Shortly before the U.S. Supreme Court affirmed the legality of using narrowly tailored, race-conscious admission policies in higher education (Gratz v. Bollinger1 and Grutter v. Bollinger2), the president of the Association of American Medical Colleges (AAMC) at that time, Jordan J. Cohen, MD, argued that “to outlaw the use of affirmative action in the admissions process would cripple the profession’s ability to achieve racial and ethnic diversity.”3 The responses to Cohen’s communication illustrate the tension within medical education regarding the goal of achieving diversity within the physician workforce. Some clearly support the continued use of race-conscious admission policies, whereas others vigorously oppose it. For example, one respondent characterized Cohen’s appeal as “a lobbying effort to persuade the reader and the government to accept quotas as a basis for admission to medical school,” and that “to use quotas as a factor in medical school admissions is to accept this lowered standard and should not be tolerated by the public, by the profession, or by the courts.”4 Another responded that “the favoring of a small number of races of necessity entails discrimination against all others.”5
Absent from this debate were suggestions for achieving racial and ethnic diversity within medical education that can be described as “race neutral.” If, as is generally agreed, the nation needs a culturally competent and diverse physician workforce, is there any way to achieve that goal without intentional efforts to identify and support qualified applicants to medical school who reflect the nation’s diversity?
The Supreme Court’s decision in Grutter v. Bollinger compels schools—including medical schools—to explore race-neutral approaches by stipulating that institutions that use or are considering using race-conscious admission policies must make a “serious, good faith consideration of workable race-neutral alternatives that will achieve the diversity the university seeks.”2(p339) Although institutions need not experiment with or implement race-neutral alternatives before implementing race-conscious admission policies, they are now obliged to consider them as they work to promote diversity within their institutions. Justice Sandra Day O’Connor, writing for the majority, observed, “We expect that 25 years from now, the use of racial preferences will no longer be necessary to further the interest approved today.”2(p343) This expectation has been widely interpreted as a sunset provision—that the affirmation of race-conscious admission programs provided by the Court is subject to review and curtailment. But, except for a flurry of reaction to the Court’s decision, there remains too little literature on the race-neutral options to adequately inform U.S. medical schools admission committees.
To date, large-scale experimentation with race-neutral alternatives to race-conscious admission approaches has followed state-level mandates (legislation, referenda, or executive or court order) outlawing race-conscious admission approaches—a context that has limited exploration of workable interactions between race-conscious and race-neutral approaches. And, overall, research into the effectiveness of such approaches is limited. What promise do current attempts at race-neutral alternatives hold for medical schools seeking a diverse and qualified student body? What more must medical schools do or consider to ensure their admission policies meet the directive contained in the Court’s decision?
There is a limited window of opportunity in which medical education can find the answers to these crucial questions. Medical schools—as they continue to vigorously employ race-conscious admission policies to achieve the benefits of diversity—must take advantage of their compliance with the Court’s decision to assess whether race-neutral approaches contribute to the diversity of their medical school classes. The purpose of this article is to inform the medical education community about race-neutral alternatives, review challenges and opportunities for implementing them, and encourage further research that will inform medical school admission policies.
Defining Race-Neutral Alternatives
Race-neutral admissions emerged in 1995 as a significant policy issue when the Regents of the University of California (UC) adopted Resolution SP-1, prohibiting the use of race, religion, sex, color, ethnicity, or national origin in the UC admission process.6 Supporters of SP-1, most notably Ward Connerly, declared race-conscious admission to be unfair, ineffective, and harmful policy.7
A focus on race-neutral alternatives in other states soon followed:
- The Hopwood v. Texas decision (1996)8 involved a ban on the consideration of race in public universities in Louisiana, Mississippi, and Texas, where the fifth Circuit U.S. Court of Appeals has jurisdiction. Texas HB 5889 established the state’s percentage-plan admission strategy in undergraduate higher education. In 2001, Texas HB 164110 articulated race-neutral factors permitted for use in graduate and professional school admission and scholarship decisions.
- Proposition 209 (1996),11 a California state ballot initiative, prohibited preferences based on race and sex in public contracting, public employment, and public education. At the same time, the Regents of the University of California (UC) formalized a “two-tiered selection process” that allowed a proportion of an admitted class to be considered on factors “supplemental” to academic criteria. In 2001, with Proposition 209 still in effect, the UC Regents rescinded SP-1 and replaced it with the “comprehensive review” admission process that considers “a single, comprehensive set of selection criteria.”12
- Initiative-200, or I-200,13 was approved by the voters of Washington State (1998) and banned consideration of race and sex for state government hiring and recruitment or admission to the state’s universities.
- “One Florida” (1999),14 an initiative of the state’s governor to replace race-conscious programs in education and business, included the Talented 20 Program, which guaranteed college admission to any high school graduate in the top 20% of his or her high school class.
In this article, we use race-neutral alternatives to refer to policies or programs that consider any combination of admission criteria with the exception of race or ethnicity. We acknowledge that some race-neutral approaches retain, as their motivation, race-conscious outcomes, blurring their categorization.15 It is beyond the scope (and usefulness) of this article, however, to sort out such subtexts. We discuss four categories of race-neutral alternatives:
- The use of traditional measures of academic performance such as standardized achievement measures (e.g., MCAT scores) and “unstandardized,” or dependent, measures such as undergraduate grade-point averages (GPAs). Such measures are the most common race-neutral alternatives presented in the literature.
