Fisher v. University of Texas at Austin, 579 U.S. __

NOTICE: This opinion is subject to formal revision beforepublication in the preliminary print of the United StatesReports. Readers are requested to notify the Reporter of Decisions,Supreme Court of the United States, Washington, D.C. 20543,of any typographical or other formal errors, in order thatcorrections may be made before the preliminary print goes topress.SUPREME COURT OF THE UNITED STATES_________________No. 14–981_________________ABIGAIL NOEL FISHER, PETITIONER v. UNIVERSITY OF TEXAS ATAUSTIN, etal.on writ of certiorari to the united states court of appeals forthe fifth circuit[June 23, 2016]Justice Kennedy delivered the opinion of the Court.The Court is asked once again to consider whether therace-conscious admissions program at the University of Texas islawful under the Equal Protection Clause.IThe University of Texas at Austin (or University) relies upon acomplex system of admissions that has undergone significantevolution over the past two decades. Until 1996, the Universitymade its admissions decisions primarily based on a measure called“Academic Index” (or AI), which it calculated by combining anapplicant’sSAT score and academic performance in high school. Inassessing applicants, preference was given to racialminorities.In 1996, the Court of Appeals for the Fifth Circuit invalidatedthis admissions system, holding that any consideration of race incollege admissions violates the Equal Protection Clause. SeeHopwood v. Texas, 78 F. 3d 932, 934–935, 948.One year later the University adopted a new admissions policy.Instead of considering race, the University began making admissionsdecisions based on an applicant’s AI and his or her “PersonalAchievement Index” (PAI). The PAI was a numerical score based on aholistic review of an application. Included in the number were theapplicant’s essays, leadership and work experience, extracurricularactivities, community service, and other “special characteristics”that might give the admissions committee insight into a student’sbackground. Consistent with Hopwood, race was not aconsideration in calculating an applicant’s AI or PAI.The Texas Legislature responded to Hopwood as well. Itenacted H. B. 588, commonly known as the Top Ten Percent Law. Tex.Educ. Code Ann. §51.803 (West Cum. Supp. 2015). As its namesuggests, the Top Ten Percent Law guarantees college admission tostudents who graduate from a Texas high school in the top 10percent of their class. Those students may choose to attend any ofthe public universities in the State.The University implemented the Top Ten Percent Law in 1998.After first admitting any student who qualified for admission underthat law, the University filled the remainder of its incomingfreshman class using a combination of an applicant’s AI and PAIscores—again, without considering race.The University used this admissions system until 2003, when thisCourt decided the companion cases of Grutter v.Bollinger,539 U.S. 306, and Gratz v.Bollinger,539 U.S. 244. In Gratz, this Courtstruck down the University of Michigan’s undergraduate system ofadmissions, which at the time allocated predetermined points toracial minority candidates. See 539 U.S., at 255, 275–276. InGrutter, however, the Court upheld the University ofMichigan Law School’s system of holistic review—a system that didnot mechanically assign points but rather treated race as arelevant feature within the broader context of a candidate’sapplication. See 539 U.S., at 337, 343–344. In upholding thisnuanced use of race, Grutter implicitly overruledHopwood’s categorical prohibition.In the wake of Grutter, the University embarked upon ayear-long study seeking to ascertain whether its admissions policywas allowing it to provide “the educational benefits of a diversestudent body...to all of the University’sundergraduate students.” App. 481a–482a (affidavit of N. BruceWalker ¶11 (Walker Aff.)); see also id., at 445a–447a. TheUniversity concluded that its admissions policy was not providingthese benefits. Supp. App. 24a–25a.To change its system, the University submitted a proposal to theBoard of Regents that requested permission to begin taking raceinto consideration as one of “the many ways in which [an]academically qualified individual might contribute to, and benefitfrom, the rich, diverse, and challenging educational environment ofthe Univer-sity.” Id., at 23a. After the board approved theproposal, the University adopted a new admissions policy toimplement it. The University has continued to use that admissionspolicy to this day.Although the University’s new admissions policy was a directresult of Grutter, it is not identical to the policy thisCourt approved in that case. Instead, consistent with the State’slegislative directive, the University continues to fill asignificant majority of its class through the Top Ten Percent Plan(or Plan). Today, up to 75 percent of the places in the freshmanclass are filled through the Plan. As a practical matter, this 75percent cap, which has now been fixed by statute, means that, whilethe Plan continues to be referenced as a “Top Ten Percent Plan,” astudent actually needs to finish in the top seven or eight percentof his or her class in order to be admitted under thiscategory.The University did adopt an approach similar to the one inGrutter for the remaining 25 percent or so of the incomingclass. This portion