The United States Supreme
Grutter v. Bollinger
Barbara Grutter sued Lee Bollinger,
then president of the University of Michigan, alleging unlawful racial preferences
in admission to the university's law school. In federal district court, March
27, 2001, Judge Bernard Friedman held the law school's race-based admissions to
be unconstitutional. Beginning at page 3 of his opinion, Judge Friedman gives
the following description of the pertinent facts.
The starting point in this case is the written admissions policy of the University of Michigan Law School, which was admitted at trial as Exhibit 4. This policy, which was adopted by the law school faculty in April 1992, was the subject of many hours of testimony during trial as well  as extensive discovery. Due to the central role the policy has played in this case, the court shall summarize the policy and highlight certain provisions.
The policy expresses the law school's desire "to admit a group of students who individually and collectively are among the most capable students applying to American law schools in a given year. . . . Collectively, we seek a mix of students with varying backgrounds and experiences who will respect and learn from each other." Exhibit 4, Admissions Policy, p. 1. The policy notes that "no applicant should be admitted unless we expect that applicant to do well enough to graduate with no serious academic problems." Id. at 2. In identifying applicants who can be expected to succeed academically, the law school's "most general measure . . . is a composite of an applicant's [Law School Admission Test] score and undergraduate gradepoint average (UGPA) (which we shall call the 'index')." Id. at 3.
Under this admissions policy, the law school pays close attention to LSAT scores and UGPA's in reviewing applications. The significance of these numbers is visually apparent from the "grid" of law school applicants, an example of which is attached to the law school's admissions policy.5 LSAT scores are shown along the horizontal axis in three- or four-point increments; UGPA's are shown along the vertical axis in quarter-point increments. Every combination of LSAT and UGPA is shown in a "cell" on this grid. In each cell, the law school reports the number of applicants with that particular combination of numerical qualifications, as well as the number of  offers of admission made to the applicants in that cell.6 Constructed in this manner, the highest combinations of LSAT scores and UPGA's are found in the upper right-hand corner of the grid. Even a cursory review of the numbers contained in this grid reveals that one's chances of being admitted increase dramatically as one moves into the upper right corner. Of the 966 offers of admission made in 1991, 843 (87%) were made to applicants who fell within the nine cells closest to this corner. In short, the numbers reflect the law school's stated policy: "Bluntly, the higher one's index score, the greater should be one's chances of being admitted. The lower the score, the greater the risk the candidate poses. . . . So we expect the vast majority of those students we admit to have high index scores." Id. at 4. See also id. at 6-7 ("The further applicants are from the upper right corner the less likely they are to be offered admission. Thus we may think of the upper right portion of the grid as indicating the combinations of LSAT and UGPA that characterize the overwhelming bulk of students admitted.")
The policy also notes, however, that admissions decisions should not be made strictly based on the index scores. A high index score may not necessarily identify an applicant who is likely to succeed in law school, and a low index score may not necessarily identify one who is likely to fail. See id. at 4-5. The policy states:
Id. at 5-6.
Thus, while the policy indicates that most offers for admission should be made to applicants with high "index" scores, the policy also states that "considerable discretion is exercised in the admissions process . . . , for many qualities not captured in grades and test scores figure in the evaluation of an application." Id. at 7. The reasons behind the exercise of this discretion are an important part of the admissions policy, and they also lie at the heart of the dispute between the parties in this case. The policy articulates two reasons why an offer of admission may be made to applicants with grades and test scores "that place them relatively far from the upper right corner of the grid." Id. at 8. The first reason is that "there are students for whom we have good reason to be skeptical of an index score based prediction." Id. As an example, the policy describes an applicant who received a poor SAT score but nonetheless went on to perform well academically in college. This applicant received an unimpressive LSAT score, which pulled down his index, but he was  admitted on the strength of his undergraduate record "with the expectation that this record would be a better predictor of [his] performance at Michigan than his LSAT score." Id. at 9.
The second reason for admitting applicants with comparatively lower index scores is that they "may help achieve that diversity which has the potential to enrich everyone's education and thus make a law school class stronger than the sum of its parts. In particular we seek to admit students with distinctive perspectives and experiences as well as students who are particularly likely to assume the kinds of leadership roles in the bar and make the kinds of contributions to society discussed in the introduction to this report." Id. at 9-10. The policy identifies these as "diversity admissions," id. at 10, and provides three examples of applicants who were admitted in 1991 under this rubric. One was born in Bangladesh, graduated from Harvard, received "outstanding references" from his professors, and had a "truly exceptional record of extracurricular activity." Id. Another was a single mother from Argentina who worked successfully in business for several years, graduated from college summa cum laude, received "glowing references," and was fluent in four languages. Id. at 10-11. The third had excellent grades and a good LSAT score; "diversity considerations" further strengthened her application because her parents were Greek immigrants, she was "immersed in a significantly ethnic home life," she had studied and traveled in Europe and was fluent in three languages. Id. at 11.
