|
Response to Diversity Distorted: How the University of Michigan Withheld Data to Hide Evidence of Racial Conflict and Polarization by Robert Lerner and Althea K. Nagai
By Dr. Lerner and Dr. Nagai charge that the University of Michigan, through the Michigan Student Study and through Patricia Gurin’s expert testimony in the two affirmative action law suits, misstated and ignored “key findings contained in the secret reports that show serious racial tension and polarization among groups on campus that get worse the longer students stay at the university.” They make these accusations based on a press release authored by Chetly Zarko and on two preliminary reports that Mr. Zarko obtained and that he claims counter Gurin’s testimony. We have already responded, charge-by-charge, to Mr. Zarko. In this document, we are responding to Drs. Lerner and Nagai, whose website report ( http://www.ceousa.org/hiddendata.pdf ) primarily focuses on the second of the two reports. It is equally incorrect. • The reports were not secret. The charges that these reports were “kept secret,” that they were “unknown to outsiders,” that they became available only when Chetly Zarko “uncovered” them from the University’s “sealed archives” are entirely false. One of the reports to which Mr. Zarko referred was provided to counsel for the Plaintiffs in the litigation challenging the University’s admissions policies, March 23, 1998. Both were presented to public gatherings. The issues discussed in both are updated and included in the final summary of the Michigan Student Study that is on the Michigan website. The report that was given to the Plaintiffs in the litigation was developed for a University of Michigan retreat in 1994 hosted by the president and provost and involving about 50 people including executive officers, deans, and other representatives from their offices. The draft report was an early analysis of the first two years of data available from the Michigan Student Study. This early analysis was not "secret." This document was shared widely with administrators, faculty, and students across the University of Michigan campus, and a number of follow-up presentations on the data were given both on campus and across the country. It was sent to counsel for the Plaintiffs on March 23, 1998 (bearing the production numbers UMA 017529-51). This is the very document that Chetly Zarko in the Wall Street Journal on May 16, 2003, describes as “consigned to the ashbin of history” and as “hidden.” This is the very same document that Center for Individual Rights refers to in its press release on May 20, 2003, as “never-released,” “secret,” and a “late disclosure.” The CIR has now retracted its press release and apologized for having made these claims. The other “report” was not even a report, but a set of 13 bar charts without accompanying narrative. The charts were used in a verbal presentation to a workshop on the social psychology of culture, contact, and conflict, sponsored by the Russell Sage Foundation, in Ann Arbor, Michigan in 1994. This was an open workshop attended by researchers from both Michigan and UCLA. There was nothing secret about these bar charts. • The reports do not undermine the claims of Professor Gurin’s report in Compelling Interest that experience with diverse peers has educational benefits for all students. None of the findings from the two reports Drs. Lerner and Nagai characterize as secret and undermining is relevant to the question Patricia Gurin examined for the litigation in the affirmative action lawsuits. The question Dr. Gurin answers in the expert report is: Does experience with diverse peers — in classrooms and in informal interactions on campus — have educational benefits for students? This relationship is central to the issue of compelling interest before the Supreme Court, and was entirely different from the focus of the two reports. All of the findings that Mr. Zarko and Drs. Lerner and Nagai use from these two preliminary reports refer to students’ perceptions of racial climate on the Michigan campus or to their attitudes about specific university affirmative action and multicultural education policies. None of them refer to students’ own educational outcomes or their own experiences with diversity. • Findings on perceptions of racial climate and attitudes toward racial policies do not address the ultimate question of how experience with diversity affects students’ educational outcomes. Dr. Gurin’s testimony did not cover perceptions of racial climate or policy attitudes because they do not tell anything about students’ educational outcomes or their own experiences with diverse peers. Patricia Gurin focused on specific educational outcomes in the three datasets that she analyzed because they follow from the theory she presented on how ethnic and racial diversity could benefit students educationally. She was perfectly clear in her expert testimony what those outcomes were: intellectual engagement, motivation for active thinking, preparation for citizenship. She was perfectly clear why they were selected and what the measures were. She was also perfectly clear why actual experience with diversity — in the classroom and in the peer environment — was crucial and what measures she used for actual experience. The Center for Individual Rights knew exactly what the Gurin Report covered and also what the overall Michigan Student Study covered because the Plaintiffs requested copies of the Michigan Student Study survey instruments. They were produced on June 2, 1999. Patricia Gurin made a judgment of what educational outcomes follow from the theory she presented