Racial diversity reconsidered
In recent years, the value of diversity has been proclaimed across the entire spectrum of American opinion. Nowhere is this more true than in our educational institutions. In defending his school’s embattled affirmative-action admissions program, former University of Michigan President Lee Bollinger recently offered a ringing endorsement of diversity:
Diversity is not merely a desirable addition to a well-run education. It is as essential as the study of the Middle Ages, of international politics and of Shakespeare. For our students to better understand the diverse country and world they inhabit, they must be immersed in a campus culture that allows them to study with, argue with and become friends with students who may be different from them. It broadens the mind and the intellect–essential goals of education.
President George W. Bush has in somewhat more prosaic terms endorsed this position: “I strongly support diversity of all kinds, including racial diversity in higher education…. A college education should teach respect and understanding and goodwill. And these values are strengthened when students live and learn with people from many backgrounds.”
It is an article of faith among American elites that a diverse environment enhances mutual understanding among students of different backgrounds and enriches the educational experience for all. Generally, it is only the means of attaining diversity, particularly racial diversity, that stirs controversy, with some supporting quotas or quota-like measures and others arguing for more informal mechanisms of outreach. Almost no one poses the question: Does diversity work? Specifically, does racial diversity on American college campuses actually produce the benefits that are claimed for it?
Of course, this question is hardly hypothetical. The question of whether racial diversity enriches the educational experience for students of all racial and ethnic backgrounds is at the heart of the twin lawsuits against the University of Michigan that will be reviewed by the Supreme Court this spring. At the University of Michigan, minorities that are considered underrepresented–blacks, Hispanics, and Native Americans-are given a 20-point bonus in a point system that rates undergraduate applicants on a scale ranging from a low of 47 points to a high of 150. Thus the 20 points awarded for minority status represent almost 20 percent of the possible variation in scores. The University of Michigan’s law school employs a similar system to ensure sufficient diversity. In 1997, two white applicants, Jennifer Gratz and Patrick Hamacher, each filed suit after being denied admission to Michigan’s undergraduate program. Another rejected white applicant, Barbara Grutter, filed a similar suit against the law school. All three argued that a dual-track admissions process violated their constitutional guarantee to equal protection under the Fourteenth Amendment.
The diversity principle
Not since its review of California v. Bakke in 1978 has the Supreme Court taken up the issue of affirmative action. Then, the Court ruled five to four that the University of California Medical School at Davis could not employ a quota system that reserved 16 to 100 places each year for minority applicants. Such systems had grown out of the civil rights legislation of the 1950s and 1960s that aimed at desegregating educational institutions. By the 1970s, the federal government required universities to account for minority enrollment, and the quickest and simplest way to boost minorities on campus was to establish quotas or quota-like policies.
But the U.C.-Davis special admissions program was challenged by a 38-year-old white engineer, Allan Bakke, who was rejected by the medical school in 1973 and 1974, despite having admissions test scores higher than many of the minority applicants who were accepted. In stating the opinion of the Supreme Court, Justice Lewis Powell, Jr., declared that under the Fourteenth Amendment, “The guarantee of equal protection cannot mean one thing when applied to one individual and something else when applied to a person of another color. If both are not accorded the same protection, then it is not equal.” The Court also judged that U.C.-Davis violated Title VI of the 1964 Civil Rights Act, which prohibits discrimination “on the grounds of race, color or national origin” in any program or institution receiving federal financial assistance.
At the same time, Justice Powell, along with four other justices, defended what Justice Brennan described as “preferential treatment of racial minorities as a means of remedying past societal discrimination to the extent that such action is consistent with the [Fourteenth] amendment.” Thus universities could use race as a factor in recruiting students as long as they did not employ a quota system. Finally, and most significantly, Justice Powell added that it was “constitutionally permissible” for U.C.-Davis to pursue diversity on account of the educational benefits that would flow from an ethnically diverse student body.
When the committee on admissions reviews the large middle group of applicants who are “admissible” and deemed capable of doing good work in their courses, the race of an applicant may tip the balance in his favor just as geographic origin or life spent on a farm may tip the balance in other candidates’ cases. A farm boy from Idaho can bring something to Harvard College that a Bostonian cannot offer. Similarly, a black student can usually bring something that a white person cannot offer.
