Affirmative action is a burning issue in the United States today. Accepted throughout our society for almost 30 years, affirmative action programs, policies, and procedures are now being systematically attacked. Opponents believe that affirmative action embraces racial preferences that threaten fundamental values of fairness and equality.1 In this article, we first review the national debate about affirmative action programs, examine the results of these programs in higher education, present data from 1995 through 1999 for underrepresented minority enrollments in U.S., California, and Texas medical schools, and briefly indicate findings showing that minority physicians serve patients of their own races and/or ethnicities, poor patients, and Medicaid patients in disproportion to these physicians' numbers. The rationale suggested by this fact, plus the goal of developing all persons to their highest potential, led, beginning in 1998, to experimentation in the admission process at the Texas A&M University Health Science Center College of Medicine, an initiative we describe and discuss in the rest of the article.
Affirmative action indicates programs, policies, and procedures initiated by the Lyndon B. Johnson Democratic administration in 1965 (Executive Order 11246) to give opportunities to minorities* and women in business, education, and public affairs. In education, the decision of the Supreme Court in the Bakke case in 19782 set the precedent for admission policies to include consideration of race as one of a number of applicant characteristics in the admission process.
Interpreting this legal decision loosely, many institutions of higher education used different—i.e., lower—quantitative standards for minority applicants than for majority applicants, and/or implemented separate admission processes. While it is true that underrepresented-minority students who are enrolled in medical schools frequently have lower standardized test scores than majority students do, the under-represented-minority students generally succeed in completing medical school and obtaining residency positions. And a 20-year study of “special-consideration-admission students” (defined as any student who had a grade-point average less than 3.0 on a 4.0 scale and average scores on the Medical College Admission Test of less than 10 on the four subscales) showed that these students experienced greater-than-average academic difficulty during the first two years of medical school, but that the differences faded in the clinical years, and there was no difference during residency training between these students and a matched control group.3
Attacks on Affirmative Action
Conservative political forces within our society have been conducting well-organized legal attacks against affirmative action throughout the past decade. A notable example of such activity is the reverse-discrimination case fought by the Center for Individual Rights, which resulted in the Hopwood vs. State of Texas decision by the Fifth Circuit Court of Appeals in 1996.4 The Center for Individual Rights sued the University of Texas (UT) on behalf of four 1992 white applicants to the law school who were denied admission. The suit claimed that minority applicants were granted admission over the white plaintiffs, who had higher grade-point averages and standardized-test scores. At that time, the UT law school based its admission decisions largely on an applicant's score on what is termed the Texas Index (TI), a composite of the applicant's undergraduate grade-point average (GPA) and his or her score on the Law School Aptitude Test (LSAT). The law school had used a separate admission committee and lower TI ranges for minority applicants in 1992 and prior years. In August 1994 a district court judge upheld the university's right to consider race as a factor in admission. However, in 1994 the law school dismantled the separate admission committee for minorities and changed its process.
The plaintiffs appealed their case to the Fifth U.S. Circuit Court of Appeals, which has jurisdiction over the states of Texas, Louisiana, and Mississippi. A panel of three judges of the Fifth Circuit Court reversed the district court judge's ruling and, furthermore, ruled that the law school failed to show a compelling state interest required to justify its race-conscious admission policy. In April 1996, seven judges on the Fifth Circuit panel wrote a dissenting opinion concerning their two colleagues' ruling. The Supreme Court refused to hear the Hopwood case, stating that because the law school had changed its process, this case did not present a live controversy. Although the three states under the Fifth Circuit's jurisdiction are legally obliged to follow the Hopwood decision, Mississippi and Louisiana continue to operate under a court order of desegregation, which is interpreted to obviate the Hopwood decision. Texas institutions of higher education were compelled to comply with the Hopwood decision in the autumn of 1998 by the formal opinion of the state attorney general.
Another example of successful dismantling of affirmative action in higher education is the action that the regents of the University of California took in 1995 prohibiting the use of race as a factor in admission decisions. This policy affected undergraduate applicants for the entering class of 1999. Following this action, the voters of California approved Proposition 209, which applied the prohibition of race as a factor in admission to the California state university system effective in 1996. In the state of Washington, Initiative 200, which prohibits consideration of race for admission, became effective in 1998.