of the class continues to be admitted based ona combination of their AI and PAI scores. Now, however, race isgiven weight as a subfactor within the PAI. The PAI is a numberfrom 1 to 6 (6 is the best) that is based on two primarycomponents. The first component is the average score a reader givesthe applicant on two required essays. The second component is afull-file review that results in another 1-to-6 score, the“Personal Achievement Score” or PAS. The PAS is determined by aseparate reader, who (1) rereads the applicant’s required essays,(2) reviews any supplemental information the applicant submits(letters of recommendation, resumes, an additional optional essay,writing samples, artwork, etc.), and (3) evaluates the applicant’spotential contributions to the University’s student body based onthe applicant’s leadership experience, extracurricular activities,awards/honors, community service, and other “specialcirc*mstances.”“Special circ*mstances” include the socioeconomic status of theapplicant’s family, the socioeconomic status of the applicant’sschool, the applicant’s family responsibilities, whether theapplicant lives in a single-parent home, the applicant’s SAT scorein relation to the average SAT score at the applicant’s school, thelanguage spoken at the applicant’s home, and, finally, theapplicant’s race. See App. 218a–220a, 430a.Both the essay readers and the full-file readers who assignapplicants their PAI undergo extensive training to ensure that theyare scoring applicants consistently. Deposition of Brian Breman9–14, Record in No. 1: 08–CV–00263, (WD Tex.), Doc. 96–3. TheAdmissions Office also undertakes regular “reliability analyses” to“measure the frequency of readers scoring within one point of eachother.” App. 474a (affidavit of Gary M. Lavergne ¶8); see alsoid., at 253a (deposition of Kedra Ishop (Ishop Dep.)). Boththe intensive training and the reliability analyses aim to ensurethat similarly situated applicants are being treated identicallyregardless of which admissions officer reads the file.Once the essay and full-file readers have calculated eachapplicant’s AI and PAI scores, admissions officers from each schoolwithin the University set a cutoff PAI/AI score combination foradmission, and then admit all of the applicants who are above thatcutoff point. In setting the cutoff, those admissions officers onlyknow how many applicants received a given PAI/AI score combination.They do not know what factors went into calculating thoseapplicants’ scores. The admissions officers who make the finaldecision as to whether a particular applicant will be admitted makethat decision without knowing the applicant’s race. Race enters theadmissions process, then, at one stage and one stage only—thecalculation of the PAS.Therefore, although admissions officers can consider race as apositive feature of a minority student’s application, there is nodispute that race is but a “factor of a factor of a factor” in theholistic-review calculus. 645 F.Supp. 2d 587, 608 (WD Tex.2009). Furthermore, consideration of race is contextual and doesnot operate as a mechanical plus factor for underrepresentedminorities. Id., at 606 (“Plaintiffs cite no evidence toshow racial groups other than African-Americans and Hispanics areexcluded from benefitting from UT’s consideration of race inadmissions. As the Defendants point out, the consideration of race,within the full context of the entire application, may bebeneficial to any UT Austin applicant—including whites andAsian-Americans”); see also Brief for Asian American Legal Defenseand Education Fund etal. as Amici Curiae 12 (thecontention that the University discriminates againstAsian-Americans is “entirely unsupported by evidence in the recordor empirical data”). There is also no dispute, however, that race,when considered in conjunction with other aspects of an applicant’sbackground, can alter an applicant’s PAS score. Thus, race, in thisindirect fashion, considered with all of the other factors thatmake up an applicant’s AI and PAI scores, can make a difference towhether an application is accepted or rejected.Petitioner Abigail Fisher applied for admission to theUniversity’s 2008 freshman class. She was not in the top 10 percentof her high school class, so she was evaluated for admissionthrough holistic, full-file review. Petitioner’s application wasrejected.Petitioner then filed suit alleging that the University’sconsideration of race as part of its holistic-review processdisadvantaged her and other Caucasian applicants, in violation ofthe Equal Protection Clause. See U.S. Const., Amdt.14,§1 (no State shall “deny to any person within its jurisdiction theequal protection of the laws”). The District Court entered summaryjudgment in the University’s favor, and the Court of Appealsaffirmed.This Court granted certiorari and vacated the judgment of theCourt of Appeals, Fisher v. University of Tex. atAustin, 570 U.S. ___ (2013) (Fisher I),because it had applied an overly deferential “good-faith” standardin assessing the constitutionality of the University’s program. TheCourt remanded the case for the Court of Appeals to assess theparties’ claims under the correct legal standard.Without further remanding to the District Court, the Court ofAppeals again affirmed the entry of summary judgment in theUniversity’s favor. 