In addition to the type of diversity that may come, for example, from an applicant's interesting or unusual employment experiences, extracurricular activities, travel experiences, athletic accomplishments, volunteer work, or foreign language fluency, the admissions policy also describes the importance of an applicant's race as a qualification which may make him a more attractive candidate for admission: 
Id. at 12.
The law school's admissions policy is further described in an October 13, 1992, memorandum from defendant Dennis Shields, who was the Director of Admissions at the law school from 1991 to 1998. This memorandum (Trial Exhibit 5), entitled "The Gospel According to Dennis," was addressed to "File Readers 1992-93," and was intended to provide guidance to his staff who assisted him in reviewing application files. Mr. Shields explained his philosophy as follows:
Mr. Shields went on to explain that in reviewing an application file he looked first at an applicant's grades and LSAT score. "My view is we will ultimately be swayed in any case by the strength of the numbers so it makes sense to know what they are before one proceeds to judge the rest of the file." "The numbers" referred to the LSAT score, the cumulative GPA, the undergraduate institution, the trend in grades, and the rank at the undergraduate institution. In addition, Mr. Shields indicated that in reviewing transcripts he looked to see if applicants chose a rigorous major, whether they pursued a liberal education, and whether they took difficult courses. After reviewing "the numbers," Mr. Shields proceeded to the essays and the letters of recommendation. The essays reveal an applicant's writing and intellectual abilities and interests, and "what the candidate might offer the academic enterprise which is legal education at Michigan." The letters of recommendation are useful because they provide information about the applicant's academic abilities, particularly "growth and improvement or other trends" and "the general rigor of the curriculum pursued by the student." Finally, Mr. Shields noted that "[t]here is a preference for those who have demonstrated academic and/or intellectual strength. Hard work and discipline evidenced in file [sic] are also important. I also like to see people who have been challenged in one way or the other and have faced up to that challenge in a positive way."
The law school's admissions policy is also succinctly described in the University of  Michigan Law School Bulletin. The bulletin for the 1996-1997 academic year, which was admitted as Trial Exhibit 6, states at page 81:7
In the 1997-1999 law school bulletin (Exhibit 8, p. 85), the first paragraph of this passage is repeated, but the second paragraph was rewritten as follows: 
At trial, a number of witnesses testified as to the admissions procedures prior to the adoption of the 1992 policy; the reasons why, and the process by which, the 1992 policy was adopted; and the manner in which the 1992 policy has been administered. This testimony assisted the court in understanding the genesis of the 1992 policy and how it works in practice.
Allan Stillwagon was the law school's Director of Admissions from 1979 to 1990. He testified that during his tenure admissions decisions were made in accordance with the policy described at pp. 85-86 of the 1988-89 Law School Announcement (Exhibit 55), which states:
Mr. Stillwagon testified that under this system, which was in effect throughout his tenure, applicants were selected from one of three groups. The first group consisted of applicants who were chosen based on "the numbers," that is, their LSAT scores and UGPA. The second consisted of a "pool" of candidates who had lower numbers but other interesting qualities. The third, known as "special admissions," was for minority candidates who did not fall within the other two groups. According to Mr. Stillwagon, approximately one-half of the minority applicants who were admitted came from the first two groups, and the other half came from the third. The "special admissions" program was adopted in order to increase minority enrollment at the law school.8 Mr. Stillwagon indicated that  the law school had a "goal" or "target" whereby 10-12% of the students of each entering class should be Black, Chicano, Native American, and mainland Puerto Rican. The law school faculty increased this percentage in the 1970s from 10% to 10-12% because they believed it necessary to increase the representation of minorities in the legal profession. Mr. Stillwagon testified that he had no discretion to disregard this policy, and that the policy was considered flexible only to the extent that the number of minority admittees could deviate by three or four students on either side of the target range. Mr. Stillwagon also testified that the 10-12% target could be achieved only through the special admissions program due to the "considerable differences" in academic credentials between the minority and non-minority applicants.
