Grutter v. Bollinger, 137 F. Supp. 2d 821 (E.D. Mich. 2001)
March 27, 2001
Lee BOLLINGER, Jeffrey Lehman, Dennis Shields, Regents of the University of Michigan, and the University of Michigan Law School, Defendants,
Kimberly James, Farah Mongeau, Jeanette Haslett, Raymond Michael Whitlow, Shabatayah Andrich, Dena Fernandez, Shalamarel Kevin Killough, Diego Bernal, Julie Fry, Jessica Curtin, James Huang, Heather Bergman, Ashwana Carlisle, Ronald Cruz, Nora Cecilia Melendez, Irami Osei-Frimpong, Gerald Ramos, Arturo Vasquez, Edward Vasquez, Vincent Kukua, Hoku Jeffrey, Karlita Stephens, by her Next Friend Karla Stephens-Dawson, Yolanda Gibson, by her Next Friend Mary Gibson, Erika Dowdell, by her Next Friend Herbert Dowdell, Jr., Agnes Aleobua, by her Next Friend Paul Aleobua, Cassandra Young, by her Next Friend Yolanda J. King, Jaasi Munanka, Jodi-Marie Masley, Shannon Ewing, Julie Kerouac, Kevin Pimentel, Bernard Cooper, Norberto Salinas, Scott Rowekamp, Russ Abrutyn, Jasmine Abdel-Khalik, Meera Deo, Winifred Kao, Melisa Resch, Oscar De La Torre, Carol *822 Scarlett, United for Equality and Affirmative Action, the Coalition to Defend Affirmative Action by any Means Necessary, and Law Students for Affirmative Action, Intervening Defendants.
United States District Court, E.D. Michigan, Southern Division.
*823 Kerry L. Morgan, Pentiuk & Couvreur, Taylor, MI, Kirk O. Kolbo, David F. Herr, Maslon, Edelman, Borman & Brand, Minneapolis, MN, Godfrey J. Dillard, Detroit, MI, for plaintiff.
Philip J. Kessler, Okemos, MI, John A. Payton, Jane Sherburne, Wilmer, Cutler & Pickering, Washington, D.C., Leonard M. Niehoff, Butzel & Long, Ann Arbor, MI, Richard A. Wilhelm, Dickinson Wright, Detroit, MI, Susan I. Leffler, Michigan Department of Attorney General, Habeas Corpus Division, Lansing, MI, Kenneth S. Geller, Mayer Brown & Platt, Washington, D.C., Jeffrey S. Silver, Ann Arbor, MI, for Defendants.
FINDINGS OF FACT AND CONCLUSIONS OF LAW
FRIEDMAN, District Judge.
On December 22, 2000, the court heard oral argument in this case on the parties' cross motions for summary judgment. The court took the motions under advisement and identified the issues for trial. Over a period of 15 days in January and February 2001, the court conducted a bench trial. In this opinion, the court shall rule on the motions and make findings of fact and conclusions of law.
Plaintiff Barbara Grutter commenced this action in December 1997. Ms. Grutter alleges that in 1996 she applied for admission to the University of Michigan Law School (hereinafter "the law school"). At first plaintiff was placed on a waiting list, but in June 1997 her application was rejected. *824 Plaintiff, who is Caucasian, alleges that her application was rejected because the law school uses race as a "predominant" factor, giving minority applicants "a significantly greater chance of admission than students with similar credentials from disfavored racial groups." Complaint ¶¶ 20, 23. In their answer to the complaint, defendants "state that they do have a current intention to continue using race as a factor in admissions, as part of a broad array of qualifications and characteristics of which racial or ethnic origin is but a single though important element." Answer ¶¶ 9, 23.
Plaintiff asserts two claims. First, she claims that defendants discriminated against her on the basis of her race, thereby violating her rights to equal protection under the Fourteenth Amendment. This claim is brought under 42 U.S.C. §§ 1981 and 1983. Second, plaintiff claims that defendants violated Title VI of the 1964 Civil Rights Act, 42 U.S.C. § 2000d, which prohibits recipients of federal funds from discriminating on the basis of race. For relief, plaintiff seeks a declaratory judgment to the effect that her rights were violated; an injunction prohibiting racial discrimination in admissions; compensatory and punitive damages; an order requiring defendants to admit her to the law school; and attorney fees and costs. The defendants are Lee Bollinger, the dean of the law school from 1987 to 1994 and president of the University of Michigan from 1997 to the present; Jeffrey Lehman, the dean of the law school from 1994 to the present; Dennis Shields, the director of admissions at the law school from 1991 to 1998; the regents of the University of Michigan; and the University of Michigan Law School.
In an opinion and order dated January 7, 1999, the court granted plaintiff's motion for class certification and for bifurcation of the trial into liability and damages phases. The class was defined as consisting of "all persons who (A) applied for and were not granted admission to the University of Michigan Law School for the academic years since (and including) 1995 until the time that judgment is entered herein; and (B) were members of those racial or ethnic groups, including Caucasian, that Defendants treated less favorably in considering their applications for admission to the Law School."
In March 1998, 41 individuals and three pro-affirmative action student groups*825 sought to intervene in the case as defendants. The individual intervenors include 21 undergraduate students of various races who currently attend the University of Michigan, Wayne State University, the University of California at Berkeley, or Diablo Valley Community College in Pleasant Hill, California, all of whom plan to apply to the law school for admission; five black students who currently attend Cass Technical High School or Northwestern High School in Detroit and who plan to apply to the law school for admission; twelve students of various races who currently attend the law school; a paralegal and a Latino graduate student at the University of Texas at Austin who intend to apply to the law school for admission; and a black graduate student at the University of Michigan who is a member of the Defend Affirmative Action Party. Motion to Intevene ¶¶ 1-41. The court initially denied the motion to intervene, but in August 1999 the court of appeals reversed and directed that the intervention be permitted.
On December 22, 2000, the court heard oral argument on the parties' cross motions for summary judgment. The court took those motions under advisement. The court indicated that the trial would focus on the following three issues: (1) the extent to which race is a factor in the law school's admissions decisions; (2) whether the law school's consideration of race in making admissions decisions constitutes a double standard in which minority and non-minority students are treated differently; and (3) whether the law school may take race into account to "level the playing field" between minority and non-minority applicants.
II. Diversity as a Rationale for Using Race as a Factor in University Admissions
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.*826 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. 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:
When the differences in index scores are small, we believe it is important to weigh as best we can not just the index but also such file characteristics as the enthusiasm of recommenders, the quality of the undergraduate institution, the quality of the applicant's essay, and the areas and difficulty of undergraduate course selection. These "soft" variables not only bear on the applicant's likely graded performance but also have the additional benefit that they may tell us something about the applicant's likely contributions to the intellectual and social life of the institution. Thus an applicant who has performed well in advanced courses in a demanding subject may have more to offer both faculty and students than an applicant with a similarly high average achieved without ever pursuing in depth any area of learning. Other information in an applicant's file may add nothing about the applicant's likely LGPA [law school grade-point average] beyond what may be discerned from the index, but it may suggest that that applicant has a perspective or experiences that will contribute to the diverse student body that we hope to assemble. The applicant may for example be a member of a minority group whose experiences are likely to be different from those of most students, may be likely to make a unique contribution to the bar, or may have had a successful *827 career as a concert pianist or may speak five languages.
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:
There is, however, a commitment to one particular type of diversity that the school has long had and which should continue. This is a commitment to racial and ethnic diversity with special reference to the inclusion of students from groups which have been historically discriminated against, like African-Americans, Hispanics and Native Americans, who without this commitment might not be represented in our student body in meaningful numbers. These students are particularly likely to have experiences *828 and perspectives of special importance to our mission.
Over the past two decades, the law school has made special efforts to increase the numbers of such students in the school. We believe that the racial and ethnic diversity that has resulted has made the University of Michigan Law School a better law school than it could possibly have been otherwise. By enrolling a "critical mass" of minority students, we have ensured their ability to make unique contributions to the character of the Law School; the policies embodied in this document should ensure that those contributions continue in the future.
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:
As one of my colleagues so elegantly stated recently "our mission is to pick winners".... To make those selections requires more than a mere review of the numbers (LSAT and GPA), credentials. This is because many of the numbers will be so close to the same as to make the candidates indistinguishable from one another on that basis. Rather we must begin with the numbers and go forward from there to scrutinize the essays and letters of recommendation (as well as considering extracurricular and work experience) to look for candidates that show intellectual talent, leadership ability, and academic acumen which augers for a lively intellectual community and important contributions to the profession.
* * * * * *
[O]f the 5,000 or more applications we will receive this year, we will offer admission to about 1,000 of the candidates. Of that thousand offers, fully 500 of the decisions are pretty easy, another 300 or are [sic] a bit tougher, and the last 200 very difficult because the numbers will be so close that in these cases we will clearly be making the most subjective of judgements.
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. *829 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:
All applications are read in their entirety, and all of the information elicited by the application is factored into the admission decision. All admissions are made with the goal of forming a class with an exciting and productive mix of students who will enhance the educational experience for each other and for the School. Law School Admission Test (LSAT) scores and undergraduate course work and performance are relied on heavily, as are comparative studies of the past performance of similar students at the Law School. Serious regard is also given to an applicant's promise of making a notable contribution to the class by way of a particular strength, attainment, or characteristic e.g., an unusual intellectual achievement, employment experience, nonacademic performance, or personal background. The guiding purpose for selection among applicants is to make the School a better a livelier place in which to learn and to improve its service to the profession and the public.
In addition to its own interest in forming a class which is strengthened by the talents and diversity of its members, Michigan recognizes the public interest in increasing the number of lawyers from groups which the faculty identifies as significantly underrepresented in the legal profession. In particular, we strongly encourage prospective students who are African American, Mexican American, Native American, or Puerto Rican and raised on the U.S. mainland to apply. Such applicants are invited to contact the Admissions Office for further information about the School's affirmative efforts to increase enrollment from among these groups. Similarly, the Law School welcomes applications from all persons without regard to their sex, religious affiliation, national origin or ancestry, age, marital status, sexual orientation, or handicap. Every Law School matriculant must be a graduate of an accredited college or university.
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:
Based on its years of teaching experience, the Michigan faculty has determined that the quality of classroom analysis and discussion is enhanced by the multiplicity of perspectives and experiences that students bring with them. Each year, the Law School enrolls a talented and diverse class of students, and the admissions office considers the ways in which a potential student will contribute to the diversity of the Law School as one of the factors in its admissions decisions. The Law School welcomes applications from all persons without regard to their sex, religious affiliation, national origin or ancestry, age, marital status, sexual orientation, or handicap. Every Law School matriculant must be a graduate of an accredited college or university.
At trial, a number of witnesses testified as to the admissions procedures prior to *830 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:
One half of the entering class is selected primarily on the basis of a prediction of their academic success in law study. The prediction is calculated on the basis of Law School Admission Test scores, undergraduate records, and studies of the past performance of previous students at the Law School....
The other half of the class will be chosen from a group of several hundred applicants whose grades and test scores qualify them for further consideration. The grades and scores necessary to gain admission to this pool vary from year to year, but the academic credentials of all who are included are invariably strong.
Selections among applicants in the pool are made for the purpose of making the School a livelier place to learn and for improving its service to the profession and the public. Specifically, the School seeks those students who are more likely to contribute affirmatively to the learning of others by reason of their unusual intellectual attainments, significant employment experience or outstanding non-academic achievements, demonstration of emotional maturity and self-discipline, unusual social background, or exceptional capacity to benefit from a particular phase of the School's program....
In administering its admissions policy, the Law School recognizes the racial imbalance now existing in the legal profession and the public interest in increasing the number of lawyers from the ethnic and cultural minorities significantly underrepresented in the profession. Minorities are strongly encouraged to apply and to call upon Assistant Dean Stillwagon for information and assistance. Black, Chicano, Native American, and many Puerto Rican applicants are automatically considered for a special admissions program designed to encourage and increase the enrollment of minorities.
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. Mr. Stillwagon indicated *831 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. 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 *832 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 American, 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.
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 *833 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. 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. 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, 98 S. Ct. 2733, 57 L. Ed. 2d 750 (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 *834 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. 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. 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 *835 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. 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:
Our goal is to have substantial and meaningful racial and ethnic diversity, but we do not wish to exhaust the positions that are open to non-grid diversity admittees in promoting racial and ethnic diversity. Nor are we insensitive to the competition among all students for admissions to this law school. Thus while we set no floors or ceilings on the numbers or proportion of students who are to be admitted as minority diversity candidates, in making judgments about admitting such candidates the Admissions Officer and Admissions Committee should be sensitive to such factors as the proportion of minority students who would be attending Michigan without a diversity admissions program and the strength of the credentials that various minority candidates bring. Also it is important to note that in the past we seem to have achieved the kinds of benefits that we associate with racial and ethnic diversity from classes in which the proportion of African American, Hispanic and Native American members has been between about 11% and 17% of total enrollees.
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 *836 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. Defendants presented Dr. Stephen Raudenbush, a professor of education at the University of Michigan. 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. 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.
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 *837 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. Dr. Larntz characterized these relative odds as "extremely large." He concluded that in 1995-2000,
membership in certain ethnic groups is an extremely strong factor in the decision for acceptance. Native American, African American, Mexican American, and Puerto Rican applicants in the same LSAT × GPA grid cell as a Caucasian American applicant have odds of acceptance that is many, many (tens to hundreds) times that of a similarly situated Caucasian American applicant.
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 *838 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. 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 *839 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. 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 "raceblind" 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.
B. Findings of Fact and Conclusions of Law
From this testimony and documentary evidence, the court makes the following findings and conclusions.