and used measures from all three datasets that represented those outcomes. To now suggest that she should have focused on some other outcomes is entirely illegitimate, as the CIR had ample opportunity to critique what she examined during the discovery process. They did not. • Perceptions of interracial friendships on the campus are not proxies for the students’ own interracial relationships. Perceptions of interracial friendships cannot be used as proxies for the students’ own experiences. Let us contrast what Drs. Lerner and Nagai conclude from students’ perceptions of interracial friendships on the Michigan campus with what students report about their actual interracial relationships. Using a question asking students their perceptions of how much friendship exists between students of color and white students, the tables presented at the 1994 research workshop show that by the end of the first year at Michigan1 72% of African American and Latino(a) students, 76% of white students, and 84% of Asian American students perceived “quite a bit” to “a great deal.” By the end of the second year, these percentages had declined to 51% for African American students, 66% for Latino(a) students, 70% for white students, and 76% for Asian American students. Sometimes Lerner and Nagai speak of these declines as what they are — declines in perceptions of interracial friendships — and at other times they imply that these declines tell something about students’ actual friendships: “Friendships among students of different races continue to decline, all the way to the fourth year.” But these percentages are not about the students’ own friendships. • Students’ own interracial friendships increased somewhat over the four years at Michigan. The great majority of students felt that their interracial friendships were positive throughout the four years. What do we know about the students’ own friendships? Patricia Gurin pointed out in her expert report (which both the Center for Individual Rights and various critics of affirmative action and of the Gurin Report had available to them) that the percentage of African American and white students who had at least one of their six best friends from a racial/ethnic background different from their own increased over the four years at Michigan. (Virtually all of the Asian American and Latino(a) students entered Michigan with at least one such friend since nearly all of them had attended predominantly white high schools, and virtually all of them still had at least one such friend at the end of the senior year.) The expert report stated: “The proportion of white students who had at least one close friend of color (among their six best friends) increased from about one third (32%) at the time they entered Michigan to almost half (46%) four years later. African American students with at least one close friend who was not African American increased from slightly less than half (47%) at time of entrance to slightly more than half (54%) when they were seniors. There was certainly no decrease in number of interracial close friends over the college years and a meaningful increase, especially for white students at Michigan. It is entirely false for Drs. Lerner and Nagai to conclude that “friendships among students of different races continue to decline, all the way to the fourth year” based on students’ perceptions of what happens on the campus rather than on their own friendships. Information on the students’ own friendships was available to Drs. Lerner and Nagai in Patricia Gurin’s expert report. Further, we know that from the end of the first year to the end of the fourth year, the vast majority of students from all racial/ethnic groups agreed: “My relationships with students from different racial/ethnic groups at the University have been positive.” There was no decline in this positive assessment of students’ own interracial relationships. Among African American students the percentages who agreed with this statement were 80% and 79% at the two points in time; among Latino(a) students 82% and 87%; among Asian American students 91% and 94%; and among white students 86% and 91%. How positive or negative a story can or should be told about the students’ own relationships at Michigan depends, of course, on the perspective one has on racial relationships in the United States. Given the fault line that race represents and the nearly totally segregated background of Michigan’s white students, we believe the picture is impressively positive. • Perceptions of amount of racial tension do not give an accurate portrayal of how much tension exists in students’ own interracial relationships. As in the friendship data, perceptions of racial tension do not tell an accurate story about actual racial tension which must be drawn from students’ own interracial relationships. If negative perceptions of the racial climate on a campus prevented students from interacting across race and ethnicity, or prevented them from having positive interracial relationships, those perceptions should be of serious concern to college administrators. This is not the case at Michigan. Very few of Michigan students’ interracial relationships were characterized by “tension and hostility.” Patricia Gurin covered this in her Expert Report. In the fourth year, the MSS asked about the quality of students’ relationships with peers from a different racial/ethnic background. In these questions we asked students to tell about both positive and negative aspects of their relationships with students in the group with which they interacted most frequently. This question was not asked earlier. It was added