Several state legislatures are substituting mandatory admission rates to undergraduate institutions for affirmative action programs. The Texas legislature required admission of all students who scored in the top 10% of their high school classes to state-supported undergraduate colleges and universities. The California legislature now guarantees admission to public colleges and universities for the top 4% of students in all public high schools in the state.5 The state of Florida has adopted a plan to replace affirmative action with a guarantee to the top 20% of graduates of every high school for admission to the public institutions of higher education. This action will affect the undergraduate entering class of 2000.6
Affirmative action programs in other levels of education are also being dismantled. The public school system of the city of San Francisco announced in February 1999 that it was abandoning race-based admission. The Boston Latin School, that city's most prestigious public school, was ordered by a three-judge panel in November 1998 to cease using racial preferences.7 Magnet schools, developed to bring about racial integration in public school systems, are being sued for reverse discrimination in Charlotte, North Carolina, and are under attack in other cities as well.8
In April 2000 the Civil Rights Commission announced its disapproval of the action of the three state legislatures of California, Texas, and Florida to require admission of a certain percentage of public high school graduates to state-supported institutions of higher learning. Thus, the issue continues to stimulate public debate and action.
Results of Affirmative Action Programs
What have been the results of affirmative action programs in United States' institutions of higher education? Have these programs given minority students more opportunities for education and development? Bowen and Bok9 document the long-term consequences of race-sensitive admissions in 28 academically select colleges and universities. The Andrew W. Mellon Foundation built the College and Beyond (C&B) database for that study, using individual students' records linked to admission records in collaboration with the participating colleges and universities. That database covers 32,000 students who began their studies at 28 C&B schools in 1989 (including more than 2,300 black students) and an additional 30,000 students who began college in the fall of 1976 (including more than 1,800 black students). An extensive survey of many of these same matriculants was conducted to obtain data beyond their college years. The authors found that the black students admitted to selective undergraduate colleges and universities had high graduation rates (79% within six years). Furthermore, a larger percentage of minority students than majority students admitted to the undergraduate colleges and universities in this study had aspirations to earn graduate or professional degrees; equal percentages of black and white students actually received degrees in law, business, or medicine. What has happened to these graduates in their careers and their lives? They are helping form the emergent middle class of minorities in our society. They are among the people who have the possibility of making a significant difference in our society for future generations. The researchers found that 20 years after entering college, the black graduates of these schools were more likely to be employed than were members of other minorities, and they had achieved high-paying and satisfying careers at young ages. Black men were earning an average annual income of $82,000, and black women were earning an average income of $58,500. The earnings of the 1976 black cohort were higher than the average earnings of the “A” students of all races who earned any U.S. college degree in 1976. The 1,900 minority graduates in the study expressed high satisfaction with their college experiences, a fact that contradicts the opinion that it is a disservice to admit minority students to elite institutions because they do not fit there. The black matriculants also have an extensive track record of service and leadership in community and civic undertakings. These results of race-sensitive admission policies to selective undergraduate institutions are widely regarded as impressive.
One study from medical education documents the long-term effects of affirmative action in the medical profession. Davidson and Lewis3 conducted a 20-year study of the special-consideration-admission students at the University of California—Irvine. These students experienced more academic difficulty during medical school, especially in the first two years, than a matched control group did. The graduation rate of the special-consideration students was 94%, compared with 98% in the control group. However, during residency training no significant difference was found between the two groups in academic difficulty or in the receipt of special honors. The residency program directors' evaluations of the residents also showed no difference. The self-described practice characteristics of the two groups were remarkably similar. However, the special-consideration doctors expressed more satisfaction with their lives than did the control doctors. Further long-term studies of the results of affirmative action in medical training are needed.
Effects of Recent Changes
What are the effects of the recent changes in policies regulating admissions to medical schools resulting from the legal decisions and legislative mandates? Table 1 presents information about the medical school entering classes of 1995 through 1999 for all U.S. medical schools and for California and Texas medical schools. Texas and California are the two states that produce the largest number of underrepresented minority students in medical schools.