The law school clearly considers an applicant's race in making admissions decisions. While the reasons for doing so have changed somewhat over time, the law school has given some degree of preference to members of particular racial groups for more than 30 years. In the early years of affirmative action, the law school extended this preference only to African American applicants. In the *840 years since then, the preference has been broadened to include Native Americans, Mexican Americans and mainland Puerto Ricans. It does not appear that any preference is given to members of any other racial groups.
The current stated reason for granting a preference to members of these groups is that certain educational benefits flow from a racially diverse student body, and members of underrepresented minorities would not be admitted in significant numbers unless race is explicitly considered. This is due to the fact that members of these groups, on average, have lower LSAT scores and lower undergraduate GPA's as compared to other applicants (i.e., Caucasians and Asians), so that comparatively few would be admitted in a system where admissions decisions were based on "the numbers."
A major bone of contention in this case has been the extent to which race is considered in the admissions process. The evidence shows that race is not, as defendants have argued, merely one factor which is considered among many others in the admissions process. Rather, the evidence indisputably demonstrates that the law school places a very heavy emphasis on an applicant's race in deciding whether to accept or reject.
This emphasis on race is apparent from the admissions policy itself, which explicitly states the law school's "commitment to racial and ethnic diversity with special reference to the inclusion of students from groups which have been historically discriminated against, like African-Americans, Hispanics and Native Americans, ..." Exhibit 4, Admissions Policy, p. 12. The admissions policy is equally explicit about the extent of this commitment: race is considered at least to the extent necessary to enroll a "critical mass" of students from these groups. While "critical mass" has proved to be a concept that has eluded precise quantification, over the years it has meant in practice that the law school attempts to enroll an entering class 10% to 17% of which consists of underrepresented minority students. The 10% figure, as a target, has historical roots going back to the late 1960s. Beginning in the 1970s, law school documents begin referring to 10-12% as the desired percentage. Professor Lempert testified that critical mass lies in the range of 11-17%. Indeed, this percentage range appeared in a draft of the 1992 admissions policy, and it was omitted from the final version despite Professor Regan's suggestion that it remain for the sake of "candor."
The actual admissions and graduation statistics confirm the law school's commitment to enroll African American, Native American and Hispanic students in the 10-17% range. For example, Exhibit 97 shows the number of students graduating (total as well as by various racial groups) for each year from 1950 to 1999. From 1973 to 1985, underrepresented minority students constituted approximately 9-10% of the graduating class, and from 1986 to 1999 students from these groups constituted approximately 12-13% of the graduating class. These percentages conclusively *841 demonstrate that the law school considers race in the admissions process because applicants from the underrepresented minority groups have, on average, considerably lower undergraduate GPA's and LSAT scores as compared to Caucasian applicants and yet, nonetheless, the percentage of applicants from these minority groups who are admitted is roughly equal to the percentage they constitute of the total applicant pool.
In 1995, for example, Native Americans constituted 1.1% of the total applicant pool and 1.2% of the admitted students; African Americans constituted 9.3% of the total applicant pool and 9.2% of the admitted students; Mexican Americans constituted 2.3% of the total applicant pool and 3.5% of the admitted students; Puerto Ricans constituted 0.5% of the total applicant pool and 0.4% of the admitted students; and Caucasians constituted 56.3% of the total applicant pool and 59.7% of the admitted students. See Exhibit 146, Table 1. That is, the representation of each group in the applicant pool roughly approximates its representation among the total admitted. Yet the median undergraduate GPA and LSAT score are generally lower even considerably lower for underrepresented minority applicants than for Caucasian applicants. See Exhibit 137, pp. 18-19 (Tables 2 and 3). The same pattern can be seen in 1996 and 1997. See Exhibit 146 (Tables 2 and 3) and Exhibit 137, pp. 18-19 (Tables 2 and 3). If race were not considered in the admissions process, one would expect to see underrepresented minority applicants admitted in much lower proportions than has been the case.
Plaintiffs' and defendants' statisticians analyzed the admissions data and both provided testimony and expert reports which assisted the court in understanding the extent to which race is considered in the law school's admissions process. Dr. Larntz' cell-by-cell analysis provided mathematically irrefutable proof that race is indeed an enormously important factor. In each year at issue in this case, Native American, African American, Mexican American and Puerto Rican applicants have been admitted in significantly greater proportions than Caucasian applicants with the same or similar undergraduate GPA's and LSAT scores. As Dr. Larntz noted, this fact is apparent on the face of the law school's admissions grids. One does not need to undergo sophisticated statistical analysis in order to see it; the statistical analysis simply confirms empirically what the grids suggest intuitively. The court specifically adopts Dr. Larntz' analysis and his conclusion that "membership in certain ethnic groups is an extremely strong factor in the decision for acceptance."
The court rejects Dr. Raudenbush's criticism of Dr. Larntz' cell-by-cell analysis. While it is true that Dr. Larntz' analysis did not include cells in which all applicants were either accepted or rejected, the court was persuaded by Dr. Larntz' explanation that these cells do not contain any comparative data. The issue in this case is whether similarly situated applicants are treated differently because of their race, and this question can be answered by examining cells in which some applicants are accepted and others rejected so that the differences in the admissions rates can be calculated. Moreover, Dr. Larntz testified that the cells he included in his analysis are the ones containing more than 95% of all admittees. Thus, the court believes that Dr. Larntz omitted very little data from his analysis and none that is relevant.
The court also rejects Dr. Raudenbush's other criticisms of Dr. Larntz' methodology. For example, Dr. Raudenbush argued that the admissions process is complex and cannot be reduced to a comparison of acceptance *842 odds based on "the numbers" and membership in various racial groups. The court agrees, and specifically finds, that the law school examines each application file individually and considers not only grades and test scores but also letters of recommendation, the applicant's personal statements, extracurricular activities, work experience, residency status, relationship to university alumni, rigor of undergraduate curriculum, trend in undergraduate grades, and reputation of undergraduate institution. Nonetheless, the court does not believe that the complexity of the admissions process undermines the validity of Dr. Larntz' analysis. When cell by cell, and year by year, underrepresented minority applicants are admitted in significantly greater proportions than their non-minority competitors with similar UGPA and LSAT scores, it is clear that the law school accords the race of the applicants a great deal of weight.
Even the testimony and reports of the law school's statistician, Dr. Raudenbush, support this conclusion. As noted above, Dr. Raudenbush testified that the number of underrepresented minority admittees would drop "sharply and dramatically" if race were not considered in the admissions process. By his calculations, the percentage of underrepresented minority applicants who are admitted would have dropped from 26% to 4% in 1995, and from 31% to 8% in 1996, under a race-blind admissions system. See Exhibit 146, Table 8. In 1997 the percentage admitted would have dropped from 33% to 8%; in 1998 the percentage admitted would have dropped from 34% to 9%; in 1999 the percentage admitted would have dropped from 37% to 8%; and in 2000 the percentage admitted would have dropped from 35% to 10%. See Exhibit 149, Table 1; and Exhibit 149, Tables 1-3. These figures, which the court has no reason to doubt, may explain even more plainly than Dr. Larntz' odds ratios and graphs the extent to which race is considered in the law school's admissions process.
Finally, the testimony of the witnesses who are familiar with the inner workings of the law school's admissions office confirmed that race is considered and that it makes a difference in admissions decisions. The current and former dean, as well as the current and former admissions directors, all testified that race is considered to the extent necessary to enroll a critical mass of underrepresented minority students. While none of these witnesses acknowledged that they have a particular number or percentage in mind as the admissions season progresses, the written and unwritten policy at the law school charges the admissions office with assembling entering classes which consist of between 10% and 17% African American, Native American, and Hispanic students. Over the years this target has been achieved, and even exceeded, despite the underrepresented minority students' generally lower LSAT scores and undergraduate GPA's. The court also finds it significant that the dean and the admissions director monitor the law school's "daily admissions reports," which classify applicants by race. These reports inform the reader how many students from various racial groups have applied, how many have been accepted, how many have been placed on the waiting list, and how many have paid a deposit. There would be no need for this information to be categorized by race unless it were being used to ensure that the target percentage is achieved.
*843 In conclusion, the court finds that the law school explicitly considers the race of applicants in order to enroll a critical mass of underrepresented minority students that is, the law school wants 10% to 17% of each entering class to consist of African American, Native American, and Hispanic students.
The court must now turn to the central issue in this case: whether the Constitution permits the consideration of race in order to achieve racial diversity. In current constitutional parlance, the question is whether the achievement of racial diversity is a compelling state interest and, if so, whether the law school's admissions policy is narrowly tailored to serve that interest.See Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 227, 115 S. Ct. 2097, 132 L. Ed. 2d 158 (1995) ("all racial classifications, imposed by whatever federal, state, or local governmental actor, must be analyzed by a reviewing court under strict scrutiny. In other words, such classifications are constitutional only if they are narrowly tailored measures that further compelling governmental interests").
In answering this question, the starting point is the Supreme Court's landmark decision in Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 98 S. Ct. 2733, 57 L. Ed. 2d 750 (1978), the only case in which the high court has ever addressed the "diversity rationale" as a justification for considering race in reviewing an application for admission to a university.
In Bakke, the plaintiff was a white male who applied in 1973 and 1974 to the University of California at Davis Medical School ("U.C.Davis"), a state institution. In both years, U.C. Davis rejected Bakke's application. After the second rejection Bakke sued the Regents of the University of California, claiming that U.C. Davis' special admissions program, which reserved 16 places in the class of 100 for members of certain minority groups, violated the California state constitution, Title VI, and the Fourteenth Amendment's Equal Protection Clause. The trial court held that U.C. Davis could not consider race in making admissions decisions. The California Supreme Court affirmed and enjoined any consideration of race in the admissions process.
On appeal, a sharply divided Supreme Court affirmed in part and reversed in part. Justice Stevens wrote an opinion in which Chief Justice Burger and Justices Stewart and Rehnquist joined. The Stevens group concluded that Title VI prohibited U.C. Davis, as a state educational institution receiving federal funding, from considering an applicant's race in making admissions decisions. Bakke, 438 U.S. at 412, 98 S. Ct. 2733. Justices Brennan, White, Marshall, and Blackmun authored a joint opinion in which they concluded that both Title VI and the Constitution permit a state educational institution to "take race into account when it acts not to demean or insult any racial group, but to remedy disadvantages cast on minorities by past racial prejudice, at least when appropriate findings have been made by judicial, legislative, or administrative bodies with competence to act in this area." Id. at 325, 98 S. Ct. 2733.
In a separate, tie-breaking opinion, Justice Powell expressed the view that a state educational institution has a compelling interest in admitting a diverse student body, and that "[e]thnic diversity ... is only one element in a range of factors a university *844 properly may consider in attaining the goal of a heterogeneous student body." Id. at 311, 314, 98 S. Ct. 2733. Because the Brennan group joined in the short portion of Justice Powell's opinion (Part V-C) which reversed the California Supreme Court's holding that an applicant's race may never be considered, a total of five Justices agreed that "the State has a substantial interest that legitimately may be served by a properly devised admissions program involving the competitive consideration of race and ethnic origin." Bakke, 438 U.S. at 320, 98 S. Ct. 2733. However, Justice Powell also found that U.C. Davis' special admissions program, which implemented a strict quota system along race lines, was not narrowly tailored to serve the interest of admitting a diverse class of students. Thus, Justice Powell joined the Stevens group, albeit on different grounds, to make a five-Justice majority holding that U.C. Davis' special admissions program was unlawful.
This court is faced with the task of piecing together the above opinions to determine Bakke's holding. Specifically, the court must determine whether Bakke held that a state institution's desire to assemble a racially diverse student body is a compelling government interest. The parties sharply disagree on this issue. In short, plaintiffs argue that although Justice Powell's opinion announced the Court's judgment, his opinion contained statements and conclusions in which no other Justice or group of Justices joined and are therefore not part of the Court's holding. Plaintiffs argue that Justice Powell's discussion of the "diversity rationale" in Part IV-D of his opinion is one such matter. Defendants argue that the Brennan group concurred with Justice Powell's conclusions regarding diversity by joining in Part V-C of his opinion, and that under Marks Justice Powell's conclusions regarding diversity are part of the Court's holding. For the reasons stated below, the court is persuaded that Bakke did not hold that a state educational institution's desire to assemble a racially diverse student body is a compelling government interest.
Of the six separate opinions issued by the Supreme Court in Bakke, Justice Powell's was the only one that considered whether a state educational institution may have a compelling interest in admitting a racially diverse class of students. While rejecting the other justifications offered for the special admissions program, Justice Powell found in Part IV-D of his opinion that "the attainment of a diverse student body ... clearly is a constitutionally *845 permissible goal for an institution of higher education." 438 U.S. at 311-12, 98 S. Ct. 2733. He cited Justice Frankfurter's statement in Sweezy v. New Hampshire, 354 U.S. 234, 263, 77 S. Ct. 1203, 1 L. Ed. 2d 1311 (1957), that universities must have the freedom to decide which students to admit, as this is relevant to providing an "atmosphere which is most conducive to speculation, experiment and creation." Justice Powell also indicated his belief that "the `nation's future depends upon leaders trained through wide exposure' to the ideas and mores of students as diverse as this Nation of many peoples." 438 U.S. at 313, 98 S. Ct. 2733. He quoted approvingly from an article by the president of Princeton University, who wrote:
[A] great deal of learning occurs informally. It occurs through interactions among students of both sexes; of different races, religions, and backgrounds; who come from cities and rural areas, from various states and countries; who have a wide variety of interests, talents, and perspectives; and who are able, directly or indirectly, to learn from their differences and to stimulate one another to reexamine even their most deeply held assumptions about themselves and their world.