Mr. Stillwagon testified that Exhibits 112 and 113 show admissions statistics for 1988-1989 and 1989-1990, respectively. These reports to the law school Committee of Visitors show significant differences in the numerical qualifications between "regular" admissions (i.e., students admitted based on LSAT and UGPA and those admitted from the "pool") and "special" admissions. In 1988, the regular admissions had a median LSAT score of 43 and a median UGPA of 3.58, whereas the special admissions had a median LSAT score of 34 and a median UGPA of  3.05. This gap in the numbers was essentially the same in 1989 and 1990. 9 On cross-examination, Mr. Stillwagon testified that explicit consideration of race was necessary since otherwise very few minority students would have been admitted.
Dennis Shields succeeded Allan Stillwagon as the director of the law school's admissions office in 1991 and he held this position until 1998. Mr. Shields testified that his memo entitled "The Gospel According to Dennis" was intended to give newcomers to his staff some guidance in reading application files. Mr. Shields did not mention race in this memo because this is not a "primary consideration" in making admissions decisions. Mr. Shields indicated that he did not tell his staff to strive to admit a particular percentage of minority students, but that an applicant's race was considered along with all other factors. Mr. Shields said that he never spoke with the law school dean about the number or percentage of minority students who should be admitted, although the dean did tell him that approximately one-third of the class should consist of Michigan residents because the law school is a state institution. Mr. Shields also testified that the minority and non-minority admittees were all well qualified for admission.
On cross-examination, Mr. Shields was asked about the manner in which he would use the so-called "daily reports," an example of which was admitted as Exhibit 10. These reports provide an overview, as of the day the report is generated, of the number of applications and their current status. That is, one can see at a glance the number of applications received to date, the number offered admission, the number rejected, the number on the waiting list, and so on. While  the first page of the report provides an overview for the total applicant pool, each of the next seven pages breaks down the information by the following racial categories: Native American, Black/African American, Caucasian/White, Mexican American, Other Hispanic, Asian merican, and Puerto Rican American. Page 9 is devoted to "Other/Non-Citizen" and page 10 is "unknown." The last four pages break down the applicants by gender and by their status as either Michigan or Non-Michigan residents. Mr. Shields testified that as an admissions season progressed, he would consult the daily reports more and more frequently in order to keep track of the racial and ethnic composition of the class. This was done in order to ensure that a "critical mass" of minority students were enrolled so as to realize the educational benefits of a diverse student body. Mr. Shields could not say what percentage was needed in order to achieve this goal. He doubted if five percent would be enough but thought that 10% might suffice. While Mr. Shields testified that he did not seek to admit a particular number or percentage of underrepresented minority students, he acknowledged that during his tenure at least 11% of each entering class consisted of African American, Hispanic and Native American students.
Mr. Shields' attention was drawn to Exhibit 15, which is the law school's admissions grids for the class entering in 1995. Like the law school's daily reports, the admissions grids are broken down by racial groups. In each cell of each grid, one can see the number of applicants who applied, the number who were accepted, and the number who enrolled for any given combination of LSAT score and undergraduate GPA. When asked why most white applicants are rejected and most black applicants are accepted in the mid-ranges of LSAT scores and UGPA, Mr. Shields  acknowledged that race does account for some of the difference.10
Erica Munzel replaced Dennis Shields as the director of the law school's admissions office in 1998. She testified that she feels bound by the law school's 1992 admissions policy, including the provision that calls for the enrollment of a "critical mass of minority students." See Exhibit 4, p. 12. Ms. Munzel testified that a "critical mass" means "meaningful numbers" or "meaningful representation," which she understands to mean a number that is sufficient so that the minority students can contribute to classroom dialog and not feel isolated. When pressed to express this concept in numerical terms, Ms. Munzel stated that there is no number or percentage, or range of numbers or percentages, which constitute critical mass. However, she stated that there must be more than a "token" number of minority students, since small numbers of students cannot contribute in the manner foreseen by the law school's diversity policy.
Ms. Munzel also indicated that she must consider the race of the applicants because a critical mass of minority students could not be enrolled if admissions decisions were based primarily on LSAT scores and UGPA's. This is apparent from Exhibit 14, which Ms. Munzel acknowledged shows the median LSAT scores and UGPA's for the students enrolled in 1994 and 1995. These figures show the same gap in the numbers between minority and non-minority  admittees as is shown in similar exhibits for previous years.11 Ms. Munzel testified that only ten minority students per entering class would be admitted if admissions decisions were driven by the numbers.