In the wake of the Bakke decision, the pursuit of diversity gradually came to supplement the effects of past societal discrimination as the basis for special admissions and hiring policies. Educational institutions began to defend affirmative action or diversity programs by attempting to demonstrate their benefits for all students, rather than as a redress for past discrimination. Of course, this meant that some otherwise qualified white students would be denied admission. When the New York Times asked former Dartmouth President James Freedman how he might justify such affirmative-action decisions, “especially when addressing a parent,” he replied: “Those kids seemed somehow to us to add qualities to the class that your son or daughter did not, which only may mean that your son or daughter was like 75 percent of the class that we’ve already admitted.”
Critics see this as “reverse discrimination.” After all, adding 20 points to a minority student’s application is the equivalent of subtracting 20 points from a white student’s score. Despite widespread public support for helping minorities in some fashion, ethnic and racial preferences in college admissions have increasingly been challenged, either through the courts or in state referenda. As a result, some states have adopted admissions policies designed to ensure substantial enrollment by black and Hispanic students without relying on explicitly racial or ethnic admissions criteria. These policies include reducing or eliminating reliance on the SAT or ACT tests (the University of Michigan’s program awards a maximum of 12 points for standardized test scores) and admitting a fixed percentile of the top students in each graduating high school class.
The District Court judgment in the 1997 case Gratz v. Bollinger, however, reaffirmed the view that diversity was important enough to justify Michigan’s preferential treatment of minorities in its admissions policy. In effect, the court accepted as fact the proposition that the presence of racial and ethnic diversity in the student body improves the quality of education, even if those students admitted under diversity criteria are not as academically proficient as those admitted under merit criteria. With the Supreme Court subsequently granting review of the undergraduate and graduate suits against Michigan, evidence as to whether a diverse student body enhances a student’s educational experience will be crucial in clarifying the relationship between affirmative action and the right to equal protection.
The argument that diversity enhances education and interracial understanding rests on a fairly substantial body of social science literature. In defending admissions policies designed to increase racial diversity on America’s campuses, advocates have relied on surveys of students (and sometimes faculty, administrators, and alumni) that have been designed to measure educational environments and intergroup relations.
In evaluating the consequences of diversity policies on campus life, it is very difficult to conduct controlled experiments and to collect data that have concrete measurable correlates. Many of the resulting studies must rely on the reported attitudes, perceptions, and memories of students, faculty, and administrators. But such evidence suffers from the subjective nature of survey responses. For example, William Bowen and Derek Bok’s frequently cited 1998 book, The Shape of the River, reports that support for campus diversity programs has been growing steadily among the alumni of elite colleges and that both blacks and whites believe that such programs help them to get along better with members of other races. But both these findings could just as easily reflect an increase in conscious or unconscious efforts to give socially appropriate answers, a result of the growing emphasis on these programs at elite schools, where they are frequently presented as highly important moral imperatives.
The problem can be illustrated by some of the actual questions that are asked. For example, an influential 1999 survey by Gary Orfield and Dean Whitla posed the following question: “Do you consider having students of different races and ethnicities to be a positive or negative element of your educational experience?” Approximately 90 percent of law students surveyed at Harvard and the University of Michigan responded that it was a positive element, while fewer than 1 percent called it a negative element. Similarly, more than two out of three said that ethnic and racial diversity enhanced their “ability to work more effectively or get along better with members of other races”; only 4 percent said diversity detracted from their ability to do so. The authors note, “In public opinion research it is very rare to find majorities of this size on any controversial issue.” This is quite true, and it suggests that the question is worded in a way that taps into what social scientists call a “valence issue,” an issue on which almost everyone agrees. In this sense, diversity is like free speech–almost everyone approves of it in the abstract, but its application in concrete situations can produce great controversy. In sum, the social science surveys that support the benefits alleged for diversity in college enrollment cannot preclude the possibility that favorable responses are products of a kind of indoctrination in the meaning of certain terms and concepts, rather than a valid reflection of real world effects. In this case, the argument that diversity is beneficial becomes circular–students are taught that diversity is valuable, asked whether diversity is valuable, and then their positive replies are seen as proof of diversity’s value.
A new approach
In order to avoid such problems in testing whether enrollment diversity programs benefit academic communities, we chose a more indirect approach that does not take survey responses at face value. In addition to asking members of the university community how they felt about campus diversity programs, we also asked them to evaluate various aspects of their general educational experience and environment, with no reference to diversity. We then correlated their attitudes with a separate empirical measure of racial diversity–the proportion of African Americans in the student body.
