Abigail Fisher, et al v. State of Texas, et al, No. 09-50822 (5th Cir. 2011)
Annotate this Case
The court issued a subsequent related opinion or order on June 17, 2011.
The court issued a subsequent related opinion or order on July 15, 2014.
The court issued a subsequent related opinion or order on November 12, 2014.
REVISED FEBRUARY 1, 2011 United States Court of Appeals Fifth Circuit IN THE UNITED STATES COURT OF APPEALS FILED FOR THE FIFTH CIRCUIT January 18, 2011 No. 09-50822 Lyle W. Cayce Clerk ABIGAIL NOEL FISHER; RACHEL MULTER MICHALEWICZ, Plaintiffs Appellants v. UNIVERSITY OF TEXAS AT AUSTIN; DAVID B. PRYOR, Executive Vice Chancellor for Academic Affairs in His Official Capacity; WILLIAM POWERS, JR., President of the University of Texas at Austin in His Official Capacity; BOARD OF REGENTS OF THE UNIVERSITY OF TEXAS SYSTEM; R. STEVEN HICKS, as Member of the Board of Regents in His Official Capacity; WILLIAM EUGENE POWELL, as Member of the Board of Regents in His Official Capacity; JAMES R. HUFFINES, as Member of the Board of Regents in His Official Capacity; JANIECE LONGORIA, as Member of the Board of Regents in Her Official Capacity; COLLEEN MCHUGH, as Member of the Board of Regents in Her Official Capacity; ROBERT L. STILLWELL, as Member of the Board of Regents in His Official Capacity; JAMES D. DANNENBAUM, as Member of the Board of Regents in His Official Capacity; PAUL FOSTER, as Member of the Board of Regents in His Official Capacity; PRINTICE L. GARY, as Member of the Board of Regents in His Official Capacity; KEDRA ISHOP, Vice Provost and Director of Undergraduate Admissions in Her Official Capacity; FRANCISCO G. CIGARROA, M.D., Interim Chancellor of the University of Texas System in His Official Capacity, Defendants Appellees Appeal from the United States District Court for the Western District of Texas No. 09-50822 Before KING, HIGGINBOTHAM, and GARZA, Circuit Judges. PATRICK E. HIGGINBOTHAM, Circuit Judge: We consider a challenge to the use of race in undergraduate admissions at the University of Texas at Austin. While the University has confined its explicit use of race to the elements of a program approved by the Supreme Court in Grutter v. Bollinger,1 UT s program acts upon a university applicant pool shaped by a legislatively-mandated parallel diversity initiative that guarantees admission to Texas students in the top ten percent of their high school class. The ever-increasing number of minorities gaining admission under this Top Ten Percent Law casts a shadow on the horizon to the otherwise-plain legality of the Grutter-like admissions program, the Law s own legal footing aside. While the Law s ultimate fate is not the fare of this suit, the challenge to the Grutter plan here rests upon the intimate ties and ultimate confluence of the two initiatives. Today we affirm the constitutionality of the University s program as it existed when Appellants applied and were denied admission. Abigail Fisher and Rachel Michalewicz, both Texas residents, were denied undergraduate admission to the University of Texas at Austin for the class entering in Fall 2008. They filed this suit alleging that UT s admissions policies discriminated against them on the basis of race in violation of their right to equal protection under the Fourteenth Amendment and federal civil rights statutes.2 They sought damages as well as injunctive and declaratory relief. Proceeding with separate phases of liability and remedy, the district court, in a thoughtful opinion, found no liability and granted summary judgment to the University. 1 539 U.S. 306 (2003). 2 Fisher v. Univ. of Tex. at Austin, 645 F. Supp. 2d 587, 590 (W.D. Tex. 2009) (citing U.S. CONST. amend. XIV, § 1, and 42 U.S.C. §§ 1981, 1983, and 2000d et seq.). 2 No. 09-50822 The procedural posture of this case defines the scope of our review. There are no class claims and both students deny intention to reapply to UT.3 It follows that Fisher and Michalewicz lack standing to seek injunctive or forwardlooking declaratory relief.4 This principle is rote. To obtain forward-looking equitable remedies, a plaintiff must show she faces imminent threat of future injury.5 Without that threat, these two applicants only have standing to challenge their rejection and to seek money damages for their injury.6 Our focus will be upon the process employed by UT to admit freshmen when Fisher and Michalewicz applied for the class entering Fall 2008, looking to earlier and later years only as they illuminate the rejection of these two applicants.7 Our task is burdened by the reality that we are examining a dynamic program administered by a large university subject to government oversight. Indeed, the first of UT s periodic five-year reviews was to begin in the fall of 2009, a review that must engage an array of variables, including an everpresent question of whether to adjust the percentage of students admitted under the two diversity initiatives. 3 Like all Texas residents, Appellants could attend UT Austin as transfer students if they first enrolled in a participating UT system school and met the standards required by the Coordinated Admissions Program, discussed in greater detail below. Instead, Appellants permanently enrolled at other institutions. 4 See Defunis v. Odegaard, 416 U.S. 312, 319 (1974) (per curiam) (dismissing for lack of standing a suit that challenged a law school admissions policy because the plaintiff would never again be required to run the gantlet of the Law School s admissions process ). 5 Adarand Constructors, Inc. v. PeÃ±a, 515 U.S. 200, 201 11 (1995); City of L.A. v. Lyons, 461 U.S. 95, 105 10 (1983). 6 See Lyons, 461 U.S. at 105 07. 7 Cf. Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 711 n.1 (relying on data from before the district court record closed, even after newer data had become available). 3 No. 09-50822 I. GRUTTER V. BOLLINGER We begin with Grutter v. Bollinger because UT s race-conscious admissions procedures were modeled after the program it approved. In rejecting constitutional challenges to the University of Michigan Law School s admissions program, Grutter held that the Equal Protection Clause did not prohibit a university s narrowly tailored use of race in admissions decisions to further a compelling interest in obtaining the educational benefits that flow from a diverse student body. 8 Mapping on Grutter, UT evaluates each application using a holistic, multi-factor approach, in which race is but one of many considerations. In granting summary judgment to UT, the district court found that it would be difficult for UT to construct an admissions policy that more closely resembles the policy approved by the Supreme Court in Grutter, and as long as Grutter remains good law, UT s current admissions program remains constitutional. 9 Laying aside the Top Ten Percent Law, that observation is indisputably sound.10 A Grutter embraced the diversity interest articulated twenty-five years earlier by Justice Powell, who wrote separately in Regents of the University of California v. Bakke.11 This vision of diversity encompassed a broad array of qualifications and characteristics where race was a single but important 8 Grutter, 539 U.S. at 343. 9 Fisher, 645 F. Supp. 2d at 612 13; see also id. at 613 ( If the Plaintiffs are right, Grutter is wrong. (internal quotation marks omitted)). 10 In practice, the admissions systems of Michigan Law School and UT differ because UT s automatic admission of the top ten percent of Texas high school seniors largely dominates [its] admissions process. Fisher, 645 F. Supp. 2d at 595. We discuss the impact of the Top Ten Percent Law in greater detail below. 11 438 U.S. 265, 269 (1978) (opinion of Powell, J.). 4 No. 09-50822 element.12 The Michigan Law School designed its admissions program to achieve this broad diversity, selecting students with varied backgrounds and experiences including varied racial backgrounds who would respect and learn from one another.13 The Court explained: [The Law School s] policy makes clear there are many possible bases for diversity admissions, and provides examples of admittees who have lived or traveled widely abroad, are fluent in several languages, have overcome personal adversity and family hardship, have exceptional records of extensive community service, and have had successful careers in other fields.14 The Law School s policy also reaffirmed its longstanding commitment to one particular type of diversity, that is, 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 [the] student body in meaningful numbers. 15 In an effort to ensure representation of minorities, the Law School sought to enroll a critical mass of minority students, which would result in increased minority engagement in the classroom and enhanced minority contributions to the character of the School. The Grutter Court endorsed this goal, holding that diversity, including seeking a critical mass of minority students, is a compelling state interest that can justify the use of race in university admissions. 16 12 See Grutter, 539 U.S. at 325 (citing Bakke, 438 U.S. at 315 (opinion of Powell, J.)). 13 Id. at 314. 14 Id. at 338 (brackets and internal quotation marks omitted). 15 Id. at 316 (internal quotation marks omitted). 16 Id. at 325; see id. at 329 30. 5 No. 09-50822 That the concept of critical mass bears a simple but deceptive label is evidenced by the division of the Justices over its meaning. In his dissent, Chief Justice Rehnquist saw critical mass as only the minimum level necessary [t]o ensure that the minority students do not feel isolated or like spokespersons for their race; to provide adequate opportunities for the type of interaction upon which the educational benefits of diversity depend; and to challenge all students to think critically and reexamine stereotypes. 17 On this view, critical mass is defined only as a proportion of the student body, and the percentage that suffices for one minority group should also suffice for another group. In contrast, Justice O Connor, writing for the Court, explained that critical mass must be defined by reference to the educational benefits that diversity is designed to produce. 18 Her opinion recognizes that universities do more than simply impart knowledge to their students. Synthesizing, we find at least three distinct educational objectives served by the diversity she envisioned: 1. Increased Perspectives. Justice O Connor observed that including diverse perspectives improves the quality of the educational process because classroom discussion is livelier, more spirited, and simply more enlightening and interesting when the students have the greatest possible variety of backgrounds. 19 In this respect, Grutter echoes Justice Powell s recognition in Bakke that it is essential to the quality of higher education that a university be able to pursue [t]he atmosphere of speculation, excitement and creation that is promoted by a diverse student body. 20 Indeed, diversity often brings not just excitement, but valuable knowledge as well. [A] student with a particular background whether it be ethnic, geographic, culturally 17 Id. at 380 (Rehnquist, C.J., dissenting). 18 Id. at 329 30 (opinion of the Court). 19 Id. at 330 (internal quotation marks omitted). 20 438 U.S. at 312 (opinion of Powell, J.) (internal quotation marks omitted). 6 No. 09-50822 advantaged or disadvantaged may bring to a [university] experiences, outlooks, and ideas that enrich the training of its student body and better equip its graduates to render with understanding their vital service to humanity. 21 2. Professionalism. The majority pointed to numerous studies showing that student body diversity . . . better prepares [students] as professionals. 22 The Court has repeatedly acknowledged the overriding importance of preparing students for work and citizenship, 23 and today s students must be prepared to work within an increasingly diverse workforce. 24 Indeed, major American businesses have made clear that the skills needed in today s increasingly global marketplace can only be developed through exposure to widely diverse people, cultures, ideas, and viewpoints. 25 A diverse student body serves this end by promot[ing] cross-racial understanding, help[ing] to break down racial stereotypes, and enabl[ing] students to better understand persons of different races. 26 3. Civic Engagement. The Court recognized that [e]ffective participation by members of all racial and ethnic groups in the civic life of our Nation is essential if the dream of one Nation, indivisible, is to be realized. 27 A diverse student body is crucial for fostering this ideal of civic engagement, because [i]n order to cultivate a set of leaders with legitimacy in the eyes of the citizenry, it is necessary that the path to leadership be visibly open to talented and qualified 21 Id. at 314. 22 Grutter, 539 U.S. at 330 (internal quotation marks omitted). 23 Id. (internal quotation marks omitted). 24 Id. (internal quotation marks omitted). 25 Id. 26 Id. (internal quotation marks and brackets omitted). 27 Id. at 332. 7 No. 09-50822 individuals of every race and ethnicity. 28 Maintaining a visibly open path to leadership demands that [a]ccess to [higher] education . . . be inclusive of talented and qualified individuals of every race and ethnicity, so that all members of our heterogeneous society may participate in the educational institutions that provide the training and education necessary to succeed in America. 29 Each member of society must have confidence in the openness and integrity of the educational institutions that provide this training. 30 Further, efforts to educate and to encourage future leaders from previously underrepresented backgrounds will serve not only to inspire, but to actively engage with many woefully underserved communities, helping to draw them back into our national fabric. B Recognizing the pursuit of diversity, including racial diversity, to be a compelling interest in higher education, Grutter endorsed the right of public universities to increase enrollment of underrepresented minorities. Grutter also cautioned that, while it accepted diversity as a compelling interest, any sorting 28 Id. 29 Id. at 332 33. The Court further explained: [E]ducation [is] pivotal to sustaining our political and cultural heritage with a fundamental role in maintaining the fabric of society. . . . [T]he diffusion of knowledge and opportunity through public institutions of higher education must be accessible to all individuals regardless of race or ethnicity. The United States, as amicus curiae, affirms that [e]nsuring that public institutions are open and available to all segments of American society, including people of all races and ethnicities, represents a paramount government objective. And, [n]owhere is the importance of such openness more acute than in the context of higher education. Id. at 331 32 (final two alterations in original; citations and some internal quotation marks omitted). 30 Id. at 332. 8 No. 09-50822 of persons on the basis of race must be by measures narrowly tailored to the interest at stake. As we read the Court, a university admissions program is narrowly tailored only if it allows for individualized consideration of applicants of all races.31 Such consideration does not define an applicant by race but instead ensures that she is valued for all her unique attributes. Rather than applying fixed stereotypes of ways that race affects students lives, an admissions policy must be flexible enough to consider all pertinent elements of diversity in light of the particular qualifications of each applicant. 32 As the Supreme Court later summarized, The entire gist of the analysis in Grutter was that the admissions program at issue there focused on each applicant as an individual, and not simply as a member of a particular racial group. 33 Thus, a university admissions policy is more likely to be narrowly tailored if it contemplates that a broad range of qualities and experiences beyond race will be important contributions to diversity and as such are appropriately considered in admissions decisions.34 Because a race-conscious admissions program is constitutional only if holistic, flexible, and individualized, a university may not establish a quota for minority applicants, nor may it evaluate minority applications on separate admissions tracks. 35 The racial-set-aside program rejected by Justice Powell in Bakke ran afoul of these related prohibitions because it reserved 16 out of 100 31 Id. at 337. 32 Id. (quoting Bakke, 438 U.S. at 317 (opinion of Powell, J.)). 33 Parents Involved, 551 U.S. at 722; see also Grutter, 539 U.S. at 337 ( The importance of this individualized consideration in the context of a race-conscious admissions program is paramount. ). 34 Grutter, 539 U.S. at 338. 35 Id. at 334 (citing Bakke, 438 U.S. at 315 16 (opinion of Powell, J.)). 9 No. 09-50822 seats for members of certain minority groups.36 A university also may not award a fixed number of bonus points to minority applicants.37 That was the lesson of Grutter s companion case, Gratz v. Bollinger, in which the Court struck down the University of Michigan s undergraduate admissions program because it automatically awarded a fixed number of admissions points to all underrepresented minority applicants, resulting in a group-based admissions boost.38 Both Bakke and Gratz firmly rejected group treatment, insisting that the focus be upon individuals and that an applicant s achievements be judged in the context of one s personal circumstances, of which race is only a part. So deployed, a white applicant raised by a single parent who did not attend high school and struggled paycheck to paycheck and a minority child of a successful cardiovascular surgeon may both claim adversity, but the personal hurdles each has cleared will not be seen to be of the same height. C Finally, Grutter requires that any race-conscious measures must have a logical end point and be limited in time. 39 This durational requirement can be satisfied by sunset provisions or by periodic reviews to reconsider whether there are feasible race-neutral alternatives that would achieve diversity interests about as well. 40 In this respect, Grutter is best seen not as an unqualified endorsement of racial preferences, but as a transient response to 36 Id. at 322; see Bakke, 438 U.S. at 289 (opinion of Powell, J.). 37 Gratz v. Bollinger, 539 U.S. 244, 271 72 (2003). 38 Id. 39 Grutter, 539 U.S. at 342. 40 Id. at 339 (quoting Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 280 n.6 (1986)). 10 No. 09-50822 anemic academic diversity. As Justice O Connor observed, We expect that 25 years from now, the use of racial preferences will no longer be necessary to further the interest approved today. 41 II. HISTORY OF THE UNIVERSITY S ADMISSIONS POLICIES Justice O Connor s vision may prove to be more aspirational than predictive. Regardless, universities will construct admissions programs wedded to their missions, which include bringing both meritorious and diverse students to campus. Each year, UT receives applications from approximately four times more students than it can enroll.42 Over the past two decades, UT has repeatedly revised its admissions procedures to reflect its calculus of educational values while navigating judicial decisions and legislative mandates. A Until 1996, UT selected students using two metrics. The first measure, still employed today, is the Academic Index ( AI ), a computation based on the student s high school class rank, standardized test scores, and the extent to which the applicant exceeded UT s required high school curriculum.43 Perceiving that AI alone would produce a class with unacceptably low diversity levels, UT considered a second element for admissions race. These measures combined resulted in UT admitting more than 90% of applicants who were ranked in the top ten percent of their high school class.44 41 Id. at 343. 42 Fisher, 645 F. Supp. 2d at 590. 43 Id. at 596. 44 Marta Tienda et al., Closing the Gap?: Admissions & Enrollment at the Texas Public Flagships Before and After Affirmative Action 52 tbl.5 (Tex. Higher Educ. Opportunity Project Working Paper), available 11 No. 09-50822 There were then no clear legal limits on a university s use of race in admissions. The Supreme Court decided Bakke in 1978 but its guidance came in a fractured decision, leaving a quarter century of uncertainty.45 The record does not detail precisely how race factored in admissions decisions during this time, but it is undisputed that race was considered directly and was often a controlling factor in admission.46 Under this race-conscious admissions policy, the freshman class entering in Fall 1993 included 5,329 students, of whom 238 at http://theop.princeton.edu/workingpapers.html. Unlike the current Top Ten Percent Law, UT s earlier policies did not mandate the admission of all top ten percent students. Thus, even though a top ranking at a predominantly minority high school would contribute to a higher AI score, the AI alone could not effectively serve as a proxy for race because, on average, minorities received lower standardized test scores. 45 Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265 (1978). Four Justices would have held that universities have broad authority to consider race in admissions in order to remedy disadvantage cast on minorities by past racial prejudice. Id. at 325 (joint opinion of Brennan, White, Marshall, and Blackmun, JJ.). Four other Justices would have held that Title VI of the Civil Rights Act of 1964 bars federally funded universities from making any admissions decisions on the basis of race. Id. at 417 18 (opinion of Stevens, J., joined by Burger, C.J., and Stewart and Rehnquist, JJ.). Justice Powell cast the decisive vote in a separate opinion not joined in full by any other Justice that invalidated the racial set-aside in the admissions program then before the Court, but reasoned that it would be constitutional for a university to consider race as one facet of diversity in a flexible review that treated each applicant as an individual. Id. at 316 19 (opinion of Powell, J.). Because none of these positions carried the support of a majority of the Court, it was not completely clear which (if any) of these rationales was controlling. See Grutter, 529 U.S. at 322 25 (2003) (recounting this history and the subsequent confusion among lower courts). 