I serve on the admissions committee of a medical school consistently ranked among the top 25 in the United States. On the surface, the process is rather straightforward: a team of academic physicians screens the thousands of applications received, we interview candidates perceived to be competitive on paper, we hold a meeting in which these candidates are vigorously discussed and assign numerical rankings, and then offers are made to the individuals we hope will matriculate. Our goal is a cohort of roughly 140 future physicians of diverse backgrounds and experiences who will provide excellent care to their patients. Although we do have an admissions philosophy (known as “holistic review,” discussed below), and often reflect on the attributes that make effective doctors, my colleagues and I would be hard-pressed to describe our work as “philosophical” in any traditional sense. We do not conceive of ourselves as debating fundamental questions of knowledge and value, but of assessing particular college graduates who aspire toward careers in ophthalmology and orthopedics.

That is not surprising: Medicine itself is far removed—and increasingly so—from its narrative history in ways that other disciplines are not. Scholars in the humanities continue to read primary texts and are expected to do so. Who would hire a Shakespeare professor who has never studied Hamlet or a historian of Ancient Athens who had not read Thucydides on the Peloponnesian War? Law students continue to slog through Marbury v. Madison, and I am told that all serious physicists, at some point during their apprenticeships, take a peek at Einstein’s Annus Mirabilis papers. In contrast, few and far between are the pediatricians who have looked up Jonas Salk’s “Studies in Human Subjects on Active Immunization Against Poliomyelitis” or the endocrinologists familiar with Banting and Best on “Pancreatic Extracts in the Treatment of Diabetes Mellitus.” Medical pioneers are often remembered only through eponyms: Osler’s nodes, Cushing’s syndrome. The Point’s brilliant columnist, Agnes Callard of the University of Chicago, recently wrote in a New York Times op-ed, “I, like Aristotle, am a philosopher….” That proposition is both true and appropriate in many contexts. Alas, in almost any context, the statement that “I, like Hippocrates, am a physician….” is both true and utterly irrelevant.

This brings me back to the philosophy of medical school admissions. From the reforms enacted in response to the Flexner Report of 1910 until the 1970s, medical students were largely selected upon a rather narrow definition of “merit” that included standardized test scores such as Medical College Admission Tests (MCATs) and undergraduate grades (with an emphasis on science courses and preference for graduates of elite institutions), possibly augmented by research experience and letters of recommendation. This system was less of a “meritocracy” than one might think: Until the 1960s, efforts were made to limit the number of Jewish matriculants; the structural barriers confronted by women and racial minorities were ignored almost entirely. More recently, medical schools—including my own—have adopted a broader lens for selecting students known as “holistic review.” We take into account the social and economic distance a student has traveled, acknowledging that the daughter of undocumented farmworkers may be unable to hire private tutors or have fewer opportunities for high-powered research during high school summers than the son of a Fortune 500 executive. Applicants are also credited for accomplishments in athletics, arts, and public service. The leadership skills required to captain a varsity sports team and the dedication called for in mastering a musical instrument are likely reflective of attributes that will transfer well to the hospital setting.

Our current selection process marks substantial progress over the metric-bound system that immediately preceded it. Yet along the spectrum of potential methods for choosing doctors, both numbers-driven and holistic quests for “merit” occupy a rather slender, adjacent range of territory. Like many other aspects of medicine, admissions may be more hidebound than those inside the bubble—myself included—realize. This is not necessarily a failure. My goal is certainly not to point fingers. Rather, I hope that by offering abstract philosophical reflections on a well-established and widely accepted practice, I can bring into relief some of the hidden ethical choices and constraints embedded in the underlying consensus. Although my focus here is on medical school admissions, I suspect similar inside-the-box thinking affects many other aspects of scientific training and institutional healthcare.

In choosing doctors, one must first ask: What is the purpose of doctors? The obvious answer, to most contemporary Americans, is to treat the sick. An advantage of practicing medicine is that one rarely has to explain or justify one’s work—a way in which physicians differ from philosophers, as well as venture capitalists and hedge fund traders. But, that is not the only purpose that doctors might serve. Another aim might be self-interested: to preserve the economic wellbeing and prestige of practitioners. Contemporary examples of work for the sake of employment abound: the state of New Jersey prohibits self-service gasoline stations in part to preserve the superfluous labor of attendants; some jurisdictions maintained in-person toll retrieval on highways long after electronic collection became possible to protect the jobs of collectors. The Oath of Hippocrates, widely perceived today as an ethics code, contains provisions to keep the trade’s secrets from outsiders and to avoid interfering in the work of surgeons, presumably so surgeons would not tread on the turf of physicians. Healing was valuable in third-century Iona, but so was the exclusivity and market share of practitioners. Cynics may see these concerns at play in efforts to prevent nurse practitioners from treating patients independently, midwives from delivering babies, and psychologists from prescribing psychiatric medication.