The measure of enrollment diversity, along with other empirical factors of the educational experience, was drawn from federal government statistics on the demographic traits of students and the academic and institutional characteristics of American colleges and universities. The evaluations of college life were taken from a cross-national survey of students, faculty, and administrators at colleges and universities in the United States and Canada. The analysis in this paper is limited to data from a sample of 140 universities and colleges in the United States, categorized by the type of institution–doctoral, comprehensive, or liberal arts schools. The schools were randomly selected from the full range of institutions in each category, with the probability of selection proportional to the number of faculty and students. The schools were then examined to ensure that they were representative with regard to such variables as region, quality, and size. Historically black colleges were excluded from the sample.
The resulting sample totaled 4,083 individuals, consisting of 1,643 students, 1,632 faculty members, and 808 administrators. Computer-assisted telephone interviews were conducted by the Angus Reid survey research firm in the spring of 1999. Response rates among those contacted were 53 percent for the student sample, 72 percent for faculty, and 70 percent for administrators. The lower response rate among students did not significantly affect the racial and ethnic representativeness of the sample. (The refusal rate among black students did not differ significantly from the refusal rate among white students.) In analyzing the results, we have focused mainly on the racial diversity provided by the presence of African-American students at predominantly white schools, since the legal and social debate has been primarily about this group. But we have also included some instructive comparisons with Hispanic and Asian enrollment.
Our survey addressed several aspects of campus opinion about diversity issues. While there was substantial agreement among students, faculty, and administrators on some issues, there were also important differences. There was widespread support for offering courses in multiculturalism on American campuses, but not for requiring them. Only one out of six students (16 percent) agreed that courses about the experience of minorities should be required, but a much larger group, 38 percent, believed that such courses should be encouraged. Hardly anyone thought that such courses should not be offered. Support for encouraging multicultural course offerings was even higher among faculty and senior administrators. Seventeen percent of faculty and administrators said such courses should be required, and 42 percent of faculty and 46 percent of administrators believed they should be encouraged.
There was also widespread endorsement of the ongoing discussion of minority issues at colleges and universities. Most members of the university community did not believe excessive attention was given to diversity issues. Over 80 percent of faculty, students, and administrators disagreed with the statement: “This university pays too much attention to minority issues.” However, this commitment to racial discourse did not appear to stem from widespread concern over racism on campus. When asked, “And do you think minority students are treated better, worse, or about the same as white students at your university?”, over 70 percent of all three groups answered “the same.” Only 18 percent of faculty, 14 percent of students, and 14 percent of administrators believed minority students were treated worse than whites.
However, a major difference of opinion separated students from staff when the means of ensuring diversity involved admissions and hiring standards. In response to the statement, “No one should be given special preference in jobs or college admissions on the basis of their gender or race,” two-thirds (67 percent) of the students strongly agreed, compared to only one-third of the faculty (34 percent) and one-quarter (26 percent) of the administrators. An overwhelming 85 percent of students either strongly or moderately agreed with the statement, as compared to 56 percent of the faculty and only 48 percent of administrators.
Similar fault lines appeared when the issue was framed in terms of relaxed admissions standards for minority students, although the differences among groups were not as great. Seventy-five percent of students disagreed with the statement: “More minority-group undergraduates should be admitted here even if it means relaxing standards.” Lower percentages (though still a majority) of faculty (57 percent) and administrators (55 percent) also disagreed with the statement.
Such widespread rejection by students of preferences or different standards for minority applicants may seem surprising. But it is clear that support for the goal of diversity diminishes where it involves preferential policies. The opportunity to take elective courses or engage in discussions about interracial and cross-cultural differences is thought to be a win-win situation for white students and minority students alike. But admissions decisions based on race create clear winners and losers. Student support for diversity dwindles in this zero-sum game between white and minority applicants.
On this matter student opinion appears to be closer than that of faculty and administrators to the views of the general public. A Newsweek poll conducted in January 2003 found that 68 percent of American adults rejected the use of racial preferences in college admissions. Although the survey questions were worded somewhat differently and the Newsweek poll is more recent than the student survey, it seems safe to say that large majorities of college students and adults oppose preferential treatment for minority student applicants.