in the senior survey to obtain a fuller picture of the qualities of the students’ own interracial relationships on campus beyond the question asked about their six best friends. What we learned about positive and negative qualities of interracial relationships on campus was included in Patricia Gurin’s expert testimony. Looking first at relationships reported by white students, we learned that only 7% said that their relationships with Latino(a) students involved “a great deal” or “quite a bit” of “tension and hostility.” The comparable percentage for white students in interactions with Asian American students was only 1% and with African American students only 4%. This is considerably fewer than the 27% of white students who perceived “a great deal” or “quite a bit” of interracial tension on campus when asked the question about the general racial climate on campus. Nor do perceptions of tensions accurately reflect tension in the actual friendships that Latino(a) and Asian American students report that they had with white students. These reports of negative experiences in their own relationships with white students were reported in Patricia Gurin’s expert testimony. Approximately l0% of both of these groups reported “quite a bit” or “a great deal” of “tension, hostility” in their relationships with white students. Again, this is fewer than the 32% of the Latino(a) and 22% of the Asian American students who perceived “quite a bit” or “a great deal” of interracial tension on the campus.3 More African American students reported negative qualities in their relationships with white students, with 15% saying that those relationships involved “quite a bit” or “a great deal” of “tension, hostility.” Among African American students, however, experiencing tension in their own relationships with white students was dramatically less frequent than the 59% who perceived “quite a bit” or “a great deal” of interracial tension on the campus.4 In summary,
• The discussion of increased perception of racial tension by Drs. Lerner and Nagai misrepresents the Michigan Student Study results. Drs. Lerner and Nagai exaggerate what the Michigan data show about perceptions of tension. One table Drs. Lerner and Nagai present indicates that the perception of “quite a bit” or “a great deal” of “interracial tension in the residence halls” increased 4% to 9% between the first and second years of college across racial/ethnic groups. Drs. Lerner and Nagai make a great deal of the increase in perception of amount of racial tension from entrance to the end of the second year, which was slightly greater than between the first and second years. But, of course, students have no basis for evaluating interracial tension in the residence halls before they have any experience in the residence halls. The questions they were asked at entrance were about what they expected to find at Michigan. Drs. Lerner and Nagai imply huge shifts — “tripling” and “doubling.” The changes that Lerner and Nagai label “tripling” are 4% at entrance, 11% at the end of the first year, and 15% at the end of the second year in white students’ perceptions of “quite a bit” or “a great deal” of tension, and 3%, 11%, and 15% for Asian Americans. While tripling is technically correct, the absolute size of these differences is not consonant with Lerner and Nagai’s characterizations of these increases as “dramatic.” This is particularly true if one considers the 4% change between the survey at the end of the first and second years, the only two that refer to the time when they were on the Michigan campus. Drs. Lerner and Nagai further state that the Michigan Student Study “deleted this question about racial tension in the dorms from their fourth year questionnaire” because “the researchers didn’t want to know the results and didn’t want anyone else to know either.” This is simply false and an example of their irresponsible attacks on our integrity. The Michigan Student Study did not include the residence hall question in the fourth year questionnaire because only 5% of Michigan students live in residence halls during the fourth year. Instead, the Michigan Student Study asked a new question to substitute for the residence hall question, one that asked for perceptions of amount of racial tension on the campus. Responses to that question were included and discussed in the final summary of the Michigan Student Study on Michigan’s website. Changing from residence hall to campus in querying perceptions of amount of racial tension doesn’t hide anything. • Perceptions of amount of racial tension are to be expected on campuses where race is salient. It is a serious, negative issue only if it prevents students from having friendships and positive relationships across race and ethnicity. One should not be surprised that perception of some racial tension exists at Michigan or that such perceptions might be expressed by more students on campuses where commitment to racial/ethnic diversity is strong. Nearly all of Michigan’s white students and half of the African American students came to Michigan (at the time of the Michigan Student Study) with practically no experience with diverse peers. They had little to anchor their perceptions. Nor is it surprising that perception of racial tension increases somewhat over the four years in college on campuses, including the University of Michigan, where the issue of race is highly salient. If perceptions of racial tension had resulted in a balkanized campus where students did not develop friendships across race, these perceptions should be a serious concern for Michigan. But, as already indicated, friendships