In deciding whether to admit an applicant, Ms. Munzel stated that she reviews the entire file. In addition to grades and test scores, she also considers the strength of the undergraduate curriculum, the college attended, the personal statements, letters of recommendation, and the applicant's background and experiences. The student's race is considered because it is relevant to achieving diversity in the entering class. Ms. Munzel stated that she was never told by the dean or by the faculty to admit a particular number or percentage of minority students, but she does consult the daily reports (such as Exhibits 11 and 12) to make sure that admissions goals, including those regarding the admission of a critical mass of minority students, are being achieved.
The court also heard testimony from Lee Bollinger and Jeffrey Lehman as to the reasons why race is considered in the admissions process. Mr. Bollinger was the dean of the law school from 1987 to 1994, and he has been the president of the University of Michigan since 1997. Mr. Lehman succeeded Mr. Bollinger as the dean of the law school in 1994, and he continues to hold this position. President Bollinger testified that in the fall of 1991 he convened a faculty committee  to review the law school's admissions policy.12 President Bollinger sought to ensure that the policy complied with the Supreme Court's ruling in Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265 (1978). He believes that it is appropriate for an applicant's race to be considered because the law school seeks a student body with diverse backgrounds and perspectives.
Dean Lehman testified that racial diversity in the student body is an important part of one's education at the law school because exposure to students of various races and perspectives helps students to understand and be sympathetic to differing points of view. He described racial diversity as "part of the general commitment to diversity." Dean Lehman agreed with the testimony offered by other witnesses to the effect that the law school seeks to admit a critical mass of underrepresented minority students, particularly those from groups which have been discriminated against historically. He was unable to quantify "critical mass" in terms of numbers or percentages, or ranges of numbers or percentages, but indicated that critical mass means "meaningful numbers," that is, numbers such that the minority students do not feel isolated or like spokespersons for their race, and feel comfortable discussing issues freely based on their personal experiences. He doubted whether critical mass would be present if only five percent of a class consisted of minority students, and he acknowledged that minority students have constituted at least 11% of every entering class since 1992.
When asked about the extent to which race is considered in admissions, Dean Lehman testified that this varies from one application file to another. In some files the applicant's race may  play no role, while in others it may be a "determinative" factor.13 Dean Lehman indicated that race is taken into consideration to the extent necessary to achieve a critical mass, although he could not quantify this in terms of numbers or percentages. While Dean Lehman reviews the daily admissions reports with the admissions director, he said he has not given any direction as to a number or percentage of minority students who should be admitted. Dean Lehman also stated that a critical mass of minority candidates cannot be admitted unless race is explicitly considered, due to the gap in LSAT scores and UGPA's between minority and non-minority students. He fears that minority enrollment would drop to "token levels" if race could not be considered, and in this context he pointed to the experience of the University of California at Berkeley, where minority enrollment dropped sharply after passage of Proposition 209.
The court also heard extensive testimony from Professor Richard Lempert, the law school professor who chaired the faculty admissions committee that drafted the 1992 admissions policy.14 Professor Lempert testified that the admissions committee was charged with examining the law school's admissions policy and ensuring that it complied with the Supreme Court's ruling in Bakke. The 1992 written policy, which was conceived in a "very deliberative process," was debated and then adopted unanimously by the full faculty. It remains in effect today. Professor Lempert emphasized that the law school seeks to admit an interesting and  dynamic class, which has a certain "synergy" that is greater than the sum of its parts. That is, the law school seeks students with a diversity of interests and backgrounds in order to enhance classroom discussion and the educational experience of students and faculty, both in and outside the classroom. In Professor Lempert's view, racial diversity is an important part of "perspective" or "experiential" diversity. A critical mass is needed so that minority students do not feel that they must be spokespersons for their race.
When asked about the policy's "commitment to racial and ethnic diversity with special reference to the inclusion of students from groups which have been historically discriminated against," see Exhibit 4, p. 12, Professor Lempert stated that this was not intended as a remedy for past discrimination, but as a means of including students who may bring to the law school a perspective different from that of members of groups which have not been the victims of such discrimination.15 Professor Lempert indicated that race is one element in the admissions decision-making process, but that some minority applicants would be admitted even if their race were not considered. He believes that race is considered only to the extent necessary to achieve critical mass.
Exhibit 34 is a draft of the admissions policy, which contains several provisions omitted from the final version. One such provision, on page 13 of the draft, states:
Professor Lempert testified that the "11% to 17%" figure, which is the range he believes constitutes critical mass, was omitted from the final version of the admissions policy because percentages were too rigid and could be misconstrued as a quota. Another provision omitted for the same reason stated on page 12 of the draft that "non-grid admittees admitted for diversity purposes shall not exceed 20% of the expected matriculants in a class." One faculty member, Don Regan, argued for retaining the "numbers on the 'target range' . . . [f]or a variety of reasons, including candor." See Exhibit 32, p. 1.