Id. at 313, 98 S. Ct. 2733 quoting Bowen, "Admissions and the Relevance of Race," Princeton Alumni Weekly 7, 9 (September 26, 1977). Justice Powell emphasized that "[e]thnic diversity, however, is only one element in a range of factors a university properly may consider in attaining the goal of a heterogeneous student body. Although a university must have wide discretion in making the sensitive judgments as to who should be admitted, constitutional limitations protecting individual rights may not be disregarded." Id. at 314, 98 S. Ct. 2733.
Having recognized the compelling nature of a university's interest in diversity, including ethnic diversity, Justice Powell went on in Parts V-A and V-B of his opinion to find that U.C. Davis' quota system was not narrowly tailored to serve this interest. He stated:
It may be assumed that the reservation of a specified number of seats in each class for individuals from the preferred ethnic groups would contribute to the attainment of considerable ethnic diversity in the student body. But petitioner's argument that this is the only effective means of serving the interest of diversity is seriously flawed. In a most fundamental sense the argument misconceives the nature of the state interest that would justify consideration of race or ethnic background. It is not an interest in simple ethnic diversity, in which a specified percentage of the student body is in effect guaranteed to be members of selected ethnic groups, with the remaining percentage an undifferentiated aggregation of students. The diversity that furthers a compelling state interest encompasses a far broader array of qualifications and characteristics of which racial or ethnic origin is but a single though important element. Petitioner's special admissions program, focused solely on ethnic diversity, would hinder rather than further attainment of genuine diversity.
438 U.S. at 315, 98 S. Ct. 2733. While rejecting a quota system, Justice Powell endorsed an admissions program such as the one adopted by Harvard College, which states that
race or ethnic background may be deemed a "plus" in a particular applicant's file, yet it does not insulate the individual from comparison with all other candidates for the available seats. The file of a particular black applicant may be examined for his potential contribution to diversity without the factor of race being decisive when compared, *846 for example, with that of an applicant identified as an Italian-American if the latter is thought to exhibit qualities more likely to promote beneficial educational pluralism. Such qualities could include exceptional personal talents, unique work or service experience, leadership potential, maturity, demonstrated compassion, a history of overcoming disadvantage, ability to communicate with the poor, or other qualifications deemed important. In short, an admissions program operated in this way is flexible enough to consider all pertinent elements of diversity in light of the particular qualifications of each applicant, and to place them on the same footing for consideration, although not necessarily according them the same weight.
438 U.S. at 317-18, 98 S. Ct. 2733.
In Part V-C of his opinion, Justice Powell concluded that
In enjoining petitioner from ever considering the race of any applicant, however, the courts below failed to recognize that the State has a substantial interest that legitimately may be served by a properly devised admissions program involving the competitive consideration of race and ethnic origin. For this reason, so much of the California court's judgment as enjoins petitioner from any consideration of the race of any applicant must be reversed.
438 U.S. at 320, 98 S. Ct. 2733.
The clearest indication that the Brennan group did not concur with Justice Powell's conclusions regarding the diversity rationale is that, although they joined in other portions of Justice Powell's opinion, they did not join in Part IV-D, the only portion of any of the Bakke opinions that specifically addressed the diversity issue. Moreover, the Brennan group apparently did not believe that the diversity rationale was before the Court, as those Justices stated that the "issue presented" in the case was "whether government may use race-conscious programs to redress the continuing effects of past discrimination." Id. at 324, 98 S. Ct. 2733. The Brennan group also stated that in their view the "central meaning of today's opinions [is that] Government may take race into account when it acts not to demean or insult any racial group, but to remedy disadvantages cast on minorities by past racial prejudice, at least when appropriate findings have been made by judicial, legislative, or administrative bodies with competence to act in this area." Id. at 325, 98 S. Ct. 2733. The Brennan group did not so much as mention the diversity rationale in their opinion, and they specifically declined to join in the portion of Justice Powell's opinion that addressed this issue. At the same time, Justice Powell specifically and vigorously disagreed with the Brennan group's conclusion that the special admissions program at U.C. Davis could be justified on the grounds that it sought to remedy general, societal discrimination. See 438 U.S. at 294-97 & nn. 34, 36, 98 S. Ct. 2588. In short, while the Brennan group and Justice Powell agreed that race may be considered in admissions (hence the joinder of the Brennan group in Part V-C of Justice Powell's opinion), they disagreed entirely as to the reasons why (hence their failure to join in Part IV-D). Thus, there is no force at all to defendants' contention that the Brennan group's joinder in Part V-C of Justice Powell's opinion may be taken as an endorsement of Justice Powell's discussion of the diversity rationale.
*847 The defendants next argue that under Marks, Justice Powell's opinion is controlling because it is the narrowest grounds which support the judgment. The court must reject this argument because a Marks analysis does not assist in interpreting Bakke.
In Marks v. United States, 430 U.S. 188, 193, 97 S. Ct. 990, 51 L. Ed. 2d 260 (1977), the Court held that "[w]hen a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, `the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds.'" In Marks this rule was applied to determine the holding of "John Cleland's Memoirs of a Woman of Pleasure" v. Massachusetts, 383 U.S. 413, 86 S. Ct. 975, 16 L. Ed. 2d 1 (1966), a pornography case in which the Supreme Court reversed a state court judgment finding a book obscene. In Memoirs, three Justices reversed on the grounds that the book had not been shown to be "utterly without redeeming social value"; two others reversed on the grounds that the First Amendment prohibits any action by government to suppress obscenity; and one Justice reversed on the grounds that the book was not "hard core pornography." In Marks the Court stated that the "governing standards" of Memoirs were those announced by the three-Justice plurality because the other Justices who concurred in the judgment did so "on broader grounds." 430 U.S. at 193, 97 S. Ct. 990.
The Court in Marks did not clearly explain what it meant by "narrow" and "broad" grounds. But in that particular case, the plurality's opinion was the "narrowest" in the sense that it was the most conservative reason for reversing the finding of obscenity and it was a reason that was subsumed within the grounds articulated by the other justices who concurred in the judgment.
The Marks framework cannot be applied to a case like Bakke, where the various Justices' reasons for concurring in the judgment are not merely different by degree, as they were in Memoirs, but are so fundamentally different as to not be comparable in terms of "narrowness." In Memoirs, the six Justices concurring in the judgment expressed three viewpoints which could be placed on a continuum from narrow to broad. The narrowest would permit a finding of obscenity if the book in question is "utterly without redeeming social value." A somewhat broader view would permit a finding of obscenity if the book is "hard cord pornography." An even broader view would not permit any book to be found obscene. On this continuum, the first view is clearly the narrowest and, therefore, articulates the binding legal standard constituting the Court's holding.
In Bakke, however, it is nonsensical to ask which opinion Justice Powell's or the Brennan group's offers the "narrowest grounds" in support of the judgment reversing the finding that race may never be considered in university admissions. The diversity rationale articulated by Justice Powell is neither narrower nor broader than the remedial rationale articulated by the Brennan group. They are completely different rationales, neither one of which is subsumed within the other. There is simply no overlap between the two rationales; in fact, as noted above, the two opinions are very much at odds with one another. Thus, the Marks framework is inapplicable because the varying positions cannot be compared for "narrowness."
For these reasons, the court concludes that Justice Powell's discussion of the diversity rationale is not among the governing standards to be gleaned from Bakke. No other member of the Court joined in *848 his view that the attainment of a diverse student class is a compelling state interest which can justify the consideration of race in university admissions. Nor can Justice Powell's discussion of this issue be understood as part of the Court's holding on the grounds that, under Marks, it is the narrowest grounds for the judgment in which he and the Brennan group joined. Therefore, this court concludes that Bakke does not stand for the proposition that a university's desire to assemble a racially diverse student body is a compelling state interest.
Further, in post-Bakke decisions the Supreme Court has indicated quite clearly that it looks upon racial classifications with even more suspicion than was the case at the time Bakke was decided. In Metro Broadcasting, Inc. v. Federal Communications Comm'n, 497 U.S. 547, 110 S. Ct. 2997, 111 L. Ed. 2d 445 (1990), the Court upheld FCC regulations which gave a preference to various racial minority groups in the awarding of broadcast licenses. Applying intermediate scrutiny, the Court upheld the regulations because "the interest in enhancing broadcast diversity is, at the very least, an important governmental objective ...." Id. at 567, 110 S. Ct. 2997. Five years later, however, in Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 227, 115 S. Ct. 2097, 132 L. Ed. 2d 158 (1995), the Court stated "that all racial classifications, imposed by whatever federal, state, or local governmental actor, must be analyzed by a reviewing court under strict scrutiny. To the extent that Metro Broadcasting is inconsistent with that holding, it is overruled." Adarand clarified that strict scrutiny applies to all racial classifications, regardless of the level of government using the classification. Here it should be noted that the Brennan group in Bakke used the less demanding standard of intermediate scrutiny in their review of U.C. Davis' admissions program. See Bakke, 438 U.S. at 359, 98 S. Ct. 2733 ("racial classifications designed to further remedial purposes `must serve important governmental objectives and must be substantially related to achievement of those objectives'"). Moreover, the Court in City of Richmond v. J.A. Croson Co., 488 U.S. 469, 493, 109 S. Ct. 706, 102 L. Ed. 2d 854 (1989), made the following significant statement: "Classifications based on race carry a danger of stigmatic harm. Unless they are strictly reserved for remedial settings, they may in fact promote notions of racial inferiority and lead to a politics of racial hostility." When read together, *849 Adarand and Croson clearly indicate that racial classifications are unconstitutional unless they are intended to remedy carefully documented effects of past discrimination. The Supreme Court has rejected various "benign" justifications for racial classifications. See, e.g., Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 106 S. Ct. 1842, 90 L. Ed. 2d 260 (1986) (finding race-conscious layoff policy unconstitutional, and rejecting "minority role model" and "societal discrimination" justifications).
The court concludes that the Supreme Court in Bakke did not recognize the achievement of racial diversity in university admissions as a compelling state interest. The court further concludes that under the Supreme Court's post-Bakke decisions, the achievement of such diversity is not a compelling state interest because it is not a remedy for past discrimination.
This is not to say that diversity does not have important educational benefits. The evidence defendants submitted on this issue, both at trial and in their summary judgment briefs, demonstrated that the educational atmosphere at the law school is improved by the presence of students who represent the greatest possible variety of backgrounds and viewpoints. Several law professors testified that classroom discussion is livelier, more spirited, and simply more enlightening and interesting when students from varying walks of life, and representing varying world views, participate. As Professor Syverud has written, "in the best classrooms, every voice is heard, and the quality of the education received is largely a function of the diversity of viewpoints and experiences among the students in the class." Report of Kent Syverud, p. 3.
However, a distinction should be drawn between viewpoint diversity and racial diversity. While the educational benefits of the former are clear, those of the latter are less so. The defendants' witnesses emphasized repeatedly that it is a diversity of viewpoints, experiences, interests, perspectives, and backgrounds which creates an atmosphere most conducive to learning. As Dean Lehman testified, it is primarily a "diversity of views" that the law school seeks.
The connection between race and viewpoint is tenuous, at best. The defendants walk a fine line in simultaneously arguing that one's viewpoints are not determined by one's race but that certain viewpoints might not be voiced if students of particular races are not admitted in significant numbers. Some of defendants' witnesses testified that classroom discussion is improved when the class is racially diverse, and some gave examples of viewpoints expressed in class by underrepresented minority students. However, these witnesses generally conceded that these *850 viewpoints might equally have been expressed by non-minority students.
Another asserted benefit of racial diversity is, essentially, that it promotes socialization. That is, cross-racial understanding is advanced, and racial stereotypes break down, in a multi-racial setting. This, it is argued, enhances all students' legal education because it enables them to better understand persons of different races and better equips them to serve as lawyers in an increasingly diverse society and an increasingly competitive world economy. See, e.g., Report of Derek Bok, pp. 3, 11, 13, 17; Report of Robert Webster, pp. 2-5; Amicus Curiae Brief of General Motors Corp., pp. 8-25. Defendants have also submitted a lengthy report by Patricia Gurin, a professor of psychology and women's studies at the University of Michigan, whose research suggests that "[s]tudents learn better in a [racially and ethnically] diverse educational environment, and they are better prepared to become active participants in our pluralistic, democratic society once they leave such a setting." Report of Patricia Gurin, p. 3.
The court does not doubt that racial diversity in the law school population may provide these educational and societal benefits. Nor are these benefits disputed by the plaintiffs in this case. Clearly the benefits are important and laudable. Nonetheless, the fact remains that the attainment of a racially diverse class is not a compelling state interest because it was not recognized as such by Bakke and it is not a remedy for past discrimination.
Even if racial diversity were a compelling state interest, defendants' use of race as an admissions factor would be constitutional only if "narrowly tailored" to serve that interest. See Adarand, 515 U.S. at 227, 115 S. Ct. 2097. For the following reasons, the court finds that defendants' use of race has not been so narrowly tailored at any time under consideration in this case.
First, defendants have indicated that they use race in admissions to the extent necessary to achieve a "critical mass" of underrepresented minority students. None of the witnesses was able to clearly define critical mass in terms of numbers or percentages. While Professor Lempert indicated that critical mass lies somewhere between 11% and 17% of the entire class, Professor Syverud stated that even one to three minority students in a particular classroom might suffice. The current director of admissions, Erica Munzel, testified that she does not know what number or percentage of the entering class would constitute critical mass, and that she relies on faculty members to keep her informed as to whether she is achieving it. "Critical *851 mass" has proved to be an amorphous concept. Apparently defendants know it when they see it, but it cannot be quantified. Narrow tailoring is difficult, if not impossible, to achieve when the contours of the interest being served are so ill-defined.