46 Records do reflect that at UT s law school during this time, minority and nonminority applicants were reviewed by separate admissions committees and were subject to different grade and test-score cutoffs. See Hopwood v. Texas, 78 F.3d 932, 935 38 (5th Cir. 1996). 12 No. 09-50822 were African-American (4.5% of the overall class) and 832 were Hispanic (15.6%).47 B Race-conscious admissions ended in 1996 with Hopwood v. Texas, when a panel of this court struck down the use of race-based criteria in admissions decisions at UT s law school.48 A majority of that panel held that diversity in education was not a compelling government interest,49 a conclusion the Texas Attorney General interpreted as prohibiting the use of race as a factor in admissions by any undergraduate or graduate program at Texas state universities.50 Beginning with the 1997 admissions cycle, UT deployed a Personal Achievement Index ( PAI ) to be used with the Academic Index. In contrast to the mechanical formulas used to calculate the AI, the PAI was meant to identify and reward students whose merit as applicants was not adequately reflected by 47 Univ. of Tex. at Austin, 1998 1999 Statistical Handbook. Minority enrollment was fairly consistent from 1989 until 1993, with some slight decreases in 1994 and 1995. UT publishes its Statistical Handbook annually, and these handbooks are cited throughout the district court record. See Univ. of Tex. at Austin Office of Admissions, Diversity Levels of Undergraduate Classes at The University of Texas at Austin 1996 2002 (2003) (Dist. Ct. Dkt. No. 96, Tab 8, Ex. B), at 5, 6; Univ. of Tex. at Austin, Proposal to Consider Race and Ethnicity in Admissions (2004) (Dist. Ct. Dkt. No. 96, Tab 11, Ex. A), at 30; Univ. of Tex. at Austin Office of Admissions, 2008 Top Ten Percent Report (Dist. Ct. Dkt. No. 94, Ex. 9), at 4 [hereinafter 2008 Top Ten Percent Report]. Handbooks dating back to 1998 are available online at http://www.utexas.edu/academic/ima/stat_handbook/. 48 78 F.3d 932 (1996). 49 Id. at 944 48. 50 See Tex. Att y Gen. Letter Op. No. 97-001 (1997). 13 No. 09-50822 their class rank and test scores. 51 Although facially race-neutral, the PAI was in part designed to increase minority enrollment; many of the PAI factors disproportionately affected minority applicants.52 UT also implemented other facially race-neutral policies that, together with the AI and PAI, remain in use today. It created targeted scholarship programs to increase its yield among minority students, expanded the quality and quantity of its outreach efforts to high schools in underrepresented areas of the state, and focused additional attention and resources on recruitment in low-performing schools.53 Despite these efforts, minority presence at UT decreased immediately. Although the 1996 admissions decisions were not affected by Hopwood, the publicity from the case impacted the number of admitted minorities who chose to enroll. In 1997, fewer minorities applied to UT than in years past. The number of African-American and Hispanic applicants dropped by nearly a quarter, while the total number of University applicants decreased by only 13%.54 This decrease in minority applicants had a corresponding effect on enrollment. Compared to 1995, African-American enrollment for 1997 dropped almost 40% (from 309 to 190 entering freshmen) while Hispanic enrollment decreased by 5% (from 935 to 892 entering freshmen). In contrast, Caucasian enrollment increased by 14%, and Asian-American enrollment increased by 20%.55 51 Fisher, 645 F. Supp. 2d at 591. 52 Id. at 591 92. 53 Id. at 592. 54 Diversity Levels of Undergraduate Classes at The University of Texas at Austin 1996 2002 (2003) (Dist. Ct. Dkt. No. 96, Tab 8, Ex. B), at 6. 55 1998 1999 Statistical Handbook. 14 No. 09-50822 C In 1997, the Texas legislature responded to the Hopwood decision by enacting the Top Ten Percent Law, still in effect.56 The law altered UT s preexisting policy and mandated that Texas high school seniors in the top ten percent of their class be automatically admitted to any Texas state university. In its first year, the Top Ten Percent Law succeeded in increasing minority percentages at UT. African-American enrollment rose from 2.7% to 3.0% and Hispanic enrollment rose from 12.6% to 13.2%. However, the absolute number of minorities remained stable as a result of a smaller freshman class. Over time, both the number and percentage of enrolled Hispanics and African-Americans increased. The entering freshman class of 2004, the last admitted without the Grutter-like plan, was 4.5% African-American (309 students), 16.9% Hispanic (1,149 students), and 17.9% Asian-American (1,218 students) in a class of 6,796 students.57 The Top Ten Percent Law did not by its terms admit students on the basis of race, but underrepresented minorities were its announced target and their admission a large, if not primary, purpose. In 2004, among freshmen who were Texas residents, 77% of the enrolled African-American students and 78% of the Hispanic students had been admitted under the Top Ten Percent Law, compared to 62% of Caucasian students.58 These numbers highlight the contribution of the Top Ten Percent Law to increasing minority enrollment, but they also reflect a 56 TEX. EDUC. CODE § 51.803 (1997). The Top Ten Percent Law was amended, during the course of this litigation, to cap the number of students guaranteed admission at UT Austin to 75% of the seats available to Texas residents. Id. § 51.803(a-1) (2010). The cap is effective starting with admissions to the Fall 2011 entering class and is currently scheduled to end with admissions to the Fall 2015 entering class. 57 2008 Top Ten Percent Report at 6 tbl.1. 58 Id. at 8; see also Fisher, 645 F. Supp. 2d at 593 (reporting statistics for total admitted applicants, both Texas and non-Texas residents). 15 No. 09-50822 trade-off implicit in the Law: the increase rested heavily on the pass from standardized testing offered by the Top Ten Percent Law. After implementation of the Law, the likelihood of acceptance for African-American and Hispanic students in the second decile of their high school class, who were without the benefits of the pass from standardized testing, declined. Meanwhile, the acceptance probability of similarly situated Caucasian students increased.59 D Hopwood s prohibitions ended after the 2004 admissions cycle with the Supreme Court s 2003 decision in Grutter.60 In August 2003, the University of Texas Board of Regents authorized the institutions within the University of Texas system to examine whether to consider an applicant s race and ethnicity in admissions in accordance with the standards enunciated in Grutter.61 As part of its examination, UT commissioned two studies to explore whether the University was enrolling a critical mass of underrepresented minorities. The first study examined minority representation in undergraduate classes, focusing on classes of participatory size, which it defined as between 5 and 24 students. UT analyzed these classes, which included most of the undergraduate courses, because they offered the best opportunity for robust classroom discussion, rich soil for diverse interactions. According to the study, 90% of these smaller classes in Fall 2002 had either one or zero African-American students, 46% had one or zero Asian-American students, and 59 Tienda et al., supra note 44, at 52 tbl.5. 60 539 U.S. 306 (2003). 61 Minutes of the Board of Regents of the University of Texas at Austin, Meeting No. 969, Aug. 6 7, 2003 (Dist. Ct. Dkt. No. 94, Ex. 19, Tab A), at 4. 16 No. 09-50822 43% had one or zero Hispanic students.62 A later retabulation, which excluded the very smallest of these classes and considered only classes with 10 to 24 students, found that 89% of those classes had either one or zero African-American students, 41% had one or zero Asian-American students, and 37% had either one or zero Hispanic students.63 In its second study, UT surveyed undergraduates on their impressions of diversity on campus and in the classroom. Minority students reported feeling isolated, and a majority of all students felt there was insufficient minority representation in classrooms for the full benefits of diversity to occur. 64 The University incorporated the findings of these two studies in its June 2004 Proposal to Consider Race and Ethnicity in Admissions.65 The 2004 Proposal concluded that diverse student enrollment break[s] down stereotypes, promotes cross-racial understanding, and prepares students for an increasingly diverse workplace and society. 66 With respect to the undergraduate program in particular, the 2004 Proposal explained that [a] comprehensive college education requires a robust exchange of ideas, exposure to differing cultures, preparation for the challenges of an increasingly diverse workforce, and acquisition of competencies required of future leaders. 67 With 62 Fisher, 645 F. Supp. 2d at 593. Classes with only one student of a given minority were thought to be just as troubling as classes with zero students of that minority because a single minority student is apt to feel isolated or like a spokesperson for his or her race. Id. at 602 603; see also Grutter, 539 U.S. at 319. 63 Lavergne Aff. (Dist. Ct. Dkt. No. 102, Tab B) ¶¶ 4 5. 64 Walker Aff. (Dist. Ct. Dkt. No. 96, Tab 11) ¶ 12. 65 Dist. Ct. Dkt. No. 96, Tab 11, Ex. A [hereinafter 2004 Proposal]. 66 Id. at 1 (internal quotation marks omitted); see also Fisher, 645 F. Supp. 2d at 603. 67 2004 Proposal at 23 (quoted in Fisher, 645 F. Supp. 2d at 602). 17 No. 09-50822 one eye on Grutter, it observed that these objectives are especially important at UT because its mission and . . . flagship role is to prepare its students to be the leaders of the State of Texas a role which, given the state s increasingly diverse profile, will require them to be able to lead a multicultural workforce and to communicate policy to a diverse electorate. 68 Citing the classroom diversity study, the 2004 Proposal explained that UT had not yet achieved the critical mass of underrepresented minority students needed to obtain the full educational benefits of diversity. Accordingly, the 2004 Proposal recommended adding the consideration of race as one additional factor within a larger admissions scoring index. This recommendation was presented as an acknowledgment that the significant differences between the racial and ethnic makeup of the University s undergraduate population and the state s population prevent the University from fully achieving its mission. 69 After more than a year of study following the Grutter decision, UT adopted a policy to include race as one of many factors considered in admissions. UT has no set date by which it will end the use of race in undergraduate admissions. Rather, it formally reviews the need for race-conscious measures every five years and considers whether adequate race-neutral alternatives exist. In addition, the district court found that the University informally reviews its admissions procedures each year.70 The current policy has produced noticeable results. One magazine dedicated to diversity in higher education ranked UT sixth in the nation in producing undergraduate degrees for minority groups. 71 In an entering class 68 Id. at 24 (quoted in Fisher, 645 F. Supp. 2d at 602). 69 Id. (quoted in Fisher, 645 F. Supp. 2d at 602). 70 Fisher, 645 F. Supp. 2d at 594. 71 Id. This particular ranking is somewhat limited in its significance, however, as the results are based on raw tabulations of the number of degrees 18 No. 09-50822 that was roughly the same size in 1998 as it was in 2008, the enrollment of African-American students doubled from 165 students to 335 students. Hispanic enrollment increased approximately 1.5 times, from 762 students to 1,228 students. Asian-American enrollment also increased nearly 10%, from 1,034 students to 1,126 students.72 By contrast, in 2004, the last year the Top Ten Percent Law operated without the Grutter plan, fall enrollment included only 275 African-Americans and 1,024 Hispanics. Because of the myriad programs instituted, it can be difficult to attribute increases in minority enrollment to any one initiative. In addition, demographics have shifted in Texas, so increases in minority enrollment likely in part reflect the increased presence of minorities statewide. III. THE CHALLENGED POLICY UT s consideration of race is one part of the complex admissions process operating when Appellants were rejected. Given Appellants challenge, we must examine the whole of the process. A UT is a public institution of higher education, authorized by the Texas Constitution and supported by state and federal funding. Accordingly, it begins its admissions process by dividing applicants into three pools: (1) Texas residents, (2) domestic non-Texas residents, and (3) international students. conferred upon minority students. Large schools, like UT, are more likely to be ranked higher simply because they graduate a greater number of students (both minorities and non-minorities). See Victor M.H. Borden, Top 100 Undergraduate Degree Producers: Interpreting the Data, DIVERSE ISSUES IN HIGHER EDUC., June 12, 2008. 72 Statistical Handbook 2004 2005, at 22 tbl.S13A; Statistical Handbook 2009 2010, at 16 tbl.S12 (data for fall enrollment only). For fall and summer numbers combined, see 2008 Top Ten Percent Report at 6. 19 No. 09-50822 Students compete for admission only against other students in their respective pool. Texas residents are allotted 90% of all available seats, with admission based on a two-tiered system, beginning with students automatically admitted under the Top Ten Percent Law and then filling the remaining seats on the basis of the Academic and Personal Achievement Indices.73 Because Appellants are Texas residents, their challenge focuses on the admissions procedures applied to in-state applicants. Texas applicants are divided into two subgroups: (1) Texas residents who are in the top ten percent of their high school class and (2) those Texas residents who are not. Top ten percent applicants are guaranteed admission to the University, and the vast majority of freshmen are selected in this way, without a confessed consideration of race. In 2008, for example, 81% of the entering class was admitted under the Top Ten Percent Law, filling 88% of the seats allotted to Texas residents and leaving only 1,216 offers of admission university-wide for non-top ten percent residents.74 The impact of the Top Ten Percent Law on UT s admissions has increased dramatically since it was first introduced in 1998, when only 41% of the seats for Texas residents were claimed by students with guaranteed admission.75 73 Admission decisions for domestic non-Texas residents and international applicants are made solely on the basis of their Academic and Personal Achievement Indices. 74 2008 Top Ten Percent Report at 8 tbl.2, 9 tbl.2b. Table 2 shows 8,984 top ten percent students were admitted in 2008. The UT Associate Director of Admissions reported that 10,200 admissions slots are available for Texas residents, leaving 1,216 slots for non-top ten percent students. Ishop Aff. (Dist. Ct. Dkt. No. 96, Tab 7) ¶ 12. 75 Id. at 7 tbl.1a. In 1998, out of a class that included 6,110 Texas residents, only 2,513 enrolled freshmen were admitted under the Top Ten Percent Law. 20 No. 09-50822 The remaining Texas applicants, who were not within the top ten percent of their high school graduating class, compete for admission based on their Academic and Personal Achievement Indices.76 The Academic Index is the mechanical formula that predicts freshman GPA using standardized test scores and high school class rank.77 Some applicants AI scores are high enough that they receive admission based on that score alone. Others are low enough that their applications are considered presumptively denied. If an application is presumptively denied, senior admission staff review the file and may, on rare occasions, designate the file for full review notwithstanding the AI score.78 The Personal Achievement Index is based on three scores: one score for each of the two required essays and a third score, called the personal achievement score, which represents an evaluation of the applicant s entire file. The essays are each given a score between 1 and 6 through a holistic evaluation of the essay as a piece of writing based on its complexity of thought, substantiality of development, and facility with language. 79 The personal achievement score is also based on a scale of 1 to 6, although it is given slightly 76 The district court found that, on relatively rare occasions, a holistic review of the entire application may result in the University admitting an applicant to the fall class even though his or her AI or PAI scores fall just shy of the official cutoff. See Fisher, 645 F. Supp. 2d at 599. 77 Fisher, 645 F. Supp. 2d at 596. The precise formulas used to calculate an applicant s Academic Index are derived by regression analysis and vary by intended major. For instance, the formula for prospective engineering majors gives greater weight to math scores, whereas the formula for prospective liberal arts majors gives somewhat greater weight to verbal scores. See 2004 Proposal at 27 & n.5. The differences in these formulas are immaterial to the present case. 78 In other words, no applicant is denied admission based purely on AI score without having her file reviewed by at least one admissions reader and her individual circumstances considered. 79 Fisher, 645 F. Supp. 2d at 597. 21 No. 09-50822 greater weight in the final PAI calculation than the mean of the two essay scores.80 This personal achievement score is designed to recognize qualified students whose merit as applicants was not adequately reflected by their Academic Index. Admissions staff assign the score by assessing an applicant s demonstrated leadership qualities, awards and honors, work experience, and involvement in extracurricular activities and community service. In addition, the personal achievement score includes a special circumstances element that may reflect the socioeconomic status of the applicant and his or her high school, the applicant s family status and family responsibilities, the applicant s standardized test score compared to the average of her high school, and beginning in 2004 the applicant s race.81 To assess these intangible factors, evaluators read the applicant s essays again, but this time with an eye to the information conveyed rather than the quality of the student s writing. Admissions officers undergo annual training by a nationally recognized expert in holistic scoring, and senior staff members perform quality control to verify that awarded scores are appropriate and consistent. The most recent study, in 2005, found that holistic file readers scored within one point of each other 88% of the time.82 None of the elements of the personal achievement score including race are considered individually or given separate numerical values to be added together. Rather, the file is evaluated as a whole in order to provide the fullest possible understanding of the student as a person and to place his or her 80 PAI = [(personal achievement score * 4) + (average essay score * 3)] / 7. Id. at 597 n.7. 81 Id. at 591 592, 597. 82 Id. at 597; see Univ. of Tex. at Austin Office of Admissions, Inter-Rater Reliability of Holistic Measures Used in the Freshman Admission Process of the University of Texas at Austin (Feb. 22, 2005) (Dist. Ct. Dkt. No. 94, Ex. 10). 22 No. 09-50822 achievements in context.83 As UT s director of admissions explained, race provides like [the] language [spoken in the applicant s home], whether or not someone is the first in their family to attend college, and family responsibilities important context in which to evaluate applicants, and is only one aspect of the diversity that the University seeks to attain. 84 Race is considered as part of the applicant s context whether or not the applicant belongs to a minority group, and so at least in theory it can positively impact applicants of all races, including Caucasian[s], or [it] may have no impact whatsoever. 85 Moreover, given the mechanics of UT s admissions process, race has the potential to influence only a small part of the applicant s overall admissions score. The sole instance when race is considered is as one element of the personal achievement score, which itself is only a part of the total PAI. Without a sufficiently high AI and well-written essays, an applicant with even the highest personal achievement score will still be denied admission.86 B Although the process for calculating AI and PAI scores is common to all parts of the University, each offer of admission to UT is ultimately tied to an individual school or major. Texas residents in the top ten percent of their high school class are guaranteed admission to the University, but they are not assured admission to the individual school or program of their choice. Most majors and colleges in the University provide automatic admission to Top Ten Percent Law applicants, but certain impacted majors including the School of Business, the College of Communication, and the Schools of 83 Fisher, 645 F. Supp. 2d at 597. 84 Walker Aff. (Dist. Ct. Dkt. No. 96, Tab 11) ¶ 15. 85 Fisher, 645 F. Supp. 2d at 597. 86 See id. at 608. 