Few would argue that the primary purpose of medical professionals today is perpetuating their own economic well-being. Yet physician reimbursement rates are directly tied to the number of practitioners in given specialties and communities. More providers mean less pay. By limiting the number of entrants into the club through caps on medical school seats and residency slots, institutional medicine ensures higher incomes for providers. In theory, higher incomes may attract those with more “merit” into medicine—as opposed to teaching or taxi driving. At the same time, choosing to graduate a class of 140 doctors, rather than 500, significantly curtails who has access to the medical power in our society. One might decide to build enough medical schools in the United States to admit any students who, with appropriate training, will prove themselves competent physicians. We do not do so. The result is that many would-be doctors travel abroad for training and many American communities go underserved.

But let us look beyond raw self-interest. I am confident that my colleagues on the admissions committee, generously volunteering their time and energy, do so in the spirit of altruism. But doctors may serve socially beneficial purposes other than treating the sick. Some may prevent the burden of disease through public health endeavors or preventive interventions. Others may pursue careers in research to develop new cures for old scourges. How should an admissions committee balance these social goods? Who, in other words, promises to become a more “effective” doctor: the pediatrician who vaccinates children or the investigator who pioneers new vaccines? At present, candidates are generally admitted based on personal attributes and accomplishments, rather than targeted future ambitions, and the chips are allowed to fall where they may. In contrast, one could determine in advance the ideal ratio between primary care providers and clinical researchers—or, for that matter, between neurologists and urologists—and admit students accordingly. With the exception of a few targeted tracks, we do not do so.

Doctors serve another purpose: they empower communities. Having a doctor in one’s family or one’s neighborhood can generate trust between the medical provider and the patient. Having a doctor who shares a similar background can increase one’s engagement with care. In addition, having knowledgeable, high-income professionals in a community provides role models for young people and generates social (as well as economic) capital. Why are not these as legitimate concerns in assembling a medical school class as any personal accomplishments? In other words, why does the selection of future doctors occur through the lens of who will best serve patients in a clinic or hospital and not that of whose presence in a profession will create a healthier society? In truth, these goals may not be contradictory—doctors from low-income and underrepresented backgrounds may well prove as effective, or more so, than those of affluence and privilege. But the focus in admissions, at present, is definitely on individual performance, albeit via holistic assessment, not on generating social capital or community building.

These considerations still occupy a narrow range. I have written elsewhere on the criteria for admission to medical schools. Why, I have asked previously, is Newtonian physics a requirement for matriculation but not fluency in Spanish? Most physicians in the United States will be called on to provide care to Spanish-speaking patients at some point in their careers. Far fewer, if any, will be asked to shoot those patients out of cannons and measure their trajectories. I suspect many of my colleagues would benefit far more from reading Harriet Washington’s Medical Apartheid, or even Macbeth, than a full year drenched in organic chemistry. But altering a course prerequisite or adding a language requirement is the sort of tinkering that avoids the larger question: What kind of human being should be a doctor? And how can we predict in advance whether a particular applicant will prove such a person?

Twelve years ago, in an op-ed in the Providence Journal, I proposed a novel solution to a challenging medical dilemma: the shortage of kidneys for transplant. Kidney donation is a relatively low-risk surgery and people can function perfectly well with only one kidney. In fact, many people are born with a solitary organ and never realize it. My proposal was to require medical school aspirants (excepting those with medical or economic contraindications) to donate a kidney prior to applying. Doing so would serve as a test of genuine selflessness, a proxy for the sacrifice required of truly effective physicians. Needless to say, this “thought exercise” did not lead to a revolution in medical school admissions, although it did provoke hostile mail. Candidly, I did not expect a merger of admissions office and transplant programs. Rather, my point was that the admissions process might be reconceptualized to address pressing societal needs (many of which do not require invasive surgery), while simultaneously favoring altruism and sacrifice over traditional measures of merit. For instance, one could require 2 years of full-time service work in low-income communities prior to admission, possibly with exceptions for those applicants who have been raised in those very communities themselves. And then, an equal or longer number of years of commitment postresidency, ensuring improved medical care for the poor. Rather than making public service a part of the admissions process, this service might become the process.