across race actually increased at Michigan over the four years. On campuses where there are few students of color and where there is little emphasis on increasing diversity, perceptions of racial tension will be low (or entirely absent). Homogeneous environments do not bring issues of race to the forefront. Nor is it surprising that perceptions of racial tension might be more widespread on campuses where affirmative action has been politicized. Wood and Sherman, two critics of affirmative action and of Patricia Gurin’s expert report, analyzed the same data collected by the Cooperative Institutional Research Program at UCLA that Gurin analyzed in the expert report and showed that perceptions of amount of racial tension are related to institutional selectivity (which they treat as a proxy for affirmative action.) We note, however, that that CIRP data did not include questions about the presence of tension in the students’ own relationships. Wood and Sherman thus provide no evidence that students’ actual relationships across race were more tense on the most selective institutions. Finally, if perception of racial tension is viewed as something to be avoided at all costs, what is the answer? Is it to return to segregated schools? The response to Senator Trent Lott indicates that the American people do not feel that the answer is to avoid the complexities of living together as a diverse society or to long nostalgically for the segregated America before the election of 1948. It is easy for people to imagine that racial tension did not exist then, and further that racial tension would simply disappear now, with no consequences to society, if diverse groups of Americans did not work and live closely together, and if affirmative action were not instrumental in bringing diverse groups together in our nation’s most selective educational institutions. The amicus brief submitted in behalf of the University respondents by former military generals and admirals challenges that point of view. Although the armed forces became integrated in 1948, there was great disparity between the officer and enlisted corps up to and through the Vietnam War. The military Amici stress how negative that disparity was for the military’s capacity to wage the Vietnam War. “In the 1960s and 1970s, however, while integration increased the percentage of African Americans in the enlisted ranks, the percentage of minority officers remained extremely low (3% of Army officers), and perceptions of discrimination were pervasive. This deficiency in the officer corps and the discrimination perceived to be its cause led to low morale and heightened racial tension. The danger that this created was not theoretical, as the Vietnam era demonstrates. As that war continued, the armed forces suffered increased racial polarization, pervasive disciplinary problems, and racially motivated incidents in Vietnam and on posts around the world…. The military’s leadership ‘recognized that its racial problem was so critical that it was on the verge of self-destruction.’ ”(pp 6-7) The Amici conclude that the military’s realization of its racial problems “set in motion the policies and initiatives that have led to today’s relative positive state of affairs.” (p 7) Through affirmative action initiatives that the military put in place — financial and tutorial assistance, recruiting programs, employing race as a factor in recruiting and admissions policies and decisions, and preparatory academies to increase the pool of qualified minority candidates — 19% of active duty officers are minority. The military is now one of America’s most integrated institutions, and it has become so through affirmative action policies. The military Amici submit that “the government’s compelling interest in promoting racial diversity in higher education is buttressed by its compelling national security interest in a cohesive military. That requires both a diverse officer corps and substantial numbers of officers educated and trained in diverse educational settings, including the military academies and ROTC programs.” (p8) The example of the military demonstrates that ignoring racial issues, however harmonious race relationships might have appeared to be until the military faced a major crisis such as the Vietnam War, actually produced profound racial tension and polarization. Affirmative action in the military has resulted in a much more realistic and healthy state of affairs. Something is clearly awry when some Americans long for the days when racial tension was not perceived but nonetheless was felt — at least by Americans of color — and largely ignored. College students need to learn how to understand perceptions of racial tension and the different perspectives on society and institutions that produce those perceptions. In fact, the assumption that students of various backgrounds have different perspectives on how much racial tension exists, and on many other matters about institutions and society, is part of why we want students of different groups to interact with each other — so they understand and learn from these differences. Societies all around the world are being threatened by racial and ethnic cleavages. Surely, college students today need to learn how to work, live, and be leaders in a diverse society lest our own heterogeneous society and democracy become as hopelessly divided as many in the world are today. • The Gurin Expert Report did not selectively distort information from the Michigan Student Study. We have dealt with Lerner and Nagai’s unwarranted conclusions from the data on perceptions of racial tension because it is the