23 No. 09-50822 Engineering, Kinesiology, and Nursing are obligated to accept only a certain number of Top Ten Percent Law applicants.87 These programs are impacted because they could fill 80% or more of their available spaces each year solely through operation of the Top Ten Percent Law. To avoid oversubscription and to allow these colleges and majors to admit some non-top ten percent applicants, UT caps the percentage of students automatically admitted to these programs at 75% of the available spaces.88 Top Ten Percent Law applicants who do not receive automatic entry to their first choice program compete for admission to the remaining spaces, and if necessary to their second-choice program, on the basis of their AI and PAI scores. The admissions office places students into matrices for each preferred school or major, with students grouped by AI score along one axis and PAI score along the other axis. Liaisons for the majors then establish a cutoff line, which is drawn in a stair-step pattern. Applicants denied admission to their firstchoice program are considered for their second choice, with cutoff lines readjusted to reflect the influx of those applicants. Any top ten percent applicants not admitted to either their first- or second-choice program are automatically admitted as Liberal Arts Undeclared majors. All other applicants not yet admitted to UT compete, again according to AI and PAI scores, for any remaining seats in the Liberal Arts Undeclared program. Although this completes the admissions process for the fall portion of the freshman class, no Texas resident who submits a timely application is denied admission. Instead, those residents not admitted to the entering fall class are 87 In addition, because of special portfolio, audition, and other requirements, the Top Ten Percent Law does not apply to the School of Architecture, the School of Fine Arts, and certain honors programs. 88 Thus, for example, the School of Business granted automatic admission only to those students who graduated in the top 4% of their high school class and selected a business major as their first choice. Ishop Dep. (Dist. Ct. Dkt. No. 96, Tab 2) at 32. 24 No. 09-50822 offered admission to either the summer program or the Coordinated Admissions Program (CAP). Marginal applicants who missed the cutoff for the fall class are offered admission to the summer program, which permits students to begin their studies at UT during the summer and then join the regularly admitted students in the fall. About 800 students enroll in the summer program each year. All remaining Texas applicants are automatically enrolled in CAP, which guarantees admission as a transfer student if the student enrolls in another UT system campus for her freshman year and meets certain other conditions, including the completion of thirty credit hours with a cumulative grade point average of 3.2 or higher. C The Academic Index and Personal Achievement Index now employed by UT have been in continuous use since 1997. The lone substantive change came in 2005, following the Grutter decision, when the Board of Regents authorized the consideration of race as another special circumstance in assessing an applicant s personal achievement score. Race like all other elements of UT s holistic review is not considered alone. Admissions officers reviewing each application are aware of the applicant s race, but UT does not monitor the aggregate racial composition of the admitted applicant pool during the process. The admissions decision for any particular applicant is not affected positively or negatively by the number of other students in her racial group who have been admitted during that year.89 Thus, it is difficult to evaluate which applicants have been positively or negatively affected by its consideration or which applicants were ultimately offered admission due to their race who would not have otherwise been offered 89 Fisher, 645 F. Supp. 2d at 598, 609. 25 No. 09-50822 admission. 90 Nevertheless, the district court found that race is undisputedly a meaningful factor that can make a difference in the evaluation of a student s application. 91 D UT undoubtedly has a compelling interest in obtaining the educational benefits of diversity, and its reasons for implementing race-conscious admissions expressed in the 2004 Proposal mirror those approved by the Supreme Court in Grutter. The district court found that both the UT and Grutter policies attempt to promote cross-racial understanding, break down racial stereotypes, enable students to better understand persons of other races, better prepare students to function in a multi-cultural workforce, cultivate the next set of national leaders, and prevent minority students from serving as spokespersons for their race. 92 Like the law school in Grutter, UT has determined, based on its experience and expertise, that a critical mass of 90 Id. at 597. 91 Id. at 597 598. 92 Id. at 603 (quoting Grutter, 539 U.S. at 319 20). More specifically, as described in the 2004 Proposal, one purpose of UT s race-conscious policy is to provide an educational setting that fosters cross-racial understanding, provides enlightened discussion and learning, and prepares students to function in an increasingly diverse workforce and society. 2004 Proposal at 25 (quoted in Fisher, 645 F. Supp. 2d at 603). Another is to produce future educational, cultural, business, and sociopolitical leaders. Id. at 24 (quoted in Fisher, 645 F. Supp. 2d at 602). And because Texas s population is uniquely diverse [i]n the near future, Texas will have no majority race tomorrow s leaders must not only be drawn from a diverse population[,] but must also be able to lead a multicultural workforce and to communicate policy to a diverse electorate. Id. at 24 (quoted in Fisher, 645 F. Supp. 2d at 602). As the state s flagship public institution, UT determined that it has a compelling educational interest to produce graduates who are capable of fulfilling the future leadership needs of Texas. Id. at 24 (quoted in Fisher, 645 F. Supp. 2d at 602). 26 No. 09-50822 underrepresented minorities is necessary to further its compelling interest in securing the educational benefits of a diverse student body. 93 UT has made an educational judgment that such diversity is essential to its educational mission, just as Michigan s Law School did in Grutter.94 Considering UT s admissions system in its historical context, it is evident that the efforts of the University have been studied, serious, and of high purpose, lending support to a constitutionally protected zone of discretion. That said, the use of race summons close judicial scrutiny, necessary for the nation s slow march toward the ideal of a color-blind society, at least as far as the government can see. IV. STANDARD OF REVIEW It is a given that as UT s Grutter-like admissions program differentiates between applicants on the basis of race, it is subject to strict scrutiny with its requirement of narrow tailoring.95 At the same time, the Supreme Court has held that [c]ontext matters when evaluating race-based governmental action, and a university s educational judgment in developing diversity policies is due deference.96 93 Fisher, 645 F. Supp. 2d at 603 (quoting Grutter, 539 U.S. at 333). 94 Grutter, 539 U.S. at 328. 95 Id. at 326, 328 (citing Adarand, 515 U.S. at 227); see also Parents Involved, 551 U.S. at 720. 96 Grutter, 539 U.S. at 327; see also id. at 328 ( The Law School s educational judgment . . . is one to which we defer. . . . Our holding today is in keeping with our tradition of giving a degree of deference to a university s academic decisions, within constitutionally prescribed limits. ). 27 No. 09-50822 A Judicial deference to a university s academic decisions rests on two independent foundations. First, these decisions are a product of complex educational judgments in an area that lies primarily within the expertise of the university, far outside the experience of the courts.97 Second, universities occupy a special niche in our constitutional tradition, with educational autonomy grounded in the First Amendment.98 As Justice Powell explained in Bakke, [a]cademic freedom . . . . includes [a university s] selection of its student body. 99 Yet the scrutiny triggered by racial classification is no less strict for taking into account the special circumstances of higher education.100 [S]trict scrutiny is designed to provide a framework for carefully examining the importance and the sincerity of the reasons advanced by the governmental decisionmaker for the use of race in [a] particular context. 101 Narrow tailoring, a component of strict scrutiny, requires any use of racial classifications to so closely fit a compelling goal as to remove the possibility that the motive for the classification was illegitimate racial stereotype. Rather than second-guess the merits of the University s decision, a task we are ill-equipped to perform, we instead scrutinize the University s decisionmaking process to ensure that its decision to adopt a race-conscious admissions policy followed from the good faith consideration Grutter requires. We presume the University acted in good faith, 97 Id. at 328. 98 Id. at 329. 99 Bakke, 438 U.S. at 312 (opinion of Powell, J.). 100 Grutter, 539 U.S. at 328. 101 Id. at 327. 28 No. 09-50822 a presumption Appellants are free to rebut.102 Relatedly, while we focus on the University s decision to adopt a Grutter-like plan, admissions outcomes remain relevant evidence of the plan s necessity a reality check. B With a nod to Grutter s command that we generally give a degree of deference to a university s educational judgments, Appellants urge that Grutter did not extend such deference to a university s decision to implement a raceconscious admissions policy. Instead, they maintain Grutter deferred only to the university s judgment that diversity would have educational benefits, not to the assessment of whether the university has attained critical mass of a racial group or whether race-conscious efforts are necessary to achieve that end. As an initial matter, this argument in its full flower is contradicted by Grutter. The majority held that, like the examination into whether the University has a compelling interest, the narrow-tailoring inquiry . . . must be calibrated to fit the distinct issues raised by the use of race to achieve student body diversity in public higher education. 103 That is, the narrow-tailoring inquiry like the compelling-interest inquiry is undertaken with a degree of deference to the University s constitutionally protected, presumably expert academic judgment. Appellants would have us borrow a more restrictive standard of review from a series of public employment and government contracting cases, in which the Supreme Court held that certain government actions to remedy past racial discrimination actions that are themselves based on race are constitutional 102 Id. at 329 ( [G]ood faith on the part of a university is presumed absent a showing to the contrary. (internal quotation marks omitted) (quoting Bakke, 438 U.S. at 318 19 (opinion of Powell, J.))). 103 Id. at 333 34. 29 No. 09-50822 only where there is a strong basis in evidence that the remedial actions were necessary. 104 The Court most recently applied this strong-basis-in-evidence standard in Ricci v. DeStefano. In Ricci, white firefighters from New Haven, Connecticut sued under Title VII, challenging the city s decision to disregard a promotions test after the results showed that white candidates significantly outperformed minority candidates.105 New Haven defended this action, arguing that if it had ratified the test results it could have faced liability under Title VII for adopting a practice that had a disparate impact on the minority firefighters.106 The white firefighters, however, argued that ignoring the test results was a violation of Title VII s separate prohibition against intentional race discrimination, or disparate treatment.107 Responding to this tension, the Supreme Court held that such intentional race-based action is not permitted by Title VII unless the employer can demonstrate with a strong basis in evidence that it would have been liable under the disparate impact provision had it not taken the action.108 The Court suggested that anything less would risk creating a de facto quota system, where an employer could disregard test results to achieve a preferred racial balance, impermissibly shifting the focus from individual discrimination 104 Ricci v. DeStefano, 129 S. Ct. 2658, 2675 (2009) (some internal quotation marks omitted) (quoting Richmond v. J.A Croson Co., 488 U.S. 469, 500 (1989), in turn quoting Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 277 (1986) (plurality)). 105 Id. at 2664. 106 Id.; see 42 U.S.C. § 2000e-2(k)(1)(A)(i) (codifying Griggs v. Duke Power Co., 401 U.S. 424 (1971)). 107 See 42 U.S.C. § 2000e-2(a)(1). 108 Ricci, 129 S. Ct. at 2664. 30 No. 09-50822 to group bias.109 Applying the strong-basis-in-evidence standard, the Supreme Court held that New Haven s fear of disparate impact liability was not adequately supported.110 The city had argued it only needed to show a fear of liability based on a good-faith belief a rough analogy to the university admissions standard. Yet the Court found that an intent-based standard could not be squared with the statutory text. The Ricci Court turned to the strong-basis-in-evidence standard as a matter of statutory construction to resolve any conflict between the disparate-treatment and disparate-impact provisions of Title VII. 111 Although Ricci did not address the firefighters equal protection claim, the Court derived its standard from Richmond v. J.A. Croson Co.,112 a government contracting case, which in turn adopted from a plurality opinion in Wygant v. Jackson Board of Education, a public employment case.113 In Wygant, the plurality concluded that defending race-based public employment decisions as responsive to present effects of past discrimination required a strong basis in evidence of the past discrimination.114 Similarly, Croson adopted this standard after observing that an amorphous claim [of] past discrimination in a particular industry cannot justify the use of an unyielding racial quota. 115 109 Id. at 2676. 110 Id. 111 Id. at 2676. We note that these statutory constraints are not present in the context of university admissions programs. 112 488 U.S. at 500. 113 476 U.S. at 277. 114 Id. at 277 278. 115 Croson, 488 U.S. at 499. 31 No. 09-50822 This recitation of history, quick as it is, makes plain that the cases Appellants cite have little purchase in this challenge to university admissions. The high standard for justifying the use of race in public employment decisions responds to the reality that race used in a backward-looking attempt to remedy past wrongs, without focus on individual victims, does not treat race as part of a holistic consideration. In doing so, it touches the third rail of racial quotas. Wygant and Croson both involved explicit quotas; in Ricci, the Court was concerned that the city s use of race threatened to devolve into a de facto quota. By contrast, Grutter recognized that universities are engaged in a different enterprise. Their holistic approach is part of a forward-looking effort to obtain the educational benefits of diversity. The look to race as but one element of this further goal, coupled with individualized consideration, steers university admissions away from a quota system. Grutter teaches that so long as a university considers race in a holistic and individualized manner, and not as part of a quota or fixed-point system, courts must afford a measure of deference to the university s good faith determination that certain race-conscious measures are necessary to achieve the educational benefits of diversity, including attaining critical mass in minority enrollment. Parents Involved in Community Schools v. Seattle School District No. 1 further supports this understanding.116 When scrutinizing two school districts race-conscious busing plans, the Court invoked Grutter s serious, good faith consideration standard, rather than the strong-basis-in-evidence standard that Appellants would have us apply.117 The Parents Involved Court never suggested that the school districts would be required to prove their plans were meticulously supported by some particular quantum of specific evidence. Rather, the Court struck down the school districts programs because they pursued racial balancing 116 551 U.S. 701 (2007). 117 See id. at 735 (quoting Grutter, 539 U.S. at 339). 32 No. 09-50822 and defined students based on racial group classifications, not on individual circumstances. In short, the Court has not retreated from Grutter s mode of analysis, one tailored to holistic university admissions programs. Thus, we apply strict scrutiny to race-conscious admissions policies in higher education, mindful of a university s academic freedom and the complex educational judgments made when assembling a broadly diverse student body. C Appellants do not allege that UT s race-conscious admissions policy is functionally different from, or gives greater consideration to race than, the policy upheld in Grutter. Rather, Appellants question whether UT needs a Grutter-like policy. As their argument goes, the University s race-conscious admissions program is unwarranted because (1) UT has gone beyond a mere interest in diversity for education s sake and instead pursues a racial composition that mirrors that of the state of Texas as a whole, amounting to an unconstitutional attempt to achieve racial balancing ; (2) the University has not given adequate consideration to available race-neutral alternatives, particularly percentage plans like the Top Ten Percent Law; and (3) UT s minority enrollment under the Top Ten Percent Law already surpassed critical mass, such that the additional (and allegedly minimal ) increase in diversity achieved through UT s Grutterlike policy does not justify its use of race-conscious measures. We will consider each of these arguments in turn. V. RACIAL BALANCING Again, diversity is a permissible goal for educational institutions, but outright racial balancing is not. Attempting to ensure that the student body contains some specified percentage of a particular racial group is patently 33 No. 09-50822 unconstitutional. 118 This concept follows from the Supreme Court s repeated emphasis that, by itself, increasing racial representation is not a sufficiently compelling interest to justify the use of racial preferences. Grutter described many important educational interests that may be sought through diversity, but steadfastly maintained that [r]acial balance is not to be to be achieved for its own sake. 119 Moreover, [t]he point of the narrow tailoring analysis in which the Grutter Court engaged was to ensure that the use of racial classifications was indeed part of a broader assessment of diversity, and not simply an effort to achieve racial balance by creating an unconstitutional quota.120 A Looking to the details of UT s race-conscious admissions policy, it is clear that administrators knew a quota system would not survive judicial review, and they took care to avoid this fatal mistake. UT s system was modeled after the Grutter program, which the Supreme Court held was not a quota. UT has never established a specific number, percentage, or range of minority enrollment that would constitute critical mass, nor does it award any fixed number of points to minority students in a way that impermissibly values race for its own sake.121 118 Grutter, 539 U.S. at 329 30 (quoting Bakke, 438 U.S. at 308 (opinion of Powell, J.)). 119 Id. at 330 (quoting Freeman v. Pitts, 503 U.S. 467, 494 (1992)). 120 Parents Involved, 551 U.S. at 723. 121 Appellants argue that UT s head-in-the-sand approach refusing to identify any specific number, percentage, or range of minority students that would constitute critical mass is an improper attempt to short circuit any inquiry into whether it can justify its policy with evidence by arguing that critical mass is a purely subjective concept that cannot be evaluated in numerical terms. Appellants claim that until UT identifies some finishing line, the use of race has no logical stopping point and is therefore too amorphous a basis for imposing a racially classified remedy. But in both Bakke 34 No. 09-50822 Further, there is no indication that UT s Grutter-like plan is a quota by another name. It is true that UT looks in part to the number of minority students when evaluating whether it has yet achieved a critical mass, but [s]ome attention to numbers, without more, does not transform a flexible admissions system into a rigid quota. 122 Whereas a quota imposes a fixed percentage standard that cannot be deviated from, a permissible diversity goal require[s] only a good-faith effort . . . to come within a range demarcated by the goal itself. 123 Indeed, UT s policy improves upon the program approved in Grutter because the University does not keep an ongoing tally of the racial composition of the entering class during its admissions process.124 UT has not admitted students so that its undergraduate population directly mirrors the demographics of Texas. Its methods and efforts belie the charge. The percentage of Hispanics at UT is less than two-thirds the percentage of Hispanics in Texas, and the percentage of African-Americans at UT is half the percentage of Texas s African-American population, while and Grutter, the controlling opinions expressly approved of policies seeking only some undefined meaningful number of minorities, see Grutter, 539 U.S. at 335; Bakke, 438 U.S. at 323 (opinion of Powell, J.), and the Court has firmly rejected the argument that diversity as a basis for employing racial preferences is simply too open-ended, ill-defined, and indefinite a ground for race-conscious university admissions policies, Gratz, 539 U.S. at 268 (internal quotation marks omitted). On the contrary, if UT were to identify some numerical target for minority enrollment, that would likely render the policy unconstitutional under Grutter. 