As noted above, the number of doctors is set artificially low. In this regard, doctors do not function in a market. Rather, medical credentials are a privilege bestowed by the state such as liquor licenses, taxi medallions, and radio waves. Much as the government can compel radio stations to provide equal time to dissenting viewpoints or require civic-minded announcements, Uncle Sam clearly has the authority to compel would-be physicians to display a commitment to the commonweal before donning a white coat. A phalanx of future physicians might cure many social ills, and a willingness to devote meaningful years of one’s prime to addressing the public welfare might be a good measure for who will prove an effective physician in the future. Of course, admissions committees look favorably on public service; an applicant who has taught with Teach for America or volunteered in a rural clinic may have a leg up. In addition, some programs do tie funding to future service obligations. To the best of my knowledge, however, no United States medical school requires a systemic, multiyear commitment to service to matriculate, nor does any formal public program exist to hire, train, assign and compensate those who would enroll in such a program. Any serious philosophical examination of the process must ask the question: Why not?

But maybe even this approach is too narrow. Access to medical education is related to access to power. Any competitive, “merit”-based system is likely to reflect, and to some degree replicate existing power structures. The very choice of how to populate admissions committees renders this more likely: Most committees consist of faculty and advanced students; few solicit input from hospital employees such as housekeepers or security guards, or lay inhabitants of the communities surrounding the hospital, or former and prospective patients. A nurse’s aide or a local activist might have a very different sense of what makes an “effective” physician from a committee composed of senior teaching faculty with graduate degrees.

Many years ago, Harvard Law Professor Duncan Kennedy suggested adoption of lottery-based admissions for law schools to address these inequities. Medical schools might consider a similar approach. I am not suggesting that names be drawn out of the phone book; we do need competent doctors, as much as we need diverse ones. (Besides, a truly random approach would render admissions committees obsolete, and few of my colleagues—no matter how altruistic—would likely vote for their own elimination.) Yet rather than deciding who will make the most effective doctors, admissions committees might instead ask the question: Is this applicant likely to be able to practice medicine effectively? Make the question one of threshold, rather than ranking. That approach would produce a much larger pool of serious candidates. One might then choose randomly from the names in the hopper, producing a class of future doctors who will serve their communities well. Objections may be raised that such an approach is unfair. But unfair to whom? To the applicants who would have received a leg up in the earlier, merit-based process as a result of the unfair luck of growing up in well-heeled suburbs or attending upscale private schools? The future patients who will continue to receive effective care? I raise these questions because “fairness” in a system predicated on underlying, pre-existing structural advantages is a rather odd mechanism for assessing efficacy and desirability.

I have raised the prospect of lottery-based admissions publicly in various settings. In my experience, prospective applicants and current students both bristle at the notion. They recognize the advantages that led many of them to a top-tier college or medical school, yet also cling to the fruits of meritocracy when it affects them personally. Everyone believes in the Revolution, it seems, until they discover that they can no longer ride in the front of the airplane. Or, to offer a stark statistic from admissions, there were 542 black male matriculants to medical school in the United States in 1978 and only 515 in 2014. Many upper-middle-class white students would describe that as a pressing concern and a social injustice. I would agree that it is both. Of course, far fewer will give up their own seat in medical school to address it. Maybe a lottery-based system for admissions troubles so many of us because it brings to mind the larger social and economic lottery that creates pathways to medicine for some and not others.

I do not mean here to advocate for a particular model for medical school admissions. In fact, a hybrid model might actually prove best: admitting some applicants to generate goodwill and social capital in communities, others based on demonstrations of altruism and service, still others randomly to foster equity and to undermine the sense of entitlement that many matriculants carry at their own achievement. What I do suggest is that the manifest considerations involved in choosing future doctors are dwarfed by the latent choices, the hidden decisions that have been rendered before any admissions committee meets or any college senior sends off an application. I suspect a similar myopia occurs in the hiring and promotion of medical faculty, the awarding of grants, the design of curricula, and many of the fundamental structures of allopathic education. We, as clinicians and as humans, see what we are trained to see.

I do not propose to know the cure. I am not even sure if there is a disease. But I do think there are enough symptoms that a wholesale examination of the patient may be in order.