The “Pipeline” Problem That Medical Schools Don’t Want to Discuss

The U.S. Department of Justice’s recent finding that Yale School of Medicine discriminated on the basis of race in admissions has reignited one of the most contentious debates in higher education. According to the DOJ, black and Hispanic applicants admitted to Yale had substantially lower median MCAT scores and GPAs than white and Asian applicants across multiple admissions cycles, with the department concluding that equally qualified black applicants had dramatically higher odds of receiving interview invitations than comparable Asian applicants.

The finding comes on the heels of an earlier DOJ determination that David Geffen School of Medicine at UCLA likewise engaged in racial discrimination in admissions, admitting black and Hispanic applicants with substantially lower academic credentials than their white and Asian peers.

For DEI critics, the Justice Department’s investigation provides confirmation that elite universities have continued racial preferences despite formal legal prohibitions. Admissions officers and other college officials argue that the Trump administration is leveraging civil rights law against diversity initiatives and misunderstanding the role of holistic admissions.

But beneath the legal and political fight lies a more fundamental question, one that receives surprisingly little attention: What is the proper way to evaluate disparities?

For decades, universities have treated demographic disparities in outcomes as self-explanatory. If black Americans make up roughly 14 percent of the total population but only about 6 percent of physicians, the observed difference is taken as evidence of wrongful underrepresentation—discrimination, in other words. 

But medical schools don’t select students from the general population. They choose from a narrow, highly filtered pool of applicants who have demonstrated sustained excellence in challenging scientific coursework over many years. If the goal is to evaluate whether admissions reflect equal standards, the relevant comparison is not the population at large but the pool of students academically prepared to succeed in elite medical training.

One of the most consequential sorting points for medicine occurs long before medical school applications are submitted—indeed, often before students even enter college. Students who eventually become competitive pre-med candidates typically begin signaling interest in, and readiness for, scientific training in high school through advanced coursework—most commonly, Advanced Placement math and science classes.

Our recent research article traces this pipeline using AP coursework, particularly AP Chemistry, one of the clearest large-scale indicators of preparation for rigorous medical training. The disparities at this stage are large, persistent, and cumulative.

Black students make up roughly 14 percent of American high school students but only about 5 percent of AP Chemistry examinees. By contrast, Asian students comprise only about 5 percent of high school students yet account for roughly 27 percent of AP Chemistry examinees. Among black students who take the exam, only about one in five earns a passing score, compared with a majority of white students and roughly two-thirds of Asian students.

At the very top of the distribution—the students most relevant for elite medical training—the differences are especially stark. Only about 2 percent of black AP Chemistry examinees earn the highest possible score of five, compared with roughly 20 percent of Asian examinees. Put differently, there are nearly 50 high-scoring Asian students for every one high-scoring black student.

And these are not isolated disparities. In AP Biology, one finds more than 25 high-scoring Asian students for every one high-scoring black student. In AP Physics, the ratio is roughly 46 to one.

These differences compound across the pipeline. Each stage—taking difficult courses, passing them, excelling in them, mastering advanced scientific material—further narrows the pool of students realistically positioned for elite medical education. By the end of high school, the distribution of academically prepared candidates is already sharply stratified.

Admissions committees do not create these disparities. Once we take the educational pipeline seriously, however, the interpretation of representation changes substantially.

If African-Americans are dramatically underrepresented among top academic performers yet appear near population parity in elite admissions, then that discrepancy itself requires explanation. It implies that admissions processes are incorporating considerations beyond conventional indicators of academic preparation and performance.

For decades, universities have treated disparities at the end of the pipeline as evidence of injustice and sought to correct them at the point of admission. But as evidence from our research suggests, many of the disparities originate much earlier—in preparation, performance, and the cumulative development of the skills required for demanding fields like medicine.

Adjusting standards at the end of the pipeline obscures that problem instead of fixes it. And this carries institutional costs. Once universities appear to subordinate meritocratic standards to demographic balancing, public trust begins to erode. Students question whether standards are applied consistently. Patients wonder whether admissions decisions are politically managed. Institutions lose legitimacy precisely because they seem unwilling to confront obvious empirical realities.

If the goal is a medical profession both excellent and broadly representative, then the solution cannot consist merely of manipulating admissions outcomes. The work must begin where the disparities first emerge—not where they become politically embarrassing.

The Department of Justice is right to force this conversation.