The final witness who testified about the law school's policy, and the reasons for the inclusion of race as a factor in admissions decisions, was Kent Syverud. He was a professor at the law school when the 1992 admissions policy was adopted, and he is now the dean of Vanderbilt Law School. He has also submitted expert reports on the educational benefits of diversity. See Exhibits 153, 154, 155. Like the other witnesses who testified on this subject, Dean Syverud believes that racial diversity is part of the diversity of perspectives needed to enhance the "classroom dynamic." Also like the other witnesses, he indicated that critical mass cannot be quantified, but that a professor knows when it is present because minority students feel free to express their views, rather than to state "expected views" or "politically correct views." Dean Syverud also indicated that when a  critical mass of minority students are present, racial stereotypes are dismantled because non-minority students see that there is no "minority viewpoint"; they see, in other words, that there is a diversity of viewpoints among minority students.
While defendants concede that race is a factor in the admissions process, they have consistently argued that race is simply one of many factors and not a "trump card." Plaintiffs, on the other hand, have argued that race is a "super factor" in the admissions process. In an attempt to quantify the extent to which race actually has been considered during the years in question, the parties presented expert testimony, and expert reports, from two statisticians. Plaintiffs presented Dr. Kinley Larntz, a professor emeritus in the Department of Applied Statistics at the University of Minnesota.16 Defendants presented Dr. Stephen Raudenbush, a professor of education at the University of Michigan.17 Both were qualified as experts in statistics and both testified at great length.
Dr. Larntz analyzed admissions data provided by the law school. This data consists of the "admissions grids" for each of the years in question (1995-2000). As noted previously, these grids show the number of applicants and the number of admittees for all combinations of  undergraduate GPA and LSAT score.18 The UGPA is presented along the vertical axis in quarter-point increments (as well as "below 2.00" and "No GPA"); the LSAT score is presented along the vertical axis in two- or three-point increments (as well as "170-Above," "120-145" and "No LSAT"). The ten UGPA rows and twelve LSAT columns in these grids produce 120 "cells" with admissions data. For each of the years in question, the law school compiled one admissions grid for all applicants, as well as separate grids for various racial groups.19
Dr. Larntz used this extensive numerical data, compiled by the law school, to make "cell-by-cell" comparisons between applicants of different races to determine by logistic regression analysis whether a statistically significant relationship exists between race and admission rates. Because the grids show the number of applicants and the number of admittees in each cell, and because different grids have been prepared for various racial groups, it is possible to make cross-racial comparisons of applicants with closely similar "academic credentials" or "numbers." To make this comparison, Dr. Larntz calculated the odds of admission for Caucasian applicants and compared them with the odds of admission for applicants of other races in order to calculate the "relative odds of acceptance" for each racial group. Caucasians were the "comparison group" — that is, each group's odds of acceptance were calculated relative to those of Caucasians. Relative odds, or an "odds ratio," greater than 1.0 would indicate that a member of the racial group in question has a greater chance of admission than does a Caucasian applicant. Relative odds less than 1.0 would  indicate the opposite. "For perspective, attaining a relative odds of 2 or 3 for cure of a disease is often the goal of a medical study. That is, a drug that doubled or tripled the odds of cure would be of great value. Double and triple digit relative odds are simply enormous!" Exhibit 137, p. 8.
Dr. Larntz calculated the relative odds of acceptance for various racial groups for each of the years in question. The results of these calculations, and the grids themselves, are presented in his expert reports. See Exhibits 137 and 138 (1995-98), 139 (1999), 141 (2000). For each of the years in question, the relative odds of acceptance for Native American, African American, Mexican American and Puerto Rican applicants were many times greater than for Caucasian applicants.20 Dr. Larntz characterized these relative odds as "extremely large." He concluded that in 1995-2000,
Dr. Larntz cautioned that the relative odds do not express the "number of times greater" a minority applicant's chances of admission are as compared to those of a Caucasian applicant. For the 1995 relative odds, for example, if a Caucasian applicant has a 6-7% chance of being admitted, an African American with a similar index score would have a 93% chance of being admitted. If a Caucasian applicant has a 10% chance of being admitted, a Mexican American applicant with a similar index score would have a 90% chance of being admitted.  Exhibit 137, p. 9; Exhibit 139, p. 7; Exhibit 141, p. 7. This conclusion remained the same even when Dr. Larntz controlled for other factors, such as Michigan residency status, gender, and whether the applicant received an application fee waiver. See Exhibit 137, p. 11; Exhibit 139, p. 8; Exhibit 141, p. 10. At trial Dr. Larntz characterized his relative odds figures as "enormous" and as showing that a "tremendous advantage" was given to applicants from these minority groups in each of the years in question.