In all U.S. medical schools there was an average of 12.66% underrepresented minorities in 1995. The proportions of black and Mexican-American students are far lower than the percentages of these groups in the population. By 1999, the proportion of underrepresented minority students in U.S. medical schools had declined to 10.65%. In Texas and California medical schools, the percentages of under-represented minorities declined throughout the five years, reaching their lowest points in both states in the entering class of 1997. That is the first year the California state university system was affected by Proposition 209 and the Texas institutions of higher education were affected by the Hopwood decision. In California medical schools in 1995, 17.5% of the students were from underrepresented minorities. Reaching the lowest point (11.6%) in 1997, the number increased slightly to 12.3% in 1999, but that is still four percentage points below the 1995 level. Notice that in the entering classes of 1997 in Texas medical schools only 13% of the students were from underrepresented minorities, in contrast to the 18% and 20% in the years preceding the Hopwood decision. The entering classes of 1998 increased to 15% underpresented minorities, but the entering classes of 1999 decreased slightly to 14% students from those groups.
Importance of Admitting Minority Applicants
Admission of minority applicants to medical school is particularly important to our society. Studies have demonstrated that minority physicians take care of a disproportionate share of patients of races and/or ethnicities similar to their own, and they take care of more indigent patients than do white doctors. Komaromy and colleagues10 published a study of physicians' practice locations and the racial and ethnic compositions and socioeconomic status of communities in California. The authors used data from the 1990 American Medical Association (AMA) Physician Masterfile, the 1990 U.S. census, a random sample of general internists, family physicians, and general practitioners, and the practice locations of 1,713 physicians who had matriculated at the University of California—San Francisco (UCSF) medical school between 1969 and 1984 who were currently practicing in California. UCSF was selected because it has the highest proportion of minority students in any U.S. medical school except the historically black medical schools. The authors found that black and Hispanic physicians practice in areas of California with higher proportions of underserved-minority people. They also care for higher proportions of patients of their own races who are uninsured or are covered by Medicaid.
A national study11 of a sample of 15,081 patients representing the total civilian U.S. population yielded similar findings. Nine measures of health status were examined. More than 14% of adult Americans who identified a usual source of care identified a nonwhite physician as that source. More than 33% of minority patients were cared for by a minority physician, compared with 11% of white patients cared for by minority physicians. Between 19 and 29% of low-income patients, Medicaid patients, and uninsured patients received care from minority physicians. Both black and Asian physicians cared for disproportionate shares of racial-minority patients and indigent patients. Furthermore, Americans who identified a nonwhite physician as their caregiver tended to be sicker than those who did not.
Another national study12 employed a questionnaire sent to a random sample of 2,600 physicians selected from the AMA Physician Masterfile. The sample comprised physicians who had graduated from medical school in 1983 or 1984 whose specialties were family practice, general internal medicine, or general pediatrics. The possible confounding variables of repaying financial aid through service, rural or inner-city background, and childhood family income of the physicians were examined. The results showed that under-represented-minority physicians reported caring for higher percentages of Medicaid and poor patients than did white and Asian physicians. These physicians were likely to serve patients of their own races and/or ethnicities.
The most thorough published study13 about the service of minority physicians to underserved populations compared physician questionnaire data (Surveys of Young Physicians) with patient data from the 1987 National Medical Expenditure Survey (NMES) and data concerning revenues derived from Medicaid. Five outcome measures were used: the percentages of each physician's patients who were uninsured, black, Hispanic, or poor, and the percentage of practice revenue derived from Medicaid. The researchers found that the physician self-reports were plausible measures of service to underserved groups. This study controlled for possible confounding variables of specialty, age, type of medical school, organization of practice, urban or rural location, educational debt, and participation in a service pay-back program. Furthermore, the authors examined changes in practice patterns over a four-year period. This well-designed study confirmed the findings of previous studies; that is, black, Hispanic, and women physicians serve high proportions of black and Hispanic patients, poor patients, and Medicaid patients. The researchers also found a relationship between the lowest socioeconomic group of physicians and underserved populations, but the socioeconomic variable was not as pronounced as the race-ethnicity and gender variables. For minority patients, the study indicated that the highest level of service is provided by physicians of the same race and/or ethnicity as their patients, regardless of gender. For poor and Medicaid patients, black and Hispanic women physicians were shown to provide the most service. Participation in financial aid pay-back programs was not associated with service to underserved populations. Little change over four years was observed in practice patterns. The authors concluded that minority physicians serve patients of their own races and/or ethnicities, poor patients, and Medicaid patients from altruistic motivation.
