I present yet another example of a scientific organization engaged in political/ideological advocacy. In this case the organization is the august body of the National Academies of Science, an organization that admits only the most accomplished of America’s scientists. It also publishes the Proceedings of the National Academies of Sciences (PNAS), a pretty well respected journal though not of the quality of Science or Nature.
In the latest issue, three authors from Harvard (and Mass General Hospital) argue, well, you can see the claim themselves from the title screenshot below. It’s a political argument, not a scientific one, and is a reaction to the likely dismantling this year of of race-based preferential admission (“affirmative action”) by the Supreme Court. The Justices have already heard a case that began in 2014, arguing that both Harvard and The University of North Carolina violated both the Constitution (the “equal protection clause” of the Fourteenth Amendment) and—for the private university of Harvard—the Title VI provision of the Civil Rights Act of 1964. (Private universities can also be sued for discrimination if they receive federal funding.)
There’s little doubt in my mind that PNAS agrees wholly with this op-ed, since it rarely publishes articles like this and also itself has a “progressive” ideological slant. I’m confident in arguing that this is probably the National Academies’ view as well as the authors’ view.
The issue here is that the paper takes an expressly ideological stance by favoring one political outcome (“equity” among doctors) over what is possibly a negative side of that outcome: worsening quality of medical care in America. This is because affirmative action in medical school had led to a substantial gap between the admissions credentials of whites and Asians on the one hand, and Hispanics and blacks on the other. And if admission credentials have ANYTHING to do with the quality of a doctor, then lowering the bar in the way that Aaron et al. suggest—achieving “equity,” or ethnic representation among doctors equal to the proportion of groups in the population—will inevitably result in the decline of the average quality of doctors.
You may, I suppose, suggest that if people have doctors who “look like them”, or if doctors treat patients who “look more like them” better, then increased equity might increase overall health quality. But that is a supposition only, and one with no empirical basis (there’s no need to mention the “dead baby” issue, which has substantial problems). It also doesn’t tell us how far we must lower the bar to achieve equity, and whether lowering the admissions bar that much can counteract any positive effect on healthcare caused by the supposed “looks like me” issue.
As I’ve said, I do favor some form of affirmative action for admission to colleges and medical schools, but more and more it’s taking the form of favoring minority candidates when you’re faced with deciding between candidates who are both qualified and nearly equally qualified. My thinking on this is evolving. But the difference in qualifications among ethnic groups among med-school applicants as well as those who are admitted is sufficiently large that my brand of affirmative action would never come close to producing equity.
Click to read (it’s free):
The paper begins with a flawed argument: that by forcing universities to not discriminate among races, the Supreme Court is practicing unwarranted “judicial intrusion into medicine” (my emphasis)
Given the likelihood that the Supreme Court will end affirmative action, medical institutions must plan how to further diversity without incurring liability. More broadly, the cases follow a pattern of judicial intrusion into the affairs of medicine and health. As with abortion, contraception, health insurance, and COVID-19, the Supreme Court has encroached on the field of medicine, denying expert guidance on what is required in order to achieve a healthy and equitable society. The affirmative action cases are emblematic of a high court intent on opposing racial progress and other forms of social change.
This is bogus. Yes, some Justices may “oppose racial progress”, but what they’ll do if they rule against affirmative action is simply enforce the law and the Constitution. That is, they will interpret the law as saying that one cannot discriminate among by race, regardless of which races are at issue. But to say that this denies doctors “expert guidance” to achieve a healthy and equitable society” is mere carping, and beside the point, for you could use the same argument to justify any sort of race-based discrimination, especially if you think feel that “equity” is an outcome more desirable than application of settled law. I find the argument for retaining “expert guidance” risible, for experts are going to differ about what societal outcome is “healthy and equitable”. What they’re really saying is that the Supreme Court is overreaching itself by disallowing discrimination on the basis of race.
Now, about the differences in qualifications of candidates. Here’s what the article says:
As of 2015, there is estimated to be a deficit of about 114,000 Black and 81,000 Hispanic doctors compared to what one would expect from proportions of the US population. This dearth is at least partially historical: Racial and ethnic minorities were excluded from attending medical school and joining medical organizations, such as the American Medical Association (AMA). In 1900, 11.6% of the US population was Black, compared with 1.3% of physicians. In 2018, 12.8% of the population was Black, but only 5.4% of physicians. Over 120 years, then, the fraction of Black physicians has increased by only 4 percentage points. This lack of representation emphasizes why the fate of affirmative action is essential to securing a racially and ethnically diverse physician body in the United States.
Affirmative action helps compensate for systemic inequities throughout childhood and young adulthood that impede the significant steps required to apply to and be admitted to medical school. The current biggest gatekeeper to medical school admission is the Medical College Admission Test® (MCAT®), whose notable racial and ethnic disparities are well documented (5–7). As of 2022, Black and American Indian/Alaska Native medical school applicants have an average MCAT® score of 497.4 and 498.7, respectively, which is about one standard deviation below the average score for White applicants of 507.9. Although these disparities do not mean that the test makers intentionally discriminate by race, they reflect the systematic disadvantage facing racial and ethnic minorities applying to medical school.
Note the claim that there is “systematic disadvantage”, which to many means “structural disadvantage”: something about the tests themselves that give racial and ethnic minorities a disadvantage. If it’s social circumstances: poverty, poor education, and so on, then that’s a different matter, though one that still must be considered. This is symptomatic of the whole failure of the authors to see the dimensionality of the problem.
What holds for MCAT scores (still required for nearly all med school admissions) also holds for grade-point averages. These recent data come from an admissions consulting service, Shemassian, which notes that race certainly does matter when applying for medical school:
And data on those who matriculate among four racial groups divided by MCAT scores and grade point averages. The differential rates of acceptance among groups is striking. Asians in particular have a hard time getting in, being accepted at only about 10% the rate of blacks in the leftmost category.