122 Grutter, 539 U.S. at 336 (citation, internal quotation marks, and brackets omitted). 123 Id. at 335 (quoting Sheet Metal Workers v. EEOC, 478 U.S. 421, 495 (1986)). 124 Cf. Grutter, 539 U.S. at 391 92 (Kennedy, J., dissenting). 35 No. 09-50822 Asian-American enrollment is more than five times the percentage of Texan Asian-Americans.125 B Appellants nevertheless argue that UT s program amounts to racial balancing because it supposedly evinces a special concern for demographically underrepresented groups, while neglecting the diverse contributions of others. These arguments do not account for the operation of UT s admissions system or the scope of the diversity interest approved by the Court in Grutter. 1 The district court expressly found that race can enhance the personal achievement score of a student from any racial background, including whites and Asian-Americans.126 For example, a white student who has demonstrated substantial community involvement at a predominantly Hispanic high school may contribute a unique perspective that produces a greater personal achievement score than a similarly situated Hispanic student from the same school. This possibility is the point of Grutter s holistic and individualized assessments, which must be flexible enough to consider all pertinent elements of diversity in light of the particular qualifications of each applicant. 127 Indeed, just as in Grutter, UT applicants of every race may submit supplemental information to highlight their potential diversity contributions, which allows 125 Fisher, 645 F. Supp. 2d at 607 n.11. 126 Id. at 606. 127 Grutter, 539 U.S. at 337 (quoting Bakke, 438 U.S. at 317 (opinion of Powell, J.)). 36 No. 09-50822 students who are diverse in unconventional ways to describe their unique attributes.128 The summary judgment record shows that demographics are not consulted as part of any individual admissions decision, and UT s admissions procedures do not treat certain racial groups or minorities differently than others when reviewing individual applications. Rather, the act of considering minority group demographics (to which Appellants object) took place only when the University first studied whether a race-conscious admissions program was needed to attain critical mass. Appellants objection therefore must be directed not to the design of the program, but rather to whether UT s decision to reintroduce race as a factor in admissions was made in good faith. 2 Appellants contend that UT revealed its true motive to be outright racial balancing when it referenced state population data to justify the adoption of race-conscious admissions measures. They insist that if UT were truly focused on educational benefits and critical mass, then there should be no reason to consult demographic data when determining whether UT had sufficient minority representation. We disagree. The University s policies and measured attention to the community it serves are consonant with the educational goals outlined in Grutter and do not support a finding that the University was engaged in improper racial balancing during our time frame of review. Both Grutter and Bakke recognized that there is of course some relationship between numbers and achieving the benefits to be derived from a diverse student body. 129 In its 128 Id. at 338; see Fisher, 645 F. Supp. 2d at 597. 129 Grutter, 539 U.S. at 336 (quoting Bakke, 438 U.S. at 323 (opinion of Powell, J.)). 37 No. 09-50822 policymaking process, UT gave appropriate attention to those educational benefits identified in Grutter without overstepping any constitutional bounds. Grutter recognized that racial and ethnic backgrounds play an influential role in producing the diversity of views and perspectives which are paramount to a university s educational mission. As Justice O Connor explained, the unique experience of being a racial minority in a society, like our own, in which race unfortunately still matters can have a significant impact on a student s views.130 The Court acknowledged that [b]y virtue of our Nation s struggle with racial inequality, [underrepresented minority students] are both likely to have experiences with particular importance to the [University s] mission, and less likely to be admitted in meaningful numbers on criteria that ignore these experiences. 131 UT properly concluded that these individuals from the state s underrepresented minorities would be most likely to add unique perspectives that are otherwise absent from its classrooms. Identifying which backgrounds are underrepresented, in turn, presupposes some reference to demographics, and it was therefore appropriate for UT to give limited attention to this data when considering whether its current student body included a critical mass of underrepresented groups. Preparing students to function as professionals in an increasingly diverse workforce likewise calls for some consideration of a university s particular educational mission and the community it serves. For instance, a nationally renowned law school draws upon a nationwide applicant pool and sends its graduates into careers in all states; therefore it is appropriate for such a school to consider national diversity levels when setting goals for its admissions program. In contrast, UT s stated goal is to produce graduates who are capable 130 Id. at 333. 131 Id. at 338. 38 No. 09-50822 of fulfilling the future leadership needs of Texas. 132 This objective calls for a more tailored diversity emphasis. In a state as racially diverse as Texas, ensuring that graduates learn to collaborate with members of racial groups they will encounter in the workforce is especially important. The 2004 Proposal concluded that a race-conscious admissions program was necessary at UT specifically because from a racial, ethnic, and cultural standpoint, students at the University [were] being educated in a less-than-realistic environment that [was] not conducive to training the leaders of tomorrow. 133 The need for a state s leading educational institution to foster civic engagement and maintain visibly open paths to leadership also requires a degree of attention to the surrounding community. A university presenting itself as open to all may be challenged when the state s minority population grows steadily but minority enrollment does not. Indeed, the 2004 Proposal expressed concern that UT appeared largely closed to nonwhite applicants and did not provide a welcoming supportive environment for minority students.134 UT was keenly aware that by sending a message that people of all stripes can succeed at UT, the University would attract promising applicants from once-insulated communities, over time narrowing the credentials gap between minority and non-minority applicants.135 After Hopwood, such applicants were dissuaded from applying to UT. But through the Top Ten Percent Law and Grutter-like plan, UT has increased its minority applicant pool in its effort to ensure that it 132 2004 Proposal at 23 (quoted in Fisher, 645 F. Supp. 2d at 602). 133 Id. at 24 25 (quoted in Fisher, 645 F. Supp. 2d at 602). 134 Id. at 14. 135 See, e.g., Mark C. Long et al., Policy Transparency and College Enrollment: Did the Texas Top Ten Percent Law Broaden Access to the Public Flagships?, 627 ANNALS AM. ACAD. POL. & SOC. SCI. 82 (2010); Kim M. Lloyd et al., Minority College Aspirations, Expectations and Applications Under the Texas Top 10% Law, 86 SOC. FORCES 1105 (2008). 39 No. 09-50822 serves as a flagship university for the entire state, not just Texans of certain backgrounds. Cultivating paths to leadership for underrepresented groups serves both the individual and the public, sustaining an infrastructure of leaders in an increasingly pluralistic society. Although a university must eschew demographic targets, it need not be blind to significant racial disparities in its community, nor is it wholly prohibited from taking the degree of disparity into account. Finally, Grutter s structure accepts that a university s twin objectives of rewarding academic merit and fostering diversity can be complementary rather than competing goals; that students rising to the top of underrepresented groups demonstrate promise as future leaders. These students relative success in the face of harmful and widespread stereotypes evidences a degree of drive, determination, and merit not captured by test scores alone. Insofar as Appellants complain that the University s limited attention to demographics was inconsistent with the legitimate educational concerns recognized in Bakke and Grutter, we conclude that their contention cannot be sustained. 3 Appellants argue that a broad approach to educational diversity is improper because critical mass must be an inward-facing concept . . . that focuses on the functioning of the student body, encompassing only that level of minority enrollment necessary to ensure that minority students participate in the classroom and do not feel isolated. While Appellants view may comport with one literal interpretation of the critical mass label, it is not the view that prevailed in Grutter. The Grutter majority defined critical mass by reference to the educational benefits that diversity is designed to produce, 136 and the educational benefits recognized in Grutter go beyond the narrow pedagogical 136 Grutter, 539 U.S. at 339. 40 No. 09-50822 concept urged by Appellants. On this understanding, there is no reason to assume that critical mass will or should be the same for every racial group or every university. We are persuaded, as was the district court, that the University adhered to Grutter when it reintroduced race into its admissions process based in part on an analysis that devoted special attention to those minorities which were most significantly underrepresented on its campus. VI. THE TOP TEN PERCENT LAW Grutter is best read as a path toward the moment when all race-conscious measures become unnecessary. To that end, Grutter requires universities that employ race-conscious admissions to seriously consider race-neutral alternatives. But [n]arrow tailoring does not require exhaustion of every conceivable race-neutral alternative, especially if the proffered alternatives would require the University to sacrifice other important interests, like its academic selectivity and reputation for excellence.137 The parties devote significant attention to the Top Ten Percent Law.138 Since the Law was first enacted in 1997, UT has seen increases in both African-American and Hispanic enrollment, but again, changing demographics and other minority outreach programs render it difficult to quantify the increases attributable to the Top Ten Percent Law.139 137 See id. at 339 340. 138 TEX. EDUC. CODE § 51.803 (1997). The precise impact UT s other raceneutral alternatives (such as scholarship and outreach programs) have had on minority enrollment is not clear, but their effect would not appear to be great enough to bear on the constitutionality of the University s race-conscious admissions policy. 139 Fisher, 645 F. Supp. 2d at 594; see also Marta Tienda & Teresa A. Sullivan, The Promise and Peril of the Texas Uniform Admissions Law 164 65 & tbl.1, in THE NEXT TWENTY-FIVE YEARS? AFFIRMATIVE ACTION AND HIGHER EDUCATION IN THE UNITED STATES AND SOUTH AFRICA 155 (David L. Featherman 41 No. 09-50822 Appellants put forward the Top Ten Percent Law as a facially race-neutral alternative that would allow UT to obtain a critical mass of minority enrollment without resorting to race-conscious admissions. As the argument goes, if the Top Ten Percent Law were able to serve the University s interests about as well as race-conscious admissions, without differentiating between students on the basis of race, then it unconstitutional.140 would render UT s current admissions program UT responds that the Top Ten Percent Law does not constitute a workable alternative to a flexible admissions system, and so it is entirely irrelevant as a matter of law in determining whether or not a university may adopt the holistic consideration of race to achieve critical mass. UT is correct that so-called percentage plans are not a constitutionally mandated replacement for race-conscious admissions programs under Grutter, although as will become apparent this realization alone does not end our constitutional inquiry. The idea of percentage plans as a viable alternative to race-conscious admissions policies was directly advocated to the Grutter Court by the United States, arguing as amicus curiae.141 In response, the Court held that although percentage plans may be a race-neutral means of increasing minority enrollment, they are not a workable alternative at least in a constitutionally significant sense because they may preclude the university from conducting the individualized assessments necessary to assemble a student body that is not just racially diverse, but diverse along all the qualities valued by the university. 142 In addition, the Court emphasized existing percentage et al. eds., 2010). 140 See Grutter, 539 U.S. at 339 (quoting Wygant, 476 U.S. at 280 n.6 (1986)). 141 The United States has since filed an amicus brief in the present case, urging us to uphold UT s current admissions program. 142 Grutter, 539 U.S. at 340 (internal citation omitted). 42 No. 09-50822 plans including UT s are simply not capable of producing a critical mass without forcing [universities] to abandon the academic selectivity that is the cornerstone of [their] educational mission. 143 That the Top Ten Percent Law is not a constitutionally-mandated alternative does not make it irrelevant. By now it is clear that the Law is inescapably tied to UT s Grutter plan, as Grutter does its work with the applicants who remain after the cut of the Top Ten Percent Law. In 2008, top ten percent applicants accounted for 8,984 of the 10,200 Texas admittees.144 Thus, with the Top Ten Percent Law in effect, UT s Grutter plan can only possibly influence the review of approximately 1,200 admitted students applications.145 In evaluating the constitutionality of an admissions program, we cannot ignore a part of the program comprising 88% of admissions offers for Texas residents and yielding 81% of enrolled Texan freshmen.146 The reality is that the Top Ten Percent Law alone does not perform well in pursuit of the diversity Grutter endorsed and is in many ways at war with it. 143 Id. 144 2008 Top Ten Report at 8 tbl.2; Ishop Aff. (Dist. Ct. Dkt. No. 96, Tab 7) ¶ 12. 145 In reality, the Grutter plan operates on even fewer applications, as many non-top ten percent students are admitted based purely on their class rank and standardized test scores, without any reference to their PAI, leaving only 841 seats in 2008 that were evaluated under the Grutter plan. See Ishop Aff. (Dist. Ct. Dkt. No. 96, Tab 7) ¶ 12. 146 2008 Top Ten Report at 7 tbl.1a; see supra note 74 and accompanying text. We also note that since it began, the Top Ten Percent Law has had an increasing impact on admissions decisions. In 1998, top ten percent candidates comprised just 41% of Texans in the freshman class. In 2004, 66% of Texan freshmen were top ten percent students, and in 2008, top ten percent students made up 81% of the Texas freshmen seats. While the legislative 75% cap on top ten percent enrollment may help alleviate some of the concerns with this plan, the fact remains that the Top Ten Percent Law operates today very differently than it did when first implemented. 43 No. 09-50822 While the Law may have contributed to an increase in overall minority enrollment, those minority students remain clustered in certain programs, limiting the beneficial effects of educational diversity.147 For example, nearly a quarter of the undergraduate students in UT s College of Social Work are Hispanic, and more than 10% are African-American. In the College of Education, 22.4% of students are Hispanic and 10.1% are African-American. By contrast, in the College of Business Administration, only 14.5% of the students are Hispanic and 3.4% are African-American.148 It is evident that if UT is to have diverse interactions, it needs more minority students who are interested in and meet the requirements for a greater variety of colleges, not more students disproportionately enrolled in certain programs. The holistic review endorsed by Grutter gives UT that discretion, but the Top Ten Percent Law, which accounts for nearly 90% of all Texas resident admissions, does not.149 147 See Univ. of Tex. at Austin Office of Info. Mgmt., Statistical Handbook 2009 2010, at 32 tbl.S27 (2010) (reporting UT enrollment by college, grade level, ethnicity, and gender); Lisa Dickson, Major Choices: Race and Gender Differences in College Major Choice, 627 ANNALS AM. ACAD. POL. & SOC. SCI. 108, 108 (2010) (analyzing UT data and finding that significant differences by gender, race, and ethnicity persist in initial college major choice even after controlling for the [SAT] score of the student and the high school class rank of the student ). 148 Statistical Handbook 2009 2010, at 31 32 tbl.S27. 149 For example, instead of admitting a minority top ten percent student from a low-performing school, UT might admit a minority student with an interest in business who is just as academically qualified (and perhaps more so), but falls outside the top ten percent of his high school class because he attends a more competitive high school. This example also demonstrates how the Top Ten Percent Law hurts academic selectivity: UT must admit a top ten percent student from a low-performing high school before admitting a more qualified minority student who ranks just below the top ten percent at a highly competitive high school. This effect, in turn, further widens the credentials gap between minority and non-minority students at the University, which risks driving away matriculating minority students from difficult majors like business or the sciences. 44 No. 09-50822 Focusing narrowly on geographic diversity, in part as a proxy for race, the Top Ten Percent Law crowds out other types of diversity that would be considered under a Grutter-like plan. By ignoring these other diversity contributions, the Top Ten Percent Law restricts the University s ability to achieve the maximum educational benefits of a truly diverse student body.150 As UT s 2003 classroom study shows, percentage plans bear little promise of producing the meaningful diverse interactions envisioned by Grutter, at least not in the classroom. For instance, the study reported that although overall enrollment of minority students at UT rose significantly between 1996 and 2002, the Fall 2002 schedule contained more classes with zero or one African American or Hispanic students than had the Fall 1996 schedule.151 Justice Ginsburg pointed out in Grutter s companion case that percentage plans create damaging incentives to the education system. She observed that [p]ercentage plans depend for their effectiveness on continued racial segregation at the secondary school level. These measures encourage parents to keep their children in low-performing segregated schools, and discourage students from taking challenging classes that might lower their grade point 150 The Top Ten Percent Law may produce diversity beyond varying hometowns, including differences in socioeconomic status and rural/urban/suburban upbringing. However, under the Top Ten Percent Law, the University does not have the opportunity to select for a wide range of diverse experiences (such as travel abroad, extra-curricular involvement, or work experience), so the Top Ten Percent Law bluntly operates as an attempt to create diversity through reliance on perceived group characteristics and segregated communities. 151 2004 Proposal at 25 & tbl.8. 45 No. 09-50822 averages. 152 Similarly, these plans create a strong incentive to avoid competitive educational institutions like magnet schools.153 Texas applicants falling outside the top ten percent group face extreme competition to gain admittance to the University. There are approximately 16,000 students competing for only 1,216 fall admissions slots. The competition is so great that, on average, students admitted from outside the top ten percent of their high school class, regardless of race, have even higher SAT scores than those granted automatic admission under the Top Ten Percent Law.154 Perversely, this system negatively impacts minority students (who nationally have lower standardized test scores) in the second decile of their classes at competitive high schools. Grutter s holistic look at race may soften this unreasonable exclusion of those second-decile minorities better qualified than many of the non-minorities bluntly swept in under the Top Ten Percent Law. But not much. It requires no empirical study to observe that those excluded under this Law have been a rich source of Texas leaders over its history and that for some applicants, admission to the flagship school of Texas is little more possible than admission to Harvard.155 152 That all of these weaknesses are Gratz, 539 U.S. at 304 n.10 (Ginsburg, J., dissenting). 153 In an effort to ameliorate this effect, a special provision of the Top Ten Percent Law provides that a high school magnet program, academy, or other special program may be considered an independent high school with its own graduating class separate from the graduating class of other students attending the high school, effectively allowing the school to certify two separate groups of Top Ten Percent Law students. See TEX. EDUC. CODE § 51.8045. 154 See 2008 Top Ten Percent Report at 12 tbl.6 (showing the average SAT range for top ten percent and non-top ten percent students); id. at 13 15 tbls.6a 6d (displaying SAT ranges based on race and top ten percent status). 