Second, there is no time limit on defendants' use of race in the admissions process. The Supreme Court has been highly critical of racial classifications which are not strictly limited in duration. See, e.g., Croson, 488 U.S. at 510, 109 S. Ct. 706 ("Proper findings ... defin[ing] both the scope of the injury and the extent of the remedy ... serve to assure all citizens that the deviation from the norm of equal treatment of all racial and ethnic groups is a temporary matter, a measure taken in the service of the goal of equality itself"); Wygant, 476 U.S. at 275, 106 S. Ct. 1842 ("The role model theory allows the Board to engage in discriminatory hiring and lay-off practices long past the point required by any legitimate remedial purpose"). The defendants have indicated that they will continue to use race as a factor in admissions for as long as necessary to admit a critical mass of underrepresented minority students, and no one can predict how long that might be. Such indefiniteness weighs heavily against a finding of narrow tailoring.
Third, by using race to ensure the enrollment of a certain minimum percentage of underrepresented minority students, the law school has made the current admissions policy practically indistinguishable from a quota system. As noted above, the law school has an unwritten policy of constituting each entering class so that at least 10-12% are students from underrepresented minority groups. This percentage has fluctuated somewhat from one year to another, but 10-12% is the approximate percentage that has been established as the minimum level needed to achieve a "critical mass" of students from these groups. While the law school has not set aside a fixed number of seats for underrepresented minority students, as did the medical school in Bakke, there is no principled difference between a fixed number of seats and an essentially fixed minimum percentage figure. Under either system, students of all races are not competing against one another for each seat, with race being simply one factor among many which may "tip the balance" in particular cases. The reservation of some seats for applicants of particular races, and the attendant lack of competition for those seats, was the principle reason Justice Powell found U.C. Davis' quota system unconstitutional. See Bakke, 438 U.S. at 315-19, 98 S. Ct. 2733. While defendants contend that the law school's admissions policy differs from U.C. Davis' in that all applicants compete against one another, the fact of the matter is that approximately 10% of each entering class is effectively reserved for members of particular races, and those seats are insulated from competition. The practical effect of the law school's policy is indistinguishable from a straight quota system, and such a system is not narrowly tailored under any interpretation of the Equal Protection Clause. See, e.g., Bakke, 438 U.S. at 307, 98 S. Ct. 2733 ("Preferring members of any one group for no reason other than race or ethnic origin is discrimination for its own sake"). It appears that the law school is engaging in simple racial balancing by focusing so carefully on admitting and enrolling a particular percentage of students from particular racial groups.
Fourth, there is no logical basis for the law school to have chosen the particular racial groups which receive special attention under the current admissions policy. The 1992 admissions policy, at page 12, identifies "African Americans, Hispanics and Native Americans" as "groups which have been historically discriminated *852 against [and] who without this commitment might not be represented in our student body in meaningful numbers." During some of the years at issue in this lawsuit, the law school bulletin indicated that special attention has been given to "students who are African American, Mexican American, Native American, or Puerto Rican and raised on the U.S. mainland." The law school has failed to offer a principled explanation as to why it has singled out these particular groups for special attention. Certainly, other groups have also been subjected to discrimination, such as Arabs and southern and eastern Europeans to name but a few, yet the court heard nothing to suggest that the law school has concerned itself as to whether members of these groups are represented "in meaningful numbers." No satisfactory explanation was offered for distinguishing between Puerto Ricans who were raised on the U.S. mainland from Puerto Ricans who were raised in Puerto Rico or elsewhere. No satisfactory explanation was offered for singling out Mexican Americans but, by implication, excluding from special consideration Hispanics who originate from countries other than Mexico. A special "commitment" is made to African Americans, but apparently none is made to blacks from other parts of the world. This haphazard selection of certain races is a far cry from the "close fit" between the means and the ends that the Constitution demands in order for a racial classification to pass muster under strict scrutiny analysis. If the law school may single out these racial groups for a special commitment today, there is nothing to prevent it from enlarging, reducing, or shifting its list of preferred groups tomorrow without any reasoned basis or logical stopping point.
A fifth and final factor the court must note in this connection is the law school's apparent failure to investigate alternative means for increasing minority enrollment. See Croson, 488 U.S. at 507, 109 S. Ct. 706 (finding that narrow tailoring was absent because, among other reasons, the city failed to consider "the use of race-neutral means to increase minority business participation in city contracting"); Wygant, 476 U.S. at 280 n. 6, 106 S. Ct. 1842 ("The term `narrowly tailored' ... may be used to require consideration of whether lawful alternative and less restrictive means could have been used.... [T]he classification at issue must `fit' with greater precision than any alternative means"). The court did hear some testimony from witnesses who indicated that underrepresented minority students cannot be enrolled in significant numbers unless their race is explicitly considered in the admissions process. However, the court heard very little *853 testimony from the authors of the 1992 admissions policy, or from those who have been involved in administering it, as to whether the deans or the faculty at the law school itself have ever given serious consideration to race-neutral alternatives. Such efforts might have included increasing recruiting efforts, decreasing the emphasis for all applicants on undergraduate GPA and LSAT scores, using a lottery system for all qualified applicants, or a system whereby a certain number or percentage of the top graduates from various colleges and universities are admitted. Even if these alternatives would not be as effective in enrolling significant numbers of underrepresented minority students, the law school's failure to consider them, and perhaps experiment with them, prior to implementing an explicitly race-conscious system militates against a finding of narrow tailoring.
For all of these reasons, the court believes that the attainment of a racially diverse student body is not a compelling state interest. Even if it were, the law school's admissions policy is not narrowly tailored to serve that interest. Accordingly, the court concludes that the law school's 1992 admissions policy violates the Equal Protection Clause of the Fourteenth Amendment and Title VI of the 1964 Civil Rights Act.
Nonetheless, having listened to extensive testimony and having reviewed all of the relevant documents, the court also concludes that the individual defendants are qualifiedly immune from damages. Under Harlow v. Fitzgerald, 457 U.S. 800, 818, 102 S. Ct. 2727, 73 L. Ed. 2d 396 (1982), "government officials performing discretionary functions generally are shielded from liability for civil damages insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known." In Anderson v. Creighton, 483 U.S. 635, 640, 107 S. Ct. 3034, 97 L. Ed. 2d 523 (1987), the Supreme Court indicated that "[t]he contours of the right must be sufficiently clear that a reasonable official would understand that what he is doing violates that right. This is not to say that an official action is protected by qualified immunity unless the very action in question has previously been held unlawful; but it is to say that in the light of preexisting law the unlawfulness must be apparent."
While the court is convinced that the law school's admissions policy is unconstitutional, the court is equally convinced that the individual defendants (Lee Bollinger, Jeffrey Lehman, and Dennis Shields) did not participate in the adoption or administration of a policy they either knew or had reason to believe was unconstitutional. The state of the law in this area is murky, to say the least. No Supreme Court decision since Bakke has addressed the constitutionality of affirmative action in university admissions, and Bakke itself, to borrow a phrase, "is a riddle wrapped in a mystery inside an enigma." Lower courts have struggled for years, with little success, to divine the legal principles to be gleaned from Bakke, and to understand how, if at all, the teachings of other affirmative action cases may apply to the higher education context. Under these circumstances, university officials understandably had difficulty formulating admissions policies that conformed with Bakke.
Moreover, it is clear from the defendants' testimony that they were concerned *854 about the constitutionality of the law school's admissions policy and that they attempted to comply with Bakke, as they interpreted the decision, while still striving to fulfill their admissions goals. President Bollinger testified that he formed the faculty Admissions Committee in the fall of 1991 because he wanted to ensure that the law school was in compliance with the law. He deliberately selected committee members who were knowledgeable about Bakke and constitutional law. Dean Lehman, who served on that committee, testified that the committee members wanted the law school's admissions policy to be lawful, especially in light of Bakke. Dennis Shields, the director of admissions from 1991 to 1998, also served on that committee. Nothing in his testimony, or in any other evidence admitted in this case, suggests that he knew or should have known that it was unconstitutional for the law school to consider race as part of its admissions policy.
In short, the "contours" of the legal standards governing affirmative action in university admissions have not been so clearly defined as to permit reasonable law school officials to know whether it was constitutional to use race in the admissions process in order to assemble a racially diverse student body. The court finds that the individual defendants in this case acted reasonably and in good faith in adopting and administering the policy in question. Accordingly, the court shall grant the motion of these defendants for summary judgment on grounds of qualified immunity.
Plaintiffs also seek, as part of their Title VI claim, money damages from the board of regents. The board argues that this damages claim is barred by the Eleventh Amendment, which generally prevents individuals from obtaining damages from a state or a state institution in federal court. See, e.g., Alabama v. Pugh, 438 U.S. 781, 782, 98 S. Ct. 3057, 57 L. Ed. 2d 1114 (1978); Edelman v. Jordan, 415 U.S. 651, 662-63, 94 S. Ct. 1347, 39 L. Ed. 2d 662 (1974).
However, the board acknowledges that in this context Congress has abrogated Eleventh Amendment immunity by conditioning the receipt of federal funds upon the duty to comply with various anti-discrimination laws. Title 42 U.S.C. § 2000d-7(a) states:
(1) A State shall not be immune under the Eleventh Amendment of the Constitution of the United States from suit in Federal court for a violation of section 504 of the Rehabilitation Act of 1973, title IX of the Education Amendments of 1972, the Age Discrimination Act of 1975, title VI of the Civil Rights Act of 1964, or the provisions of any other Federal statute prohibiting discrimination by recipients of Federal financial assistance.
(2) In a suit against a State for a violation of a statute referred to in paragraph (1), remedies (including remedies both at law and in equity) are available for such a violation to the same extent as such remedies are available for such a violation in the suit against any public or private entity other than a State.
Nonetheless, the board argues that it may be held liable in damages only if it is found to have violated "clearly established legal principles." Defendants' Memorandum of Law in Support of Renewed Motion for Summary Judgment, p. 45. Plaintiffs argue that it is inappropriate to graft a qualified immunity standard onto Title VI, and that the board is liable so long as intentional discrimination has been proven. See Plaintiffs' Memorandum in Opposition to Defendants' Motion for Summary Judgment, p. 33.
*855 The court has reviewed the cases cited by both sides and is convinced that, under Title VI, liability for damages attaches upon a showing of intentional discrimination. The right at issue need not have been "clearly established"; it suffices for a state to have violated a plaintiff's rights by intentional action. In Guardians Ass'n v. Civil Serv. Comm'n of the City of New York, 463 U.S. 582, 608 n. 1, 103 S. Ct. 3221, 77 L. Ed. 2d 866 (1983), a majority of the Court agreed "that a violation of [Title VI] requires proof of discriminatory intent." When intentional discrimination is proven, damages are among the remedies which may be awarded. See Franklin v. Gwinnett County Pub. Sch., 503 U.S. 60, 75-76, 112 S. Ct. 1028, 117 L. Ed. 2d 208 (1992). In Franklin, the Court held that damages could be awarded under Title IX to a student who allegedly had been sexually harassed by a teacher, even though the law in this area was not clearly established. See id. at 75, 112 S. Ct. 1028 (citing a Title VII sexual harassment case as sufficient authority for the proposition that sexual harassment of a student by a teacher is an "intentional action" proscribed by Title IX). The court agrees with the Fifth Circuit's view on this issue:
"Intentional discrimination," as used in this context means that a plaintiff must prove "that the governmental actor, in adopting or employing the challenged practices or undertaking the challenged action, intended to treat similarly situated persons differently on the basis of race." While we agree with the district court's conclusion that the various defendants acted in good faith, there is no question that they intended to treat the plaintiffs differently on account of their race.
Hopwood, 78 F.3d at 957 (citations omitted). Accord, Gratz, 122 F.Supp.2d at 834-36; Smith v. University of Wash. Law Sch., 2 F. Supp. 2d 1324, 1331 (W.D.Wash. 1998). See also Pederson v. Louisiana State Univ., 201 F.3d 388, 410-12 (5th Cir. 2000) (defendants liable for damages under Title IX for "intentional discrimination" where university sports program treated women differently because of their sex, although defendants were ignorant of their legal obligations. Court noted the defendants "need not have intended to violate Title IX, but need only have intended to treat women differently").
In the present case, there is no doubt that the law school's official admissions policy intentionally treats applicants differently because of their race. While the constitutionality of this policy may be debatable, the board of regents is nonetheless subject to liability in damages under Title VI insofar as it intentionally endorsed, or acquiesced in, an admissions policy that treats applicants differently because of their race. The unlawfulness of the policy need not have been clearly established for liability to attach. Accordingly, the court must deny the board's motion for summary judgment.
III. Remedying Societal Discrimination as a Rationale for Using Race as a Factor in University Admissions: The Intervenors' Case
The intervenors presented many witnesses who testified about the history and current status of racial discrimination in this country; the causes of the "achievement gap" between underrepresented minority *856 and non-minority students; the alleged cultural bias in standardized tests such as the SAT and the LSAT; and the recent experiences of some African American and Mexican American students at high schools, colleges and universities. Some of these witnesses are preeminent experts in their respective fields of history, sociology, psychology and education. Others testified about their personal experiences of the deplorable conditions at a Detroit high school, and of isolation and discrimination at predominantly white universities. The court listened intently to nearly 30 hours of testimony from the intervenors' witnesses. The court shall summarize briefly the highlights of each witness's testimony.
Four students testified. Erica Dowdell is an African American student in her junior year at the University of Michigan. She was elected to the university's student assembly on the Defendant Affirmative Action Party ticket. Ms. Dowdell grew up in Detroit and attended public schools, where all or nearly all of the students were African American. She testified to the generally shoddy conditions at her elementary, middle and high schools. Her high school, Lewis Cass Technical High School, is one of the best in Detroit, yet the building is poorly maintained and books and other resources are often lacking. Ms. Dowdell testified that in her senior year she attended a band festival at Livonia Churchill High School in the suburbs, where most of the students are white, and she was hurt and shocked at the differences in such things as the sports facilities and the quality of the students' musical instruments. As an undergraduate student at the University of Michigan, Ms. Dowdell testified that she feels isolated because African Americans are in the minority. She also testified that she experiences racism on a daily basis, and she gave examples of insensitive remarks by other students and instructors. While Ms. Dowdell is doing well academically, she feels that racism negatively affects her performance because it is discouraging and distracting.
