounds. For everyone else, a meritocracy.

    Personality seems a dumb criterion. If my alma mater used that on me, I’d still be working at McDonalds.

    1. I favor a leg up for students from disadvantaged backgrounds. For everyone else, a meritocracy.

      Exactly! Glad but not surprised I didn’t have to go any further than the second comment here to find an expression of my take on the matter.

      Were Asian-Americans admitted on the basis of academic achievement and extracurricular activities alone, regardless of ethnicity, Harvard would have a near-majority Asian student body (43%), something that Gersen says is undesirable.

      Whereas all those decades of white, WASP, moneyed, male majority classes were perfectly fine. What the hell does it say about the US when Harvard is channeling Pat Buchanan?

      Let meritocracy rule for the non-disadvantaged, and if traditionally privileged demographics get their noses out of joint, well, then perhaps they could up their game.

  3. In answer to questions – I am NOT in favor of quotas, as that reduces a complex and nuanced problem with many possible solutions to a very stupid one.

    Since Affirmative Action is by definition discriminatory, despite their academic qualifications some people are going to be excluded from the school of their choice. The world is not fair and neither should we expect university admissions. Universities should be transparent in their admissions process and if they are excluding people because they are over-represented they should just say so. Though there must be in place laws that prohibit the kind of active, hateful discrimination we have been afflicted with for so long, no one ought to feel entitled to a spot in any school.

  4. In general I feel that the argument for diversity has been put the wrong way around. While there are benefits from having a diversity of perspectives, the issue is really with exclusion, not with inclusion. As we’ve seen, the way diversity is managed doesn’t bring the institional benefits ostensibly desired, and can, as at Harvard, lead to exclusion.

    1. There is simply no way around being exclusionary. The only question a University need ask is; “who do we exclude”? Put another way – “who’s ox can we afford to gore”?

      1. True, but I think there’s a difference between merit-based selection and a conscious policy not to admit Jews or blacks.

  5. 1.) Do you favor affirmative action by ethnicity so that college student bodies mirror to some extent the composition of the country as a whole? Or do you think admission should be genetics-blind, based on criteria like scores, grades, and other achievements and activities?

    Well, neither actually. There is no easy way of achieving the former and if you implement the latter, the deck gets stacked against people from disadvantaged backgrounds.

    At this stage, there is no way you can make the composition of an incoming class even remotely similar to the composition of the country as a whole. However what can be done is to try and include as many people from disadvantaged backgrounds as possible, to spend the extra time and money needed to bring these students up to speed (certainly possible with Harvard’s endowment).

    2.) If you favor affirmative action to increase diversity of gender or ethnicity, do you favor increasing diversity of other traits, like social class, ideology, politics, and so on?

    In an ideal world, affirmative action would be based primarily on economic background with ethnicity added as a secondary factor. There is no reason why the kids of a multi-millionaire executive, sports star or entertainer should get preferential treatment because they come from a disadvantaged ethnic group and why kids from poor backgrounds get the short end of the stick just because they are white or Asian.

    4.) Do you agree with Dr. Gersen that we should strive to keep Asian students from dominating the composition of the student body?

    Full disclosure – I am South Asian with one kid who is in college and another who will be a junior in high school this fall and will be applying to colleges in the next 12-18 months or so. My biases here are obvious.

    That being said, I don’t believe it is any business of a college to keep one or other group from dominating the composition of a student body. If there is any bias, it should be towards making sure that disadvantaged groups get the opportunities they have been denied. When it comes to everyone else, the playing field should be level and the most deserving students should be selected regardless of what the eventual ethnic composition of the student body ends up being.

    – RM

  6. As long as diversity is the goal it would seem to work best by sticking with academics and percentages nation wide. In other words try to obtain percentages that come close to the national percentage. I think some companies do this in their hiring practices and even promotion practices. Nothing you do is perfectly fair.

    We know the overall lower educational systems are not equal and may never be. We also know that the ability to pay the high prices involved in education is not fair. In the end, you either have a diversity goal or you don’t.

  7. I would answer ‘yes’ to the first three questions. And since I can’t think of a better way to do this other than quotas, questions 4 and 5 are moot.

  8. Anti-affirmative action with Asians has been in place much longer than just recently (UC Berkeley and Stanford, come to mind). It is unethical, but I think the universities are justified in trying to maintain a demographic that they feel best defines their school identity.

    Maintaining diversity is the important part in the face of multiple and complex needs. No one ever said fairness was easy. I am pretty sure everyone , at some point in their life has been on the end of getting an advantage or also being denied an advantage because of either their gender, race, religion, diet, etc.

  9. I think it would be a mistake to base admission primarily on test scores and GPA. They are not perfect measures of student potential by any means.
    Colleges and universities should be using other evaluations as well, to form a balanced and interesting incoming class.
    I agree that they should be forthright about their admission process and what they consider when making admission decisions.

  10. 1.) Do you favor affirmative action by ethnicity so that college student bodies mirror to some extent the composition of the country as a whole? Or do you think admission should be genetics-blind, based on criteria like scores, grades, and other achievements and activities?