- The use of socioeconomically based measures, including, at a minimum, data about the personal or family income of the applicant. Definitions of socioeconomic status have been expanded in various ways. In one example, described by Kahlenberg, a numeric score is derived based on sociological data, such as wealth, schooling opportunities, neighborhood influences, and family structure.16
- The use of adversity indices designed to assess the effects of prior and/or current disadvantage or discrimination on applicants. Some examples of adverse circumstances are experiences with trauma within the family, linguistic challenges, and personal or environmental challenges that affected access to educational resources.17
- The use of community outreach admission strategies, where an institution’s goal of achieving diversity in admissions, as articulated in its mission statement and throughout its strategic planning, is linked to its expansion of service to its community or region.
We do not discuss a fifth category of race-neutral alternatives commonly referred to as percentage plans. These plans, which guarantee admission to college for a designated percentage of the top graduates of high schools in a state, have been implemented for undergraduate admissions in Texas, Florida, and California. They are clearly unworkable for medical schools,18 and the U.S. Supreme Court recognized the implausibility of their use by graduate and professional schools.2(p340)
Race-Neutral Alternatives in the Medical School Context
For medical schools, the key question is whether race-neutral alternatives have been shown to be effective for admitting a diverse and qualified class of medical students. Because scholarship on race-neutral admission approaches at medical schools is limited, we reviewed a broad range of resources (e.g., scholarship, descriptions, proposals, etc.) to capture the breadth of options. In some cases we have included strategies used in undergraduate settings. Traditional measures of performance and socioeconomically based approaches have garnered the most attention in the literature. Adversity indices and community-outreach-based approaches are described but have yet to be carefully evaluated. These four approaches are defined and discussed in the rest of this article; selected resources from the literature and the opportunities they describe are presented in Table 1.
Traditional measures of academic performance
Within the individualized, holistic evaluation of medical school applicants, medical schools depend on traditional measures of academic performance, specifically, undergraduate GPAs and MCAT scores. These measures are statistically valid predictors of student performance in the first three years of medical school and on Steps 1 to 3 of the United States Medical Licensing Examination (USMLE).19, 20
Although some traditional measures of academic performance are standardized (and therefore considered race neutral), traditional measures of academic performance are nonetheless inflected by race and ethnicity and other variables like socioeconomic status that may neutralize their effectiveness as race-neutral alternatives. For example, minority students tend to have lower undergraduate GPAs and MCAT scores than do their majority counterparts. The causes of these differences are myriad. Sturm and Guinier have argued that aptitude tests are poor measures for identifying students who can perform best in college; they point to research that shows a “tenuous connection between test scores and successful performance.”21 Other literature, such as Steele and Aronson’s work on stereotype threat, suggests performance on certain traditional measures may not solely be tied to academic preparation. Stereotype threat refers to the risk of being viewed through the lens of a negative stereotype, or the fear of doing something that would inadvertently confirm that stereotype.22 Steele and Aronson’s research has demonstrated that this process can lower scores on tests of academic ability for African American students, and subsequent studies demonstrate the effect on other groups.23 Thus, stereotype threat and other factors may undermine the race neutrality of tests commonly used as traditional measures of academic performance.
Experts have long cautioned against relying on traditional measures beyond the context of holistic review and thereby undermining the value of other crucial measures of ability and experience: “Standardized tests have been used in this affirmative action debate in a manner that goes far beyond the purpose for which they were created. Standardized tests were designed as a mechanism to provide an independent assessment of certain academic skills of students… they were never intended to be the central metric of merit, or of academic potential.”24 In its amici curiae brief on behalf of the University of Michigan Law School, the AAMC argued that when such a limited assessment of medical school applicants is given too much emphasis, it would significantly reduce the presence of African Americans, Hispanics, and Native Americans at U.S. medical schools.25
Dependence on traditional measures of academic performance has also come under scrutiny by those who consider its practical impact. Differences in medical school graduation rates of majority and nonmajority students are narrowing, which emphasizes that students with a broad range of standardized-test scores, rather than the narrower range defined by traditional measures, have the ability to achieve in medical school.26 Others argue that traditional measures of performance are being outpaced by the changes in the outcomes of medical education itself. Klasko poses the question:
Are GPAs and MCATs still the “gold standard” for predicting success in a future where the differential diagnosis is as close as your palm (electronic or anatomic)? Simply put: Would a candidate with a 3.2 GPA and 26 in her MCATs, with superb communications skills, coordination proven on her simulation tests, and a finely tuned eye as evidenced by her fine-arts college minor be a better practitioner in the information world of 2020 than a candidate selected by traditional means?27
These challenges demonstrate that great care should be taken when using traditional performance measures in admissions decisions.
Physicians must be capable of integrating their intellectual, social, and interpersonal skills effectively and appropriately for each patient encounter. Tools used in admissions decision making need to confirm a candidate’s potential for fulfilling this responsibility in a broad sense. Because reliance solely on undergraduate GPAs and MCAT scores in medical school admissions is unlikely and because the weight given to other measures in predicting performance varies from medical school to medical school, the important question is how best to employ these measures within a holistic review of each candidate.