758 F. 3d 633 (CA5 2014). This Court grantedcertiorari for a second time, 576 U.S. ___ (2015), and nowaffirms.IIFisher I set forth three controlling principles relevantto assessing the constitutionality of a public university’saffirmative-action program. First, “because racial characteristicsso seldom provide a relevant basis for disparate treatment,”Richmond v. J. A. Croson Co.,488 U.S. 469,505(1989), “[r]ace may not be considered [by a university] unless theadmissions process can withstand strict scru-tiny,” FisherI, 570 U.S., at ___ (slip op., at 7). Strict scru-tinyrequires the university to demonstrate with clarity that its“‘purpose or interest is both constitutionally permissibleand substantial, and that its use of the classification isnecessary ... to the accomplishment of itspurpose.’” Ibid.Second, Fisher I confirmed that “the decision to pursue‘the educational benefits that flow from student body diversity’... is, in substantial measure, an academic judgment towhich some, but not complete, judicial deference is proper.”Id., at ___ (slip op, at 9). A university cannot impose afixed quota or otherwise “define diversity as ‘some specifiedpercentage of a particular group merely because of its race orethnic origin.’” Ibid. Once, however, a universitygives “a reasoned, principled explanation” for its decision,deference must be given “to the University’s conclusion, based onits experience and expertise, that a diverse student body wouldserve its educational goals.” Ibid. (internal quotationmarks and citation omitted).Third, Fisher I clarified that no deference is owed whendetermining whether the use of race is narrowly tailored to achievethe university’s permissible goals. Id., at ___ (slip op.,at 10). A university, Fisher I explained, bears the burdenof proving a “nonracial approach” would not promote its interest inthe educational benefits of diversity “about as well and attolerable administrative expense.” Id., at ___ (slip op., at11) (internal quotation marks omitted). Though “[n]arrow tailoringdoes not require exhaustion of every conceivable race-neutralalternative” or “require a university to choose between maintaininga reputation for excellence [and] fulfilling a commitment toprovide educational opportunities to members of all racial groups,”Grutter, 539 U.S., at 339, it does impose “on theuniversity the ultimate burden of demonstrating” that “race-neutralalternatives” that are both “available” and “workable” “do notsuffice.” Fisher I, 570 U.S., at ___ (slip op., at11).Fisher I set forth these controlling principles, whiletaking no position on the constitutionality of the admissionsprogram at issue in this case. The Court held only that theDistrict Court and the Court of Appeals had “confined the strictscrutiny inquiry in too narrow a way by deferring to theUniversity’s good faith in its use of racial classifications.”Id., at ___ (slip op., at 12) The Court remanded the case,with instructions to evaluate the record under the correct standardand to determine whether the University had made “a showing thatit* plan is narrowly tailored to achieve” the educational benefitsthat flow from diversity. Id., at ___ (slip op., at 13). Onremand, the Court of Appeals determined that the program conformedwith the strict scrutiny mandated by Fisher I. See 758F.3d, at 659–660. Judge Garzadissented.IIIThe University’s program is sui generis. Unlike otherapproaches to college admissions considered by this Court, itcombines holistic review with a percentage plan. This approach gaverise to an unusual consequence in this case: The component of theUniversity’s admissions policy that had the largest impact onpetitioner’s chances of admission was not the school’sconsideration of race under its holistic-review process but ratherthe Top Ten Percent Plan. Because petitioner did not graduate inthe top 10 percent of her high school class, she was categoricallyineligible for more than three-fourths of the slots in the incomingfreshman class. It seems quite plausible, then, to think thatpetitioner would have had a better chance of being admitted to theUniversity if the school used race-conscious holistic review toselect its entire incoming class, as was the case inGrutter.Despite the Top Ten Percent Plan’s outsized effect onpetitioner’s chances of admission, she has not challenged it. Forthat reason, throughout this litigation, the Top Ten Percent Planhas been taken, somewhat artificially, as a given premise.Petitioner’s acceptance of the Top Ten Percent Plan complicatesthis Court’s review. In particular, it has led to a record that isalmost devoid of information about the students who securedadmission to the University through the Plan. The Court thus cannotknow how students admitted solely based on their class rank differin their contribution to diversity from students admitted throughholistic review.In an ordinary case, this evidentiary gap perhaps could befilled by a remand to the district court for further factfinding.When petitioner’s application was rejected, however, theUniversity’s combined percentage-plan/holistic-review approach toadmission had been in effect for just three years. While studiesundertaken over the eight years since then may be of significantvalue in determining the constitutionality of the University’scurrent admissions policy, that evidence has little bearing onwhether petitioner received equal treatment when her applicationwas rejected in 2008. If the Court were to remand, therefore,further