    There was a report recently that said that 0 of Penn’s incoming freshmen belonged to the bottom quintile in household income. What kind of affirmative action is that? That’s just a bunch of people already eating pie, arguing over who gets more pie. None of those people are hungry enough to have merited a significant adjustment in standards. I’m okay with socio-economic based adjustments, to give the truly needy some preference, but not racially based affirmative action that mostly helps wealthy people with more melanin at the expense of other people with less melanin.

    2.) If you favor affirmative action to increase diversity of gender or ethnicity, do you favor increasing diversity of other traits, like social class, ideology, politics, and so on?

    Just socio-economic.

    3.) If you favor affirmative action, should the goals be in the form of quotas?

    Don’t favor it, see above.

    4.) Do you agree with Dr. Gersen that we should strive to keep Asian students from dominating the composition of the student body?

    I think that’s a very minor point. I don’t think 43% is dominating, and if it got that way everywhere, perhaps we would all learn some very important lessons. But I don’t think dropping affirmative action would make campuses uniracial. Besides, do Chinese and Indian Americans look alike? No they don’t.

    5.) If your answer to #4 is “yes”, and Asians score the highest in all criteria save personality, is it ethical to downgrade Asians on personality scores to keep their numbers down? Do we then practice anti-affirmative action with Asians?

    Of course not. Also, as I recall, the alumni interviewers DID NOT rate Asians lower on personality. It was the Admissions people that did that.

    1. You don’t think 43% is dominating? If the actual percentage for the entire country is 5.6% what would it be?

      1. Just so it’s clear, 5.6% of Americans are Asians. We don’t know the percentage of college applicants who are Asian. IOW, the pool of college applicants is not restricted to American citizens, so the discrepancy is not as stark as the numbers you cited.

        A brief google shows >1 million foreign students applying to US schools last year, of which 37% were Chinese and 17% Indian.

      2. “You don’t think 43% is dominating? If the actual percentage for the entire country is 5.6% what would it be?”

        By that standard, how is 20%+ (the current figure) not already dominating?

        Even at 43%, it wouldn’t be a majority, and the figure would include includes ethnic Chinese, Japanese, Pakistani, Indians, Koreans (Arabs?), etc.

        Should we deny Pakistani Americans a spot unless they get a perfect score on the ACT/SAT? How about Japanese Americans?

        No, I don’t think it even makes sense to combine all those groups and complain about them being 43%.

        Would their representation be out-sized (relative to their % in the US)? Sure. So what? That would still leave more than half the spots occupied by other groups.

  11. I’ll answer 1) – 5) as ‘no’ because until you have worked out what the outcome *should* be (if no unjustified discrimination were permitted) you don’t know what actions (if any) for which potential students are appropriate.

  12. It really is a difficult question, on the whole I’d say no, one should not rate down on biased personality ratings.
    We want diversity in our colleges (not only the USA, here in South Africa the same problem exists), but you don’t want to exclude excellence because of ethnicity either. That is also deeply unfair, it appears to be a lose-lose situation, whatever you do.
    Another angle of positive discrimination is that the outstanding student from an ‘affirmative action’ background, tends to be seen as just an ‘affirmative action’ admission. That is not fair to those either. Triple lose.
    I have no patent solution, but I’d think that admission on merit is the least unfair of all unfair policies, and that manipulating the numbers should be as minimal as possible. If that means half of the students will be of Asian origins, so be it, we’ll have to learn to live with that.
    Is there something to be said for a ‘weighed lottery’? All get a chance, but if you have good scores your chances of being admitted are increased, you get more ‘tickets’ in the lottery. Definitely unfair too, but unfairness appears inevitable…

  13. The admissions process at some medical schools (including ones for which I’ve served on the admissions committee) includes a socioeconomic determination with four categories. In addition to the obvious questions about family income and highest level of education achieved by parents/family, there are considerations of 1) the primary residence (rented? owned? value?), 2) whether the applicant grew up in a rural, inner city, suburban or urban environment, 3) whether the applicant had to work to contribute to family income, 4) whether the applicant had responsibilities to care for younger siblings (and how many siblings are there), 5) whether the family was ever dependent on public assistance/free and reduced cost meals, 6) whether English was the primary language spoken at home etc. Some of these could potentially influence time available for extracurricular activities or access to MCAT prep courses. A white male who grew up on a remote cattle ranch in West Texas might be in the same socioeconomic category as an Asian-American female from an impoverished inner city neighborhood in Houston, for example. The system isn’t perfect, of course, but in my experience it has increased diversity in medical school classes, without requiring quotas.

  14. Considering the high price of Harvard I don’t think anyone should go there unless they get a full scholarship or they can pay The tuition out of pocket. I have seen too many kids come out of private schools with student loans there they could not afford.

    1. It pays to have a YUUUUGE endowment. You know what the average student graduating from Harvard owes? $6.5k.
      You know what the average student who graduates from University of Michigan owes? 26.5k.

