The Consequences of Premature Abandonment of Affirmative Action in Medical School Admissions

The Supreme Court of the United States is poised to make a momentous decision that will affect directly or indirectly virtually every feature of our society. The consequences of the Court's decision on the future of medicine could be catastrophic.

In accepting on appeal 2 cases from the University of Michigan, Gratz v Bollinger¹ and Grutter v Bollinger,² the Court has agreed to rule on the constitutionality of race-conscious decision making in higher education admissions. Although the cases before the Court speak directly only to undergraduate and law schools, respectively, there is no doubt that the rulings will govern admissions throughout all of higher education, including medical schools. Thus, the Court will either affirm or deny the right of medical schools to pursue affirmative action admission policies.

The merits of affirmative action programs in higher education in general have been widely discussed elsewhere.^{3 - 4} The case for their critical importance for medical schools can be summarized as follows: until such time that students from all racial and ethnic backgrounds emerge from the educational pipeline with an equivalent range of academic credentials, there is simply no way for medical schools to fully meet their societal obligation without using race and ethnicity as explicit factors in admissions decisions.

Medical schools have a societal obligation to select and educate the physician workforce of the future. In fulfilling that obligation, medical schools must be attentive to the future health care needs of the country and, in aggregate, produce a cadre of well-trained professionals optimally prepared to meet those needs. Medical schools would be in default on their contract with society—and could not legitimately claim public support—were they to fail to deliver on this fundamental obligation.

A key step in fulfilling this obligation is the admission process, which is the only gateway to the medical profession in this country (short of going to medical school abroad). Consequently, in selecting those who will eventually compose the physician workforce, admission committees cannot be content only with advancing the personal aspirations of a given school's most academically competitive applicants. Admission committees are not merely tally stations designed to pick the winners of the medical school sweepstakes race by determining who has amassed the most impressive academic credentials. Rather, these crucial committees have a much more difficult challenge. They are charged with constructing a balanced class that shapes the future of the medical profession by determining, in aggregate, the confines of its future membership. It is the total class balance, not merely the virtuosity of the individuals who make up the class, that defines the very objective of the admission process.

To outlaw the use of affirmative action in the admission process would be to cripple the profession's ability to achieve a vital aspect of that workforce balance—racial and ethnic diversity. The need for balancing adjustments in other aspects of the workforce (eg, primary care physicians vs orthopedic surgeons; practitioners vs investigators; cardiologists vs geriatricians) can be addressed if necessary during the process of education, training, or practice, but an inappropriate balance of racial and ethnic backgrounds among students at the initial gateway becomes a permanent fixture of that cohort and can never be ameliorated.

The practical reasons fall into 4 categories, as my colleagues and I summarized recently.⁵

1. Adequate representation among students and faculty of the diversity in our society is indispensable for quality medical education. Diverse student bodies enrich the learning experience at all levels of education.³ Diversity among medical students confers yet additional advantages, not only for the students themselves, but also for society at large. Given that our population is becoming increasingly and inexorably more multiracial and multicultural, most physicians will be called on to deal with patients from a wide variety of racial and ethnic backgrounds. The quality of health care available to future Americans, therefore, will depend not only on the scientific competence of the physician workforce but on its cultural competence as well. In acquiring the necessary skills to provide appropriate care for their diverse patients, all students, irrespective of their individual backgrounds, must gain a firm grasp on how various belief systems, cultural biases, family structures, historical realities, and a host of other culturally determined factors influence the way individuals experience illness and the way they respond to advice and treatment.

Students cannot acquire such competencies from textbooks alone. They can acquire them only by being immersed in learning environments that reflect our diverse society. That means having ongoing opportunities to interact in and out of the classroom and in various clinical settings with students from diverse racial and ethnic backgrounds and to observe how diverse mentors and fellow students deal with patients from a wide range of racial and cultural backgrounds.

2. Increasing the diversity of the physician workforce will improve access to health care for underserved populations. Abundant evidence documents the disproportionately high fraction of underrepresented minority (URM) physicians who choose to dedicate their careers to the underserved and uninsured.^{6 - 8} Given the well-documented disparities in access, as well as in health status among many minority populations and the strong preferences of many patients for racial/ethnic concordance with their physicians, increasing the proportion of minority physicians in the health care workforce constitutes a most compelling public interest.