One possibility is to use thresholds for GPAs and MCAT scores to relieve medical schools’ reliance on traditional measures of academic performance. Cohen suggests, “Rather than giving more weight to higher scores, why doesn’t each school decide for itself, from data available from its previous students, what level of GPA and MCAT performance is sufficient for predicting success in clearing the high academic hurdles of medical school—and leave it at that.”28 Such thresholds would likely be much lower than schools’ medians for accepted applicants, and other factors such as noncognitive measures of success could be used in final admission decisions for each applicant.29–31
The difficulty in establishing such thresholds, however, stems from variability in predictors, outcomes, and statistical tests used to derive the cut. Albanese, Farrell, and Dottl assessed various statistical approaches for determining thresholds by looking at GPAs, MCAT scores, and USMLE Step 1 pass rates for classes entering a large midwestern medical school from 1992 through 1998. They found that “rational and defensible” threshold measures could be set from GPAs and MCAT scores that are highly predictive of USMLE Step 1 pass rates.32(p158) However, they also note that generalizability and predictions of performance (especially clinical performance) and achievement of diversity goals remain untested.32(p158)
Investigators have also looked at methods for treating traditional performance measure data differently to promote diversity. For example, Goggin’s merit-aware index measures “the extent to which a student’s achievement [on traditional performance measures] exceed[s] what could have been reasonably expected given his or her academic background.”33(p9) Although untested in medical education, St. John and colleagues demonstrated this technique in undergraduate case studies and found that “using an index based totally on merit increased diversity in the screened-applicant pool.”34 At medical schools, Kreiter et al propose the use of “constrained optimization decision analysis” to “maximize powerful predictor variables” such as MCAT test scores and to attain diversity goals related to medical class composition.35(p118) Essentially, the analysis solves complex functions that simultaneously analyze myriad combinations of applicant variables according to constraints or composition goals. Kreiter et al note that, unlike statistical methods such as weighting methods, constrained optimization does not create a composite score for an applicant. The authors posit that its use would likely satisfy strict scrutiny* requirements.35(p118) In a controlled application of the process (using actual, but limited data), the authors modeled the selection of a racially diverse medical school class. However, as with the research on the merit-aware model, Kreiter et al acknowledge the need for considerably more research before such a method can be widely implemented.
Socioeconomically based admissions
In a 2003 report,36 Cooper addressed the socioeconomic barriers to educational success and their impact on medical school applications. Citing a statistic that 30% of black and Hispanic families exist on less than 1.25 times the income defined as the U.S. poverty level (versus 11% and 14% for whites and Asians, respectively) and the compounding effect of single-mother families among blacks and Hispanics, Cooper concludes, “It is not difficult to overlay these statistics with those of reading achievement in the third grade and to extrapolate these to the likelihood of advance placement courses in 12th grade, direct entry into a four-year college, and, ultimately, attainment of a bachelor’s degree, the gates through which students must pass en route to medical school.”36(pp873–874) It is important, therefore, to consider whether race-neutral alternatives that address the socioeconomic barriers to medical education can help build a racially and ethnically diverse medical school class.
For those who believe race-conscious admission approaches exclude members of certain underrepresented and disadvantaged populations, especially the poor, socioeconomically based admission policies are preferable for promoting equitable social change. Kahlenberg contends that race-conscious affirmative action will only create “a self-perpetuating black elite along with a white one,” whereas a socioeconomically based model can “provide a system of genuine equality of opportunity.”37 He maintains “it is possible to devise a series of fairly objective and verifiable factors that measure the degree to which a teenager’s true potential has been hidden,” and he offers some examples of where such data could be obtained.37
A major concern, however, is the use of socioeconomically based alternatives as a proxy for race and ethnicity in admission decisions. The two variables are not interchangeable. Although persons from racial and ethnic minority groups are overrepresented in the lower socioeconomic classes in the United States, the fact that there are so many more whites than blacks (for example) greatly affects the racial and ethnic representations in cohorts based on socioeconomic status. And, when unmediated consideration of traditional measures remains part of the admission process, socioeconomic status will not redress gaps in racial and ethnic underrepresentation on college campuses. Bowen and Bok examined the connection between socioeconomic status, race and ethnicity, and students’ test scores, and concluded that “[while] it is true that black students are much more likely than white students to come from families of low socioeconomic status, there are almost six times as many white students as black students who both come from low [socioeconomic status] families and have test scores that are above the threshold for gaining admission to an academically selective college or university.”38 In an unpublished study, Colburn and Jolly reported preliminary findings using AAMC student data that suggested the use of socioeconomic class or status without consideration of race and ethnicity was not effective in sustaining momentum for increasing diversity in medical school classes. In 1997, Cross and Slater predicted that “black enrollments at the nation’s highest-ranked medical schools would drop by at least 75 percent” in a strictly race-neutral admission environment.39 Although these predictions proved high, recouping the loss of diversity in medical school classes in states where race-conscious efforts have been discontinued has been painfully slow, and the impact has affected the diversity profile of the nation’s physicians, as Grumbach et al illustrate:
In California, the percentage of matriculants who were URMs [underrepresented minority students] decreased from a high of 21.9% in 1992 to 15.6% in 2000. In Texas, URMs dropped from 21.0% of matriculants to 15.6% in 2000…. These numbers are especially disturbing because of the high proportion of minorities in California and Texas. To reach population parity, California would need 40% of matriculants to be URMs, and Texas would need 43%…. Thus, most of the overall decline in URM matriculation in medical schools in the U.S. is accounted for by the decreases in California and Texas. In 1995, California and Texas were educating 18.0% of all URMs matriculating in allopathic medical schools in the U.S. By 2000, the figure was 15.5%.40
Finally, the effects of a socioeconomically based admission strategy remain unknown because a solid measure of an applicant’s socioeconomic status has not been established by medical schools.41 Without such a measure, it remains difficult to predict what the effect of socioeconomically based affirmative action would be on racial and ethnic diversity in medical schools.