      More.. University of Washington; 15k, UCLA; 16.3k, NYU 23.9K

      Source https://collegescorecard.ed.gov/

      1. I have seen people graduate from private schools owing up to $200,000. Schools with smaller endowments. Glad to see Harvard students one out better.

    2. Harvard is a terrible place to get an education (trust me, I know a bit about it), but an amazing place to get a sheepskin. A magic key that opens all doors. To someone who got a crappy college education. And now is in a position of great influence and power. Thank you, Crimson!

  15. I won’t try to address all of your questions, but I will raise one point that I think is relevant. Talking about Asians as a group has little utility except to indicate (in a gross way) the part of the world that a person comes from or whose progenitors came from. Someone from Cambodia may have little in common with someone from North Korea, for example. Also, we don’t refer usually to someone from a Central American or South American country as an American, though they are in the same way someone from Vietnam or North Korea or China is an Asian.

    In college, I had friends who were from or derived from Japan, China, Indonesia, Bolivia, Peru, Chile, and Mexico. They enriched my educational experience and I thought of each one as an individual, not a member of a broad, undifferentiated category.

    I favor having as wide a variety of ethnic and national groups as possible in the education environment. The way to achieve this is to recruit from both of these categories (ethnicity and nationality) and afford those from historically-discrinated groups special consideration if they meet minimal educational criteria for success. If this is done well and we ignore too-broad categories, we should not need to consider quotas, ever.

  16. In a rapid response, I suspect admissions should not be a meritocracy. People should pass a qualifying bar and then be randomly picked.

    The reason is that there was a paper, which I admittedly have not read and do not know quality of and how it generalizes, that suggests that research grants should be similarly allotted for best effect. If so it seems only results – of education and research – should be competed, not the resources of people and money.

    Such a lottery could be weighted for gender and ethnicity, but since the demography is rapidly changing in US I dunno why the latter would be desirable.

      1. True, but I think Torbjörn’s suggestion still has merit. Why? *Margin of error* (which is, itself, unknown). One doesn’t *know* who the best candidates are (generally speaking) and so by avoiding an “artificial ranking” one is thereby breaking a form of discrimination.

        I have no idea where to set appropriate “bars”, though.

    1. I like the idea of having a minimum bar and having a randomized selection process for everyone above the bar, given that there is not a perfect correlation between academic performance and subsequent contribution to society in a particular field, but politically I don’t think it’d work. If the minimum bar is high enough to be meaningful then certain ethnic groups would have very low representation due to the large performance differences between groups. There’d be pressure to lower the standards until you’re just picking people randomly.

    1. Iranian-Pakistan border to the Pacific, going By US census definitions. Though afaik other Anglophone countries aren’t that much different with on paper definitions, just that media representations differ.

  17. This is decades’ old in re Asian students.
    Perhaps not Asian – Americans but certainly
    in re Asians in re graduate work within the USA.

    As far back as y1993, I was in charge of providing the DOGE
    ( director of graduate education ) and her
    / his admitting committees
    with the aspiring applicants completed folders.

    For their consideration for admittance the next autumn term in to
    a public university’s economics department.
    Actually, THE premier USA university’s
    .a g r i c u l t u r a l. economics university.

    The DOGE actually told me to separate out
    the Asian – ‘sounding’ applications from
    the rest of them. To put those, the Asian – sounding ones, IN to ITS own pile. Separate from … … The One Other Pile. OF everyone else’s applications.

    Each week. For each week’s meeting in re admittance.

    AT the end of the academic year, then, IF
    there were any spots left to consider, THEN
    is when the Asian – sounding pile was … …
    even looked at. Let alone, considered. Seriously.
    y1993. And further.


    1. How does that work? How do you decide what an ‘Asian sounding’ application is?
      Names? How do you tell if Mr. Lee is Chinese or Northern Irish?
      Do Filipinos, and Goanese Indians get a free pass?
      Sounds very dodgy imo

  18. I am for pure meritocracy. If there is any exception, I’d approve it to be for donors. An university puts up with a suboptimal student in exchange for receiving money to operate. Of course, these suboptimal students could lead to patients dying and bridges collapsing, but given their rich background, this is unlikely. If a suboptimal student from a historically discriminated group is admitted to make up for past injustices, nobody gains anything except (maybe) the suboptimal student.
    Students with little income should be helped by stipends or partially or wholly waiving the tuition fee.
    About the Asian Americans – I think that if the groups currently being in the majority in the USA feel threatened by Asians, they’d better limit the Asian immigration than first let the Asians in and then engage in a conspiracy to keep them and their descendants down.

    1. ” Of course, these suboptimal students could lead to patients dying and bridges collapsing….”

      Or Twin Towers collapsing. Just sayin’.

  19. 1. Ethnicity should be a component, but it shouldn’t dominate, and it should be closer to real ethnicity, rather than US categories. A few years back Dave Cameron (oh how he seems so reasonable now) called out Ox-bridge on their diversity by pointing out under-representation of British born afro-Caribbean background students, to which they responded something along the lines of whatever we have other blacks. Treating Affluent Nigerian students who attended high-fee paying private schools on the same footing as intercity kids with Caribbean backgrounds just doesn’t make any sense to me.