3. Increasing the diversity of the research workforce can accelerate advances in medical and public health research. Achieving greater diversity among investigators is virtually certain to broaden the medical research agenda to engage appropriately many of the unsolved health problems plaguing all Americans. Progress on many of these problems, especially those rooted in social, cultural, and behavioral determinants, is hampered by a dearth of concerned investigators. Disparities in access and health status among minorities are prime examples of problems for which solutions appear particularly elusive.

Our country's research agenda is influenced significantly by those who choose careers in investigation. It is also true that individual investigators tend to research problems they see and feel. Since what people see as problems depends greatly on their particular cultural and ethnic filters, it follows that finding solutions to some of our country's most recalcitrant health problems, even being able to conceptualize what the real problems actually are, will require a research workforce that is much more diverse racially and ethnically than we now have. Creating that workforce begins with ensuring diversity among those admitted to our MD and PhD educational programs.

4. Diversity among managers of health care organizations makes good business sense. Businesses are increasingly coming to realize that their performance can be enhanced by having a managerial staff that reflects the racial and ethnic makeup of their clientele. As the "clients" of health care organizations become ever more diverse, a critical success factor will likely be the availability of an appropriately diverse physician workforce from which to recruit key leaders and managers who can best anticipate the needs of, and deal effectively with, individuals from a wide variety of racial and ethnic backgrounds.

In addition to these practical reasons for seeking a more diverse physician workforce, justification of a more principled nature is arguably even more persuasive. Medicine occupies a lofty status in US society and offers those who enter the profession many of the most challenging and rewarding career opportunities available anywhere. For this reason, many believe that the right public policy for a just society is to pry open the entry way to the profession for URMs who have been systematically and unfairly excluded for generations, both by design and circumstances, and who, in very substantial numbers, cannot yet claim all the advantages that permit other segments of our population much readier access to the privilege of a medical education.

Until the late 1960s, medical schools were as segregated as most other institutions in US society. Two historically black medical schools, Howard and Meharry, accounted for some three quarters of the relative handful of African Americans able to access the profession at that time. On average, the 85 to 90 "mainstream" medical schools during that era each enrolled 1 African American student every other year. Figures for other minority groups are not available but would doubtless reveal a similar pattern.

Although the civil rights movement of the late 1960s and early 1970s succeeded in outlawing the overt barriers that restricted minority enrollment in medical schools, the legacy of centuries of racial discrimination persisted. High rates of poverty, lack of access to educational opportunities, lower educational achievement among family members, as well as other factors conspired to limit significantly the number of URM students who chose to apply to medical school. (The term "underrepresented minority" has been used by the Association of American Medical Colleges [AAMC] and others since the early 1970s to refer, in aggregate, to African Americans, Mexican Americans, Native Americans, and mainland Puerto Ricans. The term recognized that these 4 minority groups had been barred historically from entering the medical profession by flagrant discriminatory practices. Although many other minority groups are also underrepresented among physicians [and in many cases increasingly so], the conventional meaning of the term is used here for consistency in tracking changes over time.) Even more problematic was (and is) the paucity of URM applicants whose academic credentials (eg, grade point averages [GPAs] and Medical College Admission Test [MCAT] scores) were comparable to those of white applicants.

Only by taking affirmative actions to open their doors of opportunity to minority students were medical schools able to begin rectifying the dearth of minority physicians available to our citizens after so many generations of exclusion. The effectiveness of the affirmative actions that were taken is evidenced by the data presented in Figure 1. The fraction of matriculants drawn from URM groups rose strikingly from less than 2% in 1964 to over 8% by 1971. Although the percentage of URMs commencing medical education remained roughly constant throughout the 1970s and 1980s, the absolute number continued to increase as the number of accredited medical schools increased to 126. It is noteworthy, however, that the diversity gap between physicians and the general population continued to widen during this period because of the steady rise in the percentage of URMs in the US population.

Medical schools reenergized their affirmative action programs in the early 1990s in response to a national summons to action occasioned, in part, by the AAMC launch of Project 3000 by 2000.⁹ Convinced that access to quality health care required a physician workforce that better reflected the advancing diversity of the general population, the AAMC urged all medical schools to redouble their efforts both to identify and accept more qualified minorities from the existing applicant pool and to partner with pipeline schools and colleges to expand the number of URMs who were qualified to enter the applicant pool. Once again, the effectiveness of affirmative actions is evident in the sizable rise in the percentage of URM matriculants between 1990 and 1995. Indeed, for the first time in history, the number of URMs entering medical school in 1995 exceeded 2000 and was on track to meet the goal of 3000 annual matriculants by the end of the century.