This is from the American Enterprise Institute (couldn’t they use skin-pigment neutral colors for the bars?):
Now one could argue that one standard deviation in MCAT scores doesn’t mean that, on average, a lower-scoring individual is less qualified to be a good doctor than a higher-scoring one. But if that’s the case, why use scores at all for determining who’s qualified? And if qualifications have any correlation with quality of medical care dispensed later in life, then they could have titled the article, invidiously, “Attacks on affirmative action threatens quality in medicine.”
To me, qualifications for medical school are like qualifications for being an airline pilot in one way: once you make it, you’d better be good, as you’ll have a whole lot of lives in your hands over your whole career. If you say we should lower the bar for going to med school, would you say the same thing for pilot training? If not, why not?
At any rate, the main issue with the article is that it doesn’t even discuss the possible tradeoff between the quality of American healthcare and ethnic equity among American doctors. Perhaps they feel that equity among doctors will, as I said, outweigh any bad effects of lowering the bar for med-school applicants. Or perhaps they feel that equity in a profession even outweighs the concrete benefits that the profession is supposed to confer on society. (That is, equity among doctors is more important for society than the overall quality of healthcare.) One could make both of these arguments, but the problem is the authors don’t. They simply feel that affirmative action is an affront to medical care and to American society, and that equity is a virtue that’s above American law. Indeed, opposing affirmative action is equated, throughout this article, as identical to endorsing white supremacy:
But because the Supreme Court may now deem consideration of race itself to be discrimination—even when it would benefit a racial group and society as a whole—the Supreme Court may stipulate a rule with far-reaching effects across the US healthcare industry. This would not be the first Roberts Court decision to re-interpret civil rights laws as protective of White people, as opposed to the subordinated groups these laws were meant to protect. For example, in Parents Involved in Community Schools v. Seattle School District (2007), the Supreme Court held that voluntary school desegregation plans in Seattle, WA, and Louisville, KY, violated the Equal Protection Clause. Although this clause was created to help remedy racial discrimination after the Civil War, the Supreme Court used it to impede efforts to address systemic racism. In this case, Chief Justice Roberts famously quipped, “The way to stop discrimination on the basis of race is to stop discriminating on the basis of race,” thereby imparting a “color-blind” reading on a clause aimed at racial progress. The Roberts Court’s belief that color-blindness will end racism overlooks the more structural and systemic forms of racism that are likely the biggest propagators of disadvantage today. At the same time, the color-blind approach invalidates policies aimed at racial progress like affirmative action, thus entrenching the benefits of Whiteness.
Two more points. First, the authors reply on the deeply flawed Implicit Bias Test, one that’s been rejected by most psychologists as not measuring anything meaningful and even as counterproductive in reducing bias:
Affirmative action’s loss may result in medicine and healthcare that look substantially less diverse. And any loss of diversity could be self-perpetuating. Through their presence and contributions, racial and ethnic minorities help make spaces more accommodating and inclusive to people from differing backgrounds. For example, testing of implicit bias using the well-validated Implicit Association Test has revealed that African-American physicians have far less implicit bias than White physicians Racial and ethnic prejudice, prolific throughout medicine, impact physician–patient communication, treatment decisions, and patient outcomes. A less racially diverse pool of providers could bring heightened bias toward racial and ethnic minorities, leading to a spiral effect, in which medical spaces become less varied across the board.
That, of course, is purely speculative. And remember that when affirmative action was first implemented in the Sixties, it was supposed to be a temporary expedient—a few decades at most. In the Grutter v. Bollinger decision in 2003, in which (by a 5-4 vote) the Supreme Court allowed some consideration of race in college admissions, liberal justice Sandra Day O’Connor said this:
“Finally, race-conscious admissions policies must be limited in time. The Court takes the Law School at its word that it would like nothing better than to find a race-neutral admissions formula and will terminate its use of racial preferences as soon as practicable. The Court expects that 25 years from now, the use of racial preferences will no longer be necessary to further the interest approved today.”
Note that she later clarified that she didn’t mean to set a firm deadline. But I think we know now, from the institutionalization of both affirmative action and DEI programs in colleges, that affirmative action and “racial preferences” will—are intended to—last forever.
Finally, the authors suggest the obvious: that medical schools should start trying to find ways around the upcoming changes in law to maintain a diverse medical profession:
The affirmative action cases before the Supreme Court raise serious questions about the future of a diverse medical profession and the integrity of healthcare itself. In advance of the decision, medical programs that value diversity and accept federal funds (or are government institutions) would be wise to consider alternative paths to create a racially diverse student body without explicitly considering race. With some strategizing, they can preserve some racial diversity while keeping to the letter of the law.
I’m trying to be charitable here, but it sounds like me that they want to circumvent the law in some way by “strategizing,” but that might not be possible if the Supreme Court rules that you can’t use correlates of race as proxy data for admissions. But the authors don’t like even that strategy!:
Many medical schools will be driven to alternative metrics, or proxies, to attain racial diversity. Such proxies could include family history, experiences of discrimination, socioeconomic status, and geography. However, proxies are limited for at least two reasons. First, they may poorly approximate race. For example, experiences of discrimination may seem to correlate with race, but they have frequently been claimed by White applicants, including in the Supreme Court’s historical affirmative action cases. And, second, the upcoming decision may prohibit not only consideration of race, but also similar metrics.
But back to the article. I conclude that it is not a contribution to solving the problem. It neglects important issues, brings up irrelevant ones, makes unstated assumptions, and, above all, injects ideology into science.