In addition to calculating relative odds of acceptance, Dr. Larntz also prepared graphs which plotted the probability of acceptance against the selection index. The selection index, or simply "index," is a combination of an applicant's undergraduate GPA and LSAT score. As noted above, the law school's admissions policy states that "[b]luntly, the higher one's index score, the greater should be one's chances of being admitted. The lower the score, the greater the risk the candidate poses." Exhibit 4, p. 4. Each of Dr. Larntz' graphs plots the relationship between the selection index and the probability of acceptance for Caucasian applicants and for applicants from one minority group for comparison. As one would expect, these graphs show that for all races the higher one's index score, the greater one's chances of being admitted. However, each graph shows a significant gap between the lines plotted for the Caucasian and certain minority applicants.21 Dr. Larntz concluded that "[a]ll the graphs comparing Native American, African American, Mexican American, and Puerto Rican applicants to Caucasian American applicants show wide separation indicating a much higher probability of acceptance for the particular ethnic group at a given selection index value." Exhibit 137, p. 14; Exhibit 139, pp. 9-10; Exhibit 141, pp. 9-10. Based on all of his  analysis, Dr. Larntz concluded that membership in these racial groups "is an extremely strong factor in the decision for acceptance," Exhibit 137, p. 14, and that applicants from these minority groups "are given an extremely large allowance for admission" as compared to Caucasian applicants. Exhibit 139, p. 13; Exhibit 141, p. 13. Dr. Larntz found this to be the case for each of the years in question (1995-2000).
At trial Dr. Larntz made certain cell-by-cell comparisons to highlight the difference in acceptance rates for Caucasian and minority applicants. For example, in 1995 African American and Caucasian applicants with LSAT scores of 161-163 were accepted in clearly disparate proportions at every UGPA level. Two of the three African American applicants with a UGPA of 2.5-2.74 were accepted, whereas none of the seven Caucasian applicants were accepted. All of the four African American applicants with a UGPA of 2.75-2.99 were accepted, whereas none of the 14 Caucasian applicants were accepted. Seven of the eight African American applicants with a UGPA of 3.00-3.24 were accepted, whereas two of the 42 Caucasian applicants were accepted. All of the four African American applicants with a UGPA of 3.25-3.49 were accepted, whereas five of the 126 Caucasian applicants were accepted. Five of the six African American applicants with a UGPA of 3.50-3.74 were accepted, whereas 14 of the 161 Caucasian applicants were accepted. All of the three African American applicants with a UGPA of 3.75 and above were accepted, whereas eight of the 93 Caucasian applicants were accepted. See Exhibit 143, slide 27.
Dr. Larntz highlighted similarly disparate rates of admission between African American and Caucasian applicants by holding the UGPA constant and showing the admissions figures for applicants with various LSAT scores. Dr. Larntz' Exhibit 143, slide 28, compares applicants in 1995 from these two racial groups with UGPA's of 3.25-3.49. In this UGPA range,  of six African American applicants with an LSAT score of 148-150, two were admitted; of 16 Caucasian applicants, none was admitted. Of seven African American applicants with an LSAT score of 151-153, three were admitted; of 24 Caucasian applicants, none was admitted. Of five African American applicants with an LSAT score of 154-155, four were admitted; of 51 Caucasian applicants, one was admitted. Of ten African American applicants with an LSAT score of 156-158, all were admitted; of 51 Caucasian applicants, one was admitted. Of three African American applicants with an LSAT score of 159-160, all were admitted; of 61 Caucasian applicants, one was admitted. Of four African American applicants with an LSAT score of 161-163, all were admitted; of 126 Caucasian applicants, five were admitted.