155 To reach its target class size, UT offers fall admission to 10,200 Texas applicants. Ishop Aff. (Dist. Ct. Dkt. No. 96, Tab 7) ¶ 12. For the class entering Fall 2008, after UT offered admission to top ten percent students, there were 1,216 admissions spots remaining. (The district court noted there were 841 46 No. 09-50822 apparent in the Top Ten Percent Law only make its focus upon race the plainer.156 The Top Ten Percent Law was adopted to increase minority enrollment. That it has done, but its sweep of admissions is a polar opposite of the holistic focus upon individuals. Its internal proxies for race end-run the Supreme Court s studied structure for the use of race in university admissions decisions. It casts aside testing historically relied upon, admitting many top ten percent minorities with significantly lower scores than rejected minorities and nonminorities alike. That these admitted minorities are academically able to remain in the University does not respond to the reality that the Top Ten places, but that number included the admission of so-called Group A applicants who have extremely high AI scores but are not in the top ten percent of their class. See id.) There were a total of 27,712 applicants for the fall class of 2008. Statistical Handbook 2009 2010, at 25 tbl.S21. Neither the record nor any public information released by the University disclose what portion of that total applicant pool were Texas residents, but if we assume that proportion of applicants from Texas matches the 90% of admissions slots reserved for Texas applicants, one can estimate that there were 24,940 Texas applicants. Subtracting the 8,984 students admitted under the Top Ten Percent Law yields an estimate of 15,956 applicants for 1,216 seats, or an acceptance rate of approximately 7.6%. By comparison, the overall acceptance rate at Ivy League schools for the class entering Fall 2008 ranged from 8% (Harvard) to 21% (Cornell). See The Rankings: Best National Universities, U.S. NEWS & WORLD REP., Sept. 2009, at 84 85. 156 Appellants here do not challenge the constitutionality of the Top Ten Percent Law. In fact, they endorse it as a race-neutral alternative to the Grutter plan. A court considering the constitutionality of the Law would examine whether Texas enacted the Law (and corresponding admissions policies) because of its effects on identifiable racial groups or in spite of those effects. See Pers. Adm r of Mass. v. Feeney, 442 U.S. 256 (1979); cf. Brief of Social Scientists Glenn C. Loury et al. as Amici Curiae in Support of Respondents, Grutter v. Bollinger, 539 U.S. 306 (2003), available at 2003 WL 402129, at *2, *9 *10 (noting that it is not clear that [percentage] plans are actually race-neutral and that some amici counsel in Grutter have signaled interest in moving on after this case to challenge these aspects of the Texas program ). 47 No. 09-50822 Percent Law eliminated the consideration of test scores, and correspondingly reduced academic selectivity, to produce increased enrollment of minorities. Such costs may be intrinsic to affirmative action plans. If so, Grutter at least sought to minimize those costs through narrow tailoring. The Top Ten Percent Law is anything but narrow. In short, while the Top Ten Percent Law appears to succeed in its central purpose of increasing minority enrollment, it comes at a high cost and is at best a blunt tool for securing the educational benefits that diversity is intended to achieve. We cannot fault UT s contention that the Top Ten Percent Law is plainly not the sort of workable race-neutral alternative that would be a constitutionally mandated substitute for race-conscious university admissions policies. We are keenly aware that the University turned to the Top Ten Percent Law in response to a judicial ruling. Yet we cannot agree that it is irrelevant. To the contrary, that the Top Ten Percent Law, accounting for the vast majority of in-state admissions, threatens to erode the foundations UT relies on to justify implementing Grutter policies is a contention not lacking in force. Facially neutral has a talismanic ring in the law, but it can be misleading. It is here. VII. CRITICAL MASS Appellants contend that UT s decision to reintroduce race-conscious admissions was unconstitutional because minority enrollment already met or exceeded critical mass when this decision was made, and thus any further facial consideration of race was neither warranted nor constitutional. Appellants claim the best measure of whether UT had attained the benefits of diversity is the raw percentage of minorities enrolled. As a result of the combined effects of changing demographics, targeted high school programs, and the Top Ten Percent Law, total minority enrollment had increased over the years. When the decision was made to reintroduce race-conscious admissions in 48 No. 09-50822 2004, underrepresented minorities made up 21.4% of the incoming class (4.5% African-American and 16.9% Hispanic).157 Although Texas was not constitutionally required to enact the Top Ten Percent Law, Appellants are correct that the decision to do so and the substantial effect on aggregate minority enrollment at the University places at risk UT s race-conscious admissions policies. We are confident, and hold, that a Grutter-style admissions system standing alone is constitutional. That said, whether to overlay such a plan with the Top Ten Percent Law and how to calibrate its flow presents a Hobson s choice between the minority students it contributes and the test of constitutional bounds it courts. True enough, the Top Ten Percent Law is in a sense, perhaps a controlling sense, a facially raceneutral plan. But it was animated by efforts to increase minority enrollment, and to the extent it succeeds it is because at key points it proxies for race. A Appellants propose various baseline levels of diversity which, they suggest, would fully satisfy the University s interest in attaining critical mass. They first argue that if from 13.5 to 20.1 percent minority enrollment was adjudged to be great enough diversity each year by Michigan s Law School in Grutter, then the 21.4% minority enrollment that UT had achieved prior to reintroducing race-conscious admissions must already have achieved critical mass. We find this comparison inapt for numerous reasons. Appellants comparison presumes that critical mass must have some fixed upper bound that applies across different schools, different degrees, different states, different years, different class sizes, and different racial and ethnic subcomposition. It is based on Appellants continued insistence that the concept of critical mass is defined by the minimum threshold for minority students to 157 Fisher, 645 F. Supp. 2d at 593. 49 No. 09-50822 have their ideas represented in class discussions and not to feel isolated or like spokespersons for their race. As we have discussed, Grutter firmly rejects that premise and defines critical mass by reference to a broader view of diversity. At oral argument, Appellants qualified this insistence and wisely conceded that what constitutes critical mass in the eyes of one school might not suffice at another. Grutter concerned a law school, whereas Appellants challenge UT s undergraduate program. Michigan s Law School operates on a national level, while UT focuses on recruiting and producing future leaders for Texas. The law school enrolled approximately 350 students in its first-year class, few enough students that diversity in the student body readily approximates diversity in the classroom. In contrast, UT enrolls approximately 7,000 undergraduates in its first-year class and has data showing diversity rates vary widely across individual classrooms. African-Americans and Hispanics never represented more than a combined 14.8% of the Michigan Law School s applicant pool during the examined time period,158 while those same underrepresented minorities were 28% of UT s freshman applicant pool for Fall 2008.159 Appellants point to the Supreme Court s observation in United States v. Virginia that the Virginia Military Institute could achieve at least 10% female enrollment a sufficient critical mass to provide the female cadets with a positive educational experience. 160 But this figure, even if accurate, covers only one component of the multi-faceted concept of diversity elaborated in Grutter. In any event, the claim that 10% minority enrollment is a ceiling to critical mass is confounded by Grutter. Appellants lastly note that minority enrollment at UT now exceeds the level it had reached in the mid-1990s, pre-Hopwood, when the University was 158 Grutter, 539 U.S. at 384 tbls.1 2 (Rehnquist, C.J., dissenting). 159 2008 Top Ten Percent Report at 6 tbl.1. 160 518 U.S. 515, 523 (1996) (internal quotation marks omitted). 50 No. 09-50822 free to obtain any critical mass it wanted through overtly race-based decisions. UT responds that it has consistently maintained, both in the 2004 Proposal and before this Court, that even before Hopwood it had never reached critical mass.161 While UT was making a greater use of race in that era, its pursuit of diversity was constrained by other interests, such as admitting only well-qualified students. We cannot assume that diversity levels immediately before Hopwood were indicative of critical mass. Moreover, minority enrollment in 1996 is not indicative of UT s true pre-Hopwood diversity. While admissions decisions in 1996 were not controlled by Hopwood, the case impacted enrollment, resulting in fewer minority students. If one instead compares minority enrollment from 1989 to 2004, a different picture emerges. In 2004, UT enrolled significantly fewer African-Americans than it had in 1989 (309 compared to 380). In addition, the 2004 entering class consisted of only 100 more Hispanics than the 1989 class, a low number considering the vast increases in the Hispanic population of Texas. Further, the 2004 Proposal demonstrated that the percentage of diverse classrooms had declined since 1996.162 The decrease in classroom diversity will only continue if additional minority representation is not achieved, as the University plans to increase its number of course offerings in future years. Finally, whatever levels of minority enrollment sufficed more than a decade ago may no longer constitute critical mass today, given the social changes Texas has undergone during the intervening years. Appellants proposed baselines are insufficient reason to doubt UT s considered, good faith conclusion that the University still has not reached a critical mass at the classroom level. 163 161 See, e.g., 2004 Proposal at 24 ( [R]estoration to pre-Hopwood levels is not sufficient. ). 162 Id. at 25 & tbl.8. 163 Id. at 24 (quoted in Fisher, 645 F. Supp. 2d at 602). 51 No. 09-50822 Grutter pointedly refused to tie the concept of critical mass to any fixed number. The Grutter Court approved of the University of Michigan Law School s goal of attaining critical mass even though the school had specifically abjured any numerical target.164 The Court recounted how school officials had described critical mass only through abstract concepts such as meaningful numbers, meaningful representation, and a number that encourages underrepresented minority students to participate in the classroom and not feel isolated. 165 The type of broad diversity Grutter approved does not lend itself to any fixed numerical guideposts. None of this is to say that Grutter left critical mass without objective meaning. Rather, the legally cognizable interest attaining a critical mass of underrepresented minority students is defined by reference to the educational benefits that diversity is designed to produce. 166 If a plaintiff produces evidence that calls into question a university s good faith pursuit of those educational benefits, its race-conscious admissions policies may be found unconstitutional. We are not persuaded, however, that any of the benchmarks suggested by Appellants succeed at calling that judgment into question. B As we have observed, benchmarks aside, UT s claim that it has not yet achieved critical mass is less convincing when viewed against the backdrop of the Top Ten Percent Law, which had already driven aggregate minority enrollment up to more than one-fifth of the University s incoming freshman class before less subtle race-conscious admissions were reintroduced. 164 Grutter, 539 U.S. at 318. 165 Id. 166 Id. at 330. 52 No. 09-50822 The chief difficulty with looking to aggregate minority enrollment is that it lumps together distinct minority groups from different backgrounds who may bring various unique contributions to the University environment. African-American and Hispanic students, for example, are not properly interchangeable for purposes of determining critical mass, and a university must be sensitive to important distinctions within these broad groups. In Parents Involved, the Supreme Court specifically faulted two school districts for employing only a limited notion of diversity that lumped together very different racial groups.167 One school district classified students exclusively as white or nonwhite ; another labeled them as black or other. 168 This binary conception of race runs headlong into the central teaching of Grutter and other precedents which instruct that a university must give serious and flexible consideration to all aspects of diversity.169 On this record, we must conclude that the University has acted with appropriate sensitivity to these distinctions. Although the aggregate number of underrepresented minorities may be large, the enrollment statistics for individual groups when UT decided to reintroduce race as a factor in admissions decisions does not indicate critical mass was achieved. Further, we recognize that some year-to-year fluctuation in enrollment numbers is inevitable, so statistics from any single year lack probative force; the University needs to maintain critical mass in years when yield is low just as it does when yield is high. It is also apparent that UT has given appropriate consideration to whether aggregate minority enrollment is translating into adequate diversity in the 167 Parents Involved, 551 U.S. at 723. 168 Id. at 712, 716. 169 Id. at 735. Even current labels of Hispanic, African-American, or Asian may lump very different ethnic groups into a single category. 53 No. 09-50822 classroom. Through two separate studies, the 2004 Proposal reached a serious and good faith determination that the aggregate number overstates the University s true level of diverse interaction. UT sought to obtain the full educational benefits of diversity as approved in Grutter and properly concluded that race-conscious admissions measures would help accomplish its goals. C Appellants argue that even if UT had not yet achieved critical mass under race-neutral policies, it had come close enough that the reintroduction of race-conscious measures was unwarranted. Pointing to the Supreme Court s recent decision in Parents Involved, they argue that the University s use of race is unnecessary, and therefore not narrowly tailored, because it has only a minimal effect. The district court thought this was an attempt to force UT into an impossible catch-22: on the one hand, it is well-established that to be narrowly tailored the means must be specifically and narrowly framed to accomplish the compelling interest, but on the other hand, according to [Appellants], the narrowly tailored plan must have more than a minimal effect. 170 Parents Involved does not support the cost-benefit analysis that Appellants seek to invoke. Rather, Parents Involved was primarily a critique of the school districts extreme approach that used binary racial categories to classify schoolchildren.171 The Court referred to the minimal effect sought by this policy as evidence that other, more narrowly tailored means would be effective to serve the school districts interests.172 The Court did not hold that a Grutter-like system would be impermissible even after race-neutral alternatives 170 Fisher, 645 F. Supp. 2d at 609. 171 Parents Involved, 551 U.S. at 735. 172 Id. at 733. 54 No. 09-50822 have been exhausted because the gains are small. To the contrary, Justice Kennedy who provided the fifth vote in Parents Involved wrote separately to clarify that a more nuanced, individual evaluation . . . . informed by Grutter would be permissible, even for the small gains sought by the school districts.173 VIII. CONCLUSION Mindful of the time frame of this case, we cannot say that under the circumstances before us UT breached its obligation to undertake a serious, good faith consideration before resorting to race-conscious measures; yet we speak with caution. In this dynamic environment, our conclusions should not be taken to mean that UT is immune from its obligation to recalibrate its dual systems of admissions as needed, and we cannot bless the university s race-conscious admissions program in perpetuity. Rather, much like judicial approval of a state s redistricting of voter districts, it is good only until the next census count it is more a process than a fixed structure that we review. The University s formal and informal review processes will confront the stark fact that the Top Ten Percent Law, although soon to be restricted to 75% of the incoming class, increasingly places at risk the use of race in admissions. In 1998, those admitted under the Top Ten Percent Law accounted for 41% of the Texas residents in the freshman class, while in 2008, top ten percent students comprised 81% of enrolled Texan freshmen.174 This trajectory evidences a risk of eroding the necessity of using race to achieve critical mass with accents that may, if persisted in, increasingly present as an effort to meet quantitative goals drawn from the demographics of race and a defiance of the now-demanded focus upon individuals when considering race. 173 Id. at 790 (opinion of Kennedy, J.). 174 2008 Top Ten Percent Report at 7 tbl.1a. 55 No. 09-50822 A university may decide to pursue the goal of a diverse student body, and it may do so to the extent it ties that goal to the educational benefits that flow from diversity. The admissions procedures that UT adopted, modeled after the plan approved by the Supreme Court in Grutter, are narrowly tailored procedures in some respects superior to the Grutter plan because the University does not keep a running tally of underrepresented minority representation during the admissions process. We are satisfied that the University s decision to reintroduce race-conscious admissions was adequately supported by the serious, good faith consideration required by Grutter. Finally, it is neither our role nor purpose to dance from Grutter s firm holding that diversity is an interest supporting compelling necessity. Nor are we inclined to do so. The role of black athletes in the southern universities forty years ago presents diversity s potential better than can we, although at that early juncture, it was ability overcoming a barrier of race.175 The judgment of the district court is AFFIRMED. 175 See David K. Wiggins & Patrick B. Miller, THE UNLEVEL PLAYING FIELD: A DOCUMENTARY HISTORY OF THE AFRICAN AMERICAN EXPERIENCE IN SPORT 443 (2003) (quoting Roy Wilkins, who wrote in the 1930s that black athletes carry more interracial education than all the erudite philosophy ever written on race (internal quotation marks omitted)). 56 No. 09-50822 KING, Circuit Judge, specially concurring: I concur in the judgment and in the analysis and application of Grutter in Judge Higginbotham s opinion. No party challenged, in the district court or in this court, the validity or the wisdom of the Top Ten Percent Law. We have no briefing on those subjects, and the district court did not consider them. Accordingly, I decline to join Judge Higginbotham s opinion insofar as it addresses those subjects. 