Medical educators need to be concerned about training minority physicians for the 21st century because minority populations are growing more rapidly than the majority population. In the state of California, minority and ethnic groups already outnumber whites. Demographic projections for the state of Texas indicate that by the year 2030, Hispanics will outnumber Anglos (the term for whites in Texas and other areas of the Southwest). Figure 1 shows projections of the Texas population by racial and ethnic groups for the years 1990 through 2030.14 These data suggest that the population of Texas will continue to grow rapidly, due primarily to the growth of the Hispanic and other minority populations (not including the black population). From 1990 to 2030, the percentage of change for Hispanics is projected to be 257.6%; for other minorities, the projection is 648.4%. That contrasts with the projected rate of 20.4% for the Anglo population and 62.0% for the black population. Eighty-seven and a half percent of the growth of the Texas population will be due to minority racial and/or ethnic groups between the years 1990 and 2030.
These projections, considered together with the facts revealed by the studies of minority physicians currently in practice, illustrate concretely why medical schools need to recruit qualified minority applicants and educate them as physicians. It is in the economic self-interest of the United States to produce minority physicians in numbers proportionate to the expected growth of the minority population during the 21st century. Furthermore, it is in the humanitarian interests of our society to develop all persons to their highest potential.
The question, then, in medical school admission is Can race-neutral admission policies succeed in admitting underrepresented minority students in sufficient numbers to fulfill the physician workforce needs of the U.S. population, including underserved segments of the population? The rest of this article describes recent attempts by the Texas A&M University Health Science Center College of Medicine to answer that question, and the preliminary results of these attempts.
ONE SCHOOL'S EFFORTS
The Admission Process When Race Was a Criterion
In 1991, using what was then a state-of-the-art concept, the Texas A&M University Health Science Center College of Medicine developed new criteria for admission. At that time algorithms had wide appeal because algorithms increased processing efficiency and seemed objective. Race was one criterion among many criteria considered in the admission process.
The first algorithm calculated each applicant's “Academic Score” by combining the applicant's undergraduate (and graduate, if applicable) grade-point averages (GPAs) with his or her Medical College Admission Test (MCAT) score. In considering the MCAT scores of applicants, majority applicants' scores were compared with national majority applicants' average scores. For underrepresented-minority students, comparisons were made with national underrepresented-minority applicants' average scores. Applicants were invited for interviews based on the ranking of their Academic Scores. There was provision to “screen” applicants if unusual circumstances indicated the need of additional consideration. Screening an applicant was considered an exception; consequently, only a few applicants were screened each year.
During that year, the admission committee, with the assistance of a national consultant well known in medical education, had given considerable attention and effort to refining the interviewing process. The committee's strong consensus was that the interview should evaluate noncognitive factors that are attributes of the profession, such as communication skills, motivation, and record of social service. The consultant trained the committee members in asking questions and evaluating applicants during the interview. The consultant and the committee also developed materials to train new committee members each year. During the interview season each year, the admission committee met every week and carefully reviewed every applicant's three interview evaluations. Discussion followed and then the committee members voted their opinions of each applicant on a scale of 1 (lowest rating) to 10 (highest rating). All of the committee members' ratings were averaged to calculate the applicant's Committee Score.
A second algorithm, which computed the final score for applicants, weighted the Committee Score and the Academic Score equally, counting each as 50% of the final score. Applicants were ranked by their final scores, and acceptances were offered according to those rankings until the class was filled. While developing the algorithms, the committee had discussed the relative weighting of the Academic Score and the Committee Score for the final score. However, no empirical study was done at that time to determine what the weighting should be.
Changes in the Admission Process
As a result of the Hopwood decision in 1996, the admission criteria and process had to be changed. The national average MCAT score for minority students could no longer be used as a benchmark separate from the majority average MCAT score. During 1997, strategic planning for the medical college took place, and the mission statement of the medical college was revised to include explicitly the goal of producing physicians to practice medicine in rural and underserved areas of Texas. Therefore, selecting applicants who are likely to practice in those areas became a priority. Also, the admission committee sought to select and enroll applicants who were well qualified academically and who had demonstrated altruism. Discussion about possible changes in the criteria and process were initiated by the dean of the medical college and the chairman of the committee.