focus of their critique, and because alarms about supposed racial tension on the nation’s college campuses have formed a major critique by the opponents of affirmative action. The Gurin Expert Report provided data, as we noted above, that speak directly to the issue of racial tension — data from the students’ own experiences. It is also important to point out, once again, that the only data that we used from the Michigan Student Study for the Expert Report were those relevant to the regression analyses linking the students’ own diversity experiences to their own educational outcomes, as predicted from our theory. The Michigan Student Study contained many questions on other matters that were not relevant as measures of the students’ experiences with diversity or as measures of their educational outcomes. We excluded all of them, whether or not they might have produced findings that could be construed as “negative” or “positive” in their implications for the lawsuits. • Racial differences in perspectives provide opportunities for students to learn. In addition to the findings on perceptions of racial tension, Lerner and Nagai present and discuss findings from the bar charts on students’ attitudes toward affirmative action, their evaluations of Michigan’s institutional commitment to diversity, and their views about how much white faculty respect students of color. There are two important points to make about this aspect of the Lerner and Nagai charges.
Lerner and Nagai view these differences as evidence of racial polarization and as “negative consequences” of “diversity achieved through racial preference policies.” We view these differences as examples of racially and ethnically-based perspectives in society from which students learn by their interaction with each other. Far from “suppressing” these findings, the Gurin Expert Report stresses different racial/ethnic perspectives as a major reason that actual interaction across race and ethnicity is important for student learning. Moreover, we stress the fact that these differences appear in the students’ responses at the time of entrance to Michigan. Those differences indicate the large racial divide in the wider society that has affected students growing up in largely racially segregated worlds. Some of these entrance-level differences become larger; some become smaller; some stay the same. The point of discussion across race is not that students will end up agreeing with each other or that diversity only works when differences become smaller. The point is for students to understand and learn from the different perspectives. • Counsel for the Plaintiffs and critics of the Gurin Expert Testimony have shown virtually no interest in the Michigan Student Study until an investigative reporter claimed to have unearthed a “secret” document — only weeks before the decision of the Supreme Court. Since the attacks on Gurin’s Expert Report began almost two years ago, the critics from the National Association of Scholars have paid no attention to the Michigan Student Study. They have focused exclusively on our analyses of the national data on colleges and universities collected by the Cooperative Institutional Research Program (CIRP) at UCLA, because the critics have consistently argued that we had to prove that percent minority on a campus has direct impact on students’ educational outcomes. Percent minority is a constant at a single institution. An example of previous inattention to the Michigan Student Study is revealed dramatically in two reports by Drs. Richard Lerner and Althea Nagai. The most recent one, posted on their website May 15, 2003, shows great interest in the Michigan Student Study data. But in their previous critique of Patricia Gurin’s testimony (http://www.ceousa.org/Gurin1.pdf), they claimed that the Michigan Student Study was irrelevant to the lawsuits. On the very first page of the executive summary of that critique, they claimed that: “The bulk of Professor Gurin’s analysis is based on the Cooperative Institutional Research Program (CIRP) data set,” and then noted dismissively: “There are two additional surveys of students from the University of Michigan. There is no school that serves as a comparison (i.e., control) group. These findings should be ignored.” Consistent with that statement, the earlier critique by Drs. Lerner and Nagai dealt only with the results from the national database provided by CIRP. For similar reasons, Wood and Sherman representing the National Association of Scholars Amicus Brief to the Supreme Court also dealt only with the results from CIRP. The omission of the Michigan results reported above and in the Gurin Expert Report is particularly dramatic in Wood and Sherman ( http://nas.org/rhe2.pdf ). They devote a section to the supposed negative effects of interracial tensions on campuses (including Stanford University, UCLA through the eyes of John McWhorter) and quote from national polls but never mention the information on tension in the Michigan students’ own relationships that was available in the Gurin expert testimony. To our knowledge, none of the other Amici in behalf of the Plaintiffs has dealt with the information made available on Michigan students in the Gurin Expert Report. Now these critics are questioning our conclusions about the generally positive quality of interracial interactions on Michigan’s campus but they appear to be uninformed that these conclusions came from information that was part of the Gurin Expert Report Why do these critics now suddenly feel that the Michigan Student Study is so relevant to the lawsuits and that they completely undermine Patricia