57 No. 09-50822 EMILIO M. GARZA, Circuit Judge, specially concurring: Whenever a serious piece of judicial writing strays from fundamental principles of constitutional law, there is usually a portion of such writing where those principles are articulated, but not followed. So it goes in Grutter, where a majority of the Court acknowledged strict scrutiny as the appropriate level of review for race-based preferences in university admissions, but applied a level of scrutiny markedly less demanding. To be specific, race now matters in university admissions, where, if strict judicial scrutiny were properly applied, it should not. Today, we follow Grutter s lead in finding that the University of Texas s race-conscious admissions program satisfies the Court s unique application of strict scrutiny in the university admissions context. I concur in the majority opinion, because, despite my belief that Grutter represents a digression in the course of constitutional law, today s opinion is a faithful, if unfortunate, application of that misstep. The Supreme Court has chosen this erroneous path and only the Court can rectify the error. In the meantime, I write separately to underscore this detour from constitutional first principles. I The Equal Protection Clause of the Fourteenth Amendment provides that no State shall deny to any person within its jurisdiction the equal protection of the laws. U.S. CONST. amend. XIV. One of the Amendment s core principles is to do away with all governmentally imposed discriminations based on race, Palmore v. Sidoti, 466 U.S. 429, 432 (1984), and to create a Nation of equal citizens in a society where race is irrelevant to personal opportunity and achievement. Richmond v. J.A. Croson Co., 488 U.S. 469, 505 06 (1988). This is why [r]acial and ethnic distinctions of any sort are inherently suspect and . . . call for the most exacting judicial examination. Miller v. Johnson, 515 U.S. 900, 904 (1995) (quoting Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 291 (1978) (opinion of Powell, J.)). It matters not whether the racial preference is 58 No. 09-50822 characterized as invidious or benign: strict scrutiny applies regardless of the race of those burdened or benefitted by a particular classification. Shaw v. Reno, 509 U.S. 630, 650 51 (1993) (quoting Croson, 488 U.S. at 494). To survive such exacting scrutiny, laws classifying citizens on the basis of race must be narrowly tailored to achieving a compelling state interest. Miller, 515 U.S. at 904. In Grutter, the majority acknowledged these fundamental principles, see Grutter v. Bollinger, 539 U.S. 306, 326 27 (2003), but then departed and held, for the first time, that racial preferences in university admissions could be used to serve a compelling state interest. Id. at 328. Though the Court recognized that strict scrutiny should govern the inquiry into the use of race in university admissions, id. at 326, what the Court applied in practice was something else entirely. A The Grutter majority asserts that [s]trict scrutiny is not strict in theory, but fatal in fact. 539 U.S. at 326 (quoting Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 237 (1995)). But since the Court began applying strict scrutiny to review governmental uses of race in discriminating between citizens, the number of cases in which the Court has permitted such uses can be counted on one hand.1 The Court has rejected numerous intuitively appealing justifications 1 See Grutter, 539 U.S. at 328 (recognizing racial diversity in the context of higher education as compelling); Freeman v. Pitts, 503 U.S. 467, 494 (1992) (remedying the effects of past intentional discrimination a compelling governmental interest); Korematsu v. United States, 323 U.S. 214, 216 (1944) ( [P]ressing public necessity may sometimes justify the existence of [racial discrimination]; racial antagonism never can. ). In Fullilove v. Klutznick, 448 U.S. 448 (1980), the Court upheld a federal law that set aside public works monies for minority-owned businesses. Although Fullilove has not been expressly overruled, it is unlikely that its holding survives the Court s later Equal Protection decisions. See ERWIN CHEMERINSKY, CONSTITUTIONAL LAW: 59 No. 09-50822 offered for racial discrimination, such as remedying general societal discrimination, see Croson, 488 U.S. at 496 98 (plurality opinion); enhancing the number of minority professionals available to work in underserved minority communities, see Bakke, 438 U.S. at 310 11 (opinion of Powell, J.); and providing role models for minority students, see Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 275 76 (1986) (plurality opinion). In all of these cases, the Court found that the policy goals offered were insufficiently compelling to justify discrimination based on race. In those rare cases where the use of race properly furthered a compelling state interest, the Court has emphasized that the means chosen must work the least harm possible, Bakke, 438 U.S. at 308 (opinion of Powell, J.), and be narrowly tailored to fit the interest with greater precision than any alternative means. Grutter, 539 U.S. at 379 (Rehnquist, C.J., dissenting) (quotation omitted). Moreover, the failure to consider available race-neutral alternatives and employ them if efficacious would cause a program to fail strict scrutiny. See Wygant, 476 U.S. at 280 n.6 (plurality opinion) (the term narrowly tailored . . . requires consideration of whether lawful alternative and less restrictive means could have been used. ); see also Adarand, 515 U.S. at 237 38; Croson, 488 U.S. at 507; Fullilove v. Klutznick, 448 U.S. 448, 537 (1980) (Stevens, J., dissenting) ( Racial classifications are simply too pernicious to permit any but the most exact connection between justification and classification. ). Beyond the use of race-neutral alternatives, the Court, pre-Grutter, had considered several other factors in determining whether race-conscious programs were narrowly tailored. Programs employing a quota system would fail this inquiry, as would programs of unlimited duration. See Bakke, 438 U.S. at PRINCIPLES AND POLICIES § 9.3.5, at 738, 742 43 (3d ed. 2006). Korematsu s authority is likewise suspect. 60 No. 09-50822 315 18; Croson, 488 U.S. at 498. The Court looked to a program s flexibility and its capacity for individualized consideration. See United States v. Paradise, 480 U.S. 149, 177 (1987) (plurality opinion); Croson, 488 U.S. at 508. The Court also considered the relationship between the numerical goal and the percentage of minority group members in the relevant population, and whether the means chosen were likely to be overinclusive. See Croson, 488 U.S. at 506 10. Finally, the Court considered the program s burden on innocent third parties. See, e.g., Metro Broad., Inc. v. FCC, 497 U.S. 547, 630 (1990) (O Connor, J., dissenting) (programs should not unduly burden individuals who are not members of the favored racial and ethnic groups ); Bakke, 438 U.S. at 308 (opinion of Powell, J.). Grutter changed this. After finding that racial diversity at the University of Michigan Law School ( Law School ) was a compelling governmental interest, the Court redefined the meaning of narrow tailoring. See Grutter, 539 U.S. at 387 (Kennedy, J., dissenting) ( The Court, however, does not apply strict scrutiny. By trying to say otherwise, it undermines both the test and its own controlling precedents. ); see generally Ian Ayres & Sydney Foster, Don t Tell, Don t Ask: Narrow Tailoring After Grutter and Gratz, 85 TEX. L. REV. 517 (2007). The Court replaced narrow tailoring s conventional least restrictive means requirement with a regime that encourages opacity and is incapable of meaningful judicial review under any level of scrutiny. Courts now simply assume, in the absence of evidence to the contrary, that university administrators have acted in good faith in pursuing racial diversity, and courts are required to defer to their educational judgments on how best to achieve it. Grutter, 539 U.S. at 328 29. What is more, the deference called for in Grutter seems to allow universities, rather than the courts, to determine when the use of racial preferences is no longer compelling. See id. at 343 ( We take the Law School at its word that it would like nothing better than to find a race-neutral admissions formula and will terminate its race-conscious admissions program 61 No. 09-50822 as soon as practicable. ). This new species of strict scrutiny ensures that only those admissions programs employing the most heavy-handed racial preferences, and those programs foolish enough to maintain and provide conclusive data, will be subject to exacting judicial examination. Miller, 515 U.S. at 904. Others, like the University of Michigan in Grutter, and the University of Texas here, can get away with something less. B Setting aside for a moment Grutter s finding that racial diversity within the Law School was a compelling state interest, see infra Sections I.D and III, I find troubling the Court s treatment of whether the Law School s chosen means using race as a plus factor was narrowly tailored to achieving that end. The Court discussed five hallmarks of a narrowly tailored race-conscious admissions program in answering this question: (1) the absence of quotas; (2) a program that does not unduly harm any racial group; (3) serious, good-faith consideration of race-neutral alternatives; (4) a program that contains a sunset provision or some logical end point; and (5) individualized consideration of all applicants. See 539 U.S. at 335 43. The Court s opinion effectively emptied at least three of these criteria of their probative content, leaving the first and fifth as determinative in any narrow tailoring inquiry. See Ayres & Foster, 85 TEX. L. REV. at 543. First, Grutter defined a quota as reserving a fixed number or percentage of opportunities for certain minority groups, and insulating individuals from those groups from competition with all other candidates for available seats. Id. at 333 36. These prohibitions were clear well before Grutter. See Bakke, 438 U.S. at 317; Croson, 488 U.S. at 496. Only those programs with overt numerical set-asides or separate minority admissions tracks would fail this requirement. Next, the Court found that race-conscious admissions programs do not unduly burden innocent third parties so long as they provide individualized 62 No. 09-50822 consideration. Grutter, 539 U.S. at 341 ( [I]n the context of its individualized inquiry into the possible diversity contributions of all applicants, the Law School s race-conscious admissions program does not unduly harm nonminority applicants. ). Here, the Court collapsed the second narrow tailoring criterion into the fifth. Grutter also held that there were no workable race-neutral alternatives at the Law School, such as using a lottery system or decreasing the emphasis for all applicants on undergraduate GPA and LSAT scores. Id. at 340. The Court likewise rejected the United States argument that the Law School s plan was not narrowly tailored because race-neutral alternatives that had proven effective elsewhere (i.e., the percentage plans utilized in California, Florida, and Texas) were available and would deliver the educational benefits the Law School was seeking. Id. The Court held that [n]arrow tailoring does not require exhaustion of every conceivable race-neutral alternative. . . . Narrow tailoring does, however, require serious, good faith consideration of workable race-neutral alternatives that will achieve the diversity the university seeks. Id. at 339. After Grutter, universities are no longer required to use the most effective raceneutral means. So long as admissions officials have given serious, good faith consideration to such programs, they are free to pursue less effective alternatives that serve the interest about as well. Id. (citing Wygant, 476 U.S. at 280 n.6 (plurality opinion)). Thus, this third criterion is now essentially without meaning. Given the deference that universities educational judgments are to be afforded post-Grutter, serious, good faith consideration is a peculiarly low bar that will be satisfied in most every case. Compare id. at 339 (narrow tailoring require[s] serious, good faith consideration of workable race-neutral alternatives ), with id. at 329 ( [G]ood faith on the part of a university is presumed absent a showing to the contrary. ) (citation and internal quotation marks omitted). 63 No. 09-50822 Finally, while the Court acknowledged that race-conscious admissions programs must be limited in time, such as by sunset provisions or periodic reviews to determine whether the preferences remain necessary, the Court suspended application of this criterion for twenty-five years. Id. at 343 ( We expect that 25 years from now, the use of racial preferences will no longer be necessary to further the interest approved today. ). In doing so, the Grutter majority simply accepted the Law School s promise that it would terminate its race-conscious policies as soon as possible. See id. at 343 ( We take the Law School at its word that it would like nothing better than to find a race-neutral admissions formula and will terminate its race-conscious admissions program as soon as practicable. ). The Court s approval here is remarkable given the constitutional gravity of this experiment (i.e., the Law School s allocation of preferences along racial lines). This fourth criterion will now be considered satisfied with little or no showing on the part of university administrators, at least until 2028. And thus, all that truly remains of strict scrutiny s narrow tailoring inquiry post-Grutter is the requirement of individualized consideration. But what does this term mean specifically? Grutter never tells us. Moreover, the weight given to race as part of this individualized consideration is purposefully left undefined, making meaningful judicial review all but impossible. C In Grutter, the University of Michigan Law School sought to achieve a student body that was both academically strong and diverse along several dimensions, including race. There, the Court endorsed the Law School s highly individualized, holistic review of each applicant s file, giving serious consideration to all the ways an applicant might contribute to a diverse educational environment. Id. at 337. The Court noted approvingly that the Law School had no policy . . . of automatic acceptance or rejection based on any 64 No. 09-50822 single soft variable. Id. The Grutter majority permitted the use of race and ethnicity as plus factors within the Law School s holistic review, but this simply raises the question: how much of a plus?2 Grutter did not say. Instead, the Court implicitly forbade universities from quantifying racial preferences in their admissions calculus. Contrasting the admissions system found unconstitutional in Gratz, the Grutter majority noted that the Law School awards no mechanical, predetermined diversity bonuses based on race or ethnicity. Id. (citing Gratz v. Bollinger, 539 U.S. 244, 271 72 (2003)). On this view, rigid point systems that allocate preference points for racial and ethnic status are unconstitutional because they preclude admissions counselors from conducting the type of individualized consideration the Court s opinion in Grutter requires. Gratz, 539 U.S. at 277 (O Connor, J., concurring) (citation omitted). But it is not clear, to me at least, how using race in the holistic scoring system approved in Grutter is constitutionally distinct from the point-based system rejected in Gratz.3 If two applicants, one a preferred minority and one 2 The Court s discussion of race as a plus factor takes place in the context of strict scrutiny s narrow tailoring inquiry. Whether race should be considered at all is a separate, more fundamental, matter. See infra Section III. 3 Although I do not believe the government s use of race in university admissions can ever serve a compelling interest, assuming that it can, there is no reason why a well-designed point system could not account for an applicant s race, among other variables, and yet still provide meaningful, individualized consideration. See Ayres & Foster, 85 TEX. L. REV. at 566 70; see also Gratz, 539 U.S. at 295 (Souter, J., dissenting) ( [I]t is hard to see what is inappropriate in assigning some stated value to a relevant characteristic, whether it be reasoning ability, writing style, running speed, or minority race. Justice Powell s plus factors necessarily are assigned some values. The college simply does by a numbered scale what the law school accomplishes in its holistic review ; the distinction does not imply that applicants to the undergraduate college are denied individualized consideration . . . . (citation omitted)). 65 No. 09-50822 nonminority, with application packets identical in all respects save race would be assigned the same score under a holistic scoring system, but one gets a higher score when race is factored in, how is that different from the mechanical groupbased boost prohibited in Gratz? Although one system quantifies the preference and the other does not, the result is the same: a determinative benefit based on race. Grutter s new terminology like individualized consideration and holistic review tends to conceal this result. By obscuring the University of Michigan s use of race in these diffuse tests, the Court allowed the Law School to do covertly what the undergraduate program could not do overtly. See Gratz, 539 U.S. at 270 76. This much is clear and has been discussed elsewhere.4 I write separately to add my view, confirmed while deciding this appeal, that Grutter s narrow tailoring inquiry now reduced to testing for individualized consideration is incapable of meaningful judicial review. Traditionally, strict scrutiny required that the overall benefits of programs employing racial classifications justified the overall costs.5 See Ayres & Foster, 85 TEX. L. REV. at 526 & n.38. In Grutter, not only did the Court fail to conduct such an analysis, it rejected the only means for measuring the constitutionally relevant costs and benefits. Id. Although I disagree with the Court that raceconscious policies can ever serve a compelling interest in university admissions, by prohibiting race and ethnicity from being quantified at all, Grutter eliminated 4 See, e.g., Larry Alexander & Maimon Schwarzschild, Grutter or Otherwise: Racial Preferences and Higher Education, 21 CONST. COMMENT. 3 (2004); CHEMERINSKY, CONSTITUTIONAL LAW 744. 5 For example, a race-conscious admissions policy that added just one, three, or five members of a preferred minority group to an enrolling class of 6,700 would fail to be narrowly tailored. Such a program would have an intolerably high cost for little return. See infra Section II. 66 No. 09-50822 any chance for courts to critically evaluate whether race is, in fact, the defining feature of an admissions packet. Post-Grutter, there is no way to assess how much of a plus race gets as a plus-factor in any admissions system. And without the ability to measure the number of but-for admits (i.e., admitted minority students for whom race was the decisive factor), courts cannot meaningfully evaluate whether a university s use of race fits its asserted interest narrowly. See id. at 527 41, 575 82.6 In short, it is impossible to subject such uses of race to strict scrutiny. Grutter rewards admissions programs that remain opaque. Even assuming the Court s educational benefits of diversity justification holds true, see infra Section I.D, there are far more effective race-neutral means of screening for the educational benefits that states like Michigan and Texas ostensibly seek. To the degree that state universities genuinely desire students with diverse backgrounds and experiences, race-neutral factors like specific hardships overcome, extensive travel, leadership positions held, volunteer and work experience, dedication to particular causes, and extracurricular activities, among many other variables, can be articulated with specificity in the admissions essays.7 These markers for viewpoint diversity are far more likely 6 See also id. at 528 n.42 (citing, inter alia, WILLIAM G. BOWEN & DEREK BOK, THE SHAPE OF THE RIVER: LONG-TERM CONSEQUENCES OF CONSIDERING RACE IN COLLEGE AND UNIVERSITY ADMISSIONS 31 39 (1998)) (observing that the degree of racial preferences can be measured by examining the number of but-for admits and the qualification differentials between but-for admits and nonpreferred applicants who would have been admitted in the absence of affirmative action). 7 In addition to the two essays that UT requires as part of each application packet, the University considers several of the factors described above in determining an applicant s personal achievement score. See Fisher v. Univ. of Tex. at Austin, 645 F. Supp. 2d 587, 597 (W.D. Tex. 2009) ( The third [Personal Achievement Index] element is the personal achievement score, which is based on an evaluation of the file in its entirety by senior members of the admissions 67 No. 09-50822 to translate into enhanced classroom dialogue than a blanket presumption that race will do the same. Moreover, these markers represent the kind of life experiences that reflect industry. Race cannot. While race inevitably colors an individual s life and views, that facet of race and its impact on the individual can be described with some precision through an admissions essay. We should not presume that race shapes everyone s experiences in the same ways and award preference (or a bonus, or a plus ) accordingly. Such a policy, however labeled, is not narrowly tailored. Finally, the Court s unusual deference to educators academic judgments that racial diversity is a compelling interest, coupled with the deference allegedly owed to their determination of when the use of race is no longer necessary, see id. at 343, would appear to permit race-based policies indefinitely. For example, notwithstanding that a university s race-conscious policies had achieved 25% African-American and 25% Hispanic enrollment in the student body generally, that university could still justify the use of race in admissions if these minority students were disproportionately bunched in a small number of classes or majors. In fact, the majority s application of Grutter today reaches just such a result. Despite Top Ten Percent s demonstrable impact on minority enrollment at the University of Texas, the majority opinion holds that the University s use of race in admissions can be justified by reference to the paucity of minority students in certain majors: While the [Top Ten Percent] Law may have contributed to an increase in overall minority enrollment, those minority students remain clustered in certain programs, severely limiting the staff. The evaluators conduct a holistic review considering the applicant s demonstrated leadership qualities, extracurricular activities, awards and honors, work experience, service to the school or community, and special circumstances. ). 