Dr. Larntz showed similar discrepancies in the admissions rates between Caucasians and members of other minority groups in each of the years in question. See, e.g., Exhibit 143, slides 47- 51. He concluded that the law school gives an "incredibly large allowance" to Native American, African American, Mexican American and Puerto Rican applicants, as compared to Caucasian applicants with similar undergraduate GPA's, LSAT scores, and residency status. Dr. Raudenbush testified as defendants' statistician. He suggested that Dr. Larntz' analysis of the admissions data is flawed because it did not consider the effect of "unquantifiable" factors such as applicants' letters of recommendation and essays, or the reputation of the applicants' undergraduate institutions. In addition, Dr. Raudenbush criticized Dr. Larntz' odds ratio analysis because it disregarded cells in which all applicants were accepted, or all were rejected, and this resulted in the loss of information.22 He also suggested that because the odds ratios vary from one  cell to another, an overall or "composite" odds ratio is not informative. In addition, Dr. Raudenbush was suspicious of the odds ratios because they vary widely from one year to another, whereas the actual percentage of applicants admitted (at least for African Americans and Caucasians) has remained relatively stable. See Exhibit 194.
Aside from criticizing Dr. Larntz' analysis, the primary focus of Dr. Raudenbush's own analysis and testimony was on the predicted effect of eliminating race as a factor in the law school's admissions process. In Dr. Raudenbush's view, a "race-blind" admissions system would have a "very dramatic," negative effect on minority admissions but only a slight effect on non-minority admissions, due to the vastly greater number of non-minority applicants. In the year 2000, 35% of underrepresented minority applicants and 40% of non-minority applicants were admitted. See Exhibit 187. Dr. Raudenbush predicted that if race were not considered, then only 10% of underrepresented minority applicants and 44% of non-minority applicants would be admitted. If correct, this would mean that in the year 2000 only 46 underrepresented minority applicants would have been admitted (instead of 170 who actually were admitted), of whom only 16 would enroll (instead of 58 who actually enrolled). Under this scenario, underrepresented minority students would have constituted 4% of the entering class in 2000, instead of 14.5% as actually occurred. See Exhibit 189.
5. For easy reference a copy of this grid, which shows admissions information for 1991, is attached to this opinion as Exhibit A. [return to text]
6. For example, in 1991 there were 499 applicants with an LSAT score between 38 and 41and a UGPA between 3.50 and 3.74; and of these, 36 were offered admission. In later years, the LSAT was scored on a scale between 120 and 180 points. The grids for the years at issue in this case (1995 to the present) still show UGPA in quarter-point increments on the vertical axis, as was done in earlier years. But the LSAT scores, which are still shown along the horizontal axis, are presented in the following increments: no LSAT, 120-145, 146-147, 148-150, 151-153, 154-155, 156-158, 159-160, 161-163, 164-166, 167-169, and 170-above. [return to text]
7. The identical statement appears at page 81 of the law school bulletin for 1995-1997. See Trial Exhibit 7. [return to text]
8. Trial Exhibit 53 is a document entitled "The History of Special Admissions at the University of Michigan Law School, 1966-1981." This document traces the history of the law school's efforts to enroll a certain percentage of students from particular, identified minority groups during this time period. Beginning in 1966, the law school faculty became concerned about the low numbers of black students. For the first time, "those who are Negroes or from disadvantaged backgrounds" who were on the waiting list for admission were given preference. Id. at 5. In 1970, the dean of admissions indicated he would seek to admit black and Mexican-American students "who fall below the admission standards regularly applied" in sufficient (continued from next page) numbers to constitute 10% of the entering class. Id. at 16. Exhibit 53 shows that from the late 1960s to the early 1980s the law school faculty frequently debated the issue of special admissions — the reasons for the policy, how it should be administered, the minority groups to which it was directed, and the "target" percentage the law school should aim to achieve. Over the years, the law school faculty apparently reached a consensus that black and Hispanic students should constitute between 10% and 12.5% of the entering class. See id. at 16, 19, 22, 27, 31, 34, 45, 48-50, 57. In 1975 the law school faculty formally adopted a special admissions policy that identified "Blacks, Chicanos, American Indians, and Puerto Rican Americans" as the groups which "have been substantially underrepresented in the student body and the legal profession" and directed that members from these groups constitute 10-12% of the entering class. See id. at 48-49. The dean of the law school, Terrance Sandalow, reaffirmed the 10-12% goal in a memorandum to the faulty in 1978. See id. at 53-55. [return to text]