Additional evidence for the effectiveness of race-conscious decision making in medical school admission can be found in the sharp decline in URM matriculants in the aftermath of the anti-affirmative action ballot initiatives (California passed Proposition 209 in 1996¹⁰ and Washington passed Initiative 200 in 1998¹¹ ) and lower court decisions in the latter half of the 1990s (the US 5th Circuit court, in Hopwood v University of Texas,¹² banned the use of race-conscious admission decision in Texas, Mississippi, and Louisiana in 1996) (Figure 1). As shown in the Table 1, the decline was particularly stunning in California, Texas, Mississippi, Louisiana, and Washington, the 5 states now legally prohibited from giving any consideration to the race or ethnicity of applicants at any of their higher education institutions.

The Supreme Court's decision in Regents of the University of California v Bakke,¹³ which is the decision currently under review by the Supreme Court, has since 1978 provided the putative legal basis for the admission policies and practices of most medical schools as they have attempted to increase diversity within the ranks of physicians. Justice Powell, in the decisive opinion in the Bakke case, legitimized the use of race and ethnicity as "plus factors" in deciding which of 2 otherwise equally qualified applicants could be admitted without violating the constitutional rights of the rejected applicant. Some have likened the application of this legal authority to placing "a thumb on the scale" to tip the advantage to a minority applicant over another student with an arguably comparable record.

In actuality, medical school admission committees have been forced to place a much heavier weight on an applicant's race and ethnicity than Justice Powell envisioned to achieve even the modest narrowing in the diversity gap depicted in Figure 1. The reason is that, in reality, MCAT scores and GPAs of the typical URM applicant are distinctly lower than those of the typical white or Asian applicant. As shown in Figure 2, which depicts data for all applicants to US medical schools in 2001, very few URM applicants had combinations of MCAT scores and GPAs that were competitive with those of the vast majority of other applicants. Consequently, in considering URM applicants for acceptance, medical schools have, of necessity, had to place less weight on these easily quantifiable measures of academic achievement, while taking care to admit no one who is judged unprepared to do the rigorous work required to handle the medical school curriculum. The MCAT scores and GPAs of all the URM students accepted by US medical schools in 2001 are also depicted in Figure 2 and compared with those of all other students accepted that year. As shown, the "academic credentials" of many accepted students from URM backgrounds were outstanding, but most fell into ranges lower than was characteristic of the great majority of other accepted students.

In this context, it is critically important to recognize that admission committees scrutinize all applicants, minority and nonminority alike, for evidence of a host of attributes less easily quantifiable than GPAs and MCAT scores, but no less important than academic achievement as qualities deemed essential for future physicians. Hence, in deciding who merits acceptance, admission committees routinely look for evidence of, among other things, leadership, overcoming adversity, capacity for hard work, participation in service-oriented extracurricular activities, willingness to serve others, willingness to serve the underserved, compassion, sensitivity, empathy, and ability to communicate with others.

Medical school admission committees have become remarkably adept over the years at identifying URM applicants who, despite less impressive GPAs and MCAT scores (Figure 2), have succeeded in medical school and have gone on to productive careers. For example, 90% of the URM students who matriculated in 1990 cleared all the academic hurdles of medical school and were awarded the MD degree. (Only 7.5% withdrew or were dismissed for academic reasons.) Although the comparable graduation rate for white matriculants that year was higher at 96%, the remarkably high success rate of URM students is testament to the ability of medical schools, not only to identify qualified applicants without overreliance on standardized, numerical criteria, but also to maintain uncompromisingly high academic standards in the process.^{14 - 15}

Importantly, figures such as these also belie concerns that use of affirmative action in medical school admissions allows "unqualified" or "incompetent" individuals to become physicians. Such concerns are understandably disturbing to those unfamiliar with the requirements that must be met before anyone can become a practicing physician in this country. To the extent that the medical school admission process misjudges a student's academic potential and admits someone—irrespective of racial or ethnic background—who proves unable to manage the rigors of medical school, the multiple evaluation procedures used to monitor student progress through the curriculum serve to greatly minimize the possibility of an "unqualified" individual receiving the MD degree. In addition, the rigorous, multistep medical licensing process provides the public with yet another safeguard that only highly qualified individuals gain the privilege of practicing medicine.¹⁶

To answer this question requires agreement first on the meaning of the word "qualified" as applied to medical school admission. Clearly, exceptional intellectual abilities are required of anyone hoping to meet the demanding academic standards of the medical school curriculum and intending to fulfill the high expectations of a practicing physician or scientist. No one would argue for admitting a student to medical school who did not evidence the intelligence and academic potential necessary for completing the MD degree. Such an admission policy would not only violate our oath to patients, but it would also constitute a gross disservice to individual students. That is why admission committees choose to use MCAT scores and GPAs as elements in their overall assessment of students.