    2. Yes. If anything this is more important in some universities, there should be flexibility, different universities are gonna need different strategies, say Kingston or South Bank might want to try and attract more kids from affluent backgrounds whilst helping some mode disadvantaged but capable London students go to a Russell group Uni. Perhaps some kind of Erasmus for within the UK is in order, frankly I’m surprised something like this doesn’t exist already.

    3. In general, no. I mean, should some London Universities exclude minority/non-white-British students to better reflect the national balance? Have targets where there are obviously discrepancies, but rigid quotas are a bit meh for me.

    4. Don’t see why its necessary. If anything, Harvard is a is a rich New England family Uni no? If having them dominate isn’t a problem, why should it be a problem if someone else ‘dominates’? And what does ‘Asian’ mean in this context, USians? or Overseas students?

    5. Are personality tests appropriate for teenagers? They are still developing socially, an introvert who was an odd ball in his school could open up and become more sociable in the right company. Personality tests to me smack of vague subjective ‘cultural fit’ criteria, which I suspect is at the heart of the issue in Havard’s admissions.

    I attended international schools as a teenager, which, compared to by experiences at British Unis, were subjectively much more diverse (some even promoted themselves on the basis of having almost a 100 nationalities), people mingled more, we learned bits of each other’s languages, and had the maddest feasts. Its a great experience. Was disappointed to discover this was lacking at uni 🙁
    I picked my college on the basis of the number of foreign students in the hope I’d meet more international schooled students who were ‘like me’ (having found Oxford Brookes to be dull), short answer, I did, but in doing so I unknowingly chose one of the worst Unis in the UK. So it goes.

      1. If you’re aiming to boost the presence of people from disadvantaged backgrounds, you can hardly ignore it.
        But like I said, it shouldn’t dominate, you don’t mark a French person up or down just because their French, you take into account where they grew up, what their family situation is, what kind of school they went too. In short, you take a more, ahem, intersectional (see what I did there) approach.

        1. Ok, I hear you.
          In some disciplines it make sense, teachers, doctors, police that works in the communities.
          My problem however is that insisting on demographic representations in human endeavours is totalitation and ignores population differences (cultural and generic)

    1. Nice post, Gareth.

      And I certainly agree that ‘personality tests’ are suspect (I won’t say completely bogus) – they probably favour insurance salesmen over nerds. “If you can fake sincerity, you’ve got it made”.


  20. I am in favour of socio-economic affirmative action. Race is certainly a proxy for SES (on average), but I think there are far better measurements that can and should be used. For instance: household income (at multiple time points), single or double parent households, and highest degree/diploma obtained by each parent. Preferential admission for students from disadvantaged socio-economic backgrounds is a good thing for both the university at large and the ultimate goal of equal opportunity and SES for future generations.

    1. Also, preferential admissions for alumni, big donors, and athletics need to be eliminated. Without this, any attempt at affirmative action feels hypocritical at best.

  21. “I am in favor of affirmative action to rectify the historical discrimination against underrepresented and oppressed minorities”

    I thing it is a short-sighted policy that hurts the people you want to help in the long run.
    I support bursaries for poor students on academic merit.

    If I remember correctly, in Hungary about 85% of graduates were Jewish in the 1920’s and 1930’s. Would you have supported affirmative action for the Hungarians?

    This is of course the problem with “multi-cultural” societies if there are infact average group IQ differences and you want to force equality of outcome.

    Maybe, just maybe Asian and Jewish students outperform other groups on merit, not because of structural inequalities.

    The only sane policy I believe is to help smart students from poor backgrounds irrespective of their ethnicty and then just accept that for either cultural or biological reasons some groups will not be reprecentitve in fields like theoretical physics!

    1. then just accept that for either cultural or biological reasons some groups will not be reprecentitve in fields like theoretical physics!

      I’d rather not ‘just accept’ that, thanks. What biological reasons can you think of that would prevent an specific group of humans to be underrepresented in theoretical physics?

      1. “What biological reasons can you…”
        I have this radical theory that the brain is biological and that it relates to intelligence.

        What is your explanation why for example Ashkenazi Jews are so over-represented in many diciplines that requires high abstract reasoning?

        Surely this will be a combination of cultural/environmental and genetic factors.
        I see no reason why australian aborigines cannot be superior in visual spatial memory and Ashkenazi’s in verbal reasoning. (on average)

        (Unless you postulate that the human brain is exempt from natural selection/evolution)

    2. “in Hungary about 85% of graduates were Jewish in the 1920’s and 1930’s. Would you have supported affirmative action for the Hungarians?”

      Maybe. This is indeed thorny. Having an elite which is seen by the majority population as being nearly 100% outsiders is a dangerous situation.

      If one has a choice (by nonviolent means!) it would be good not to land in 1920-Hungary’s position. While most of the contemporary immigration discussion focuses on low-skill immigration (Gypsies, in our analogy) this is also a point which deserves attention. For lots of examples, I highly recommend Amy Chua, “World on Fire”.

