• Affirmative action

    My friend Hank Aaron (yes the economist, not the baseball guy) sent along a nice review essay on this subject, simply titled: “What Should the Supreme Court Do About Affirmative Action?”

    Hank’s email happened to arrive at the end of our admissions season. I’ve read more than 100 admissions folders this year in my capacity chairing our school’s MA-level admissions committee. This is a weighty responsibility. Ethical, legal, and operational tensions surrounding university admissions are intimately familiar to me.

    I continue to be struck that most of the best arguments, on almost every side, were presented with greatest eloquence, sincerity, and rigor forty years ago in briefs and public commentary occasioned by the Bakke case, arguably the most famous Supreme Court decision of the 1970s. In 1977, McGeorge Bundy of all people provided one of the very best arguments in that case. As Bundy noted, immediate pressures for race-conscious admissions came from many sources, but its

    deeper and more durable cause was the growing conviction that there was a fundamental contradiction between an asserted opposition to racism and the maintenance, by whatever process of selection, of essentially all-white colleges and professional schools.

    Bundy’s essay is especially striking when one considers that Bakke was closer in time to the passage of the Civil Rights Act of 1964 than we now are to the date of George W. Bush’s reelection.

    We still know surprisingly little about what really matters, statistically, in predicting which of our applicants would most benefit from our academic programs to do great things in the world. One partial exception arises in medicine. There is some evidence that medical school applicants with higher MCAT scores are more technically proficient and display better clinical reasoning than otherwise similar peers, and this matters.

    Even in clinical medicine, the evidence is far from definitive.  Performance on the biological and physical science MCAT portions significantly predicts grades and board scores in these areas. The links between other aspects of the MCAT and medical school performance, not to mention links to actual clinical proficiency, are much more murky.

    The reality, for so many different reasons, is that we cannot provide excellent medical care to African-American or Latino patients and communities if the medical and skilled nursing workforces do not represent men and women of color in their ranks. This holds to an even greater degree in social services and public health.

    It’s not merely that patients and clients want to “see people who look like them,” though this sentence crudely captures one human reality in the helping professions. More than that, one needs to know something about–have some tactile familiarity with–the life circumstances of the people one is trying to help. An endocrinologist whose teachers, classmates, and peers are all products of upper-middle-class meritocratic suburbia will be poorly-equipped to care for many of his patients.

    This isn’t merely a matter of race or ethnicity. For years, medical and public health authorities struggled to address an HIV epidemic exploding within populations and communities that were worlds away from the personal experience and biography of most medical and public health practitioners. Crucial time was lost, and the HIV prevention was less effective than it should have been.

    Much of the available evidence on admissions concerns who will do well in our coursework, which of course is a very different question from who will make a real difference in the world. Hank Aaron, in his nice essay, suggested that the “environment in law school is ruthlessly meritocratic to an extent true of few undergraduate programs.”  In one narrow sense that is true. Yet in another way, that’s definitely untrue. At a minimum, the criteria through which law-school “merit” is judged deserve greater scrutiny. Much of the legal curriculum is famously unrelated to actual legal practice. Many personal skills and knowledge relevant to excellent lawyering are basically unaddressed and untaught in elite legal education.

    In my own teaching, students with higher math GRE scores perform much better in my economics and medical cost-effectiveness courses. That clearly matters. Yet our MA students go on to many social welfare, public health, and social work roles.  For many of my students, technical skills probably matter more within our own building than these skills matter after students walk out of our door.

    There is also a weirdly revealing way we try to read people’s resumes backwards to establish their proper ranking in our meritocratic order. Donald Trump wants to inspect President Obama’s college transcripts, presumably to see whether the President benefitted from affirmative action. Leaving aside the racialized insult, it’s as if Barack Obama’s subsequent achievements leading Harvard Law Review, serving on the University of Chicago faculty, rocketing to the United States Senate and to the presidency would somehow be invalidated if it turned out that he had received a B- in freshman calculus or didn’t ace the LSAT.

    Counselors in our BAM intervention reduced violent offending among participating Chicago youth by about 44%. Almost every one of those counselors was an African-American or Latino man from a community similar to those we served. Some are SSA graduates. Others might struggle in my microeconomics class. They are good at their jobs.

    Race-conscious admissions aren’t my favorite. They don’t address earlier, more fundamental obstacles to upward mobility. These policies can bring genuine costs, most notably the genuine human costs imposed on worthy rejected applicants such as Allan Bakke.

    A policy that results in large and visible differences in entry qualifications would be especially costly and unwise. The achievement gap is a real problem. There is something dishonest and undignified about seeking to explain it away or in criticizing generally-reasonable specific content of the tests that yield the most disturbing race/ethnic differences in mean scores.

    The fact that educational institutions pursue affirmative action policies in a furtive, euphemistic way makes it hard to apply standard policy analysis to scrutinize and improve these policies over time, to understand which students are helped or hurt by such policies, or to understand how to implement these policies consistently and well. I would like to see greater attention paid to class-based differences in opportunity and to enlarging the scope of opportunity so that people aren’t so often subjected to a rat race for competitive prestige when they seek basic investments in their human capital.

    For all its problems, race-conscious affirmative action remains necessary. I haven’t seen credible alternatives that aren’t worse. Abolishing affirmative action, or dramatically curtailing it, could easily bring great harm.

    • California banned affirmative action in 1996. We can see the results of that decision today. With the exception of one or two “elite” institutions, California university campuses are more diverse than before the ban and the graduation rates of minority students is much higher. Here is a review from the Cato Institute:


      See also this book by UCLA law professor and civil rights activist Richard Sander and journalist Stuart Taylor:


    • “…one needs to know something about–have some tactile familiarity with–the life circumstances of the people one is trying to help.”