68 No. 09-50822 beneficial effects of educational diversity. For example, nearly a quarter of the undergraduate students in UT s College of Social Work are Hispanic, and more than 10% are African-American. In the College of Education, 22.4% of students are Hispanic and 10.1% are African-American. By contrast, in the College of Business Administration, only 14.5% of the students are Hispanic and 3.4% are African-American. It is evident that if UT is to have diverse interactions, it needs more minority students who are interested in and meet the requirements for a greater variety of colleges, not more students disproportionately enrolled in certain programs. Ante at 42. If this is so, a university s asserted interested in racial diversity could justify race-conscious policies until such time as educators certified that the elusive critical mass had finally been attained, not merely in the student body generally, but major-by-major and classroom-by-classroom. Given the large-scale absence of African-American and Hispanic students from thousands of classes at the University of Texas, Fisher, 645 F. Supp. 2d at 607, today s decision ratifies the University s reliance on race at the departmental and classroom levels, and will, in practice, allow for race-based preferences in seeming perpetuity. Such a use of race has no logical stopping point and is not narrowly tailored. See Croson, 488 U.S. at 498 (citing Wygant, 476 U.S. at 275). Allowing race-based social engineering at the university level is one thing, but not nearly as invasive as condoning it at the classroom level. I cannot accept that the Fourteenth Amendment permits this level of granularity to justify dividing students along racial lines. D The same imprecision that characterizes Grutter s narrow tailoring analysis casts doubt on its discussion of racial diversity as a compelling state interest. Grutter found that the Law School had a compelling interest in securing the educational benefits of a diverse student body, and that achieving a critical mass of racially diverse students was necessary to accomplish that 69 No. 09-50822 goal. Id. at 333. The Law School defined critical mass as a number that encourages underrepresented minority students to participate in the classroom and not feel isolated . . . or like spokespersons for their race. Id. at 318 19. The Court clarified: critical mass is defined by reference to the educational benefits that diversity is designed to produce. Id. at 330. Justice O Connor s majority opinion identified three such constitutionally relevant benefits: (i) increased perspective in the classroom; (ii) improved professional training; and (iii) enhanced civic engagement. Id. at 330 33. The first element is based on Justice Powell s focus in Bakke on the campus-level benefits of diversity. The second two are new.8 My difficulty with Grutter s educational benefits of diversity discussion is that it remains suspended at the highest levels of hypothesis and speculation. And unlike ordinary hypotheses, which must be testable to be valid, Grutter s thesis is incapable of testing. Justice O Connor s majority opinion rests almost entirely on intuitive appeal rather than concrete evidence. 1 8 See Robert C. Post, The Supreme Court, 2002 Term Forward: Fashioning the Legal Constitution: Culture, Courts, and Law, 117 HARV. L. REV. 4, 59 60 (2003) ( Although Grutter casts itself as merely endorsing Justice Powell s opinion in Bakke, Grutter s analysis of diversity actually differs quite dramatically from Powell s. Powell conceptualized diversity as a value intrinsic to the educational process itself. He regarded diversity as essential to the quality of higher education, because education was a practice of enlightenment, of speculation, experiment, and creation, that thrived on the robust exchange of ideas; characteristically provoked by confrontation between persons of distinct life experiences. . . . [Grutter] instead conceives of education as instrumental for the achievement of extrinsic social goods like professionalism, citizenship, or leadership. . . . Grutter s justifications for diversity thus potentially reach far more widely than do Powell s. ); see also Ayres & Foster, 85 TEX. L. REV. at 578 n.215 (citing commentary). 70 No. 09-50822 The first constitutionally relevant benefit that makes up Grutter s compelling interest is racial diversity s direct impact in the classroom. Here, the Court concluded that diverse perspectives improve the overall quality of education because classroom discussion is livelier, more spirited, and simply more enlightening and interesting when students have the greatest possible variety of backgrounds. Id. at 330 (internal quotation marks omitted). This rationale conforms to Justice Powell s opinion in Bakke that universities should pursue [t]he atmosphere of speculation, excitement and creation that is promoted by a diverse student body. 438 U.S. at 312 (opinion of Powell, J.).9 I question the validity of this surmise. Nonetheless, assuming a critical mass of minority students could perceptibly improve the quality of classroom learning, how would we measure success? By polling students and professors, as the University of Texas has done?10 How would we know whether the substantial social harm we are tolerating by dividing students based on race is worth the cost? That classroom discussion is, in fact, being enhanced? How can a party prove that it is? How can an opposing party prove that it is not? My concern with allowing viewpoint diversity s alleged benefits to justify racial preference is that viewpoint diversity is too theoretical and abstract. It cannot be proved or disproved. Sure, the Grutter majority cited to expert reports and amicus briefs from corporate employers as evidence that student body diversity improves educational outcomes and better prepares students for the 9 Justice Powell s opinion in Bakke conspicuously avoided claiming a categorical educational benefit of diversity, asserting only the potential for such benefits. See 438 U.S. at 314 (opinion of Powell, J.). 10 Every measure of social benefit or harm would be subjective and, at worst, capable of manipulation through framing biases. 71 No. 09-50822 workforce. Id. at 330. But this support can be easily manipulated.11 If all a university need do is find . . . report[s], studies, or surveys to implement a raceconscious admissions policy, the constraints of the Equal Protection Clause will, in effect, have been rendered a nullity. Croson, 488 U.S. at 504; see also J.E.B. v. Alabama ex rel. T.B., 511 U.S. 127, 139 n.11 (1994) ( [C]lassifications that rest on impermissible stereotypes violate the Equal Protection Clause, even when some statistical support can be conjured up for the generalizations. ). Grutter permits race-based preferences on nothing more than intuition the type that strict scrutiny is designed to protect against. See 539 U.S. at 327 ( Absent searching judicial inquiry into the justification for such race-based measures, we have no way to determine what classifications are benign or remedial and what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics. ) (citation and internal quotation marks omitted). Grutter and Bakke err by simply assuming that racial diversity begets greater viewpoint diversity. This inference is based on the assumption that members of minority groups, because of their racial status, are likely to have unique experiences and perspectives incapable of expression by individuals from outside that group. But as the Court has recognized elsewhere, the Constitution prohibits state decisionmakers from presuming that groups of individuals, whether classified by race, ethnicity, or gender, share such a quality collectively. See Miller, 515 U.S. at 914 (the Equal Protection Clause forbids the demeaning notion that members of the defined racial groups ascribe to certain minority views that must be different from those of other citizens. ) (citation omitted). There is no one African-American or Hispanic viewpoint,12 and, in fact, Grutter 11 See Alexander & Schwarzschild, 21 CONST. COMMENT. at 5 n.9 (criticizing the Court s undue reliance on amicus briefs from corporate employers). 12 For example, life experiences differ significantly if a Hispanic student s ethnicity originates in Mexico as opposed to Spain, or, for that matter, any of 72 No. 09-50822 approved the Law School s diversity rationale precisely because of the role that racial diversity can play in dispelling such falsehoods. See id. at 320 (citing expert testimony suggesting that when a critical mass of underrepresented minority students is present, racial stereotypes lose their force because nonminority students learn there is no minority viewpoint but rather a variety of viewpoints among minority students. ); and id. at 333 ( [D]iminishing the force of such stereotypes is a crucial part of the Law School s mission, and one that it cannot accomplish with only token numbers of minority students. ). Grutter sought to have it both ways. The Court held that racial diversity was necessary to eradicate the notion that minority students think and behave, not as individuals, but as a race. At the same time, the Court approved a policy granting race-based preferences on the assumption that racial status correlates with greater diversity of viewpoints. 2 Grutter s second asserted educational benefit of diversity relates to improved professional training. Here, Justice O Connor writes that diversity promotes cross-racial understanding, helps to break down racial stereotypes, and enables students to better understand persons of different races. Id. at 330 (internal quotation marks and brackets omitted). Such training is essential, the argument goes, for future leaders who will eventually work within and supervise a racially diverse workforce. Id. at 330 31. various Central and South American countries. Likewise, an African-American student whose roots come from Nigeria would be distinct in culture and ethnicity from a student whose ancestry originated in Egypt or Haiti. This same principle applies for students from non-preferred racial classes. For example, secondgeneration students from English, Irish, Scottish, or Australian ancestry would come with very different cultural experiences, and yet all of these students would be grouped together as White in racial classification systems like the one used at the University of Texas. 73 No. 09-50822 State universities are free to define their educational goals as broadly as needed to serve the public interest. We defer to educators professional judgments in setting those goals. Grutter, 539 U.S. at 328 ( Our holding today is in keeping with our tradition of giving a degree of deference to a university s academic decisions, within constitutionally proscribed limits. ). My concern, discussed throughout this opinion, is not that Grutter commands such deference, but that it conflated the deference owed to a university s asserted interest with deference to the means used to attain it. See id. at 388 (Kennedy, J., dissenting) ( The Court confuses deference to a university s definition of its educational objective with deference to the implementation of this goal. ). There is, however, one aspect of the Court s improved professional training rationale that I find especially troubling. While Grutter made much of the role that educational institutions play in providing professional training, see id. at 331 ( We have repeatedly acknowledged the overriding importance of preparing students for work and citizenship ), the cases the Court relied on involved primary and secondary schools. See id. (citing Plyer v. Doe, 457 U.S. 202, 221 (1982) (describing education as pivotal to sustaining our political and cultural heritage ) and ibid. (citing Brown v. Bd. of Educ., 347 U.S. 483, 493 (1954) ( education . . . is the very foundation of good citizenship. )). I question whether these cases apply with equal force in the context of higher education, where academic goals are vastly different from those pursued in elementary and secondary schools. Moreover, a university s self-styled educational goals, for example, promoting cross-racial understanding and enabling students to better understand persons of different races, could just as easily be facilitated in many other public settings where diverse people assemble regularly: in the workplace, in primary and secondary schools, and in social and community groups. See Grutter, 539 U.S. at 347 48 (Scalia, J., dissenting). I do not believe that the university has a monopoly on furthering these societal goals, or even 74 No. 09-50822 that the university is in the best position to further such goals. Notwithstanding an institution s decision to expand its educational mission more broadly, the university s core function is to educate students in the physical sciences, engineering, social sciences, business and the humanities, among other academic disciplines. 3 Finally, Grutter articulated a third benefit of racial diversity in higher education: enhancing civic engagement. Here, the Court wrote that: Effective participation by members of all racial and ethnic groups in the civic life of our Nation is essential if the dream of one Nation, indivisible, is to be realized. ... In order to cultivate a set of leaders with legitimacy in the eyes of the citizenry, it is necessary that the path to leadership be visibly open to talented and qualified individuals of every race and ethnicity. All members of our heterogeneous society must have confidence in the openness and integrity of educational institutions that provide this training. . . . Access to [higher] education . . . must be inclusive of talented and qualified individuals of every race and ethnicity, so that all members of our heterogeneous society may participate in the educational institutions that provide the training and education necessary to succeed in America. Id. at 332 33. Unlike the first two educational benefits of diversity, which focused on improving classroom discussion and professional training, this third claimed benefit plainly has nothing to do with the university s core education and training functions. Instead, Grutter is concerned here with role that higher education plays in a democratic society, and the Court suggests that affirmative action at public universities can advance a societal goal of encouraging minority participation in civic life.13 This proposition lacks foundation. 13 See Lani Guinier, The Supreme Court, 2002 Term Comment: Admissions Rituals as Political Acts: Guardians at the Gates of Our Democratic 75 No. 09-50822 If a significant portion of a minority community sees our nation s leaders as illegitimate or lacks confidence in the integrity of our educational institutions, as Grutter posits in the block quote above, see id., 539 U.S. at 332, I doubt that suspending the prevalent constitutional rules to allow preferred treatment for as few as 15-40 students, see infra Section II, is likely to foster renewed civic participation from among that community as a whole.14 Grutter replaced Bakke s emphasis on diversity in educational inputs with a new emphasis on diversity in educational outputs. By expanding Justice Powell s original viewpoint diversity rationale to include diversity s putative benefits in the workforce and beyond (i.e., inspiring a sense of civic belonging in discouraged minority communities), the Court has endorsed a compelling interest without bounds. Post-Grutter, it matters little whether racial preferences in university admissions are justified by reference to their potential for improved discussion in individual classrooms, or even at the university generally. Now such preferences can be justified based on their global impact. By removing the focus of attention from diversity s educational value at the campus level, the Court has ensured that the educational benefits of diversity will accommodate all university affirmative action plans as compelling. E Ideals, 117 HARV. L. REV. 113, 174 76 (2003). 14 This is not to criticize universities, like the University of Texas, for implementing policies that seek to increase minority representation, not merely for its educational benefits on campus, but also for the secondary benefits that such increases in minority enrollment can have in the workplace and in society generally. A university degree confers professional and leadership opportunities unavailable otherwise, and ensuring that all segments of society have meaningful access to public institutions of higher education represents a paramount government objective. Grutter, 539 U.S. at 331 32 (citing Brief of United States as Amicus Curiae 13). I do not question this goal, but rather the constitutionality of using race to attain it. 76 No. 09-50822 Finally, by using metaphors, like critical mass, and indefinite terms that lack conceptual or analytical precision, but rather sound in abject subjectivity, to dress up constitutional standards, Grutter fails to provide any predictive value to courts and university administrators tasked with applying these standards consistently. And notwithstanding the Court s nod to federalism, Grutter s ambiguity discourages States from experimenting or departing from the one accepted norm. See id. at 342 (citing United States v. Lopez, 514 U.S. 549, 581 (1995) (Kennedy, J., concurring) ( [T]he States may perform their role as laboratories for experimentation to devise various solutions where the best solution is far from clear. )). In the absence of clear guidance, public universities nationwide will simply model their programs after the one approved in Grutter rather than struggle with the risks and uncertain benefits of experimentation. That is exactly what has occurred here. With one exception the Top Ten Percent law the race-conscious admissions policy that we review today is identical to the program used at the Law School. II As mentioned at the outset, I concur in the opinion because I believe today s decision is a faithful application of Grutter s teachings, however flawed I may find those teachings to be. I am compelled to follow the Court s unusual deference towards public university administrators in their assessment that racial diversity is a compelling interest, as well as the Court s refashioned narrow-tailoring inquiry. See 28 U.S.C. § 453. My difficulty is not necessarily with today s decision, but with the one that drives it. Nonetheless, there is one aspect of Judge Higginbotham s thoughtful opinion that gives me pause about whether Grutter compels the result we reach today. Ultimately, and regrettably, 77 No. 09-50822 I recognize that the deference called for by Grutter may make this concern superfluous. As today s opinion notes, the University of Texas s race-conscious admissions policy is nearly indistinguishable from the program approved by the Supreme Court in Grutter.15 Ante at 2, 4, 25. As such, the majority opinion summarily finds that, like the Law School in Grutter, the University of Texas has a compelling interest in obtaining the educational benefits of diversity in its undergraduate program. Id. at 26. After affording all deference due, today s decision focuses on the efficacy of the University s race-conscious admissions policy. Id. at 29 ( [W]hile we focus on the University s decision to adopt a Grutter-like plan, admissions outcomes remain relevant evidence of the plan s necessity a reality check. ). In my view, the efficacy of the University s racebased admissions policy is more than merely relevant, it is dispositive. The plaintiffs here argue that the University of Texas s interest in obtaining a racially diverse student body is not compelling because the University has already achieved critical mass by way of Texas s Top Ten Percent law. See TEX. EDUC. CODE § 51.803 (1997). The University disagrees. This claim is difficult to evaluate. The University refuses to assign a weight to race or to maintain conclusive data on the degree to which race factors into admissions decisions and enrollment yields. See Fisher, 645 F. Supp. 2d at 608 09 ( At no point in the process is race considered individually or given a numerical value; instead, the file is evaluated in its entirety in order to provide a better understanding of the student as a person and place her achievements in context. ). Whether the University of Texas s use of race is narrowly tailored turns on whether its chosen means using race as a plus factor in the 15 As a result, UT s policy suffers from all the same defects as the Law School policy evaluated in Grutter and discussed previously in this opinion. See supra Section I. 78 No. 09-50822 University s holistic scoring system are effective, not just in theory, but also in practice. If, apart from the Top Ten Percent law, the University of Texas s raceconscious admissions program added just three-to-five African-American students, or five-to-ten Hispanic students, to an entering freshman class of 6,700, that policy would completely fail to achieve its aims and would not be narrowly tailored. See Ayres & Foster, 85 TEX. L. REV. at 523 n.27 ( At least as a theoretical matter, narrow tailoring requires not only that the preferences not be too large, but also that they not be too small so as to fail to achieve the goals of the relevant compelling government interest. ). The marginal benefit of adding just five or ten minority students to a class of this size would be negligible and have no perceptible impact on the educational benefits that diversity is designed to produce. Grutter, 539 U.S. at 330 ( [C]ritical mass is defined by reference to the educational benefits that diversity is designed to produce. ).16 This is especially so, if, as the district court suggests, the largescale absence of African-American and Hispanic students from thousands of classes indicates UT has not reached sufficient critical mass for its students to benefit from diversity and illustrates UT s need to consider race as a factor in admissions in order to achieve those benefits. Fisher, 645 F. Supp. 2d at 607 16 See Bakke, 438 U.S. at 316 (opinion of Powell, J.) (noting the necessity of including more than a token number of black students ). See also Patricia Gurin et al., Diversity and Higher Education: Theory and Impact on Educational Outcomes, 72 HARV. EDUC. REV. 330, 360 61 (2002) (enrolling significant numbers of students of various groups is necessary to enable students to perceive differences both within groups and between groups ); Kathryn R.L. Rand & Steven Andrew Light, Teaching Race Without a Critical Mass: Reflections on Affirmative Action and the Diversity Rationale, 54 J. LEGAL. EDUC. 316, 332 34 (2004) (noting that under a cost-benefit analysis it may be more difficult to justify an affirmative action program when a university is unable to enroll a critical mass of minority applicants). 