      That’s one avenue. The other is assimilation — if the importance of whatever marker separates the elite can be diminished, that also calms things. Unfortunately affirmative action (and our whole political mood) seems to be strengthening such divisions.

      1. “Having an elite which is seen by the majority population as being nearly 100% outsiders is a dangerous situation.”

        The problem is that there are many examples in history of minority groups either overperforming or underperforming in relation to the majority which causes serious social issues. Jews in Europe, Chinese in Malaysia, australian aborigines in Australia, Irish Travellers in Ireland, Asians and African Americans in the US,
        When a minority is underperforming we call it structural discrimination, but how do we explain the reverse?

        In Europe both the Roma and Jews were discriminated against. Are both cultural and genetic factors at play?
        In the current climate we cannot even ask these questions.

        1. Right. But even without answering any questions of why, these differences are an observed fact. And one which we ought not to ignore when discussing immigration / assimilation / identity issues.

          Many wish this fact would go away, and stridently tell us that it will not apply to “new soviet man” / will disappear as soon as we send the police on implicit bias training / can be fixed by affirmative action. These are extraordinary claims, and the onus is on their supporters to produce good evidence. Not on those of us who think that things will stay much the same.

  22. Why should we care about superficial characteristics such as skin hue and eye shape in admissions? If the goal is to remedy the effects of American slavery, then it should only be for American blacks. It doesn’t follow that anyone with dark skin should get preference.

    Diversity may be important, but not the kind usually talked about. Diversity initiatives don’t seem very interested in intellectual or cultural or political diversity, which seem more important in solving problems and providing an enriching educational environment.

    I worry about the continual lowering of standards in order to effect this diversity. Replacing mathematical problem-solving in the hard sciences with subjective group projects because certain groups have a hard time with math does not train better scientists. Eliminating tests for math and reading for teachers because teachers from certain groups have trouble with numeracy and literacy does not improve our children’s education. Ignoring the MCAT scores for certain groups because they have trouble passing the test does not make our health care system better. Requiring large amounts of grant money to be set aside for researchers with the right skin color does not increase our research output. Eliminating fitness and knowledge tests for firemen because some applicants are physically weak and others don’t understand how fire spreads doesn’t improve our fire safety.

    I know it’s a common saying that “diversity is our strength”, but I don’t believe it – certainly not the type of diversity commonly referred to. Everyone knows that the “diverse” peoples are the ones who tend to perform poorly (which is why Asians often “don’t count”). Lowering standards throughout education and discriminating against people who perform well in favor of people who perform poorly may be necessary costs to achieve some worthy social goal. They may be costs that we’re capable and willing to pay. But I’ve no doubt that our scientific and technological output, and our collective ability as a nation, are weakened as a result.

    Whether it’s college applications or fire fighting, raise people up to the standard if you can, perhaps with additional resources, but don’t lower standards, is what I would say. And if they simply can’t meet the minimum standard despite the extra help, then they should probably be somewhere else.

    1. Eliminating tests for math and reading for teachers because teachers from certain groups have trouble with numeracy and literacy

      Does this happen somewhere? And what group of teachers would you be referring to?

      1. A recent example for New York State: https://www.nytimes.com/2017/03/13/nyregion/ny-regents-teacher-exams-alst.html

        I remember a reading of a similar test for teachers in Florida, which required the ability to do arithmetic at the 8th grade level and to read at the 10th grade level. IIRC, 25% of white applicants failed, which, given that they all had bachelor’s degrees, should already be a national disgrace, but 79% of black applicants failed. So there was discussion about eliminating the test to reduce the “disparate impact”. I don’t know what happened to it, though.

        1. Thanks for the link. Interesting article, and I thought this bit might add context for those hearing about it just from your comment:

          Michael Middleton, dean of the Hunter College School of Education, said in an interview on Monday that the battery of exams currently required of teacher candidates — four, in most cases — was onerous and expensive, and that eliminating the ALST was appropriate.

          “We already know that our licensure candidates have a bachelor’s degree, which in my mind means they have basic literacy and communication skills,” Dr. Middleton said.

          I’m always curious whether there is evidence that the scores on these tests are indicative of how successful a teacher will be.

          1. High scores are not an indicator of raw teaching ability, but low scores would indicate an insufficient grasp of the basic material to be taught.

            1. I have to agree.

              Thinking back to my schooldays, the best teachers were not necessarily the near-geniuses, but the ones who had a good and broad grasp of their subject and could communicate it in an interesting way.

              That means they had to be *good* at their subject. And good literacy was essential for that, even for maths/science subjects.

          2. I wonder why we bother trying to get everyone through high school, if 8th grade math and 10th grade reading are not necessary even for teaching. Should we hire math teachers directly out of middle school then?

    2. Since the subject of teachers came up, surely ‘personality’ tests should be adjusted to suit the intended profession of the applicant.

      By that I mean, a prospective teacher should be a good communicator, able to put across concepts in a way that engages the interest of the listener. Painstaking thoroughness and ability in sifting out errors – not so important. For a prospective engineer the criteria might be tilted the other way.