Research indicates that socioeconomically based alternatives may work best when they complement race-conscious efforts. For example, using data from the National Center for Education Statistics’ National Education Longitudinal Study of 1988 and High School and Beyond, Carnevale and Rose analyzed simulations of alternative admission strategies and the role that socioeconomic status and race and ethnicity (among other factors) play in college admission. They found that “there is a need for much more vigorous use of economic affirmative action” and that “a credible procedure that can reproduce the level of diversity that exists in society today without purposely singling out African Americans and Hispanics at some point in the selection process has yet to be found.”42 They conclude, “Politically speaking, the best way to pursue economic affirmative action is as a supplement to, rather than a replacement for, racial affirmative action.”42
One promising approach is the capacity of U.S. medical schools to engage within their communities in efforts to widen and strengthen the educational pipeline, because the negative impact of socioeconomic disadvantage on educational development can be measured from the very start of a child’s public education. Cooper observed that family income, parental education, and similar variables can be responsible for differences in educational progress in ways that are similar to those of race and ethnicity, and he recommends large-scale changes in K–12 education to remedy the situation, so that more individuals from low-income and minority backgrounds will graduate from high school adequately prepared for college.36
Many academic health centers already participate actively in such partnerships.43 Patterson and Carline have detailed the multiple benefits of such partnerships, including the potential for extended, tailored, and comprehensive interventions that target specific local populations.44,45 The commitment to such programs, both institutional and financial, can waver, however, because to date there is no clear measure with which to evaluate success. For example, Patterson and Carline examined 12 programs, nine of which had program evaluations. None, however, was able to assess “long-range outcomes of students [sic] career choice.”44(p19) Of concern, too, is whether such long-term partnerships should concentrate broadly on academic enrichment or specifically on increasing the number of health professions applicants from among their graduates.45(p3) Recently, Grumbach and Chen have found that the UC postbaccalaureate premedical program, which admits California residents from disadvantaged backgrounds, has achieved success in increasing medical school enrollments of minority and disadvantaged students.46 Their report demonstrates that a race-neutral pipeline initiative can affect enrollments by underrepresented minorities. Their findings also strengthen the argument that academic preparation is an essential component of application to medical school and, as such, should remain the focus of such programs.
In 2004, Harvard College announced an initiative designed expressly to increase socioeconomic diversity among its students. The plan includes an intensive recruiting effort to encourage talented students from families of low and moderate income to attend Harvard College.47 It also includes significant increases in financial aid so that parents with incomes of less than $40,000 will no longer be expected to contribute to the cost of their children’s attendance at Harvard. The Harvard plan includes increased efforts to recruit students from a wide range of backgrounds, regardless of income. The plan reemphasizes an admission process that takes note of applicants who have remarkable accomplishments despite limited resources at home or in their local schools and communities. Harvard Medical School has also established an extensive web of intensive summer pipeline programming for academically talented K–12 students in Boston and Cambridge designed to target financially disadvantaged students.48
The use of adversity indices is an attempt to broaden the criteria used in the assessment of students’ achievement. In addition to socioeconomic status, adversity indices allow for the consideration of students’ achievements in overcoming educational and personal challenges in admission decisions or in the awarding of scholarships as part of recruitment efforts. Many consider students’ merit in such achievements in much the same way that Goggin defines merit achievement beyond expectation: “… there has been a deeply held belief, shared by virtually all Americans, that hard work, citizenship, and extraordinary effort—that is, meritorious behavior, especially in the face of grim circumstance—occurs also among those who have experienced over a century of discrimination.”33(pp8–9) Adversity indices allow for a review of achievement within the context of the circumstances of that achievement.
Most examples of adversity indices pertain to undergraduate institutions, and their descriptions do not reveal the extent to which they maintain or increase diversity. Interestingly, adversity indices are being used at public universities in those states where race-conscious admission practices have been outlawed (e.g., California, Texas, Washington) or where judicial scrutiny of admission practices have required changes (e.g., Michigan).
The UC system employs its “Comprehensive Review” approach that weighs adversity factors in addition to grades and test scores. UC’s approach illustrates the breadth of considerations that can be made (when using adversity indices) in its assessment of “academic accomplishments in light of an applicant’s life experiences and special circumstances, such as disabilities, low family income, first generation to attend college, need to work, disadvantaged social or educational environment, difficult personal and family situations or circumstances, and refugee or veteran status.”12 While UC’s Comprehensive Review has been in effect, both student-run, race-conscious outreach efforts and a UC-run outreach initiative have tried to bolster the diversity of its applicant pool.49 Amid these recruitment, financial aid, and retention programs, it is difficult to evaluate the success of Comprehensive Review as a truly race-neutral program.