Gurin’s testimony? It seems obvious that this sudden interest in the Michigan Student Study is tied to the charge that material on perception of amount of tension was “secret” and “hidden.” In their retraction about the charge of “secrecy,” the CIR still maintains that findings from the Michigan Student Study that they once considered irrelevant nonetheless indicate that we are selectively distorting evidence. • Drs. Lerner and Nagai leave out important information from the material they critiqued. Drs. Lerner and Nagai report selectively from the bar charts in the 1994 hand-out. The hand-out from the research workshop contained 13 bar charts comparing entrance (or end of the first year) responses with senior year responses to questions on perceptions of racial climate and attitudes toward affirmative action and diversity. Lerner and Nagai report on only six of the 13 charts, which they turn into tables. Four of the six findings that they report for white students change in what they consider a “negative” direction; none change in a “positive” direction. In the omitted seven tables, four indicate a significant increase in “positive” responses to affirmative action and what students have learned from their diversity experiences. One of these seven tables showed a “negative” change. One of the omitted bar charts is especially important because it covers attitudes toward affirmative action policies that affect the admissions process. The findings indicate that from entrance to the senior year white students show a significant increase in agreeing that “a high priority should be given to see that students of color receive financial aid for education after high school.” (Another Michigan Student Study finding is relevant here. Although it was not included in the bar charts on which Lerner and Nagai were commenting, it was presented and discussed in the four year report on the Michigan website to which Lerner and Nagai had access and actually referred to in writing their current critique. This finding also indicates a significant increase from entrance to fourth year in white students’ agreement that “different admissions criteria with respect to SAT and ACT scores may be justified for some students of color.”)5 Conclusion The task that Patricia Gurin undertook was to examine the educational impact of these interracial relationships for students at Michigan. It is clear from the analyses that she carried out, and reported both in the Expert Report and two peer-reviewed journal articles that actual experience with diverse peers fostered learning and democracy outcomes for all groups of students. None of the findings in the tables presented and discussed by Drs. Lerner and Nagai contradict or undermine this conclusion. Footnotes 1 We are emphasizing the percentages at the end of the first year of college since that provides some basis for students to have perceptions about friendships on the campus. Their expectations at time of entrance are interesting, and were included in the 1994 bar charts that Drs. Lerner and Nagai use, but they are expectations and not perceptions of campus life. 2 These figures vary slightly (1-2% percentage points) from the figures given in Table 16 of the final report of the Michigan Student Study on Michigan’s website. These figures come from the longitudinal sample of students who answered both the end of the first year and end of the fourth year questionnaires. The figures in the Michigan Student Study final report come from all students who answered the fourth year survey.3 As indicated in the Gurin Expert Report, the proportions of white, Asian American, and Latino(a) students who reported tension in their own interracial relationships were dramatically outweighed by the proportions (ranging from 39% to 85%) who saw these relationships as involving extensive cooperation (“studied together”) and intimacy (“shared personal feelings and problems”). 4 As indicated in the Gurin Expert Report, the positive and negative qualities of the friendships African American students had with white students were more nearly equivalent. We never hid — in any reports or in the Gurin Expert Report — that from the perspective of African American students, relationships with white students were somewhat more tense and somewhat more distant than were the relationships of Asian American and Latino(a) students with white students. However, white students did not experience their relationships with African American students as being more tense than their relationships were with other students of color. 5 These findings on white students’ increasing support for affirmative action admissions policies over their four years at Michigan were not reported in the Gurin Report. The Expert Report did not report any findings not directly related to the analyses between students’ own diversity experiences and their educational outcomes — regardless of their positive or negative implications for the lawsuits. Bibliography Matlock, John, Gurin, Gerald and Wade-Golden, Katrina. Michigan Student Study final report, available at http://www.umich.edu/~oami/mss/ Lerner, Robert, and Nagai, Althea K. A Critique of the Expert Report of Patricia Gurin in Gratz v. Bollinger, (pdf) available at http://www.ceousa.org/Gurin1.pdf Lerner, Robert, and Nagai, Althea K. Diversity Distorted: How the University Of Michigan Withheld Data to Hide Evidence of Racial Conflict and Polarization, May 15, 2003, available at http://www.ceousa.org/hiddendata.pdf Wood, Thomas E. and Sherman, Malcolm J. Supplement to Race and Higher Education, January 2003, (pdf) available at http://nas.org/rhe2.pdf |