Perhaps more surprisingly, substantial opposition to racial preferences was found among minorities as well as whites. The Newsweek poll found that 56 percent of all minorities surveyed said that blacks should not receive preferential treatment in the admissions process, while 71 percent rejected gender or racial preferences in college admissions altogether. Among white students the level of opposition to gender or racial preferences reached 89 percent. Fifty-four percent of minority faculty also expressed opposition. When the issue was framed in terms of “relaxing normal academic standards of admissions,” only 38 percent of minority students favored such a policy, a figure that dropped to 21 percent among white students. Once again, a majority (52 percent) of minority faculty opposed such a lowering of standards.
Affirmative action’s effects
Our study enters new territory by using survey data in conjunction with external evidence to test empirically the claims made for the benefits of enrollment diversity. Our survey included questions on perceptions of the educational environment and on perceptions of discrimination and the treatment of minorities.
Respondents were asked four questions on educational environment: how satisfied they were with their university experience (asked of students only); how well the school educates its students; how hard students work at their studies (all groups); and how well-prepared academically students were upon entering (faculty and administrators only). Respondents were also asked three questions on minorities and discrimination: whether minority students were treated better, worse, or about the same as white students; the extent to which racial discrimination was a problem at their institutions; and whether they personally had been treated unfairly because of their race, ethnicity, gender, sexual orientation, religious beliefs, or political views. These last three questions were asked of all groups.
We used the responses to these questions to test specific hypotheses about diversity on campus. These hypotheses predict that increases in the proportion of black student enrollment will be associated with favorable responses to the corresponding survey items–greater satisfaction and more favorable ratings of educational quality, higher levels of preparation and commitment of students, less racial discrimination and better treatment of minority students, and fewer reported encounters with discrimination.
Data on black, Hispanic, and Asian-American student enrollment, along with information on other student traits and characteristics of educational institutions, were obtained from the National Center for Education Statistics (NCES), the federal government’s primary center for collecting and analyzing data related to education, and from U.S. News and World Report’s 1998 college ratings. We then determined how the proportion of black student enrollment correlates with survey responses about educational quality and racial discrimination on campus. The relationships were analyzed separately for the student, faculty, and administrator samples. (Of course, the most important of these is the student sample, since their attitudes and experiences are the principle focus of the debate over academic diversity.)
It is commonly believed that increases in black enrollment will produce positive assessments from students about their educational experience. But in fact the correlations went in the opposite direction. As the proportion of black students rose, student satisfaction with their university experience dropped, as did their assessments of the quality of their education and the work ethic of their peers. In addition, the higher the enrollment diversity, the more likely students were to say that they personally experienced discrimination. The same pattern of negative correlations between educational benefits and increased black enrollment appeared in the responses of faculty and administrators. Both groups perceived decreases in educational quality and academic preparation as the number of black students increased. Faculty members also rated students as less hard-working as diversity increased.
The results were more mixed with regard to discrimination and minority relations. Among faculty and administrators, higher minority enrollment was significantly associated with perceptions of less campus discrimination and, among administrators, more positive treatment of minority students. But these findings were offset by the absence of similar results among students, who reported more personal victimization as diversity increased.
Of course, correlation is not causation, and it is possible that these correlations might be masking the effects of other factors. For example, the association of diversity with negative educational experiences might be due to racial discrimination that traps black students in less desirable schools. In order to control for these possible factors, we used the statistical tool of multiple regression, which isolates the separate effect of each potential causal agent from all others. Each evaluation of college life that produced a significant correlation with enrollment diversity was tested by a regression that included a host of background variables taken from the most recently available NCES and U.S. News data. These included: first, individual demographic traits such as race, gender, religion, age, marital status, citizenship, income, and parental education levels; second, academic or institutional factors such as public vs. private schools, selectivity (proportion of applicants admitted), student-faculty rat io, faculty experience and academic success, proportion of full-time vs. part-time faculty, proportions of faculty disciplines and student majors; and third, number of student organizations and proportion of students living on campus. In this way, we were able to tease out the remaining associations of enrollment diversity with educational outcomes and racial relations. (1)
The regression analysis showed that, even after controlling for these demographic, academic, and institutional factors, enrollment diversity still contributed significantly to students’ evaluations of college life. Once again, the greater the school’s diversity, the less students were satisfied with their own educational experience. In addition, greater diversity was associated with perceptions of less academic effort among students and a poorer overall educational experience. Finally, enrollment diversity was positively related to students’ experience of unfair treatment, even after the effects of all other variables were controlled. (As the proportion of black students grew, the incidence of these personal grievances increased among whites. Among blacks, however, there was no significant correlation. Thus diversity appears to increase complaints of unfair treatment among white students without reducing them among black students.)