factfinding would be limited to a narrow 3-year sample,review of which might yield little insight.Furthermore, as discussed above, the University lacks anyauthority to alter the role of the Top Ten Percent Plan in itsadmissions process. The Plan was mandated by the Texas Legislaturein the wake of Hopwood, so the University, like petitionerin this litigation, has likely taken the Plan as a given since itsimplementation in 1998. If the University had no reason to thinkthat it could deviate from the Top Ten Percent Plan, it similarlyhad no reason to keep extensive data on the Plan or the studentsadmitted under it—particularly in the years before Fisher Iclarified the stringency of the strict-scrutiny burden for a schoolthat employs race-conscious review.Under the circ*mstances of this case, then, a remand would donothing more than prolong a suit that has already persisted foreight years and cost the parties on both sides significantresources. Petitioner long since has graduated from anothercollege, and the University’s policy—and the data on which it firstwas based—may have evolved or changed in material ways.The fact that this case has been litigated on a somewhatartificial basis, furthermore, may limit its value for prospectiveguidance. The Texas Legislature, in enacting the Top Ten PercentPlan, cannot much be criticized, for it was responding toHopwood, which at the time was binding law in the State ofTexas. That legislative response, in turn, circ*mscribed theUniversity’s discretion in crafting its admissions policy. Thesecirc*mstances refute any criticism that the University did not makegood-faith efforts to comply with the law.That does not diminish, however, the University’s continuingobligation to satisfy the burden of strict scrutiny in light ofchanging circ*mstances. The University en-gages in periodicreassessment of the constitutionality, and efficacy, of itsadmissions program. See Supp. App. 32a; App. 448a. Going forward,that assessment must be undertaken in light of the experience theschool has accumulated and the data it has gathered since theadoption of its admissions plan.As the University examines this data, it should remain mindfulthat diversity takes many forms. Formalistic racial classificationsmay sometimes fail to capture diversity in all of its dimensionsand, when used in a divisive manner, could undermine theeducational benefits the University values. Through regularevaluation of data and consideration of student experience, theUniversity must tailor its approach in light of changingcirc*mstances, ensuring that race plays no greater role than isneces-sary to meet its compelling interest. The University’sexamination of the data it has acquired in the years sincepetitioner’s application, for these reasons, must proceed with fullrespect for the constraints imposed by the Equal Protection Clause.The type of data collected, and the manner in which it isconsidered, will have a significant bearing on how the Universitymust shape its admissions policy to satisfy strict scrutiny in theyears to come. Here, however, the Court is necessarily limited tothe narrow question before it: whether, drawing all reasonableinferences in her favor, petitioner has shown by a preponderance ofthe evidence that she was denied equal treatment at the time herapplication was rejected.IVIn seeking to reverse the judgment of the Court of Appeals,petitioner makes four arguments. First, she argues that theUniversity has not articulated its compelling interest withsufficient clarity. According to petitioner, the University mustset forth more precisely the level of minority enrollment thatwould constitute a “critical mass.” Without a clearer sense of whatthe University’s ultimate goal is, petitioner argues, a reviewingcourt cannot assess whether the University’s admissions program isnarrowly tailored to that goal.As this Court’s cases have made clear, however, the compellinginterest that justifies consideration of race in college admissionsis not an interest in enrolling a certain number of minoritystudents. Rather, a university may institute a race-consciousadmissions program as a means of obtaining “the educationalbenefits that flow from student body diversity.” Fisher I,570 U.S., at ___ (slip op., at 9) (internal quotation marksomitted); see also Grutter, 539 U.S., at 328. As thisCourt has said, enrolling a diverse student body “promotescross-racial understanding, helps to break down racial stereotypes,and enables students to better understand persons of differentraces.” Id., at 330 (internal quotation marks and alterationomitted). Equally important, “student body diversity promoteslearning outcomes, and better prepares students for an increasinglydiverse workforce and society.” Ibid. (internal quotationmarks omitted).Increasing minority enrollment may be instrumental to theseeducational benefits, but it is not, as petitioner seems tosuggest, a goal that can or should be reduced to pure numbers.Indeed, since the University is prohibited from seeking aparticular number or quota of minority students, it cannot befaulted for failing to specify the particular level of minorityenrollment at which it believes the educational benefits ofdiversity will be obtained.On the other hand, asserting an interest in the educationalbenefits of diversity writ large is insufficient. A university’sgoals cannot be elusory or amorphous—they