Hanson and Burt describe how the University of Texas at Austin redesigned its scholarship criteria after the Hopwood decision with the goal of “embrac[ing] the University’s commitment to a diverse campus” without using racial or ethnic background as a selection criterion.50 The school’s “Adversity Index” considers “economic indicators of the students’ parents specifically and their high school in general, their parents’ educational level, and the percentage of students from their high school that applied to the University of Texas at Austin in the past. The Adversity Index also compares students academically with peers from their own high school.”51
After the 2003 Grutter and Gratz decisions, the University of Michigan revised its undergraduate admission procedures (then based on a point system) to comply with the Court’s rulings. To promote diversity, the undergraduate school implemented the individualized holistic review of each applicant. Although traditional academic factors—such as high school grades, the quality of the high school curriculum, the competitiveness of the high school, and scores on standardized tests—continue to be the most important criteria in considering an applicant for admission, the university also takes into account a range of additional factors.52 These include race and ethnicity, socioeconomic and educational background, geographic considerations, and gender, among others.53 Such factors help demonstrate the ways each applicant might contribute to the overall diversity of the student body, but no single factor, including race or ethnicity, has a fixed weight in the admission process. Although this process is still race conscious, each factor—including race—is considered flexibly in the context of an applicant’s entire file.
Roithmayr describes an experimental adversity measure called “direct measures” that allows for preferences in admission decisions if an “application demonstrated that she had suffered from the effects of racial discrimination, that she could contribute an important and under represented viewpoint to the classroom on issues of social and racial justice, and/or that she likely would provide resources to underserved communities.”54(p6) Defined in this way, direct measures classify applicants “directly on the basis of their experience of racial discrimination” rather than on the basis of their race or ethnicity.54(p6) In theory, direct measures admission may allow a medical school to “directly address important institutional goals, without triggering the Court’s apparently fatal disapproval of racial classifications [in 2001] as the means to achieve those goals.”54(p7) There is no literature to indicate that this approach has been implemented or measured for effectiveness, either in an undergraduate or medical school setting.
At medical schools where adversity and disadvantage, in addition to race and ethnicity, are taken into account during holistic, individualized reviews of applicants, the admission process can be intentionally connected to a school’s mission to serve its community or region. Well-crafted mission statements are crucial to pursuing both race-conscious and race-neutral admission policies because mission statements that describe a school’s educational and diversity goals help satisfy the requirements of strict scrutiny. According to Coleman and Palmer, “With respect to diversity goals, in particular, there must be clarity regarding what kind of student body the institution wants to attract (and why) and how the institution conceptualizes (or defines) its goals and objectives.”15 For example, the Howard University College of Medicine’s mission statement explains that the school “recognizes as its primary obligation the provision of a quality medical education for any student, irrespective of race, creed, sex, or national origin, but with emphasis upon the provision of educational opportunities for those socially, economically and culturally disadvantaged students, particularly black students, who may not otherwise have an opportunity to acquire an education.”55(p150) Likewise, the University of Texas Medical School at Houston states its mission to be “to provide the highest quality of education and training of future physicians for the state of Texas, in harmony with the state’s diverse population, and to conduct the highest caliber of research in the biomedical and health sciences.”55(p340) Within this context, admission efforts based in community outreach can help a medical school to address diversity in admissions and meet local or regional health needs.
Historically, community-based outreach to bolster minority admissions to medical school involved race-conscious programs focusing on the public education pipeline. Most notably, the Health Professions Partnership Initiative (HPPI), directed by the AAMC and funded by the Robert Wood Johnson and W.K. Kellogg Foundations, supported medical school–public school partnerships in 16 communities from 1996 to 2005. Although many succeeded,56 analysis of the partnerships revealed important organizational challenges. Carline and Paterson found that “discontinuities between higher education and the public schools can lead to problems in setting up partnerships and can engender continuing distrust; if these difficulties are not overcome, failure ultimately occurs.”57 Limited evidence of these programs’ effectiveness in increasing diversity in medical school applications and admissions and recent cuts in federal funding for Title VII of the Public Health Service Act have made it difficult to adequately secure resources for these complex programs.58
Increasingly, medical schools have begun to leverage their community and regional presence to align their student recruitment diversity goals with their mission to meet local health care needs. Acosta and Olsen describe the University of Washington’s efforts within tribal communities to address regional health needs and workforce diversity in Washington, Wyoming, Alaska, Montana, and Idaho (WWAMI). Their initiative takes advantage of “the uniqueness of the WWAMI program, the mission and commitment of the medical school, and the demographics of the region.”59 Acosta and Olsen report that 477 American Indian and Native Alaskan students have participated in the multifaceted outreach, academic enrichment, and student support programs, and 102 of these students have matriculated at U.S. medical schools (34 at the University of Washington School of Medicine).59(p868)
Thomson et al describe a similar race-neutral program to attract medical school students from medically underserved areas to programs that address the health care needs of their own communities60 The Premedical Honors College is an eight-year, BS–MD program created by Baylor College of Medicine and the University of Texas–Pan American to “increase the number of physicians addressing health care needs of underserved populations in Texas.”60(p454) The program targets south Texas, a predominately Hispanic, medically underserved 13-county area. The program has developed a successful cadre of physicians who themselves are predominately Hispanic and who are serving minority and medically underserved populations in Texas and the nation.