These quantitative measures are known to be statistically valid predictors of a student's academic performance in medical school, especially in the basic science portions of the curriculum, and of a student's ability to pass Step 1 of the United States Medical Licensing Examination (USMLE).^{14 - 15} Indeed, the likelihood that a student will encounter academic difficulty of any kind increases strikingly for those with MCAT scores and GPAs well below the range achieved by the vast majority of matriculants from all backgrounds.¹⁷ By the same token, no evidence exists that high GPAs or MCAT scores are guarantees of stellar academic performance or, more to the point, are indicative of the quality of health care a student eventually provides as a practitioner. Decades of experience have convinced medical educators that the attributes patients seek in their physicians extend far beyond those that can be identified by objective measures of academic qualifications. That is why admission committees seek students from all backgrounds who exhibit not only academic potential, but also a variety of other attributes.

So, to be more "qualified" than someone else for admission to medical school is not simply a matter of having higher grades or MCAT scores. In this sense, race-conscious decision making to achieve diversity does not deny admission to more qualified students. Acknowledging this reality, however, should not obscure an important fact. Compared with other students, a higher percentage of URM students prove unable to make the grade academically and fail to graduate; in addition, those who do graduate take somewhat longer, on average, to do so.¹⁵ Put another way, given the inherent limitations of the admission process, and despite the best efforts of admission committees, the price of pursuing the important goal of narrowing the diversity gap in medicine is to accept that a small portion of the limited capacity available in medical schools will be lost to potentially more qualified applicants. But in the judgment of medical educators, the benefits of constructing a balanced class far outweigh this cost.

As an instructive analogy, consider the challenge of putting together a first-rate symphony orchestra. To achieve a properly balanced sound, a conductor may well have to trade off more individual virtuosity in one section of the orchestra in favor of that available in the market of qualified instrumentalists needed in another section. Whether an analogous trade-off to achieve racial and ethnic diversity in medicine is in the public interest is precisely the decision now facing the Supreme Court.

Opponents of affirmative action argue that applicants for admission to medical school should be selected without regard to—or even knowledge of—their race or ethnicity, ie, attending only to race-blind "qualifications." Consider what would happen to minority medical school enrollment if URM students were required to have the same level of MCAT scores and grades as white students. For purposes of this exercise, assume that the typical URM applicant, in comparison with the typical white applicant, exhibits no greater degree of altruism, compassion, industriousness, leadership qualities, or any other sought-after nonacademic attribute. Using data from all applicants to the 119 nonminority medical schools (ie, excluding the 3 Puerto Rican and 3 historically black medical schools) in 2001, the number of students who would have been accepted under these hypothetical conditions can be calculated and compared with the actual number admitted (the algorithm used to make this calculation was developed to analyze the consequences of abandoning race-conscious decision making in law school admissions).¹⁸ In that year, 1697 URM students were offered admission. Under the "no affirmative action" scenario, only 513 would have made the cut, some 70% fewer! Under this scenario, barely 3% of all medical students would be URMs, approximating the circumstance prevailing in the early 1960s before affirmative actions were taken to overcome the legacy of a frankly discriminatory past.

The answer to this question is not precisely known, but educators speculate that a variety of interrelated factors is responsible. Among those factors are fewer well-trained teachers and less well-equipped schools available for minority students; stereotypic lower expectations of teachers for minority students; stereotypic lower expectations of minority students for themselves; lower levels of academic achievement among parents of URM students; a paucity of academically accomplished role models to emulate; and lack of domestic and social support for the studious behavior seemingly valued by many whites and Asians.