must be sufficientlymeasurable to permit judicial scrutiny of the policies adopted toreach them.The record reveals that in first setting forth its currentadmissions policy, the University articulated concrete and precisegoals. On the first page of its 2004 “Proposal to Consider Race andEthnicity in Admissions,” the Univer-sity identifies theeducational values it seeks to realize through its admissionsprocess: the destruction of stereotypes, the “‘promot[ion of]cross-racial understanding,’” the preparation of a studentbody “‘for an increasingly diverse workforce andsociety,’” and the “‘cultivat[ion of] a set of leaderswith legitimacy in the eyes of the citizenry.’” Supp. App.1a; see also id., at 69a; App. 314a–315a (deposition of N.Bruce Walker (Walker Dep.)), 478a–479a (Walker Aff. ¶4) (settingforth the same goals). Later in the proposal, the Universityexplains that it strives to provide an “academic environment” thatoffers a “robust exchange of ideas, exposure to differing cultures,preparation for the challenges of an increasingly diverseworkforce, and acquisition of competencies required of futureleaders.” Supp. App. 23a. All of these objectives, as a generalmatter, mirror the “compelling interest” this Court has approved inits prior cases.The University has provided in addition a “reasoned, principledexplanation” for its decision to pursue these goals. FisherI, supra, at ___ (slip op., at 9). The Univer-sity’s39-page proposal was written following a year-long study, whichconcluded that “[t]he use of race-neutral policies and programsha[d] not been successful” in “provid[ing] an educational settingthat fosters cross-racial understanding, provid[ing] enlighteneddiscussion and learning, [or] prepar[ing] students to function inan increasingly diverse workforce and society.” Supp. App. 25a; seealso App. 481a–482a (Walker Aff. ¶¶8–12) (describing the“thoughtful review” the University undertook when it faced the“important decision ... whether or not to use race inits admissions process”). Further support for the University’sconclusion can be found in the depositions and affidavits fromvarious admissions officers, all of whom articulate the same,consistent “reasoned, principled explanation.” See, e.g.,id., at 253a (Ishop Dep.), 314a–318a, 359a (Walker Dep.),415a–416a (Defendant’s Statement of Facts), 478a–479a, 481a–482a(Walker Aff. ¶¶4, 10–13). Petitioner’s contention that theUniversity’s goal was insufficiently concrete is rebutted by therecord.Second, petitioner argues that the University has no need toconsider race because it had already “achieved critical mass” by2003 using the Top Ten Percent Plan and race-neutral holisticreview. Brief for Petitioner 46. Petitioner is correct that auniversity bears a heavy burden in showing that it had not obtainedthe educational benefits of diversity before it turned to arace-conscious plan. The record reveals, however, that, at the timeof petitioner’s application, the University could not be faulted onthis score. Before changing its policy the University conducted“months of study and deliberation, including retreats, interviews,[and] review of data,” App. 446a, and concluded that “[t]he use ofrace-neutral policies and programs ha[d] not been successful inachieving” sufficient racial diversity at the University, Supp.App. 25a. At no stage in this litigation has petitioner challengedthe University’s good faith in conducting its studies, and theCourt properly declines to consider the extrarecord materials thedissent relies upon, many of which are tangential to this case atbest and none of which the University has had a full opportunity torespond to. See, e.g., post, at 45–46 (opinion of Alito, J.)(describing a 2015 report regarding the admission of applicants whoare related to ‘‘politically connected individuals’’).The record itself contains significant evidence, bothstatistical and anecdotal, in support of the University’s position.To start, the demographic data the University has submitted showconsistent stagnation in terms of the percentage of minoritystudents enrolling at the Univer-sity from 1996 to 2002. In 1996,for example, 266 African-American freshmen enrolled, a total thatconstituted 4.1 percent of the incoming class. In 2003, the yearGrutter was decided, 267 African-American studentsenrolled—again, 4.1 percent of the incoming class. The numbers forHispanic and Asian-American students tell a similar story. SeeSupp. App. 43a. Although demographics alone are by no meansdispositive, they do have some value as a gauge of the University’sability to enroll students who can offer underrepresentedperspectives.In addition to this broad demographic data, the University putforward evidence that minority students admitted under theHopwood regime experienced feelings of loneliness andisolation. See, e.g., App. 317a–318a.This anecdotal evidence is, in turn, bolstered by further, morenuanced quantitative data. In 2002, 52 percent of undergraduateclasses with at least five students had no African-Americanstudents enrolled in them, and 27 percent had only oneAfrican-American student. Supp. App. 140a. In other words, only 21percent of undergraduate classes with five or more students in themhad more than one African-American student enrolled. Twelve percentof these classes had no Hispanic students, as compared to 10percent in 1996. Id., at 74a, 140a. Though a college mustcontinually