9. In 1989, regular admissions had a median LSAT score of 43 and a median UGPA of 3.60, while special admissions had a median LSAT score of 35 and a median UGPA of 3.06. In 1990, regular admissions had a median LSAT score of 45 and a median UGPA of 3.60, while special admissions had a median LSAT score of 38 and a median UGPA of 3.16. [return to text]
10. Mr. Shields also acknowledged that, at his deposition, he testified that race "generally" explains the difference in admissions rates between minority and non-minority groups. The comparison during Mr. Shields' cross-examination was the difference in admission rates between white and black applicants with LSAT scores between 154 and 169 and with UGPA's between 3.25 and 4.00. In these cells nearly all of the African American applicants (48 of 52) were accepted, whereas a much smaller percentage of the Caucasian applicants (379 of 1437) were accepted. In the same cluster of cells, 30 of 40 Mexican American applicants were accepted; 7 of 14 Native American applicants were accepted; and 3 of 5 Puerto Rican applicants were accepted. [return to text]
11. See, e.g., Exhibit 112 (1988-1989), Exhibit 113 (1989-1990), Exhibit 114 (1990-1991), Exhibit 115 (1992), Exhibit 116 (1993). Exhibit 14 shows that in the 1994 entering class, white students had a median LSAT score of 168 and a median UGPA of 3.57, while the corresponding figures were 157 and 2.97 for African American students, and 162 and 3.26 for Mexican American students. In the 1995 entering class, white students had a median LSAT score of 167 and a median UGPA of 3.59, while the corresponding figures were 155 and 3.18 for African American students, and 159 and 3.35 for Mexican American students. [return to text]
12. It was this committee's report and recommendation that became the law school's official admissions policy when it was adopted, unanimously, by the law faculty in April 1992. See Exhibit 4 (cover sheet). [return to text]
13. Dean Lehman conceded that the different admission rates for different racial groups is "partly indicative" of the extent to which race is considered. Referring to Ex. 15 (the law school's 1995 grids), Dean Lehman acknowledged that all African American applicants with an LSAT score of 159-160 and a UGPA of 3.00 and above were admitted, whereas only one of 54 Asian applicants and four of 190 Caucasian applicants with these qualifications were admitted. [return to text]
14. The other members of the committee were Don Herzog, Jeffrey Lehman, Don Regan, Ted Shaw and Dennis Shields. See Exhibit 4, p. 14. [return to text]
15. Professor Lempert indicated that other groups, such as Asians and Jews, have also been discriminated against, but they were not mentioned in the law school's admissions policy because members of these groups are already being admitted to the law school in significant numbers. [return to text]
16. Dr. Larntz' expert reports were admitted as Exhibits 137-142. Exhibit 143 consists of the tables and charts of Dr. Larntz' "powerpoint presentation." [return to text]
17. Dr. Raudenbush's expert reports were admitted as Exhibits 145-150. [return to text]
18. The law school grids also indicate the "yield," that is, the number of admittees who enrolled. [return to text]
19. For example, separate grids are compiled for Native Americans, African Americans, Caucasian Americans, Mexican Americans, Hispanic Americans, Asian/Pacific Island Americans, Puerto Rican Applicants, and for other groups as well. [return to text]
20. In 1995, the relative odds of acceptance were 61.37 for Native Americans, 257.93 for African Americans, 81.90 for Mexican Americans, and 37.86 for Puerto Ricans. In 1996, the relative odds were 29.81 for Native Americans, 313.59 for African Americans, 81.46 for Mexican Americans, and 45.40 for Puerto Ricans. In 1997, the relative odds were 37.37 for Native Americans, 53.49 for African Americans, 17.55 for Mexican Americans, and 32.78 for Puerto Ricans. In 1998, the relative odds were 23.98 for Native Americans, 132.16 for African Americans, 23.53 for Mexican Americans, and 17.84 for Puerto Ricans. See Exhibit 137, pp. 23- 26. In 1999, the relative odds were 32.05 for Native Americans, 206.45 for African Americans, 43.77 for Mexican Americans, and 41.71 for Puerto Ricans. See Exhibit 139, p. 21. And in 2000, the relative odds were 24.61 for Native Americans, 443.26 for African Americans, 16.99 for Mexican Americans, and 28.63 for Puerto Ricans. [return to text]
21. See Exhibit 137, Figures 9, 10, 11, 14, 17, 18, 19, 22, 25, 26, 27, 30, 33, 34, 35, 38; Exhibit 139, Figures 3, 4, 5, 8; Exhibit 141, Figures 3, 4, 5, 8. [return to text]
22. Dr. Larntz testified that he disregarded these cells because they do not contain "comparative information." That is, if all applicants in a particular cell are accepted, or if all are (continued from next page) rejected, then there is no basis for calculating the odds ratio. [return to text]
Copyright by the Regents of the University of Michigan