In a similar vein, Blumenthal61 argues that new medical schools should be placed in areas that are currently undersupplied with physician training resources. Pointing out that some areas of the country with large areas of underserved and often minority populations are also in states that are undersupplied with physician-training resources, such as California and Texas, building new medical schools in those regions could develop physicians that are more likely to address the health care needs of those local communities.61
Of the four categories of race-neutral alternatives reviewed in this article, none is currently a superior alternative to race-conscious admission programs. It is clear that an overreliance on traditional measures, such as MCAT scores, and considerations of socioeconomic status that do not include considerations of a candidate’s race or ethnicity, will have a negative impact on current levels of diversity in medical schools. Socioeconomically based admissions hold promise, but they require further study to determine their actual effect when combined with race-conscious admission programs. Adversity indices, although innovative, remain unproven in their ability to increase diversity. And community-outreach-based strategies are promising but understudied, even though they are necessary to protect schools with race-conscious admission strategies. Thus, despite the critical ramifications of underrepresentation of racial and ethnic minority students in the medical education pipeline, and the looming potential of a sunset provision for race-conscious admission programs, there is a great deal of work to be done to adequately understand and create best practices for building a diverse and qualified medical student body.
The literature highlights three areas for further study. First, there is a significant lack of literature regarding the effectiveness of race-neutral alternatives used in combination with race-conscious alternatives. Confounding factors such as parallel race-conscious outreach and recruitment efforts, refinements in financial aid distribution, etc., obfuscate the role of race-neutral alternatives in advancing the diversity of medical school classes. Research must be designed to account for the discrete effects of all approaches.
Second, research that builds on our current understanding of the benefits of diversity needs to be expanded to lend an evidentiary basis for tailoring race-neutral alternatives to race-conscious admission programs. For example, accepting that diversity contributes to the educational environment, what constitutes the “critical mass”† of diversity capable of producing those benefits? More specifically, what constitutes critical mass in light of any individual school’s objectives for admission as they relate to the school’s institutional mission? This question of critical mass is important for two reasons. At its most crude (and incorrect), resistance to the concept of critical mass is often based on its comparison with quotas. Translation of the idea of critical mass into practice remains elusive. Research data that demonstrate how and why critical mass is distinguished from the label of quotas and how to satisfy the Court’s direction that critical mass be “defined by reference to the educational benefits that diversity is designed to produce”2(p330) are essential. Further, the Court requires institutions to conduct periodic reviews to assess the continuing necessity of race-conscious admission programs.15 Applying an evidence-based definition of critical mass will inform these efforts.
And third, what potential exists for new ideas (e.g., geographical targeting) to have an impact on the ability of racial and ethnic minority students to more readily access medical education, and is it possible these initiatives can function as “race-neutral” alternatives? In addition to research, there is a need for literature that describes the implementation of these innovations and their relationship to medical school admissions.
Our review of the literature has limitations. First, our review was informal and collected relevant and representative literature, rather than being a structured review. Second, the body of literature on race-neutral alternatives to race-conscious admission remains small. In particular, literature that pertains directly to admission policies at U.S. medical schools is scant. It is difficult to design research on admission approaches that is capable of teasing out confounding influences; also, resources for research remain scarce. Third, admission policies across the country continually shift and are frequently affected by legislation, court decisions, referenda, or institutional mandate. It would be difficult to determine whether changes in admission policies more frequently follow the sway of politics or the development of empirical evidence. Despite these limitations, we believe the literature is essential for informing the medical education community and moving forward the discussion about future innovation and assessment. We encourage schools to share information about what they are doing.
Diversity in the medical profession is recognized implicitly and explicitly as a public good and an integral component of quality of care.62 With the U.S. Supreme Court’s rulings in Gratz and Grutter, the ability of medical schools to use race-conscious admission policies for promoting diversity has been affirmed and preserved. At the same time, however, medical schools that employ race-conscious admission programs must also consider race-neutral alternatives. Additional research is needed at both the institutional and national levels to assess how race-neutral strategies fare over the long term, and more research is needed to determine which race-neutral strategies, when employed in tandem with existing race-attentive approaches, bolster racial and ethnic diversity. Such research may broaden our understanding of diversity beyond race and ethnicity and more fully and beneficially capture the complex interactions of race, ethnicity, and class in U.S. society. Nonetheless, of chief concern is any strategy’s effectiveness at producing the benefits of diversity within medical education, their compatibility with existing institutional missions, and their potential to achieve a more diverse health care workforce.
The authors thank Arthur L. Coleman, JD, and Robert Sabalis, PhD, for their thoughtful comments and helpful guidance.
1 Gratz v Bollinger. 539 US 244 (2003).
2 Grutter v. Bollinger. 539 US 330, 332–3 (2003).
3 Cohen JJ. The consequences of premature abandonment of affirmative action in medical school admissions. JAMA. 2003;289:1143–1149.
4 Miller HC. Affirmative action in medical school admissions. JAMA. 2003;289:3085–3086.
5 Heider MJ. Affirmative action in medical school admissions. JAMA. 2003;289:3086.
6 Regents of the University of California. Policy Ensuring Equal Treatment Admissions (SP-1). Oakland, Calif: University of California; 1995.