reassess its need for race-conscious review, here thatassessment appears to have been done with care, and a reasonabledetermination was made that the University had not yet attained itsgoals.Third, petitioner argues that considering race was not necessarybecause such consideration has had only a “‘minimal impact’in advancing the [University’s] compelling interest.” Brief forPetitioner 46; see also Tr. of Oral Arg. 23:10–12; 24:13–25:2,25:24–26:3. Again, the record does not support this assertion. In2003, 11 percent of the Texas residents enrolled through holisticreview were Hispanic and 3.5 percent were African-American. Supp.App. 157a. In 2007, by contrast, 16.9 percent of the Texasholistic-review freshmen were Hispanic and 6.8 percent wereAfrican-American. Ibid. Those increases—of 54 percent and 94percent, respectively—show that consideration of race has had ameaningful, if still limited, effect on the diversity of theUniversity’s freshman class.In any event, it is not a failure of narrow tailoring for theimpact of racial consideration to be minor. The fact that raceconsciousness played a role in only a small portion of admissionsdecisions should be a hallmark of narrow tailoring, not evidence ofunconstitutionality.Petitioner’s final argument is that “there are numerous otheravailable race-neutral means of achieving” the University’scompelling interest. Brief for Petitioner 47. A review of therecord reveals, however, that, at the time of petitioner’sapplication, none of her proposed alternatives was a workable meansfor the University to attain the benefits of diversity it sought.For example, petitioner suggests that the University couldintensify its outreach efforts to African-American and Hispanicapplicants. But the University submitted extensive evidence of themany ways in which it already had intensified its outreach effortsto those students. The University has created three new scholarshipprograms, opened new regional admissions centers, increased itsrecruitment budget by half-a-million dollars, and organized over1,000 recruitment events. Supp. App. 29a–32a; App. 450a–452a(citing affidavit of Michael Orr ¶¶4–20). Perhaps moresignificantly, in the wake of Hopwood, the University spentseven years attempting to achieve its compelling interest usingrace-neutral holistic review. None of these efforts succeeded, andpetitioner fails to offer any meaningful way in which theUniversity could have improved upon them at the time of herapplication.Petitioner also suggests altering the weight given to academicand socioeconomic factors in the University’s admissions calculus.This proposal ignores the fact that the University tried, andfailed, to increase diversity through enhanced consideration ofsocioeconomic and other factors. And it further ignores thisCourt’s precedent making clear that the Equal Protection Clausedoes not force universities to choose between a diverse studentbody and a reputation for academic excellence. Grutter, 539U.S., at 339.Petitioner’s final suggestion is to uncap the Top Ten PercentPlan, and admit more—if not all—the University’s students through apercentage plan. As an initial matter, petitioner overlooks thefact that the Top Ten Percent Plan, though facially neutral, cannotbe understood apart from its basic purpose, which is to boostminority enrollment. Percentage plans are “adopted with raciallysegregated neighborhoods and schools front and center stage.”Fisher I, 570 U.S., at ___ (Ginsburg, J., dissenting)(slip op., at 2). “It is race consciousness, not blindness to race,that drives such plans.” Ibid. Consequently, petitionercannot assert simply that increasing the University’s reliance on apercentage plan would make its admissions policy more raceneutral.Even if, as a matter of raw numbers, minority enrollment wouldincrease under such a regime, petitioner would be hard-pressed tofind convincing support for the proposition that college admissionswould be improved if they were a function of class rank alone. Thatapproach would sacrifice all other aspects of diversity in pursuitof enrolling a higher number of minority students. A system thatselected every student through class rank alone would exclude thestar athlete or musician whose grades suffered because of dailypractices and training. It would exclude a talented young biologistwho struggled to maintain above-average grades in humanitiesclasses. And it would exclude a student whose freshman-year gradeswere poor because of a family crisis but who got herself back ontrack in her last three years of school, only to find herself justoutside of the top decile of her class.These are but examples of the general problem. Class rank is asingle metric, and like any single metric, it will capture certaintypes of people and miss others. This does not imply that studentsadmitted through holistic review are necessarily more capable ormore desirable than those admitted through the Top Ten PercentPlan. It merely reflects the fact that privileging onecharacteristic above all others does not lead to a diverse studentbody. Indeed, to compel universities to admit students based onclass rank alone is in deep tension with the goal of educationaldiversity as this Court’s cases have defined it. SeeGrutter, supra, at 340 (explaining that percentageplans “may preclude the university from conducting theindividualized assessments necessary to assemble a student bodythat