The regression analysis of the faculty sample also showed that enrollment diversity contributed independently to variation in all three measures of the educational environment. With all other variables controlled, enrollment diversity was inversely related to faculty satisfaction with the quality of education, the work effort of the student body, and the academic readiness of students. The administrators’ judgments of student preparation and the quality of the educational experience were similar. On the other hand, the association of diversity with more positive faculty perceptions of the treatment of minorities, and with both faculty and administrators’ perceptions of less campus discrimination, held true. It is also notable that the respondent’s race contributed independently to evaluations of academic readiness. White faculty members were found to have a more positive evaluation of students’ academic skills. This would seem to preclude an explanation of the findings as the product of negative stereotyping on the part of whites.
Is diversity beneficial?
For historical and legal reasons the idea of diversity has been embraced on campuses most frequently to defend enrollment preferences for African-American students. But it has become far more general in its application. Colby College President William Adams is representative in his claim that it is a “fundamental truth” about American higher education that students “learn more and more powerfully, in settings that include individuals from many different backgrounds and perspectives … [and that] produce the kind of conversations and thinking our students will need to be successful in the world and to contribute to it in meaningful ways.”
To test this broader application of the model, we applied our statistical analysis to Asian-American and Hispanic enrollment. A few questions produced significant correlations with increasing Hispanic enrollment: Students were less likely to express satisfaction with the education they were receiving, and faculty and administrators were less likely to perceive discrimination at their schools. When these correlations were tested by the more rigorous method of multiple regression, the findings for students and administrators remained statistically significant.
Increases in the proportion of Asian-American students were associated with favorable perceptions among students, faculty, and administrators. Among students, the only significant correlation was between higher proportions of Asian-American students and perceptions of a hard-working student body. For faculty and administrators, increases in Asian-American enrollment were correlated with positive perceptions both of the work ethic on campus and of student preparation. For both samples, however, these correlations disappeared when the impact of other variables was factored in. Finally, after the effects of all other variables were controlled, faculty members were less likely to perceive discrimination on campus as the proportion of Asian-American students rose.
This study grew out of our concern that many in the social sciences proclaim the benefits of racial diversity without having subjected these supposed relationships to appropriate empirical tests. When student evaluations of the educational and racial atmosphere were correlated with the percentage of black students enrolled at a college or university, the predicted positive associations with educational benefits and interracial understanding failed to appear. The statistically significant associations that did appear were the opposite of those predicted. These results were clear and consistent as to student impressions of education quality. They were echoed by equally consistent patterns of responses among faculty and administrators. The findings on race relations were more mixed, but certainly insufficient to support the generally held consensus. Among faculty and administrators, diversity brought perceptions of better race relations, but among students, increased diversity was accompanied by more personal ex periences of discrimination.
U.S. Commission on Civil Rights Chairman Mary Frances Berry recently averred, “It seems to me that if racial diversity is a worthy goal, rather than people squirming around to address race, they should acknowledge there is nothing wrong with giving a preference here.” To the contrary, our findings suggest that not all forms of diversity are created equal. The increased presence of black and Hispanic students has not led to the expected improvements. Meanwhile, the increased presence of Asian Americans seems to have at least some positive impact.
Since higher percentages of black and Hispanic students are produced in part by affirmative action, while the same is not true for Asian-American students, it may be that affirmative action places students in academic environments for which they are unsuited, leading to tension and dissatisfaction all around. The results we have found should lead to further study of interracial relations on campus, the educational backgrounds of minority students, and the academic effects of affirmative action. Obscuring existing problems by trumpeting ever louder the benefits of diversity is no way to help students of any race.
(1.) Supporters of diversity often argue that the presence of multicultural curricula will mitigate academic deficits of minority students by helping students of different races to learn to work together and more fully understand each other. However, when we controlled for the presence of such curricula, such as black studies departments or programs, no change in the findings occurred.
STANLEY ROTHMAN is Mary Huggins Gamble Professor of Government Emeritus at Smith College and co-author of The Least Dangerous Branch? (Praeger Press, 2002). SEYMOUR MARTIN LIPSET is the Hazel Professor of Public Policy at George Mason University and author of It Didn’t Happen Here (W.W. Norton, 2000). NEIL NEVITTE is professor of political science at the University of Toronto and editor of Value Change and Governance (University of Toronto Press, 2002).