Concepcion Escobar is a Mexican American and Native American student at the University of Michigan Law School. She testified primarily about her experiences at a predominantly black public high school in Chicago and then at Amherst College, where most of the students are white and from privileged backgrounds. In her high school there were very few college counselors, and Ms. Escobar was not given any information about scholarships for which she may have been eligible. At Amherst, she abandoned her plans to pursue a pre-med curriculum because her high school had not adequately prepared her for college-level math and science classes. Ms. Escobar indicated that she did not often speak in class at Amherst because the student population was predominantly white and she did not want to play the role of representing her race. She also testified about some of the difficulties she and her classmates had in understanding one another, as they came from such completely different backgrounds.
Crystal James is an African American student in her second year of law school at the University of California at Los Angeles. Ms. James entered UCLA in 1999, three years after the passage of Proposition 209, which banned state-sponsored affirmative action and racial preferences in California. She testified that she was shocked to discover that she was one of only two African American law students in her entering class of 300. Ms. James feels isolated and believes that students and teachers expect her to represent the "black viewpoint" when racial issues are discussed. She also has experienced a loss of self-confidence and optimism, and a decline *857 in her academic performance, which she attributes to her racial isolation and to subtle forms of racism on campus. The latter phenomenon has taken the form of students making anti-affirmative action comments, teachers failing to direct classroom discussion of racial issues in a positive manner, and teachers neglecting to call on her in class because, she believes, they assume she would not be able to respond. In sum, Ms. James testified that the loss of affirmative action in California has resulted in far fewer underrepresented minority students being admitted to prestigious state universities such as UCLA and Berkeley, and that those who are admitted feel isolated and defeated.
The last student to testify was Tania Kappner, an African American graduate of the master's of education program at the University of California at Berkeley. Like Ms. James, Ms. Kappner testified about her feelings of isolation, since the number of black students in her program dropped significantly following the passage of Proposition 209. Ms. Kappner teaches English at a high school near Berkeley, and she testified that the abolishment of affirmative action has had a demoralizing effect on her minority students.
The court also heard testimony from several university professors who, as noted above, are highly respected scholars in their various fields of endeavor. Gary Orfield is a professor of education and social policy at Harvard University. He testified extensively about racial segregation and affirmative action. His expert report, which is a collection of articles he has published on these issues, was admitted as Exhibit 167.
Professor Orfield testified that racial segregation in American public schools is particularly pronounced in the northeast and midwest sections of the country, and that Michigan is among the four most highly segregated states. Of all the states, Michigan has the highest percentage (64%) of black students who attend schools whose student populations are 90-100% minority. See Exhibit 196. Detroit has the most highly segregated schools of any major metropolitan area in the country. Over 90% of the children enrolled in Detroit public schools are African American. Professor Orfield testified that racial segregation in the schools is related to segregation in housing, and he characterized housing in Detroit as "hyper-segregated." Further, Professor Orfield indicated that segregation in schools is associated with high levels of poverty which, in turn, are associated with poor resources and decreased educational opportunities. And as a rule, the poorest schools are the ones with the highest minority population. See Exhibit 197. Two-thirds of African American and 70% of Hispanic schoolchildren attend segregated schools. Professor Orfield testified that, as a result, most minority students do not receive a public education that prepares them for college. He noted, for example, that the University of Chicago has routinely rejected applicants who are Chicago valedictorians, even those from magnet schools, because city schools do not compare with schools in the suburbs. Professor Orfield has written that "[c]ollege admissions officers have long known that class rank is hardly comparable from one high school to the next. The top students in many high-poverty schools are woefully unprepared for college." Exhibit 167 D, G. Orfield and E. Miller, Chilling Admissions 10 (1998).
Professor Orfield was also asked several questions about the effects of Hopwood v. State of Texas, 78 F.3d 932 (5th Cir.1996), *858 which banned the consideration of applicants' race in university admissions in the Fifth Circuit; and of Proposition 209, which in 1996 eliminated all racial preferences, including in university admissions, in California. Professor Orfield noted that following Hopwood and Proposition 209, enrollment of underrepresented minority students dropped sharply in Texas and California. For example, at the University of Texas School of Law at Austin, whose admissions system was challenged in Hopwood, the percentage of the entering class that is African American dropped from 5.8% (29 students) in 1996 to 0.9% (4 students) in 1997. Native American enrollment at that law school dropped from 1.2% (six students) in 1996 to 0.2% (one student) in 1997. Hispanic enrollment dropped from 9.2% (46 students) in 1996 to 6.7% (31 students) in 1997.See Exhibit 131.
Professor Orfield also cited statistics showing that similar declines in law school enrollments occurred in California following the passage of Proposition 209. In 1996, for example, 89 Hispanics, 43 African Americans and 10 Native Americans were enrolled as first year students at the top three public California law schools. In 1997, these numbers fell to 59, 16, and 4, respectively. At Berkeley, only one African American enrolled in the freshman law class in 1997, whereas 20 had been enrolled in the freshman class the year before. See Exhibit 132.
Part of the reason for the drop in the number of underrepresented minorities in law schools, following the elimination of affirmative action, can be seen in Exhibits 199 and 200. These exhibits are graphs showing the distribution of LSAT scores by race in 1997-1998. Exhibit 199 shows that the test scores of each racial group formed a similar bell-shaped curve, but that the curve for whites is located to the right of those of all underrepresented minority groups. This confirms the data from other exhibits, which reveal that the median LSAT score of white test-takers is significantly higher than the median underrepresented minority score. See Exhibit 137, p. 19; Exhibit 139, p. 20; Exhibit 141, p. 17. Exhibit 200 shows the distribution of LSAT scores by race, but indicates the absolute numbers of tests taken. The curve of the white test-takers absolutely dwarfs the curves of the minority test-takers; this graph dramatically illustrates the fact that the vast majority of test-takers are white and that the vast majority of high LSAT scores are obtained by white test-takers. In Professor Orfield's words, "hardly any" minority test-takers receive scores in the 169-180 range from which the best law schools expect their applicants to come.
Professor Orfield concluded that affirmative action is needed in order to achieve racial diversity at law schools because the academic averages between underrepresented minority and non-minority groups is such that a "very segregated outcome" would result if race were not considered in the admissions process. He believes there is no double standard in doing so, since most minority children are so disadvantaged in their public school education.
*859 Professor Orfield also testified that the elimination of affirmative action in university admissions leads to a phenomenon known as "cascading," whereby minority students no longer gain admittance to the most prestigious institutions and are instead relegated to less selective ones. In his view, this is undesirable because the more prestigious universities are "leadership training institutions."
The court also heard extensive testimony from another distinguished sociologist, Dr. Walter Allen, a professor of sociology at the University of California at Los Angeles. His expert reports were admitted as Exhibits 156-158.
Professor Allen testified that affirmative action in the law school admissions process is needed in order to counterbalance the negative effects of racism on the academic performance of underrepresented minority students in college. Professor Allen believes that underrepresented minorities suffer systematic deprivations in both housing and educational opportunities, which account for much of the "GPA gap" that exists between minority and non-minority college students. Professor Allen indicated that underrepresented minority students at predominantly white undergraduate institutions often experience a racially hostile environment, which depresses their confidence and desire to succeed academically. He stated that African American students perform better at historically black colleges and universities (HBCU's) than at predominantly white institutions because they feel validated and encouraged to succeed in the friendlier and more supportive atmosphere at HBCU's. In contrast, at predominantly white institutions, African Americans and other underrepresented minority students often feel isolated, disliked and marginalized.
Professor Allen studied the racial climate at four of the undergraduate institutions which supply a portion of the students who apply to the University of Michigan Law School. The "feeder schools" selected for this research were the University of Michigan, Michigan State University, the University of California at Berkeley, and Harvard University. Professor Allen also studied the racial climate at the law school itself. His research was conducted at these campuses principally by studying the responses of minority and non-minority students in focus groups and surveys.
Professor Allen concluded that the students at the four feeder campuses "described the overall racial climate of their campuses as a place where they feel their presence is questioned and belittled." Exhibit 157, p. 11. The underrepresented minority focus group participants described various items which contributed to this climate, including insensitive or racist remarks by faculty and students, a feeling of isolation due to the low number of fellow students of the same race, avoidance of racial issues in classroom discussions, the *860 lack of minority role models such as faculty members, exclusion from white study groups, and unequal treatment by campus police. Some of the focus group participants also indicated that the negative racial climate caused them to feel alienated and discouraged and that this harmed their academic performance.
Professor Allen also surveyed 200 students at the feeder institutions. Thirty-five percent felt discriminated against at their campus because of their race, and 47% felt they had been unfairly graded. See Exhibit 157, p. 51. However, 75% felt they had made the right choice in attending their undergraduate institution, and 84% "would definitely recommend their university to other students who want to attend college." Id. pp. 51, 55.
The court also heard testimony from Professor John Hope Franklin, a renowned historian with special expertise in the history of race relations in the United States. His expert report was admitted as Exhibit 163. Professor Franklin described his personal experiences as the victim of racial discrimination as an undergraduate student in Nashville, as a graduate student doing research in North Carolina, and as a young history professor seeking housing in Brooklyn. Professor Franklin testified that even now, as a distinguished educator and historian, he experiences racism. For example, in recent years he has been approached more than once by a white person in a hotel lobby or private club who asked him to fetch her coat or car. Professor Franklin believes that great strides have been made in improving race relations, but that much more remains to be done. Professor Franklin believes that active effort is needed to eliminate the race problem, and that the abolishment of affirmative action at universities is a step toward resegregation. However, he also expressed the belief that academic standards should not be lowered for minority students, and that all people should be judged on their individual merits.
Three witnesses testified on the subject of bias in standardized testing. This was important testimony, given the significance of such tests in undergraduate and graduate admissions. The first such witness was Jay Rosner, the executive director of the San Francisco office of the Princeton Review Foundation (PRF). The PRF is affiliated with the Princeton Review, an organization that provides LSAT preparation courses to approximately 10,000 students each year. The PRF focuses on providing such courses to underrepresented minority students.
Mr. Rosner's expert reports were admitted as Exhibits 168 and 169. He testified that the LSAT is developed and administered by the Law School Admission Council (LSAC), which claims that the LSAT score helps predict students' first year grades in law school. However, according to Mr. Rosner, the actual correlation is only 16-20%, which is to say that 80-84% of first year law school grades are not predicted by the LSAT. Mr. Rosner stated that test preparation courses like those offered by the Princeton Review, or by its competitor Kaplan, Inc., generally improve one's LSAT score by approximately seven points.
Mr. Rosner testified that the PRF engages in "targeted outreach" to offer its test preparation services to African American and other minority students. In his experience, white students are aware of the benefits of taking such a course, but minority students do not share this awareness. For example, at Howard University, *861 a predominantly African American institution, Mr. Rosner once had difficulty filling the 15 seats in an LSAT preparation course, despite the fact that the customary $1,000 fee was reduced to $200. He also testified that despite PRF's outreach efforts, the vast majority of the students who take an LSAT preparation course are white, and that this fact accounts for some of the test-score gap between minority and non-minority test-takers.
Martin Shapiro, a professor of psychology at Emory University, testified on the issue of bias in standardized testing. He also submitted an expert report, which was admitted as Exhibit 170. Professor Shapiro testified that standardized admission tests, by the manner in which new test questions are "pretested," tend to perpetuate bias against groups which have performed poorly on the tests in the past. He also noted the large difference between the average LSAT score for whites and blacks, and the weak correlation, of about 27%, between performance on the LSAT and first year law school grades.
On the issue of bias in standardized testing, the court also heard testimony from David White, director of Testing for the Public, a group which offers test preparation courses for women and minority students. His expert report was accepted as Exhibit 173. White provided information similar to that provided by Dr. Larntz regarding the gaps in LSAT scores and UGPA's between whites and underrepresented minorities. Exhibit 219 shows that while 46% of white law school applicants in 1996-1997 scored at or above 155 on the LSAT, only 8% of black applicants did so. While 46% of white applicants had a UGPA of 3.25 or above, only 17% of black applicants had a UGPA in this range. And while 27% of white applicants had both qualifications, only 3% of black applicants had both. Whites on average score higher on the LSAT than any other racial group at all socioeconomic levels. See Exhibits 220, 221. Even among applicants who attend the same undergraduate institution and have the same undergraduate GPA, the LSAT gap as compared to white applicants is 4.0 points for Native Americans, 6.8 points for Hispanics, and 9.2 points for African Americans. See Exhibit 223.
The court also heard testimony from Dr. Eugene Garcia, the Dean of the School of Education at the University of California at Berkeley. His expert report was admitted as Exhibit 164. Dr. Garcia testified about the effects of Proposition 209 on student admissions at the eight campuses of the University of California system. From 1995 to 2000, underrepresented minority admissions dropped by 1% overall. See Exhibit 213. Underrepresented minority admissions at the most selective campuses, Los Angeles and Berkeley, have dropped by 45% and 42%, respectively, whereas the admission rates at Santa Cruz and Riverside, which are less selective, have increased by 27% and 87%, respectively. Dr. Garcia attributed these *862 changes to the fact that underrepresented minority applicants generally have lower GPA's and test scores. Their lower average GPA is due in large part to the generally poorer quality of the schools attended by underresented minority students. And their lower average test scores are to some extent due to the fact that standardized tests, like the SAT and LSAT, are "heavily loaded with academic English." This disadvantages many Hispanics because English is their second language. African Americans are also disadvantaged because many 60% by his estimate speak "Black English." Dr. Garcia indicated that test preparation courses do improve test scores, but that these courses are expensive, which adds to the disadvantage against poorer groups. Dr. Garcia characterized the loss of underrepresented minority students at Los Angeles and Berkeley, and the corresponding gain of underrepresented minority students at Santa Cruz and Riverside, as a "redistribution" within the University of California system. This is the phenomenon which Professor Orfield called "cascading."