is not just racially diverse, but diverse along all thequalities valued by the university”); 758 F. 3d, at 653 (pointingout that the Top Ten Percent Law leaves out students “who felloutside their high school’s top ten percent but excelled in uniqueways that would enrich the diversity of [the University’s]educational experience” and “leaves a gap in an admissions processseeking to create the multi-dimensional diversity that [Regentsof Univ. of Cal. v. Bakke,438 U.S. 265 (1978),]envisions”). At its center, the Top Ten Percent Plan is a bluntinstrument that may well compromise the University’s own definitionof the diversity it seeks.In addition to these fundamental problems, an admissions policythat relies exclusively on class rank creates perverse incentivesfor applicants. Percentage plans “encourage parents to keep theirchildren in low-performing segregated schools, and discouragestudents from taking challenging classes that might lower theirgrade point averages.” Gratz, 539 U.S., at 304, n. 10(Ginsburg, J., dissenting).For all these reasons, although it may be true that the Top TenPercent Plan in some instances may provide a path out of povertyfor those who excel at schools lacking in resources, the Plancannot serve as the admissions solution that petitioner suggests.Wherever the balance between percentage plans and holistic reviewshould rest, an effective admissions policy cannot prescribe,realisti-cally, the exclusive use of a percentage plan.In short, none of petitioner’s suggested alternatives—nor otherproposals considered or discussed in the course of thislitigation—have been shown to be “available” and “workable” meansthrough which the University could have met its educational goals,as it understood and defined them in 2008. Fisher I,supra, at ___ (slip op., at 11). The University has thus metit* burden of showing that the admissions policy it used at thetime it rejected petitioner’s application was narrowlytailored.* * *A university is in large part defined by those intangible“qualities which are incapable of objective measurement but whichmake for greatness.” Sweatt v. Painter,339 U.S.629,634 (1950). Considerable deference is owed to a university indefining those intangible characteristics, like student bodydiversity, that are central to its identity and educationalmission. But still, it remains an enduring challenge to ourNation’s education system to reconcile the pursuit of diversitywith the constitutional promise of equal treatment and dignity.In striking this sensitive balance, public universities, likethe States themselves, can serve as “laboratories forexperimentation.” United States v. Lopez,514U.S. 549,581 (1995) (Kennedy, J., concurring); see alsoNew State Ice Co. v. Liebmann,285 U.S. 262,311(1932) (Brandeis, J., dissenting). The University of Texas atAustin has a special opportunity to learn and to teach. TheUniversity now has at its disposal valuable data about the mannerin which different approaches to admissions may foster diversity orinstead dilute it. The University must con-tinue to use this datato scrutinize the fairness of its admis-sions program; to assesswhether changing demographics have undermined the need for arace-conscious policy; and to identify the effects, both positiveand negative, of the affirmative-action measures it deemsnecessary.The Court’s affirmance of the University’s admissions policytoday does not necessarily mean the University may rely on thatsame policy without refinement. It is the University’s ongoingobligation to engage in constant deliberation and continuedreflection regarding its admissions policies.The judgment of the Court of Appeals is affirmed.It is so ordered.Justice Kagan took no part in the consideration or decision ofthis case.

SUPREME COURT OF THE UNITED STATES_________________No. 14–981_________________ABIGAIL NOEL FISHER, PETITIONER v.UNIVERSITY OF TEXAS AT AUSTIN, etal.on writ of certiorari to the united statescourt of appeals for the fifth circuit[June 23, 2016]Justice Alito, with whom The Chief Justice andJustice Thomas join, dissenting.Something strange has happened since our priordecision in this case. See Fisher v. University of Tex.at Austin, 570 U.S. ___ (2013) (Fisher I). In thatdecision, we held that strict scrutiny requires the University ofTexas at Austin (UT or University) to show that its use of race andethnicity in making admissions decisions serves compellinginterests and that its plan is narrowly tailored to achieve thoseends. Rejecting the argument that we should defer to UT’s judgmenton those matters, we made it clear that UT was obligated (1) toidentify the interests justifying its plan with enough specificityto permit a reviewing court to determine whether the requirementsof strict scrutiny were met, and (2) to show that thoserequirements were in fact satisfied. On remand, UT failed to dowhat our prior decision demanded. The University has still notidentified with any degree of specificity the interests that itsuse of race and ethnicity is supposed to serve. Its primaryargument is that merely invoking “the educational benefits ofdiversity” is sufficient and that it need not identify any metricthat would allow a court to determine whether its plan is needed toserve, or is actually serving, those interests. This is nothingless than theplea for deference that we emphatically rejected inour prior decision. Today, however, the