      But if I’m going for an operation or an airline flight I hope like hell the surgeon or the aeronautical engineers who designed the plane were selected on ability and not some also-rans who got an easy ride because it was socially beneficial.


  23. I disagree with any form of discrimination affirmative or otherwise. If you are smart enough to get in then anything else is superfluous. After all why stop at ethnicity? You could start including gender, the ever popular sexual identity/ orientation, vegetarian, conservative background, religous belief/atheism, level of critical thinking in their homes, school popularity and the list will go on.

  24. The problem is that all schools act on quotas. A different one (for private schools but public colleges are not allowed) is discrimination on sex. By balancing sexes 1:1 they discriminate women who are more abundant in the pool and have higher scores… In my college changes of a woman getting in are 17% compared to 25% men. The majority of the minorities are women, which effectively reduces even more the changes that a non-minority woman gets in. I am in favor of affirmative action, but don’t know how it can be done without any side-effects.

  25. I do strongly think “admission should be genetics-blind, based on criteria like scores, grades, and other achievements and activities”.

    You don’t right past wrong by making new ones.

        1. Whether or not they be, test the talents, not the genetics.

          (Unless in a universe where genetics is an easier and more reliable predictor of talents than a direct test of talent. That’s not our universe, however).

    1. If I steal a painting from you, get away with it, pass it on to my kid, and then long after I’m dead our families figure out what happened, should my kid be forced to give it back? On one hand, he didn’t steal it and as you say ‘you don’t right past wrongs by making new ones.’ On the other hand, if he doesn’t have to give it back you’ve just basically said theft is legal as long as the thief can get away with it for a generation or two.

      That, I think is the analogous reasoning for affirmative action. Families that are now wealthy because they profited off the labor of others vs. families whose labor was taken to serve others; do we try to give back some equivalent to what was taken? The wealthy high school kid is not ‘guilty’ of such taking. And nobody is saying that they are. Nevertheless, the fact that he legally inherited his painting painting doesn’t necessarily mean the other family deserves nothing for their loss.

      1. There is such a thing as “durée de prescription” in French law. I’m sure there is an American equivalent. So yes, basically, “theft *is* legal as long as the thief can get away with it for a generation or two.”

        It’s not even the thief getting away with it. The thief dies at some point. That’s people who committed no crime “getting away” with something someone else did before they were even born.

        Multi-decades / century grudge holding at the scale of large groups is even more asinine IMO.

        My grand-father and great-uncle died in a concentration camp at the hands of the Germans (well, 6 months later for my great-uncle, who looked like a walking skeleton when released).

        Should I track the guard’s grandkids down and demand retribution ? Am I morally better than they are because I share blood with victims ? Do they owe me anything ? I think not. Neither they nor I have anything to do with what happened then, save a few DNA markers.

        We can at most speculate idly about what life would look like now if our ancestors had been on different paths, but it’s pretty pointless.

        [Nota: In an ideal world I’d prefer to do away with the concept of inheritance entirely, but that’s another debate.]

        1. Oops — I meant *great*-grand father, not grand-father. That one died pretty uneventfully, by all accounts.

  26. PCC(e) asks a final question separately from the five he enumerates – what good is a college where you don’t confront people with different views?

    So, to what extent should this be a goal of college? To what extent does diversity as currently practised contribute to this goal?

    My impression from news reports (so probably wildly distorted)is that there is a good deal of self-segregation, wilful fleeing from some views, and much more confrontation than debate. By contrast, my memory (probably rose-tinted) of university life 50 years ago and well before diversity became an intellectual idol is that there was vigorous debate about all sorts of issues as part of student life.

    A further question: to what extent does diversity benefit those students who gain admission to college by reason of diversity rather than academic performance? Now and then there are research claims that these students experience considerable frustration, much higher failure rates and leave with little to show but debt, part of which paid the salaries of diversity administrators who enticed them to enter college.

    1. What you allude to at the end – the mismatch effect whereby minorities get accepted to colleges that they’re not prepared for rather than ones where they can genuinely succeed – is a way that these diversity mandates hurt the people they’re trying to help. In almost any activity, you’re more likely to get discouraged and give up when you’re competing with people far above you, and the performance gap is bigger at the elite schools. Better to graduate with some confidence in your ability and a decent degree from a second-rate school than to be discouraged and fail in a first-rate school…

  27. JFK’s original visions for “affirmative action” was to remove artificial barriers to opportunity. The Carter administration fundamentally changed that to the erection of artificial quotas.

    Two wrongs don’t make a right. Quotas are racism.

    Yes, the SC has ruled that inchoate ‘diversity’ is a compelling interest. But a university should be a place of ideas, so the diversity it seeks should be wrt perspectives, not melanin content.

    Given my druthers, there’d be no ethnic quotas, no personality filters; just let the chips (i.e., test scores & grades)fall as they may.

    1. then how about “blind auditions,” like they used to do when auditioning to play for the symphony. (Not sure if that’s still the practice.)