Dr. Garcia testified that the University of California has been unable to admit underrepresented minority students in significant numbers by any means other than by explicitly considering race in admissions. He favors race-conscious admissions policies because they enable universities to consider historical discrimination against particular racial groups. Dr. Garcia testified that a decrease in racial diversity at universities decreases the quality of the educational experience for all students. He also expressed the belief that institutions of higher education prepare students to take leadership roles later in life, and that part of this preparation should include exposure to people of other races, particularly in light of the fact that American society is itself becoming increasingly diverse.
On cross-examination, Dr. Garcia acknowledged that at the school of education of which he is the dean, 28% of this year's class consists of underrepresented minority students. Dr. Garcia indicated that the school of education was able to achieve this level of diversity by decreasing reliance on the Graduate Record Examination (GRE), and by expending greater effort in recruiting new students.
The court also heard testimony from Faith Smith, the president of Native American Educational Services (NAES) College; and Frank Wu, an associate professor of law at Howard University. Ms. Smith testified about the educational difficulties which face Native Americans, and the importance of increasing the number of Native American lawyers. Her expert report was admitted as Exhibit 171. Professor Wu testified about the historical and present societal discrimination against Asian Americans, particularly in the areas of employment and housing. His expert report was admitted as Exhibit 175.
Finally, the intervenors called Professor Richard Lempert. As noted above, Professor Lempert has served for many years on the faculty of the law school as well as in the sociology department at the University of Michigan. While the defendants called him to testify about the genesis of the law school's admissions policy, the intervenors called him to testify about a survey study he and others performed which examined the career success of underrepresented minority graduates of the law school. Professor Lempert's study, and a condensed version thereof, were admitted as Exhibits 165 and 166.
While Professor Lempert's report is quite lengthy, he summarized the main conclusions as follows:
The results reveal that almost all of Michigan Law School's minority graduates pass a bar exam and go on to have *863 careers that appear successful by conventional measures. In particular, the survey indicates that minority graduates (defined so as to include graduates with African American, Latino, and Native American backgrounds) are no less successful than white graduates, whether success is measured by the log of current income, self-reported satisfaction, or an index of service contributions. Also, although an admissions index that combines LSAT scores and undergraduate grade-point average is a significant predictor of law school grades, it does not predict career success on any of our three outcome measures.
Exhibit 166, p. 1. Professor Lempert also wrote:
What we find is a strong, statistically significant relationship between LSAT and UGPA, on the one hand, and grades at the end of three years of law school on the other. But we find no significant relationship between the LSAT or UGPA and what matters more the achievement of students after graduation. Drawing on work done in connection with the affirmative action lawsuit against the University of Michigan Law School, we can also say that had the LSAT and the UGPA been the only criteria for admissions at Michigan, few of Michigan's minority graduates would have been admitted to the school, even though their career success since law school is similar to the career success of Michigan's white graduates and consistent with the aspirations Michigan has for all students it admits.
Id. at 33-34. In his testimony, Professor Lempert also indicated that the law school's race-conscious admissions policy has helped to integrate the legal profession. He testified that if affirmative action in admissions did not exist, the number of underrepresented minorities admitted to any law school in the country would be reduced by 75%. Not only would underrepresented minority applicants "cascade down" to less selective law schools, but they probably would, if admitted, simply displace other minority applicants. Professor Lempert believes that the LSAT is a valid predictor of academic success in law school (though not of later success in the profession), and he does not believe there are any "hard data" showing any racial bias against minorities in terms of the LSAT's predictive capacity. See also Exhibit 166, p. 34 ("LSAT scores and an index based on LSAT scores and UGPAs are significantly correlated with both first-year and final law school grade-point averages").
B. Findings of Fact and Conclusions of Law
From this extensive body of evidence, the court makes the following findings and conclusions.
First, there is no question about the long and tragic history of race discrimination in this country. At different times in history, Native Americans have been deprived of land and forced onto reservations; African Americans have been held in slavery; and Asian Americans have been detained in internment camps. Many ethnic and racial groups including, but not limited to, the underrepresented minority groups identified by the law school's admissions policy have been victims of discrimination in housing, employment, education, public accommodations, and elsewhere.
*864 While American public policy has sought to eradicate race discrimination, its lingering effects are apparent. As Professor Wu has written, "[F]or many people of color, racism has decreased the amount and value of economic, social, and cultural capital inherited from our ancestors. Not only did we receive less material wealth, we also received less `insider knowledge' and fewer social contacts so instrumental to one's educational and professional advancement in America." Exhibit 175, p. 131. The lingering effects can be seen, for example, in lower education rates and higher poverty rates among some minority groups as compared to whites.
The evidence presented at trial also made clear that underrepresented minority groups have, on average, lower undergraduate grade-point averages than whites. Among applicants accepted by the University of Michigan law school from 1995 through 2000, the median UGPA of every underrepresented minority group has been lower than the median UGPA of Caucasians by approximately one-tenth to three-tenths of a point. A greater gap exists in LSAT scores. From 1995 to 2000, the median LSAT score of the underrepresented minority groups has been lower than the median LSAT score of Caucasians by approximately seven to nine points.
The reasons for the "GPA gap" and the "LSAT gap" are complex. While one must be cautious in making generalizations, the evidence at trial clearly indicates that much of the GPA gap is due to the fact that disproportionate numbers of Native Americans, African Americans, and Hispanics live and go to school in impoverished areas of the country. It should not surprise anyone that students who attend schools where books are lacking, where classrooms are overcrowded, and where Advanced Placement or other higher level courses are not offered are at a competitive disadvantage as compared with students whose schools do not suffer from such shortcomings. An educational deficit in the K-12 years will, for most students, have a negative ripple effect on academic performance in college. Ms. Escobar, for example, had to drop out of the pre-med program in college because her high school, a large public school in Chicago, failed to provide her with the necessary *865 mathematics and science background. Of course, students of any race will have difficulty "catching up" in college if they have had a poor K-12 education.
The intervenors have also argued that the GPA gap is due in part to the racially hostile environment at college and university campuses. The court does not doubt that many underrepresented minority students find the racial climate hostile at the law school's "feeder" institutions. Nor does the court doubt that such a climate can have a negative effect on these students' academic performance. Nonetheless, the court is unable to give any weight to Professor Allen's study of this issue, due to the small number of students who participated in the focus groups and surveys and due to the manner in which the students were selected.
Regarding the small numbers, the court notes that a total of 65 students participated in Professor Allen's focus groups. Of these, 18 were white and Asian students; only 47 were members of the underrepresented minority groups at issue in this case. Of these 47 students, 20 participated in four focus groups at the University of Michigan; 15 participated in three focus groups at Berkeley; eleven participated in two focus groups at Harvard; and one participated in one focus group at Michigan State University. See Exhibit 157, Figure 2. Similarly small numbers of underrepresented minority students participated in the surveys. Of the 200 total respondents, 140 were African Americans, Hispanics or Native Americans, 47 of whom were participants in the focus groups. See Exhibit 157(2), Tables 3, 6, 9 and 12. At trial, Professor Allen acknowledged that it is difficult to make generalizations based on results from such small sample sizes. The court seriously doubts whether any reliable generalizations can be made about the racial climate at the feeder institutions, whose total student population well exceeds 125,000, based on the information gathered from 47 focus group participants and 140 survey respondents.
The reliability of Professor Allen's conclusions is further undermined by the manner in which students were solicited for inclusion in the research project. While Professor Allen supervised the focus groups and the composition of the survey, he relied on student volunteers to recruit the study's participants. Several of these solicitations, in the form of electronic mail and facsimile transmissions, were co-authored by United for Equality and Affirmative Action, which is one of the intervenors in this case. These solicitations called upon students "to contribute to a legal case that will impact the educational opportunities of minority students for generations to come. It is our `Brown v. Board of Education' and we must do everything we can to ensure a victory." Exhibit 176. Other solicitations began with the heading, "URGENT PLEASE PARTICIPATE IN A STUDY FOR THE U OF M AFFIRMATIVE ACTION CASES." Exhibit 177. These solicitations also indicated that "[t]he student intervenors in the watershed University of Michigan affirmative action cases are seeking students ...." Id. Professor Allen acknowledged that the wording of these solicitations "possibly" could have biased *866 the selection process. The court believes not only that such bias may have occurred, but that it surely did occur, because the solicitations clearly communicated a desire to attract research participants who shared a particular viewpoint.
The court has three other difficulties with Professor Allen's research results. The first is that they fail to quantify the degree to which the academic performance of underrepresented minority students is reduced by the racially hostile environment at the feeder institutions. Nor did the evidence show that the academic performance of all underrepresented minority students is reduced to the same extent, or whether the effect varies from one individual to another. Thus, even if it were constitutionally permissible for the law school to consider the race of underrepresented minority applicants in order to compensate for the reduction in their grades and test scores attributable to a racially hostile environment in college, the court has no means of knowing the permissible extent of such compensation, either for particular individuals or for all underrepresented minority law school applicants.
Second, even if the effect of a racially hostile environment on academic performance could be quantified, no testimony was offered as to the percentage or numbers of underrepresented minority students at the University of Michigan Law School who come from the four "feeder" institutions. Nor was any testimony or other evidence offered to show whether the racial climate at other undergraduate institutions from which underrepresented minority law students have graduated is hostile to the same degree as at the four feeder schools. Thus, it is impossible to know whether Professor Allen's findings apply to most or only some of the underrepresented minority law school applicants.
Third, Professor Allen's research failed to inquire into the effects of racial hostility on the academic performance of minority groups other than African Americans, Latinos, Asian Americans and Native Americans. It seems quite likely that Arab students, or those from eastern or southern Europe, or adherents of a religious minority, may also to one degree or another be subjected to a hostile undergraduate environment. Given the narrow focus of Professor Allen's research project, the court and the law school admissions office can only guess as to precisely which applicants are entitled to a "boost."
The gap in LSAT scores between underrepresented minorities and Caucasians is even more difficult to explain than the gap in undergraduate GPA's. The intervenors contend that this gap is due to various factors, including cultural bias in the LSAT and a psychological phenomenon experienced by underrepresented minority test-takers known as "stereotype threat." The court is unable to find that anything in the content or design of the LSAT biases the test for or against any racial group. If such a bias exists, it was not proved at trial. Jay Rosner gave some testimony about "white preference" and "black preference" questions on the SAT (not the LSAT). Three of these questions appeared in newspaper articles in 1997 and 1999, while Mr. Rosner found others in unspecified "`88-`89 ETS data." See Exhibits 202-210. The court is unable to give *867 any weight to this testimony, as Mr. Rosner is not an expert in test design. Nor did he claim to have studied, or to have the expertise to study, the issue of bias in standardized tests. Professor Shapiro testified generally as to how pre-testing procedures can be used to perpetuate a testscore gap, but he did not testify that such procedures are used in connection with the LSAT. Professor Shapiro and David White testified about the existence of the LSAT gap, but they did not demonstrate to the court's satisfaction that the gap is due to the selection of the test questions or the design of the test.
The intervenors have also suggested that underrepresented minority students do less well than Caucasians on the LSAT, and on other standardized tests, because of a psychological phenomenon known as "stereotype threat." No witnesses testified directly on this subject. However, defendants did file a report by Dr. Claude Steele, a psychology professor at Stanford University who has done some research in this area. See Appendix in support of defendants' motion for summary judgment on grounds of qualified immunity, volume 3. Professor Steele writes that stereotype threat
refers to the experience of being in a situation where one recognizes that a negative stereotype about one's group is applicable to oneself. When this happens, one knows that one could be judged or treated in terms of that stereotype, or that one could inadvertently do something that would confirm it. In situations where one cares very much about one's performance or related outcomes as in the case of serious students taking the SAT this threat of being negatively stereotyped can be upsetting and distracting. Our research confirms that when this threat occurs in the midst of taking a high stakes standardized test, it directly interferes with performance.
Steele Report, p. 7. Professor Steele indicates that he once conducted an experiment where black and white Stanford University sophomores were given a difficult verbal test, using questions from the advanced Graduate Record Examination in literature. When the students were told that their ability was being tested, the black students performed "dramatically worse" than did the white students. But when the students were told that the test was a "problem-solving task," not a measure of ability, black and white performance was the same. Id. at 8. Professor Steele indicates that he has performed other research which "shows stereotype threat to be a very general effect, one that is undoubtedly capable of undermining the standardized test performance of any group negatively stereotyped in the area of achievement tested by the test." Id. at 9.
Due to the sparseness of the evidence on this issue, the court is unable to determine whether stereotype threat explains any part of the gap between Caucasian and underrepresented minority LSAT scores. Professor Steele's report describes his research only in the most general terms. He reports the results of only one experiment he performed using the GRE, and he does not indicate when the experiment was done, how many students participated, whether the results were tested for statistical significance, or whether the results were published and subjected to peer review. Nor has Professor Steele provided any evidence, by way of survey data for example, to show that the members of any particular racial group perceive themselves as being the object of a stereotype that expects underachievement. Professor Steele does not quantify the effect of stereotype threat; nor, at least according to this report, has he performed any research on the LSAT. If there is evidence showing that stereotype threat accounts for some of *868 the LSAT gap, it was not produced in this case.