Several attempts were made during the summer and autumn of 1997 to modify the existing admission process. The admission officers of the medical college consulted with the university legal counsel and the director of admissions for the undergraduate university, who had extensively analyzed factors relevant to admission such as high school data, socioeconomic status, parental background, and others. Valuable information and advice were obtained during these discussions. Beginning in August 1997, committee members screened a larger number of applicants than in previous years, trying to give more consideration to factors such as obstacles to education in decision making about whom to invite to interview. Two of us (JCE and FGM) conducted a study examining the empirical effects of weighting the interview more heavily in the process. Weighting the interview more heavily than the quantitative algorithm, we reasoned, would enable the committee to select applicants who were interested in practicing medicine in rural and underserved areas of the state and those who were altruistic. The interview part of the admission process had been developed thoroughly and carefully, and the admission committee members gave a large amount of time and effort to conducting the interviews. Because the Academic Score had previously been averaged in at every step in the process, the GPA and the MCAT scores counted more than 50% in the total process. Therefore, it seemed appropriate to increase the weight of the interview in the process.
Our study examined 439 applicants who were interviewed during the 1996-97 term. The findings showed that increasing the weight of the interview to 60% and decreasing the Academic Score weight to 40% had almost no effect on the final score. However, increasing the Committee Score to 70% and decreasing the Academic Score to 30% would have affected 32 of the applicants such that three additional members of underrepresented minorities would have been offered acceptances; two of these three claimed to be disadvantaged. Two additional Asian applicants also would have been accepted.15 However, the results of this study did not persuade the committee members to change the process. Committee members searched to find variables to use in selection formulas so that the admission process would not take additional time and effort. Admission data were analyzed to determine whether the size of the undergraduate institution was one of the factors associated with academic achievement in medical school.
By the spring of 1998, it became clear that substantive changes had to be made. The entering class of 1997 had only five (8%) underrepresented-minority students, in contrast to the 1996 entering class, 15% of whom were minority students. The admission committee members had worked harder than ever before. These efforts had not succeeded in increasing or maintaining the desired diversity.
A major new plan
Therefore, a major new plan was developed. Three subcommittees of the admission committee were appointed and charged to make recommendations regarding the criteria for interviewing applicants, the interview protocol, and the admission committee deliberations. These subcommittees worked throughout the summer to complete their tasks for the beginning of the new admission cycle in August. A full committee meeting was held in early August, at which time each subcommittee presented its recommendations and the committee made decisions. The national consultant who had worked with the committee in 1991 was brought in again to conduct a workshop about interviewing in early August. The workshop consisted of information presentation, discussion of pertinent issues, and the practice of interviewing skills.
Early in 1998, an administrative decision was made to join the American Medical College Application Service (AMCAS) for the 1998-99 admission cycle, the first time in the history of the school. We thought that the college would obtain more applicants, including more minority applicants, by joining AMCAS. The electronic application was also an attractive feature of AMCAS. The college continued to require a secondary application.
The admission committee accepted the recommendations of the subcommittees and implemented these changes for the 1998-99 admission cycle. The Academic Score for each applicant was computed according to the same algorithm that had been used since 1991. However, the committee decided to screen the large group of applicants whose academic scores fell above the minimum Academic Score to determine which applicants to invite to interview. The members felt that the screening would allow them to make better-informed judgments based on more information than simply using the quantitative algorithm. Three persons (the associate dean, the assistant dean, and the committee chair) conducted a study of the screening criteria and process used in the 1997-98 cycle. The results of this study were presented to the subcommittee. The screening instrument was revised after members of the subcommittee discussed the items and scoring of the instrument.