Lower levels of income within many minority households is another likely contributing factor, but it is important to note that, on average, minority students from middle-class families also score lower on standardized tests and present lower GPAs in comparison to whites and Asians.^{19 - 20} For example, as depicted in Figure 3, the average total MCAT score for URM students from families with incomes greater than $80 000 is lower than that of non-URM students whose families earn less than $30 000. Such disparities in test scores are sobering reminders that powerful social, economic, cultural, and educational forces still operate along racial and ethnic lines in this country. The precise mechanisms may not be clear, but the resulting educational inequalities that accumulate disproportionately for minority children from kindergarten through high school and beyond are all too evident.

Seemingly successful adaptations to the legal restrictions on the use of affirmative action that exist in some jurisdictions might be interpreted as cause for optimism that race-conscious decision making in medical school admissions may not be necessary to achieve adequate diversity. Unfortunately, such optimism is not warranted. Approaches such as "percentage plans," in which a fixed percentage of a high school's top graduates is guaranteed admission to a state institution of higher education, obviously cannot work for medical school or for any other postbaccalaureate program.^{21 - 22} Moreover, those who challenge the consitutionality of affirmative action are already challenging these plans on the same grounds.²³

What about other surrogate markers of diversity, such as living in a low-income ZIP code, coming from a disadvantaged family background, having overcome adversity, or expressing a willingness to serve the underserved?^{24 - 26} Although superficially attractive, these alternatives could be characterized as patently transparent contortions intended to achieve the same outcome—more racial and ethnic diversity—as the straightforward, honest application of affirmative action. This kind of semantic gymnastics is surely an unwholesome strategy for the medical profession to pursue. Moreover, given that many more white applicants than URM applicants can reasonbly claim to possess these surrogate markers, using such proxies would not achieve the desired diversity in any event. Finally, as with percentage plans, legal challenges to their use for this purpose are inevitable.²⁷

The answer depends entirely on how quickly our country can remedy the unequal educational opportunities available to many URM children, and on how quickly the United States can rid itself of the other more subtle forms of discrimination that conspire to thwart the academic achievements—and aspirations—even of those who have access to a quality education. For those underrepresented in medicine, the sizable gap that still exists between the legal protections of individual rights and the denial of those rights in everyday practice is a sure sign that much more time is needed before abandonment of affirmative action in medical education can be righteously defended.

In the near term, there is simply no alternative to the use of race-conscious decision making in medical school admission if our society is to have the benefit of a reasonably diverse physician workforce. No amount of rhetoric can avoid the demographic reality of a burgeoning URM population that, for a variety of reasons, has on average significantly lower levels of academic achievement. If we are precluded from using race-conscious decision making in medical school admissions, the nation must accept the reality of still more decades in which the physician workforce is incapable of providing an otherwise achievable quality of health care for large segments of the American people.

Some opponents of affirmative action argue that allowing minority students with lower academic credentials than required of whites to gain admission to medical schools is itself a form of racism, contributing to defeatism among the very students we are trying to help.²⁸ They conjecture that only by maintaining equivalent objective standards for everyone will minority students, over time, be sufficiently motivated to excel on their own. To the extent that this line of reasoning recognizes the responsibility of parents and advocates of minority students, and of the students themselves, to reduce the barriers to academic achievement, no one would argue. To the extent that it envisions a sufficiently rapid reversal of deeply rooted societal and cultural norms to repair the unhealthy diversity gap in medicine, it is at best unrealistic.

Under our constitutional system, it is the courts—and ultimately the Supreme Court—that must judge when de jure constitutional protections can be depended on to outweigh de facto abridgement of civil rights. The courts have the responsibility for deciding when race can be used in conferring or denying a public benefit, and when it cannot. It would be ironic if the legal provisions designed to protect minorities from oppression become, in the courts' hands, the very instruments for denying society the needed participation of qualified minorities in medicine. Prematurely foreclosing opportunities only recently made available to URMs in the name of protecting individual rights would be to deny future generations of Americans the right to better health.

We are all eager to see the day when race no longer matters in our country—when race ceases to be a relevant, much less a compelling credential for access to anything. Unfortunately, for access to the medical profession, that day has not yet come. For the foreseeable future, the use of race-conscious decision making in medical school admissions is the only way medicine can meet its obligations to everyone in our society. At this point in our history, only by reaching out affirmatively to URM students who have not acquired competitive academic credentials, but who possess all of the qualities of mind and spirit required to succeed as medical students, can our profession achieve the diversity needed to (1) provide a quality education for all students; (2) supply a balanced cohort of practitioners, investigators, and health care managers for an increasingly diverse population; and (3) help fulfill our country's ideals of fairness, justice, and equity.