The intervenors' witnesses offered two explanations for the LSAT gap which the court does find plausible. The first is that the LSAT, and other standardized tests, are written in what Dr. Garcia characterized as "academic English." Evidence at trial indicated that this causes a problem for underrepresented minority students, many of whom do not speak English, or do not speak standard English, as their first language. The second is that underrepresented minority students are less likely to take an LSAT preparation course, which could increase one's score. Mr. Rosner testified that underrepresented minority students enroll in such courses only in "token numbers." This may well be due to the cost of taking a test preparation course, although Mr. Rosner testified that in his experience underrepresented minority students are simply less aware than are Caucasian students about the benefits such a course can provide.
The intervenors have argued strenuously that the law school's race-conscious admissions policy does not apply lower academic standards to underrepresented minority applicants than are applied to non-minority applicants. In this view, there is no "double standard" in the admissions process because the UGPA gap and the LSAT gap are attributable to past and present discrimination against underrepresented minorities in the form of segregation in housing and schools, a racially hostile environment at undergraduate institutions, cultural bias in standardized tests, and stereotype threat. That is to say, underrepresented minority applicants are actually performing at a higher academic level than is suggested by their undergraduate grade-point averages and LSAT scores because of the more difficult conditions under which they were obtained. Therefore, the intervenors argue, the law school's race-conscious admissions policy simply "levels the playing field," or compensates for the differing levels of adversity, by giving some degree of preference to underrepresented minority applicants.
The court has made its findings on the factual issues relevant to this inquiry and, as noted above, the court concludes that the comparatively lower grades and test scores of underrepresented minorities is attributable, at least in part, to general, societal racial discrimination against these groups. While the court may agree with some of the factual underpinnings of the intervenors' argument, the legal conclusion they draw therefrom is flawed both as a matter of logic and as a matter of constitutional law.
The logical flaw in the argument is that it assumes all members of the underrepresented minority groups have suffered adversity entitling them to some degree of upward adjustment in their UGPA and LSAT scores. Conversely, the intervenors' argument assumes that no members of non-minority groups have suffered any such adversity which would entitle them to a similar adjustment in their grades and scores. Of course, neither assumption is correct. Every law school applicant is an individual whose personal history is unique. Some applicants, regardless of their race, may have been the victims of racial discrimination or have attended a substandard school or have grown up in poverty, while others will have been more fortunate. There is no basis in logic or in the evidence for assuming that all members of some racial groups are victims of adverse circumstances or, conversely, that all members of other racial groups are beneficiaries of privilege.
The legal flaw in the intervenors' conclusion is even more daunting, and it is this: *869 the Supreme Court has held that the effects of general, societal discrimination cannot constitutionally be remedied by race-conscious decision-making. For example, in Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 274, 106 S. Ct. 1842, 90 L. Ed. 2d 260 (1986), the Supreme Court stated that it "never has held that societal discrimination alone is sufficient to justify a racial classification. Rather, the Court has insisted upon some showing of prior discrimination by the governmental unit involved before allowing limited use of racial classifications in order to remedy such discrimination."See also id. at 276, 106 S. Ct. 1842 ("Societal discrimination, without more, is too amorphous a basis for imposing a racially classified remedy.") In the present case there has been no evidence, or even an allegation, that the law school or the University of Michigan has engaged in racial discrimination. All of the evidence submitted by the intervenors relates to discrimination on the part of society at large or by entities other than the law school or the University of Michigan. As a matter of federal constitutional law, the law school is not permitted to "level the playing field" in the manner urged by the intervenors. That is, the law school may not consider the race of applicants in order to compensate for the effects of discrimination by others or by society generally.
The intervenors' other argument in support of the law school's race-conscious admissions policy is that it is necessary in order to prevent "resegregation." The intervenors fear that if the law school is prohibited from considering race in admissions, then the number of underrepresented minority students will fall dramatically, perhaps to mere "token" levels. They point to the law schools at Austin and Berkeley, where underrepresented minority enrollment fell sharply following the elimination of affirmative action, as examples of what may occur here.
The court agrees that it would be unfortunate if the number of students from any racial group would decline at the University of Michigan Law School. It is the court's sincere hope that such consequences can be avoided, and some evidence adduced at trial suggested that they can be. Dr. Garcia testified that Berkeley's school of education has succeeded in constituting a class that is 28% minority by increasing its recruiting efforts and decreasing its reliance on standardized test scores. At the most selective law schools in the University of California system, the underrepresented minority population increased *870 from 9.3% of the entering class in 1999 to 10.5% in 2000. See Exhibit 132. Throughout the University of California system, underrepresented minority admissions decreased just one percent from 1995 to 2000, although one must recognize the significant shift of the underrepresented minority student population from the more to the less selective campuses during this time period. See Exhibit 213.
Regardless of one's analysis of the admissions data in California and Texas, the constitutionality of the current admissions system at the University of Michigan Law School does not turn on the predicted consequences of instituting a race-blind admissions system. The current system is either constitutional or it is not. The court is unaware of any precedent for the proposition that a constitutional challenge to a voluntarily adopted racial classification may be defeated by the argument that elimination of the classification will or may have undesirable consequences, be they political, social, economic or otherwise. If undesirable consequences are likely or even certain to occur, the answer is not to retain the unconstitutional racial classification but to search for lawful solutions, ones that treat all people equally and do not use race as a factor.
One such solution may be to relax, or even eliminate, reliance on the LSAT. The evidence presented at trial indicated that the LSAT predicts law school grades rather poorly (with a correlation of only 10-20%) and that it does not predict success in the legal profession at all. If, as its admissions policy states, the law school seeks students who "have substantial promise for success in law school" and "a strong likelihood of succeeding in the practice of law," one must wonder why the law school concerns itself at all with an applicant's LSAT score. Defense counsel, and counsel for the intervenors, asserted at trial that the American Bar Association (ABA) and the Association of American Law Schools (AALS) will not accredit law schools which do not require applicants to submit an LSAT score. This is not entirely accurate. Standard 503 of the ABA's Standards for the Approval of Law Schools states:
A law school shall require all applicants to take an acceptable test for the purpose of assessing the applicants' capability of satisfactorily completing its education program. A law school that is not using the Law School Admission Test sponsored by the Law School Admission Council shall establish that it is using an acceptable test.
Section 6-2 of the AALS bylaws, which covers law school admissions, states:
a. A member school shall admit only those applicants who appear to have the capacity to meet its academic standards.
b. In order that appropriate intellectual rigor may be maintained, a member school shall admit to its first professional degree program only those applicants *871 who have the level of intellectual maturity and accomplishment normally demonstrated by the award of an undergraduate degree.
c. A member school shall deal fairly with applicants for admission.
Exhibit 31. Thus, neither accrediting organization requires law schools to require applicants to take the LSAT. While the ABA does require law schools to "require all applicants to take an acceptable test," such as the LSAT, it does not require that law schools give the test results any particular weight.
Another solution may be for the law school to relax its reliance on undergraduate GPA. The law school's admissions policy acknowledges that, even in combination, the LSAT score and undergraduate GPA are "far from perfect" predictors of success in law school. In fact, the policy asserts that the correlation between the index score and first-year law school grades is merely 27%. See Exhibit 4, Admissions Policy, p. 3. The policy also notes the obvious fact that high undergraduate grades may overstate an applicant's academic achievements or promise, and that low grades may understate them. To determine the significance of an applicant's undergraduate GPA, the law school's admissions officers say they consider such factors as the difficulty of the course work and the reputation of the undergraduate institution. The law school may be able to counterbalance some of the negative effect of the UGPA gap by also determining whether individual applicants have had to overcome any particularly challenging or difficult obstacles. So long as the law school acknowledges that such obstacles may confront an applicant of any race, consideration might be given to such things as growing up in difficult family circumstances, attending underfunded public schools, or learning English as a second language.
Another solution may be for the law school to reduce or eliminate the preference now given to the sons and daughters of University of Michigan alumni. The current admissions director testified that applicants with a family connection to alumni are given "careful consideration," although this factor adds nothing in terms of diversity. Common sense would suggest that a preference of this nature perpetuates past imbalances and has no connection to any measure of "merit."
In this regard, the court would make the obvious observation that, ultimately, the law school student population naturally will become racially diverse under a race-blind admissions system when the gaps in LSAT scores, undergraduate GPA's, and other measures of academic performance are eliminated by investing greater educational resources in currently under-performing primary and secondary school systems. Dr. Garcia testified to some of the efforts being made in this regard in California. This is a social and political matter, which calls for social and political solutions. The solution is not for the law school, or any other state institution, to prefer some applicants over others because of race.
Whatever solution the law school elects to pursue, it must be race neutral. The focus must be upon the merit of individual applicants, not upon assumed characteristics of racial groups. An admissions policy that treats any applicants differently from others on account of their race is unfair and unconstitutional. As a matter of constitutional law, such a system cannot be justified on the grounds that certain races are at a greater competitive disadvantage *872 than others because of discrimination or other societal conditions which may have created an "uneven playing field." Nor can a race-conscious system be upheld based on the predicted consequences of moving to a race-blind system. In sum, the court must reject the arguments raised by the intervenors in defense of the law school's current admissions policy.
The Supreme Court has often stated that "distinctions between citizens solely because of their ancestry [are] odious to a free people whose institutions are founded upon the doctrine of equality." Wygant, 476 U.S. at 273, 106 S. Ct. 1842, quoting Hirabayashi v. United States, 320 U.S. 81, 100, 63 S. Ct. 1375, 87 L. Ed. 1774 (1943). In our history, such distinctions generally have been used for improper purposes. Even when used for "benign" purposes, they always have the potential for causing great divisiveness. For these reasons, all racial distinctions are inherently suspect and presumptively invalid. This presumption may be overcome only upon a showing that the distinction in question serves a compelling state interest, and that the use of race is narrowly tailored to the achievement of that interest. It does not suffice for the interest in question merely to be important, beneficial, or laudable; the interest must be compelling.
For the reasons stated in this opinion, the court concludes that the University of Michigan Law School's use of race as a factor in its admissions decisions is unconstitutional and a violation of Title VI of the 1964 Civil Rights Act. The law school's justification for using race to assemble a racially diverse student population is not a compelling state interest. Even if it were, the law school has not narrowly tailored its use of race to achieve that interest. Nor may the law school's use of race be justified on the alternative grounds urged by the intervenors to "level the playing field" between applicants of minority and non-minority racesbecause the remedying of societal discrimination, either past or present, has not been recognized as a compelling state interest. Accordingly,
IT IS ORDERED that plaintiffs' request for declaratory relief is granted. The court finds and declares that the University of Michigan Law School's use of race in its admissions decisions violates the Equal Protection Clause of the Fourteenth Amendment and Title VI of the Civil Rights Act of 1964.
IT IS FURTHER ORDERED that plaintiffs' request for injunctive relief is granted. The University of Michigan Law School is hereby enjoined from using applicants' race as a factor in its admissions decisions.
IT IS FURTHER ORDERED that the parties' various motions for summary judgment are denied, except that the motion of the individual defendants for summary judgment on grounds of qualified immunity is granted.
IT IS FURTHER ORDERED that the clerk schedule the damages phase of the trial.
*873 EXHIBIT A
 Unless indicated otherwise, the court uses the terms "minorities," "minority groups," and "underrepresented minorities" interchangeably in this opinion to refer to African American, Native American, Mexican American and mainland Puerto Rican students, as these are identified in University of Michigan Law School documents as the groups which receive special attention in the admissions process.
 Section 1 of the Fourteenth Amendment to the Constitution states:
All persons born or naturalized in the United States, and subject to the jurisdiction thereof, are citizens of the United States and of the State wherein they reside. No State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any State deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws.
 42 U.S.C. § 2000d states: "No person in the United States shall, on the ground of race, color, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance."
 These groups are United For Equality and Affirmative Action (UEAA), the Coalition to Defend Affirmative Action By Any Means Necessary (BAMN), and Law Students for Affirmative Action (LSAA).
 For easy reference a copy of this grid, which shows admissions information for 1991, is attached to this opinion as Exhibit A.
 For example, in 1991 there were 499 applicants with an LSAT score between 38 and 41 and 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.
 The identical statement appears at page 81 of the law school bulletin for 1995-1997. See Trial Exhibit 7.
 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 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.
 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.
 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.
 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.
 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).
 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.
 The other members of the committee were Don Herzog, Jeffrey Lehman, Don Regan, Ted Shaw and Dennis Shields. See Exhibit 4, p. 14.
 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.
 Dr. Larntz' expert reports were admitted as Exhibits 137-142. Exhibit 143 consists of the tables and charts of Dr. Larntz' "power-point presentation."
 Dr. Raudenbush's expert reports were admitted as Exhibits 145-150.
 The law school grids also indicate the "yield," that is, the number of admittees who enrolled.
 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.
 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.
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.
 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.
 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 rejected, then there is no basis for calculating the odds ratio.
 Professor Lempert has written that the law school has considered applicants' race since 1966. See Exhibit 166, p. 2.
 This is the reason stated in the policy itself, and it is the reason given by defendants' witnesses and counsel. However, the court notes that as recently as 1996-1997 the law school's bulletin indicated that one reason for the affirmative action policy has been to further "the public interest in increasing the number of lawyers from groups which the faculty identifies as significantly underrepresented in the legal profession." Exhibit 6, p. 81.
 The percentages range from a low of 5.8% in 1974 to a high of 12.5% in 1977, but the mean percentage of underrepresented minority students from 1973 to 1985 was 9.7%.
 The percentages range from a low of 5.4% in 1998 to a high of 19.2% in 1994, but the mean percentage of underrepresented minority students from 1986 to 1999 was 12.6%.
 From the graduation years 1986 to 1999, underrepresented minorities constituted at least 9.8% (1999) and as much as 19.2% (1994) of the class, except in 1998 when the percentage dipped to 5.4%. See Exhibit 97.