      It’s like how Obama can relate to another AA living in the projects JUST BECAUSE they’re both black. All blacks are the same according to our blog host. If you’re black and your patient is black, we can conclude that you have “some tactile familiarity with the life circumstances” of the patient. Do you realize how racist you are? Please tell us how to stereotype our black patients, if you have some time.

    • Racial quota systems are and shoudl be ruled unconstitutional.

      Colleges/schools dont want to invest resources at early stages of education to beef up minority students — they want an easy racial quota system which doesnt cost any money and is put into place at the time of admission.

      This is wrong. It is time to stop letting colleges scapegoat their responsibility.

    • “The reality, for so many different reasons, is that we cannot provide excellent medical care to African-American or Latino patients and communities if the medical and skilled nursing workforces do not represent men and women of color in their ranks.”

      Aside from affirmative action, this point calls for a better way to match patients with the right doctors. For example, in his book, Dr. Stephen E. Goldstone, himself a gay man and medical doctor, mentioned how he once developed Bell’s palsy, which in rare cases can be associated with AIDS. He sought out a supposedly gay-friendly doctor to examine it, but the doctor went through the whole exam without asking about sexuality or other risk factors for AIDS, concluding with “And of course you’re not gay” and proceeding with some anti-gay slurs.

      Even though, in principle, any doctor should be capable of treating a gay patient, the reality is that many straight doctors aren’t. Few LGBT patients voluntarily come out to their doctors, and even when doctors are both knowledgeable and accepting, there is often a tendency to judge a patient’s sexuality by stereotypes, as happened in Goldstone’s case, while failing to discuss LGBT health concerns with presumed heterosexual patients.

      This is important because not only is sexuality an important indicator to be taken into account for a variety of diagnoses, but there are many preventive health measures that apply only to some sexual orientations and not to others. For example, while gay men who practice anal sex are recommended to get anal pap smears, this would be a waste of time and money for a strictly heterosexual man who has never had anal sex.

      And as Goldstone’s case illustrates, LGBT patients can’t really take their doctor at their word if they claim to be “LGBT-friendly.” One solution some LGBT patients use is to simply seek out doctors who are also LGBT, which brings us back to the point that the composition of med schools needs to match the composition of society.

    • Class conscious affirmative action (as opposed to race conscious affirmative action) is in everybody’s best interest as rising inequality, the growing underclass of all races, and the enormous differences in opportunity for graduates of the elite colleges and universities, especially in law and finance, are ingredients of economic and social instability. The deck is stacked against the lower and middle class, far more than in my day when a relatively inexpensive undergraduate and graduate education guaranteed a well-paying and challenging career. Today, the cost of that same education is prohibitively expensive and the opportunities upon graduation are modest at best, often leaving the graduate with enormous debt and very little extra income, if any, to repay it. In my day, nobody incurred student debt. I recently visited an elite law school with my good friend who works at the school. What impressed me more than anything were the “student services”: one office devoted exclusively to assisting students obtain judicial clerkships, another devoted exclusively to obtaining summer clerkships at prestigious (i.e., high income) law firms (the predicate for full-time employment after graduation), another devoted exclusively to assisting students obtain positions with not for profit organizations, another devoted exclusively to obtaining academic positions. Sure, annual tuition is over $50,000, but good grief. In my day, “student services” consisted of one person who would schedule the on-campus interviews and make sure the interviewers had plenty of coffee. But then, the annual tuition was about $1,000.

    • In many years of hiring decisions in corporate America, one clear learning was this: If two people apply with equal qualifications, the one who is “most like me” will receive the higher evaluation. As humans, we simply aren’t able to apply equal criteria to such assessment.

      So what do you do about this? In the laboratory, if I knew that a given instrument gave biased results in some situations, I would apply a correction factor to zero out that bias.

      I think this is exactly what affirmative action programs attempt to do. We can certainly discuss where to apply such corrections, how large the corrections should be, etc. But the one thing I am certain of is that a correction factor of zero is always and forever wrong.

    • While I think your argument has some merit, ultimately, we don’t get to a society where the content of your character (or mind) matter more than the color of you skin by sorting people into categories based on skin color and treating them differently based on shade.

      In America, the request that you officially categorize yourself by skin color is ubiquitous.

    • @ David C.

      David, what seems sensible to you is not borne out by the evidence. There is no evidence that affirmative action programs have increased the wages or incomes of blacks, women, Hispanics, or any other group for that matter. See my review of the scholarly evidence here:


      And see the very excellent book on this whole subject by former CBO director June O’Neill and her husband, David O’Neill:


      BTW, we should all welcome these findings. Granting favors based on skin color is not only disgusting, it is a clear violation of the 14th Amendment. Turns out it doesn’t even help the people it is intended to help.

    • @Dr. Pollack

      Fantastic post!

      Regarding your conclusion that “I haven’t seen credible alternatives that aren’t worse.”

      I’ve always thought a better solution would rely on SES rather than race. I’m curious why you judge that this route would be worse or not credible.

    • Do you have a citation for this evidence? Thanks!

      “There is some evidence that medical school applicants with higher MCAT scores are more technically proficient and display better clinical reasoning than otherwise similar peers, and this matters.”

    • As long as we
      1. Hire police and firefighters through civil service exams
      2. Have absurd licensing of nurses and Doctors.
      3. Have white and black population that see themselves as distinct.

      we will need AA in government schools, jobs and licensing. We do not need AA in the private realm. Only Government would be so stupid as to hire all white police and firefighters to operate in majority black districts.

      BTW AA for Hispanics is absurd (BTW my wife is Hispanic).