      You’re behind a curtain with a number. Your gender, ethnicity, age, your parents income, whatever, is hidden. And your voice can be disguised.

      Would that help?

      1. Blind auditions and the like are wonderful solutions to removing artificial barriers! When employed by symphonies, it resulted in a large spike in women being hired. Interestingly, when done in a recent study with resumés of job applicants, it resulting in less women being selected. But fair is fair.

  28. Maybe the first question is: why does Harvard exist? I mean, why isn’t it just another Caltech — also elite, but very different?

    My understanding is that, roughly, they aim to have students who will end up in the country’s leading class. In the 1920s it was clear that most leaders would be WASPs, some with famous names. So that’s who they admitted. In the 2010s, it’s clear (to me) that the next generation’s leaders will (for various reasons) contain more blacks and fewer asians than the top 1% of SAT scores. So that’s who they admit.

    Caltech, on the other hand, admits the brightest. This is a great strategy for producing future professors, and Nobelists. But not leaders. The smart jewish students Harvard rejected in the 1920s made great lawyers and scientists, but not so many secretaries of state. It’s fine to wish this were different, and to lament the reasons… but my argument is independent of that. Harvard made an accurate prediction about what America was going to be like.

    I wish they could be honest about this. The personality ratings are obviously nonsense, and actually seem more offensive to me than honestly making predictions. But that’s not the world we live in. Perhaps because honest predictions seem offensive, as the world is not as nice as we would like to think, and we’re inclined to hate people who point this out.

    (And there are legalistic reasons. The laws seem pretty contradictory, to me, reflecting our ambivalence about various kinds of discrimination.)

  29. Terrible behaviour from the regressive SJWs who put this racist policy into action.

    But remember, these are the “anti-racists”. Lol.

  30. 1 (Favor AA?) Sure, in moderation. Like many policies or tools, I think it can do good but that there certainly can be corrupt, abusive or just plain bad implementations of it too.

    2 (Extend to other traits?) Maybe…I think the more traits you try to select for, the more likely a school will slip into the ‘corrupt, abusive, or just plain bad’ camp.

    3. (Quotas?) It’s a somewhat moot point since SCOTUS has declared them unconstitutional. However, I think a transparent and clearly communicated quota is more ethical than fixing/rigging a subjective personality score.

    4. (Asian dawn?) No, I don’t think striving to keep a group out is a good goal. Granted, it might end up being an inevitable consequence of trying to diversify the student body, but even if that’s the case, I think it’s good to keep goal statements and policy statements focused on the positive outcome you want to achieve. That way, if and when someone smarter comes along and starts thinking up new and elegant ways to achieve the goal you set in stone, their new ways actually help the school achieve its real goal rather than merely achieve some proxy measure.


    I have a question back for you, PCC. What do you think of the idea of some sort of lottery or partial lottery for all applicants who meet a school’s criteria? Let’s face it, many schools get far more wonderful applicants than they can accept. The schools then play this game of trying to decide whether the valedictorian varsity athlete who started their own company selling their own music is more worthy than the valedictorian varsity athlete who started their own company selling their own paintings. It’s baloney; they are both worthy, both wonderful, and both exceed any reasonable college entry criteria. A lottery acknowledges that after a certain point of achievement, there’s no sense in trying to rank or prioritize students. Pros: 1. it’s fair. 2. It’s transparent. 3. It would cut down on the ‘academic arms race’ among high schoolers, probably giving them more free time to pursue their actual interests rather than do things to pad their resume. 4. If the criteria aren’t extreme (valedictorian varsity etc…), this would put a lot more poor kids in the running with wealthier kids. Cons: 1. a random draw is only going to be as diverse as the pool. Low diversity in pool -> low diversity in the accepted class. If 80% of the people who apply are males, then without tweaking the lottery your class would be 80% male. So schools would probably have to do *some* weighting if they have diversity as a goal. 2. A lottery would do nothing and might even exacerbate the stress students feel of applying to colleges. 3. It wouldn’t work (well) for rewarding legacies, scholarship athletes, and some other specialized entries. 4. With a random draw, there’s always some chance of getting weird swings, though this chance is probably pretty low if you’re picking thousands from tens of thousands.

    …but now I’m writing too much. What do you think of the base idea?

      1. There are many ways to weight a draw. Two examples. One: you could have separate pools and draw a certain amount from each one. For Harvard, for example, you have a legacy pool. I suspect most schools would want equal numbers of men and women, so another would be to pool women and men separately and then select the same number from each. Two: you don’t give each candidate the same probability of being drawn. Let’s say the school really wants West Virginians, but only a few applied. If an equal probability would give each of 10,000 candidates a 10% chance of getting in, you instead give each of your West Virginians a 15% chance of being drawn and reduce the odds of each of the much larger cohort of non-West Virginians to 9.5% chance.

        There’s probably many other ways of tweaking a lottery. Those are just the first two that sprung to mind.