 While plaintiffs claim that the law school's policy violates both the Fourteenth Amendment and Title VI of the 1964 Civil Rights Act, 42 U.S.C. § 2000d, the legal analysis under the two claims is the same. See Bakke, 438 U.S. at 287, 98 S. Ct. 2733; Johnson v. Board of Regents, 106 F. Supp. 2d 1362, 1366 (S.D.Ga.2000).
 Justice Powell found U.C. Davis' special admissions program invalid under the Fourteenth Amendment, see Bakke, 438 U.S. at 320, 98 S. Ct. 2733; while the Stevens group avoided the constitutional issue and found that the program violated Title VI, see id. at 421, 98 S. Ct. 2733.
 The difficulty of this task was perhaps best summarized by the Fifth Circuit in United States v. City of Miami, 614 F.2d 1322, 1337 (5th Cir.1980), which stated: "We frankly admit that we are not entirely sure what to make of the various Bakke opinions. In over one hundred and fifty pages of United States Reports, the Justices have told us mainly that they have agreed to disagree."
 See Marks v. United States, 430 U.S. 188, 97 S. Ct. 990, 51 L. Ed. 2d 260 (1977).
 U.C. Davis argued that its special admissions program served four purposes: "(i) reducing the historic deficit of traditionally disfavored minorities in medical schools and in the medical profession; (ii) countering the effects of societal discrimination; (iii) increasing the number of physicians who will practice in communities currently underserved; and (iv) obtaining the educational benefits that flow from an ethnically diverse student body." Bakke, 438 U.S. at 306, 98 S. Ct. 2733.
 The Brennan group stated that it "join[ed] Parts I and V-C of our Brother POWELL's opinion and three of us agree with his conclusion in Part II that this case does not require us to resolve the question whether there is a private right of action under Title VI." 438 U.S. at 328, 98 S. Ct. 2733.
 The court acknowledges that the status of Justice Powell's opinion has been the subject of much debate and disagreement. Some courts have concluded that the diversity rationale is part of Bakke's holding. See, e.g., Smith v. University of Wash. Law Sch., 233 F.3d 1188, 1200-1201 (9th Cir.2000); Gratz v. Bollinger, 122 F. Supp. 2d 811, 819-21 (E.D.Mich.2000). Others have reached the opposite conclusion. See, e.g., Hopwood v. State of Texas, 78 F.3d 932, 944-45 (5th Cir. 1996); Johnson v. Board of Regents of the Univ. Sys. of Georgia, 106 F. Supp. 2d 1362, 1368-69 (S.D.Ga.2000). Other courts have avoided deciding the issue. See, e.g., Eisenberg v. Montgomery County Pub. Sch., 197 F.3d 123, 130 (4th Cir.1999) ("whether diversity is a compelling governmental interest remains unresolved, and in this case, we also choose to leave it unresolved"); Wessmann v. Gittens, 160 F.3d 790, 796 (1st Cir.1998) ("we need not definitively resolve this conundrum today").
 This statement appears in Part III-A of Justice O'Connor's opinion, in which Justices Rehnquist, White and Kennedy joined. Justice Scalia, in a concurring opinion, wrote that "[a]t least where state or local action is at issue, only a social emergency rising to the level of imminent danger to life and limb ... can justify an exception to the principle embodied in the Fourteenth Amendment that `[o]ur Constitution is color-blind, and neither knows nor tolerates classes among citizens.'" 488 U.S. at 521, 109 S. Ct. 706 (citations omitted). Thus, four members of the Court expressed the view that racial classifications are permissible only to remedy past discrimination, and a fifth expressed the view that they are permissible only in cases of social emergency.
 For example, race-based hiring or promotion programs have been upheld where necessary to correct a racial imbalance caused by documented race discrimination by a particular agency or employer. See, e.g., United States v. Paradise, 480 U.S. 149, 107 S. Ct. 1053, 94 L. Ed. 2d 203 (1987); Franks v. Bowman Transp. Co., 424 U.S. 747, 96 S. Ct. 1251, 47 L. Ed. 2d 444 (1976). Race-based student assignments have been upheld where necessary to correct a racial imbalance, or a misallocation of school resources, caused by de jure racial segregation. See, e.g., Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1, 91 S. Ct. 1267, 28 L. Ed. 2d 554 (1971).
 Professor Syverud's report is among those contained in volume 3 of the appendix filed by defendants in support of their renewed motion for summary judgment on grounds of qualified immunity.
 Professor Terrance Sandalow, former dean of the law school and long-time law professor, has written in this connection:
Students learn from one another in different ways. In the course of discussion, whether in the classroom or in dormitory `bull sessions,' participants are likely to be exposed to unfamiliar ideas. My own experience and that of colleagues with whom I have discussed the question, experience that concededly is limited to the classroom setting, is that racial diversity is not responsible for generating ideas unfamiliar to some members of the class. Students do, of course, quite frequently express and develop ideas that others in the class have not previously encountered, but even though the subjects I teach deal extensively with racial issues, I cannot recall an instance in which, for example, ideas were expressed by a black student that have not also been expressed by white students. Black students do, at times, call attention to the racial implications of issues that are not facially concemed with race, but white and Asian-American students are in my experience no less likely to do so.
Terrance Sandalow, "Identity and Equality, Minority Preferences Reconsidered," 97 Mich. L.Rev. 1874, 1906-1907 (1999).
 In this connection it is worth noting that the law school, like U.C. Davis, "does not purport to have made, and is in no position to make such findings. Its broad mission is education, not the formulation of any legislative policy or the adjudication of particular claims of illegality." Bakke, 438 U.S. at 309, 98 S. Ct. 2733.
 See the University of Michigan Law School Bulletin, 1996-1997 (Exhibit 6), p. 81; the University of Michigan Law School Bulletin, 1995-1997 (Exhibit 7), p. 81. No such statement appears in the University of Michigan Law School Bulletin, 1997-1999 (Exhibit 8), p. 85.
 A close fit is lacking for another reason namely, that there is no basis for believing that a particular member of any underrepresented minority group is "particularly likely to have experiences and perspectives of special importance to our mission." Exhibit 4, 1992 Admissions Policy, p. 12. One cannot assume, for example, that a particular African American applicant has been the victim of race discrimination or that a particular Mexican American applicant has grown up in an economically depressed neighborhood.
 Winston Churchill, in a radio broadcast in October 1939. J. Bartlett, Familiar Quotations 620 (1968).
 Of course, whether plaintiffs actually should be awarded any damages is a question to be resolved in the next phase of the trial. Now that defendants' liability has been established, defendants may still defeat plaintiffs' claim for compensatory damages by showing that plaintiffs would not have been admitted under a race-blind admissions policy. See Hopwood, 78 F.3d at 955-57 (discussing burden of proof under Mt. Healthy).
 In this connection Professor Orfield meant racial separation, not de jure segregation.
 The percentage of African American students increased to 1.8% (nine students) in 1998 but fell slightly to 1.7% (nine students) in 1999.
 The percentage of Native American students increased to 1.0% (five students) in 1998 but fell to 0.4% (two students) in 1999.
 The percentage of Hispanic students increased to 7.6% (37 students) in 1998 and to 8.1% (42 students) in 1999.
 By this Professor Orfield was referring to the University of California at Berkeley, the University of California at Davis, and the University of California at Los Angeles.
 Professor Orfield has written:
The reversal of affirmative admissions in higher education can drastically reduce black, Latino, and American Indian enrollment on highly selective campuses. The increased use of tests and grades as entrance standards will tend to exacerbate the existing inequities in U.S. society. If affirmative action is outlawed nationally, as it has been in Texas, the impact on access to leading public and private universities would be enormous. Many of our most able students would find themselves on campuses overwhelmingly dominated by white and Asian students. The severe isolation characteristic of our more affluent suburbs would become the rule in the institutions that train the leaders of our society and our professions. This threatens critical educational functions of universities and their ability to fully serve their communities.
Exhibit 167D, G. Orfield and E. Miller, Chilling Admissions 14 (1998).
 Additional testimony about this history of racial discrimination in the United States was provided by Dr. Eric Foner, a professor of American history at Columbia University.
 The University of California has campuses at Los Angeles, Berkeley, San Diego, Irvine, Davis, Santa Barbara, Santa Cruz, and Riverside.
 Exhibit 214 shows applications, admissions, admissions rates, and enrollment figures for the years 1995 to 2000 at all of the University of California campuses by the race of the applicants. It appears that the most significant declines in admission rates occurred among African Americans, Native Americans, Chicanos and Latinos at Berkeley, Los Angeles and San Diego. The court notes that admission rates of whites and Asian Americans also dropped at most of the campuses over the same time period, albeit less dramatically. At San Diego, for example, the admission rate for Asian Americans fell from 62.8% in 1995 to 46.1% in 2000. At Santa Barbara, the admission rate for whites fell from 85.3% in 1995 to 48.9% in 2000.
 Professor Lempert found that the law school's "admissions index" (combination of LSAT score and UGPA) was an equally poor predictor of career success for both minority and white alumni. See Exhibit 166, pp. 94-99 (Tables 29A, 29B, 29C, 30A, 30B, 30C).
 While not introduced at trial, defendants did submit a report from Dr. Thomas Sugrue, a professor of history and sociology at the University of Pennsylvania, in support of their motion for summary judgment on grounds of qualified immunity. Professor Sugrue cites census data showing the significant disparities along racial lines in educational attainment, unemployment rates, income levels, poverty rates, and household net worth. See Sugrue Report, pp. 50, 53-57.
 The median undergraduate GPA for accepted applicants, by race, is shown in Exhibit 137, p. 21; Exhibit 139, p. 19; and Exhibit 141, p. 16. The smallest difference between the Caucasian UGPA and an underrepresented minority group UGPA was .05 (Caucasian vs. Puerto Rican in 1996, and Caucasian vs. Native American in 1998). The largest difference was .41 (Caucasian vs. African American in 2000). Averaging the figures over the six years, one sees that the gap between the Caucasian UGPA was .21 for Native Americans, .32 for African Americans, .14 for Mexican Americans, and .22 for Puerto Ricans.
 The median LSAT scores for accepted applicants, by race, is shown in Exhibit 137, p.22; Exhibit 139, p. 20; and Exhibit 141, p. 17. The smallest difference between the Caucasian LSAT score and an underrepresented minority group LSAT score was three points (Caucasian vs. Native American in 2000). The largest different was 12 points (Caucasian vs. African American in 2000). Averaging the figures over the six years, one sees that the point gap between the Caucasian LSAT score was 6.8 for Native Americans, 9.6 for African Americans, 7 for Mexican Americans, and 7.6 for Puerto Ricans.
 All 47 of these students were African American or Hispanic. No Native Americans participated in the focus groups. See Exhibit 157, Figure 2.
 At trial Professor Allen indicated that focus groups 1, 2, 5, 8, 9 and 11 were conducted at The University of Michigan; focus groups 4, 6 and 10 were conducted at Berkeley; focus groups 3 and 7 were conducted at Harvard; and focus group 12 was conducted at Michigan State University.
 The magnitude of this flaw is suggested by Exhibit 18, which identifies all of the undergraduate schools which graduated students who entered the law school's 1995 class. Of the 337 students in that class, only 85(25%) came from the four "feeder" institutions studied by Professor Allen. The remaining 252 students (75%) came from 106 other colleges and universities which were not included in the study.
 This statement is from Part I of Justice Powell's opinion, which was joined by Justices Burger and Rehnquist. Justice O'Connor filed a concurring opinion in which she specifically joined Part I of Justice Powell's opinion, see 476 U.S. at 294, 106 S. Ct. 1842 and "agree[d] with the plurality that a governmental agency's interest in remedying `societal' discrimination, that is, discrimination not traceable to its own actions, cannot be deemed sufficiently compelling to pass constitutional muster under strict scrutiny." 476 U.S. at 288, 106 S. Ct. 1842. Justice White also concurred in the judgment and wrote that "[n]one of the interests asserted by the Board, singly or together, justify this racially discriminatory layoff policy and save it from the strictures of the Equal Protection Clause." 476 U.S. at 295, 106 S. Ct. 1842. Thus, five members of the court indicated that the Fourteenth Amendment does not permit government to make racial classifications in order to remedy societal discrimination.
 This statement is from Part III-A of Justice Powell's opinion, in which Justices Burger, Rehnquist and O'Connor joined. Justice White, in a concurring opinion, tacitly agreed with the statement by indicating that "[n]one of the interests asserted by the Board, singly or together, justify this racially discriminatory layoff policy and save it from the structures of the Equal Protection Clause." 476 U.S. at 295, 106 S. Ct. 1842.
 This is not a case where a challenge to a desegregation order is opposed on the grounds that the order must be kept in place lest the school district revert to a racially segregated state. See, e.g., Belk v. Charlotte-Mecklenburg Bd. of Educ., 233 F.3d 232 (4th Cir.2000) (opinion vacated). Nor, with all due respect to the intervenors' counsel, can the present admissions policy be defended on the grounds that its elimination would result in the creation of an all-white campus. In this connection the intervenors have raised the specter of Sweatt v. Painter, 339 U.S. 629, 70 S. Ct. 848, 94 L. Ed. 1114 (1950), in which the Supreme Court struck down the de jure exclusion of blacks from the University of Texas Law School and the creation of a separate law school for black students. There is simply no comparison between the de jure segregation at issue in Sweatt and a race-blind admissions policy which may admit fewer underrepresented minority students than are admitted by a race-conscious policy.
 The court did not hear any testimony as to whether the University of Michigan Law School has considered reducing its reliance on the LSAT or whether it has considered challenging the ABA's requirement that applicants take "an acceptable test."
 The court notes again that no party in this case has alleged, or offered any evidence to suggest, that the law school or the University of Michigan has committed any acts of discrimination against any minority group which might warrant a race-based remedy.