The revised instrument had the following four categories, whose scores could total to a maximum of 100 points:
- ▪ Academic performance/intellectual capacity, with a maximum of 30 points
- ▪ Humanism: dedication to service and capacity for effective interactions, with a maximum of 40 points made up of subscores on the student's participation in community service, desire to help others, leadership in community and school organizations, effective communication, honesty, healthcare-related experiences, degree of having a responsible, professional standard of behavior, personal statements (strength of character, originality, clarity, and depth), evaluation of self, motivation for medicine as a career, concern for welfare of others, and compassion
- ▪ Special life experiences, with a maximum of 15 points made up of subscores on the student's supportive letters of evaluation from professors, evidence of interest in a primary care, rural, and/or underserved area of the state, and awareness and knowledge of cultural factors as these may affect health care
- ▪ Other compelling factors, with a maximum of 15 points (this was a generic category; no examples of the factors were given)
There was space for comments. The evaluator could request an interview or decide not to interview regardless of the number of points given, and the evaluator could also request that a second person screen the application. Seven members of the admission committee agreed to be responsible for screening the applicants whose Academic Scores were above the minimum. Each application was screened by one admission committee member or the assistant dean for admissions. The form and screening process provided structure but also flexibility to the persons doing the screening.
The committee desired to interview as many applicants as possible to make judgments about the suitability of each applicant to the mission of the college and to evaluate the character and motivation of the applicant. Another change was designed to give careful deliberation to applicants deemed acceptable after the first algorithm computed the Academic Score. We analyzed the records of the 1995, 1996, and 1997 applicant pools to determine how many applicants in those years might possibly have been considered High Academic applicants (HAA) and how many might have been quickly recognized as unacceptable, thus allowing more time and effort to focus on the large, middle pool of acceptable applicants. Based on analyses of the Academic Scores and the Committee Scores (after interviews), we found that between 9% and 19% of the High Academic applicants had enrolled in our college. The committee had been devoting as much or more time to this pool of High Academic applicants as they had to the larger percentage of applicants from the middle of the pool who had enrolled in the college. The committee anticipated an increased applicant pool from AMCAS, approximately 1,500 applicants instead of 1,400, and set a goal of interviewing 700 applicants during the 1998-99 admission cycle. In 1997-98, 479 applicants had been interviewed from a total pool of 1,420 applicants. Our analysis had also revealed that we could anticipate rejecting approximately 500 applicants. Therefore, we decided to devote more time and effort to the middle pool of approximately 700 applicants.
Considerable debate ensued among the committee members about the number of interviews to give each applicant. Giving three interviews, as had been done in prior years, meant that 2,100 interviews (1,050 hours of interviews) would have to be given. This seemed an impossible task considering the small size of the basic science faculty and the pressure on the clinical faculty to generate revenue. A proposal to give only one interview to High Academic applicants was not accepted; members did not feel that adequate evaluation could be made in only one interview. In the end, the members decided to give two one-half hour interviews to every applicant who was selected for an interview.
The subcommittee charged to design the process of the committee deliberations after the interview recommended that applicants who obtained average interview scores of 9 or 10 on the evaluation scale of 1 to 10 be accepted in the committee meeting by acclamation. They also recommended that applicants with average scores of 5 or less be rejected by acclamation. They expected about 285 of these applicants. These decisions cleared the agenda of the weekly meetings for longer, more careful deliberation about the applicants with acceptable interview scores in the range of 6 to 8 (the range of scores received by the majority of those who were interviewed) and for discussion of any applicant who had received scores with three points or more of discrepancy between interviewers. Committee members had the flexibility to present and discuss any candidate accepted or rejected by acclamation if they desired. This option was used only a few times throughout the committee meetings.
The committee made the substantive change of using the Committee Score (that is, the score of 1 to 10 that was the average of the voting members' scores following the presentation and discussion of each applicant) to rank applicants for the “Alternate List.” Previously, the Committee Score had been averaged with the Academic Score to rank order applicants. Using the Committee Score in this new way gives greater weight to the interview evaluations and the deliberations of the committee. The committee also decided to offer “rolling acceptances” instead of waiting until all applicants had been interviewed and ranked to offer acceptances. We hoped to attract desirable candidates to enroll by offering them acceptance quickly after their interviews. Each week after the committee meeting, letters of acceptance were sent out to those applicants who had committee scores of 8, 9 or 10; letters of rejection were also sent out each week. Applicants who had committee scores of 6 or 7 were ranked in priority order by the committee members and placed on the Alternate List each week. The entire Alternate List was examined and prioritized at each meeting. Beginning in December, the chair of the committee and the assistant dean for admissions made offers to applicants according to the priorities on the Alternate List.