79 No. 09-50822 (citing statistics showing that in 2002, the University offered over 5,631 classes, 79% of which (4,448) had just one or zero African-American students; 30% of classes (1,689) had zero or one Hispanic students).17 So, the controlling question is, Is the University of Texas s race-conscious policy effective? And by effective, I do not mean that every statistically insignificant gain (i.e., adding one, three, or five students at the margin) qualifies. The constitutional inquiry for me concerns whether the University s program meaningfully furthers its intended goal of increasing racial diversity on the road to critical mass. I find it does not. In the 2008 admissions cycle, 29,501 students applied to the University of Texas. See Fisher, 645 F. Supp. 2d at 590. Less than half, 12,843, were admitted and 6,715 ultimately enrolled.18 Id. Of these enrolled students, 6,322 came from Texas high schools.19 See Implementation and Results of the Texas 17 These statistics represent all classes at UT with five or more students, including large lecture courses. For classes with five to 24 students the most likely to foster the vibrant discussion described in Grutter and Bakke the figures are more revealing. In 2002, UT offered 3,616 classes with five to 24 students. Of these, 90% had one or zero African-American students and 43% had one or zero Hispanic students. See Proposal to Consider Race and Ethnicity in Admissions, June 25, 2004 at 26, Table 8. 18 Today s decision, like the district court s, alternates between using statistics from admitted and enrolled students. If realizing the educational benefits of diversity is the University s asserted interest, only the data for enrolled students is relevant to our review. 19 In the discussion that follows, I use the number of enrolled Texas residents (6,322) as a baseline rather than the aggregate enrollment for firsttime freshman (6,715). There are two reasons for this. First, this case asks us to decide the necessity of UT s race-conscious admissions policy in light of Texas s Top Ten Percent law. I find this question is evaluated most effectively by comparing enrollment data for Texas residents, which include precise figures for Top 10% and Non-Top 10% enrollees. Second, as the majority opinion recognizes, ante at 44 n.155, the record does not include data showing what portion of the total applicant pool were Texas residents and what portion came from out-of-state. This is problematic. We know, for example, that the 2008 80 No. 09-50822 Automatic Admissions Law (HB 588) at the University of Texas at Austin, October 28, 2008 at 7 ( 2008 Top Ten Percent Report ). 5,114 (80.9% of enrolled Texas residents) of these students were a product of Top Ten Percent, meaning that, at most, 1,208 (19.1%) enrolled non-Top Ten Percent Texas residents had been evaluated on the basis of their AI/PAI scores. Id. Of the 363 African-American freshmen from Texas high schools that were admitted and enrolled (6% of the 6,322-member enrolling class from Texas high schools), 305 (4.8%) were a product of Top Ten Percent, while 58 (0.92%) AfricanAmerican enrollees had been evaluated on the basis of their AI/PAI scores.20 See 2008 Top Ten Percent Report at 7. For the 1,322 (21%) total enrolled in-state Hispanic students, 1,164 (18.4% of enrolled in-state students) were a product of Top Ten Percent, while 158 (2.5%) had been evaluated on the basis of their AI/PAI scores. Id. We know that in some cases an applicants AI score is high enough that the applicant is granted admission based on that score alone. But we do not have data to show how many of these 58 African-American and 158 Hispanic students were admitted automatically based on their AI scores, which are race-neutral, and how many were admitted after factoring in the students PAI scores, which use the University s Grutter-like holistic evaluation plan and entering freshman class included 375 African-American and 1,338 Hispanic students, and that 363 and 1,322 of these students, respectively, were Texas residents. See 2008 Top Ten Percent Report at 6 7. So, although we know that the 2008 enrolling freshman class included 12 African-American and 16 Hispanic students from out-of-state, we cannot intelligently discuss the potential impact of UT s race-conscious policy on this data set without also having total application and admissions information available for non-Texas residents. This does not affect my conclusions the number of non-Texas African-American and Hispanic students enrolled in the freshman class is statistically insignificant. 20 In this section, I often refer to a raw number followed by a percentage listed in parentheses. E.g., 305 (4.8%). This percentage figure (__%) is calculated by dividing number of students cited by 6,322, the number of enrolled Texas residents in the 2008 freshman class. 81 No. 09-50822 include consideration of an applicant s race as one of seven special circumstances. Nonetheless, assuming that all 58 and 158 African-American and Hispanic students, respectively, were admitted on the basis of their combined AI and PAI scores (i.e., that none of these minority students gained admission on the basis of their race-neutral Academic Index score alone), the question is whether the University s use of race, which is a highly suspect tool, Croson, 488 U.S. at 493, as part of the PAI score contributes a statistically significant enough number of minority students to affect critical mass at the University of Texas. We do not know, because the University does not maintain data, the degree to which race influenced the University s admissions decisions for any of these enrolled students or how many of these students would not have been admitted but-for the use of race as a plus factor. But assuming the University gave race decisive weight in each of these 58 African-American and 158 Hispanic students admissions decisions, those students would still only constitute 0.92% and 2.5%, respectively, of the entire 6,322-person enrolling in-state freshman class. And this is assuming a 100%, unconstitutional use of race, not as a plus factor, but as a categorical condition for guaranteed admission. See Grutter, 539 U.S. at 329 30 (making race an automatic factor in admissions would amount to outright racial balancing, which is patently unconstitutional. ). Assume further, that such a prohibited use of race was employed in only half of the University s admissions decisions. This would still only yield 29 (0.46%) African-American and 79 (1.25%) Hispanic students. Now assume that the University s use of race is truly holistic; that given the multitude of other race-neutral variables the University considers and values sincerely, race s significance is limited in any individual application packet. See Fisher, 645 F. Supp. 2d at 608 ( UT considers race in its admissions process as a factor of a factor of a factor of a factor. As described in exhaustive 82 No. 09-50822 detail above, race is one of seven special circumstances, which is in turn one of six factors that make up an applicants personal achievement score. ). Lastly, assume that in this system, the University s use of race results in a but-for offer of admission in one-quarter of the decisions. A twenty-five percent but-for admissions rate seems highly improbable if race is truly limited in its holistic weighting, but the unlikelihood of the assumption proves my point. Even under such a system, the University s proper use of race holistically would only yield 15 (0.24%) African-American and 40 (0.62%) Hispanic students. African- American students, for example, admitted and enrolled by way of this holistic system would still only constitute two-tenths of one percent of the University of Texas s 2008 entering freshman class. Such a use of race could have no discernable impact on the classroom-level educational benefits diversity is designed to produce or otherwise influence critical mass at the University of Texas generally. Such a plan exacts a cost disproportionate to its benefit and is not narrowly tailored. This is especially so on a university campus with, for example, 4,448 classes (out of 5,631) with zero or one African-American students, and 1,689 classes with zero or one Hispanic students. Fisher, 645 F. Supp. 2d at 607. More importantly, if the figures above are reasonably accurate, the University s use of race also fails Grutter s compelling interest test as a factual matter. See 539 U.S. at 333 ( [D]iminishing the force of [racial] stereotypes is both a crucial part of the Law School s mission, and one that it cannot accomplish with only token numbers of minority students. ). From its inception immediately following Grutter, the University s race-conscious admissions policy was described as essential to the University of Texas s educational mission: [T]o accomplish [UT s] mission and fulfill its flagship role . . . the undergraduate experience for each student must include classroom contact with peers of differing racial, ethnic, and cultural backgrounds. The proposal to consider race in the admissions 83 No. 09-50822 process is not an exercise in racial balancing but an acknowledgment that significant differences between the racial and ethnic makeup of the University s undergraduate population and the state s population prevent the University from fully achieving its mission. Fisher, 645 F. Supp. 2d at 602 (citing Proposal to Consider Race and Ethnicity in Admissions, June 25, 2004 at 24). If the University s use of race is truly necessary to accomplish its educational function, then as a factual matter, the University of Texas s race-conscious measures have been completely ineffectual in accomplishing its claimed compelling interest. In contrast, Top Ten Percent was responsible for contributing 305 and 1,164 African-American and Hispanic students, respectively, to the entering 2008 freshman class using entirely race-neutral means. These students represent 4.8% and 18.4% of the entering in-state freshman class. In addition, of the 58 African-American and 158 Hispanic enrolled students evaluated on the basis of their AI and PAI scores, if the University s use of race was truly holistic, the percentage of these students for whom race was a decisive factor (i.e., but-for admits) should be minimal. In other words, the vast majority of these 58 and 158 students were admitted based on objective factors other than race. That is, the University was able to obtain approximately 96% of the African-American and Hispanic students enrolled in the 2008 entering in-state freshman class using race-neutral means. And although the University argues that this number still does not qualify as critical mass, one thing is certain: the University of Texas s use of race has had an infinitesimal impact on critical mass in the student body as a whole. As such, the University s use of race can be neither compelling nor narrowly tailored. I do not envy the admissions officials at the University of Texas. In 1997, in response to our decision in Hopwood v. Texas, 78 F.3d 932 (5th Cir. 1996), the people of the State of Texas determined, through their elected representatives, 84 No. 09-50822 that something needed to be done to improve minority enrollment at Texas s public institutions of higher education. Texas s Top Ten Percent law was intended to effectuate that desire. We take no position today on the constitutionality of that law.21 Instead, we are asked to scrutinize the legality of the University s race-conscious policy designed to complement Top Ten Percent. Even with the limited data available, I cannot find that the University of Texas s use of race is narrowly tailored where the University s highly suspect use of race provides no discernable educational impact. In my view, the University s program fails strict scrutiny before or after Grutter. See, e.g., Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 790 (2007) (Kennedy, J., concurring) ( [I]ndividual racial classifications employed in this manner may be considered legitimate only if they are a last resort to achieve a compelling interest. ) (citation omitted) (emphasis added). Before Grutter, it is unlikely that the Supreme Court would have found that the University of Texas s means were narrowly tailored to the interest it asserts. Nonetheless, narrow tailoring in the university admissions context is not about balancing constitutional costs and benefits any longer. Post-Grutter, universities need not inflict the least harm possible so long as they operate in good faith. And in assessing good faith, institutions like the University of Texas need not even provide the type of metrics that allow courts to review their affirmative action programs. As long 21 In assessing whether the University s use of race is narrowly tailored, today s majority opinion finds that Top Ten Percent is not a race-neutral alternative that serves the University s asserted interest about as well as its Grutter-like plan. See ante at 39 46. My concurrence should not be read to approve or reject the constitutionality of percentage plans like Top Ten Percent. That issue remains open. I write separately to underscore the minimal effect that the University s use of race has had on critical mass in light of Top Ten Percent, and why the University s use of race would not, therefore, be narrowly tailored applying traditional strict scrutiny principles before Grutter. I recognize that Grutter appears to swallow this concern. 85 No. 09-50822 as these public institutions remain sufficiently opaque in their use of race, reviewing courts like ours will be hard pressed to find anything short of good faith and narrow tailoring. In the world post-Grutter, courts are enjoined to take universities at their word. III The Supreme Court s narrow tailoring jurisprudence has been reliably tethered, at least before 2003, to the principle that whenever the government divides citizens by race, which is itself an evil that can only be justified in the most compelling circumstances, that the means chosen will inflict the least harm possible, see Bakke, 438 U.S. at 308 (opinion of Powell, J.), and fit the compelling goal so closely that there is little or no possibility that the motive for the classification was illegitimate racial prejudice or stereotype. Croson, 488 U.S. at 493; see also Kimel v. Fla. Bd. of Regents, 528 U.S. 62, 84 (2000) ( [W]hen a State discriminates on the basis of race . . . , we require a tighter fit between the discriminatory means and the legitimate ends they serve. ). Grutter abandoned this principle and substituted in its place an amorphous, untestable, and above all, hopelessly deferential standard that ensures that race-based preferences in university admissions will avoid meaningful judicial review for the next several decades. My disagreement with Grutter is more fundamental, however. Grutter s failing, in my view, is not only that it approved an affirmative action plan incapable of strict scrutiny, but more importantly, that it approved the use of race in university admissions as a compelling state interest at all. The idea of dividing people along racial lines is artificial and antiquated. Human beings are not divisible biologically into any set number of races.22 A 22 See Alexander & Schwarzschild, 21 CONST. COMMENT. at 6 & n.10 ( There is broad scholarly support for this proposition. See, e.g., NAOMI ZACK, PHILOSOPHY OF SCIENCE AND RACE 58 62 (2002); JOSEPH L. GRAVES, JR., THE EMPEROR S NEW 86 No. 09-50822 world war was fought over such principles. Each individual is unique. And yet, in 2010, governmental decisionmakers are still fixated on dividing people into white, black, Hispanic, and other arbitrary subdivisions. The University of Texas, for instance, segregates student admissions data along five racial classes. See, e.g., 2008 Top Ten Percent Report at 6 (reporting admissions data for White, Native-American, African-American, Asian-American, and Hispanic students). That is not how society looks any more, if it ever did. When government divides citizens by race, matters are different.23 Government-sponsored discrimination is repugnant to the notion of human equality and is more than the Constitution can bear. See Grutter, 539 U.S. at 388 (Kennedy, J., dissenting) ( Preferment by race, when resorted to by the State, can be the most divisive of all policies, containing within it the potential to destroy confidence in the Constitution and the idea of equality. ). There are no de minimis violations of the Equal Protection Clause, and when government undertakes any level of race-based social engineering, the costs are enormous. Not only are race-based policies inherently divisive, they reinforce stereotypes that groups of people, because of their race, gender, or ethnicity, think alike or have common life experiences. The Court has condemned such class-based presumptions repeatedly. See, e.g., United States v. Virginia, 518 U.S. 515, 533 (1996) ( Supposed inherent differences are no longer accepted as a ground for CLOTHES: BIOLOGICAL THEORIES OF RACE AT THE MILLENNIUM (2001); Joshua M. Glasgow, On the New Biology of Race, 100 J. PHIL. 456 (2003). ). 23 See Alexander & Schwarzschild, 21 CONST. COMMENT. at 6 7 ( [W]hen the government classifies people racially and ethnically, and then makes valuable entitlements such as admission to a university turn on those classifications, . . . that very fact encourages people to think that races are real categories, not bogus ones, and that one s race is an exceedingly important rather than a superficial fact about oneself and others. In other words, it encourages people to pay close attention to race and to think in racial terms. ). 87 No. 09-50822 race or national origin classifications. ); Shaw, 509 U.S. at 647 (rejecting the notion that members of the same racial group regardless of their age, education, economic status, or the community in which they live think alike, share the same . . . interests, or have a common viewpoint about significant issues); Wygant, 476 U.S. at 316 (Stevens, J., dissenting) (the premise that differences in race, or the color of a person s skin, reflect real differences . . . is utterly irrational and repugnant to the principles of a free and democratic society ). I do not see how racial discrimination in university admissions is any less repugnant to the Constitution. If anything, government-sponsored discrimination in this context presents an even greater threat of long-term harm.24 24 Professor Cohen succinctly describes some of the effects of racial and ethnic preferences in higher education: 1. preference divides the society in which it is awarded; 2. it establishes a precedent in excusing admitted racial discrimination to achieve political objectives; 3. it corrupts the universities in which it is practiced, sacrificing intellectual values and creating pressures to discriminate by race in grading and graduation; 4. . . . 5. it obscures the real social problem of why so many minority students are not competitive academically; 6. it obliges a choice of some few ethnic groups, which are to be favored above all others; 7. . . . 8. it removes incentives for academic excellence and encourages separatism among racial and ethnic minorities; 9. it mismatches students and institutions, increasing the likelihood of failure for many minority students; and 10. it injures race relations over the long haul. CARL COHEN & JAMES P. STERBA, AFFIRMATIVE ACTION & RACIAL PREFERENCE 109 (2003). 88 No. 09-50822 For the most part, college admissions is a zero-sum game. Whenever one student wins, another loses. The entire competition, encouraged from age five on, is premised on individual achievement and promise.25 It is no exaggeration to say that the college application is 18 years in the making and is an unusually personal experience: the application presents a student s best self in the hopes that her sustained hard work and experience to date will be rewarded with admission. Race-based preferences break faith with this expectation by favoring a handful of students based on a trait beyond the control of all. See Bakke, 438 U.S. at 361 (opinion of Brennan, White, Marshall & Blackmun, JJ.) ( [A]dvancement sanctioned, sponsored, or approved by the State should ideally be based on individual merit or achievement, or at least on factors within the control of the individual . . . . ). Given the highly personal nature of the college admissions process, this kind of class-based discrimination poses an especially acute threat of resentment and its corollary entitlement. More fundamentally, it assures that race will always be relevant in American life, and that the ultimate goal of eliminating entirely from governmental decisionmaking such irrelevant factors as a human being s race will never be achieved. Croson, 488 U.S. at 495 (citation and internal quotation marks omitted). Yesterday s racial discrimination was based on racial preference; today s racial preference results in racial discrimination. Changing the color of the group discriminated against simply inverts, but does address, the fundamental problem: the Constitution prohibits all forms of government-sponsored racial 25 For example, in the School of Architecture, the School of Fine Arts, and certain honors programs, where aptitude is essential, the University requires special portfolio, audition, and other requirements. See ante at 23 n.87. In these and other impacted programs where student demand outstrips available space, the University recognizes and uses merit as the decisive consideration in admission. I do not see why excellence and merit warrant less consideration in the University s other disciplines. 89 No. 09-50822 discrimination. Grutter puts the Supreme Court s imprimatur on such ruinous behavior and ensures that race will continue to be a divisive facet of American life for at least the next two generations. Like the plaintiffs and countless other college applicants denied admission based, in part, on government-sponsored racial discrimination, I await the Court s return to constitutional first principles. 90