7 Connerly W. My fight against race preferences: a quest toward -creating equal.’ Chron High Educ. 2000;46:B6–B8.
8 Hopwood v Texas. 78 F3d 932 (5th Cir 1995).
9 State of Texas. Texas House Bill 588; 1997.
10 State of Texas. Texas House Bill 1641; 2001.
11 State of California. California Civil Rights Initiative (Proposition 209); 1996.
12 Office of Strategic Communications. Facts about the University of California: Comprehensive Review: November 2001. Available at: (http://www.ucop.edu/news/cr
). Accessed October 19, 2006.
13 Washington State Civil Rights Initiative (I-200), 1998. Chapter 49.60 RCW. Available at: (http://apps.leg.wa.gov/rcw/default.aspx?cite=49.60
). Accessed October 19, 2006.
14 Florida Governor’s Office. “One Florida” Initiative, 2000. Available at: (http://www.myflorida.com/myflorida/government/governorinitiatives/one_florida/index.html
). Accessed October 19, 2006.
15 Coleman AL, Palmer SR. Admissions and Diversity after Michigan: The Next Generation of Legal and Policy Issues. Washington, DC: The College Board; 2006.
16 Kahlenberg R. The Remedy: Class, Race, and Affirmative Action. New York, NY: Basic Books; 1996.
17 University of Washington. University of Washington Admission Process. Available at: (http://depts.washington.edu/gpsspub/admitpol.pdf
). Accessed August 8, 2006.
18 Lakhan SE. Diversification of U.S. medical schools via affirmative action implementation. BMC Med Educ. 2003;17:6. Available from: BioMed Central Ltd, London, UK. Accessed March 1, 2006.
19 Koenig J, Huff K, Julian E. Predictive Validity of the Medical College Admission Test. Washington, DC: Association of American Medical Colleges; 2002. MCAT Monograph 8.
20 Julian ER. Validity of the medical college admission test for predicting medical school performance. Acad Med. 2005;80:910–917.
21 Sturm S, Guinier L. The future of affirmative action: reclaiming the innovative ideal. Calif Law Rev. 1996;84:953–1036.
22 Steele CM. Thin ice: “stereotype threat” and black college students. The Atlantic Monthly. 1999;284(2):44–47,50–54.
23 Steele CM, Aronson J. Stereotype threat and the intellectual test performance of African Americans. J Pers Soc Psychol. 1995;69:797–811.
24 Nickens HW. Questions and Answers on Affirmative Action in Medical Education. Washington, DC: Association of American Medical Colleges; 1998.
25 Brief of Amici Curiae Association of American Medical Colleges, et al, Grutter v. Bollinger, 123 S.Ct. (2003). Available at: (http://www.aamc.org/diversity/amicusbrief.pdf
). Accessed October 19, 2006.
26 Robinson L, ed. Table B5: Enrollment of Selected Race and Ethnic Groups in First-Year Classes of Medical Schools and Table B8: US Medical School Graduates from Selected Race and Ethnic Groups. In: AAMC Data Book 2005. Washington, DC: Association of American Medical Colleges; 2005.
27 Klasko S. Viewpoint: what they don’t teach you in business school. AAMC Reporter. 2005;15:3.
28 Cohen JJ. Our compact with tomorrow’s doctors. Acad Med. 2002;77:475–480.
29 Albanese MA, Snow MH, Skochelak SE, Huggett KN, Farrell PM. Assessing personal qualities in medical school admissions. Acad Med. 2003;78:313–321.
30 Koenig JA, Sireci SG, Wiley A. Evaluating the predictive validity of MCAT scores across diverste applicant groups. Acad Med. 1998;73:1095–1106.
31 Eva KW, Reiter HI, Rosenfeld J, Norman GR. The ability of the multiple mini-interview to predict preclerkship performance in medical school. Acad Med. 2004;79(10 suppl):S40–S42.
32 Albanese MA, Farrell P, Dottl SL. A comparison of statistical criteria for setting optimally discriminating MCAT and GPA thresholds in medical school admissions. Teach Learn Med. 2005;17:149–158.
33 Goggin WJ. A “merit-aware” model for college admissions and affirmative action. Postsecondary Education Opportunity Newsletter. 1999;May:6–12.
34 St. John EP, Simmons AB, Musoba GD. Merit-aware admissions in public universities. The NEA Higher Education Journal. 2001–2001;Winter:35–46.
35 Kreiter CD, Stansfield B, James PA, Solow C. A model for diversity in admissions: a review of issues and methods and an experimental approach. Teach Learn Med. 2003;15:116–122.
36 Cooper RA. Impact of trends in primary, secondary, and postsecondary education on applications to medical school. II: considerations of race, ethnicity, and income. Acad Med. 2003;78:864–876.
37 Kahlenberg R. Class, not race. The New Republic. 1995;212(April 3);21–27.
38 Bowen WG, Bok DC. The Shape of the River: Long-Term Consequences of Considering Race in College and University Admissions. Princeton, NJ: Princeton University Press; 1998.
39 Cross T, Slater RB. Special report: why the end of affirmative action would exclude all but a very few blacks from America’s leading universities and graduate schools. Journal of Blacks in Higher Education. 1997;17(Autumn):8–17.