  31. I would note that a lot of these issues are really driven by a perceived shortage of slots at high-quality schools. Everybody wants to go to Harvard, because that sheepskin will provide career benefits beyond the actual quality of the education gained (as someone discussed in depth above). If there were sufficient slots at high-quality colleges for everyone, AND if attending any such high-quality college and getting good grades were perceived to be equivalent to attending Harvard and getting good grades, all of this fighting over slots would disappear. So my solution to this would be: high-quality college slots should not be a scarce resource. I’m not sure how to fix that problem, besides simply throwing money at it, but I would suggest that maybe that is the problem that really needs to be solved.

    It’s similar to how scientists all want to publish in Nature or Science. That’s not because those journals are actually objectively better; there are lots of journals with just as good editors and reviewers, and that IMHO publish less nonsense than Nature and Science do (Nature and Science sometimes publish articles of dubious scientific merit but that they think will generate lots of citations; I have seen this sausage being made from the inside, and it’s not pretty). The desire to publish in Nature and Science, then, is really only because of the current societal bias toward Nature and Science – their unearned and undeserved prestige, which gives those who publish in those journals a boost in getting grants, getting tenure, etc. Similarly, the root problem underlying all this debate about affirmative action is perhaps Harvard’s unearned and undeserved prestige. You wouldn’t (I hope) try to fix the Nature/Science issue with complicated racial quotas and so forth; instead, you would try to build more open-access journals, and raise the perceived prestige of those alternatives, and talk publicly about how Nature and Science really aren’t as good as they are held to be, and how there ought to be no shame in publishing in PLoS One instead as long as the actual paper is good, and so forth. I would point toward the same strategy for colleges. It won’t get fixed overnight, and probably affirmative action and racial balancing and so forth are needed until then, and I’m fine with that (I could give my answers to Jerry’s questions but this post is too long already); but the long-term goal should be a more healthy academic ecosystem with more healthy attitudes about what it means to attend Harvard versus, e.g., my alma mater of San Jose State University. I think I got a better education there than I would have gotten at Harvard; so why aren’t all the top Asian students fighting to get into SJSU? That, I think, is the root of the problem. Fix that and you fix the whole thing.

  32. “We are the Borg. Your biological and technological distinctiveness will be added to our own. Resistance is futile.”

    I see a lot of fear driving decisions. Too bad.

  33. I find it so odd that racial preferences are the norm in some areas of life (college admissions, company hiring practices) but would be unthinkable in other settings: imagine if the PA rang out in the supermarket “No waiting on aisle 6 if you’re black.”

    What a strange world we live in.

  34. Soccer is 100% meritocracy: Iceland and Panama are in the World Cup, not China or USA. Why would law school be different?

  35. You can’t pretend that the sub-standardly educated prospective student is prepared for Harvard. Most won’t do well.

    We should providing quality elementary and high school educations for all. It isn’t done because it costs money. Affirmative action does not solve this problem.

  36. Hypothetical question:

    How would you react if you achieve a fair society and students are enrolled on merit but post graduates are dominated by specific groups, not proportional demographically.

    What do you do if for example Asians and Jewish males keep out performing other groups in areas like mathematics and physics by a huge uncomfortable margin?
    (b.t.w, Richard Feynman faced quotas in his day)

    I think the big problem is that there is an a priori assumption that humans are all the same.

    While it is certainly true that not eveyone has the same opportunity and we should address that, we set ourselves completey unrealistic goals that are not based on human diversity.

  37. We should believe in equal opportunity but not in equal outcome. The most able people in zany society are going to rise to the top and be rewarded better. If Harvard does not easy to accept the best students their reputation will decline and other universities with better students will surpass them. This will take a few years but it will happen.

  38. What I’m going to do to is start a college where all of the people who were denied entrance to top universities by racist allotment systems will be allowed into my University, by merit and only merit. In a couple of generations my meritocratous hordes will rule the world and I will be the de-facto leader.

  39. An economic status correction to GPA, etc. I think has some importance. In the US I see the biggest barriers being social class, which of course often correlates but is not identical to race.

    However, as a colleague of mine from McGill pointed it out, this has to be done with some care. The McGill “needs based” scholarships, for example, were (according to him) not taking into account the *relationship* between a student and his family. In his case, he had lived and worked on his own for a few years before coming to Montreal and McGill. He was denied a needs-based because his *parents* were wealthy. Now, that may have given him a heads up earlier in life but wasn’t a *current* factor. The earlier vs. later is where things get really messy.

  40. It should be based on merit decided by SAT score, IQ, and personality tests appropriately weighted for optimal interactions in seminars, regardless of race — however that is defined.

  41. Every institution in the United States was built on white supremacy and despite recent well-meaning efforts to level the playing field among traditionally marginalized groups (i.e. Affirmative Action), it would be impossible to have a meritocratic system in education without completely disproving the lie that “everybody has been given the same opportunities.” The solution is repairing the damage that’s been done to oppressed populations and dismantling and rebuilding the education system to begin with. Until then, you’re all fighting over who gets to sit at the top of a hill made up of lies and human suffering. Anyway, good luck with that.

Leave a Reply