The substantive changes made to the admission criteria and process for the 1998-99 admission cycle placed more responsibility on the admission committee members and less reliance on quantitative algorithms. The criteria, while retaining a sound basis of academic qualifications, were broadened to include consideration of applicants who fit the mission of the college to practice medicine in rural and other underserved areas of the state and who had demonstrated altruism. The process involved more thorough perusal of the applications and more extensive deliberation about the suitability of the applicants who had acceptable interview evaluations.
Table 2 presents basic data about the applicant pools and the enrolled entering classes of 1995 through 1999. It is notable that the applicant pool and percentages of underrepresented minorities enrolled climbed higher in 1995 and 1996 and then dropped in 1997, the first year that Texas medical schools were affected by the Hopwood decision. Prior to 1995, the numbers of underrepresented minorities who were applicants and who had enrolled had been approximately half of the 1995 numbers for those groups. The percentages of underrepresented minority applicants remained stable at 10% of the total applicant pool from 1995 through 1998; the increase to 11% in 1999 is encouraging.
The members of the admission committee increased their efforts, devoting more time and thought to the admission process for the entering classes of 1998 and 1999 to select applicants, including underrepresented minorities, who fit the mission and ethos of our college. The percentages of underrepresented minorities who were interviewed, were offered acceptances, and matriculated dropped steeply in 1997 upon implementation of the Hopwood decision. The lowest percentage of underrepresented matriculants (6.3%) occurred in the entering class of 1998. During the admission cycle of 1997-98, when the college was struggling to comply with the Hopwood decision, minor changes were implemented. The alarming results galvanized the admission committee into examining its process (as described earlier in this article) and making the major changes discussed above, the results of which are reflected in the entering class of 1999. In all categories in 1999, the numbers of underrepresented minorities increased from 1998. The increases, however, were small—one percentage point, generally.
Preliminary results of the 1999-2000 admission cycle indicate that approximately 10% of the entering class of 2000 will be underrepresented-minority students. Although it is encouraging to see an upward trend again, that trend is far below the rate at which we need to produce minority physicians to serve the growing minority population. Population projections for Texas indicate that by the year 2010, approximately 51% of the state population will be non-Anglo, that is, black, Hispanic, and other minorities.14 The 1999 entering medical classes of all medical schools in Texas contained only 14% of underrepresented-minority students. Assuming graduation from medical school in four years and three years of graduate medical education, these physicians will begin medical practice in 2006. That means that in 2010 in Texas, fewer than 14% of the physicians will be minority physicians, while 50% the state population will be composed of members of minorities. In California currently, the total minority population exceeds the white population. Projections for the state of California indicate that by 2010, no one ethnic group will be a majority. Studies cited earlier9–12 clearly demonstrate that physicians from minority racial and/or ethnic groups care for disproportionate shares of patients from such groups, and for poor patients. These figures, therefore, have enormous implications for physicians and for the burden on states to finance health care for the future.
A CALL TO ACT IN CONCERT
Are other medical colleges contemplating or experimenting with changes in admission criteria and process? If so, what are the changes and what do the early results indicate? As we have described in this article, affirmative action programs are under broad attack across the United States. Programs in Texas and California, the two largest producers of minority students, were dismantled first, and now programs in other states are being dismantled. Producing minority physicians is a critical issue in the national political agenda for health care. Selecting applicants who are a good fit for the mission of each particular medical college is also a concern for admission committees. Most medical college admission committees have concerns about their existing criteria. Experimentation may be going on in a number of medical colleges. Thinking, discussing, and acting in concert can improve the likelihood that we medical educators will develop legal and effective criteria and processes for selecting applicants who will fulfill the complex needs of the future.
We wrote this article in the spirit of stimulating national, ongoing dialog about the admission criteria and processes that can effectively select applicants who fit the mission of each medical college, among whom will be a sufficient number who are likely to care for patients from minority groups. Economics, as well as humanitarian interests, compel all medical schools to find solutions lest the generations of this new millennium curse us for our intellectual myopia.