40 Grumbach K, Coffman J, Muñoz C, Rosenoff E, Gándara P, Sepulveda E. Strategies for Improving the Diversity of the Health Professions. San Francisco, Calif: The California Endowment; 2003.
41 Jolly P. Medical School Tuition and Young Physician Indebtedness. Washington, DC: Association of American Medical Colleges; 2004.
42 Carnevale AP, Rose SJ. Socioeconomic status, race/ethnicity, and selective college admissions. In: Richard D. Kahlenberg, ed. America’s Untapped Resource: Low-Income Students in Higher Education. New York, NY: Century Foundation Press; 2004:101–156.
43 Johnson LM, ed. Enrichment Programs. Minority Student Opportunities in United States Medical Schools. Washington, DC: Association of American Medical Colleges; 2005.
44 Patterson D, Carline J. Literature Review on Partnerships Compiled for the Health Professions Partnership Initiative. Washington, DC: Association of American Medical Colleges; 2004.
45 Patterson DG, Carline JD. Promoting minority access to health careers through health profession—public school partnerships: a review of the literature. Acad Med. 2006;81(6 suppl):S5–S10.
46 Grumbach K, Chen E. Effectiveness of University of California postbaccalaureate premedical programs in increasing medical school matriculation for minority and disadvantaged students. JAMA. 2006;296:1079–1085.
47 Harvard announces new initiative aimed at economic barriers to college. Harvard University Gazette. August 12, 2004.
48 Office of Diversity and Community Partnerships. Harvard Medical School. DCP K–12 Programs. Available at: (http://www.hms.harvard.edu/dcp/k12/index.htm
). Accessed October 19, 2006.
49 Oakes J, Rogers J, Lipton M, Morrell E. The social construction of college access: confronting the technical, cultural, and political barriers to low-income students of color. In Tierney WG, Hagedorn LS. Increasing Access to College: Extending Possibilities for All Students. Albany, NY: State University of New York Press: 2002:105–122.
50 Hanson GR, Burt L. Responding to Hopwood: Using Policy Analysis Research to Re-design Scholarship Criteria. Available at: (http://www.utexas.edu/student/admissions/research/hopwood.html
). Accessed October 31, 2005.
51 Scholarship Provider of the Year Award Presented to The University of Texas at Austin. Available at: (http://www.utexas.edu/opa/news/04newsreleases/nr_200411/nr_scholarship041123.html
). Accessed October 31, 2005.
52 University of Michigan News Service. New U-M undergraduate admissions process to involve more information, individual review. Available at: (http://www.umich.edu/news/?Releases/2003/Aug03/admissions
). Accessed October 19, 2006.
53 University of Michigan. Evaluation Categories for the Admissions Review Process 2004–2005. Available at: (http://www.admissions.umich.edu/process/review/categories
). Accessed October 19, 2006.
54 Roithmayr D. Direct measures: an alternative form of affirmative action. Urbana-Champaign, Ill: University of Illinois College of Law. Law and Economics Working Paper Series 2001. Working Paper No. 00-23.
55 Association of American Medical Colleges. 2006–2007 Medical School Admission Requirements (MSAR). Washington, DC: Association of American Medical Colleges; 2005.
56 Cleveland EF, Steinecke A. Lessons learned from the Health Professions Partnership Initiative (HPPI), 1996–2005. Acad Med. 2006;(6 suppl):S1–S60.
57 Carline JD, Patterson DG. Characteristics of health professions schools, public school systems, and community-based organizations in successful partnerships to increase the numbers of underrepresented minority students entering health professions education. Acad Med. 2003;78:467–482.
58 Harris S. Budget cuts threaten Title VII programs. AAMC Reporter. 2006;April:4–10.
59 Acosta D, Olsen P. Meeting the needs of regional minority groups: the University of Washington’s programs to increase the American Indian and Alaskan Native physician workforce. Acad Med. 2006;81:863–870.
60 Thomson WA, Ferry PG, King JE, Martinez-Wedig C, Michael LH. Increasing access to medical education for students from medically underserved communities: one program’s success. Acad Med. 2003;78:454–459.
61 Blumenthal D. Toil and trouble? Growing the physician supply. Health Aff (Millwood). 2003;22:85–87.
62 Kington R, Tisnado D, Carlisle DM. Increasing racial and ethnic diversity among physicians: an intervention to address health disparities? In: Smedley BD, Stith AY, Colburn L, Evans CH, eds. The Right Thing To Do, The Smart Thing To Do: Enhancing Diversity in the Health Professions. Washington, DC: National Academy Press; 2001:57–90.
*Coleman and Palmer explain: “Strict scrutiny is the most rigorous standard of judicial review. It is applicable to race- and ethnicity-conscious decisions that confer opportunities or benefits because distinctions based on race and ethnicity are ‘inherently suspect’ under federal law. To satisfy strict scrutiny, institutional policies must serve a ‘compelling interest’ and be ‘narrowly tailored’ to achieve that interest.”15
†Coleman and Palmer note that, in the Grutter decision, Justice O’Conner defines “critical mass” only as “‘meaningful numbers’ or ‘a number that encourages underrepresented minority students to participate in the classroom and not feel isolated.’”15 As a result, the concept remains contentious.