Court inexplicably grantsthat request.To the extent that UT has ever moved beyond aplea for deference and identified the relevant interests in morespecific terms, its efforts have been shifting, unpersuasive, and,at times, less than candid. When it adopted its race-based plan, UTsaid that the plan was needed to promote classroom diversity. SeeSupp. App. 1a, 24a–25a, 39a; App. 316a. It pointed to a studyshowing that African-American, Hispanic, and Asian-Americanstudents were underrepresented in many classes. See Supp. App. 26a.But UT has never shown that its race-conscious plan actuallyameliorates this situation. The University presents no evidencethat its admissions officers, in administering the “holistic”component of its plan, make any effort to determine whether anAfrican-American, Hispanic, or Asian-American student is likely toenroll in classes in which minority students are underrepresented.And although UT’s records should permit it to determine withoutmuch difficulty whether holistic admittees are any more likely thanstudents admitted through the Top Ten Percent Law, Tex. Educ. CodeAnn. §51.803 (West Cum. Supp. 2015), to enroll in the classeslacking racial or ethnic diversity, UT either has not crunchedthose numbers or has not revealed what they show. Nor has UTexplained why the underrepresentation of Asian-American students inmany classes justifies its plan, which discriminates againstthose students.At times, UT has claimed that its plan is neededto achieve a “critical mass” of African-American and His-panicstudents, but it has never explained what this term means.According to UT, a critical mass is neither some absolute number ofAfrican-American or Hispanic students nor the percentage ofAfrican-Americans or Hispanics in the general population of theState. The term remains undefined, but UT tells us that it will letthe courts know when the desired end has been achieved. See App.314a–315a. This is a plea for deference—indeed, for blinddeference—the very thing that the Court rejected inFisherI.UT has also claimed at times that the race-basedcomponent of its plan is needed because the Top Ten Percent Planadmits the wrong kind of African-American and Hispanicstudents, namely, students from poor families who attend schools inwhich the student body is predominantly African-American orHispanic. As UT put it in its brief in Fisher I, therace-based component of its admissions plan is needed to admit“[t]he African-American or Hispanic child of successfulprofessionals in Dallas.” Brief for Respondents, O. T. 2012, No.11–345, p.34.After making this argument in its first trip tothis Court, UT apparently had second thoughts, and in the latestround of briefing UT has attempted to disavow ever having made theargument. See Brief for Respondents 2 (“Petitioner’s argument thatUT’s interest is favoring ‘affluent’ minorities is a fabrication”);see also id., at 15. But it did, and the argument turnsaffirmative action on its head. Affirmative-action programs werecreated to help disadvantaged students.Although UT now disowns the argument that theTop Ten Percent Plan results in the admission of the wrong kind ofAfrican-American and Hispanic students, the Fifth Circuit majoritybought a version of that claim. As the panel majority put it, theTop Ten African-American and Hispanic admittees cannot match theholistic African-American and Hispanic admittees when it comes to“records of personal achievement,” a “variety of perspectives” and“life experiences,” and “unique skills.” 758 F.3d 633, 653(2014). All in all, according to the panel majority, the Top TenPercent students cannot “enrich the diversity of the student body”in the same way as the holistic admittees. Id., at 654. AsJudge Garza put it in dissent, the panel majority concluded thatthe Top Ten Percent admittees are “somehow more hom*ogenous, lessdynamic, and more undesirably stereotypical than those admittedunder holistic review.” Id., at 669–670 (Garza, J.,dissenting).The Fifth Circuit reached this conclusion withlittle direct evidence regarding the characteristics of the Top TenPercent and holistic admittees. Instead, the assumption behind theFifth Circuit’s reasoning is that most of the African-American andHispanic students admitted under the race-neutral component of UT’splan were able to rank in the top decile of their high schoolclasses only because they did not have to compete against white andAsian-American students. This insulting stereotype is not supportedby the record. African-American and Hispanic students admittedunder the Top Ten Percent Plan receive higher college grades thanthe African-American and Hispanic students admitted under therace-conscious program. See Supp. App. 164a–165a.It should not have been necessary for us togrant review a second time in this case, and I have no greaterdesire than the majority to see the case drag on. But that need nothappen. When UT decided to adopt its race-conscious plan, it hadevery reason to know that its plan would have to satisfy strictscrutiny and that this meant that it would be its burden toshow that the plan was narrowly tailored to serve compellinginterests. UT has failed to make that showing. By all rights,judgment should be entered in favor of petitioner.But if the majority is determined to give UT yetanother chance, we should reverse and send this case back to theDistrict Court. What the majority has now done—awarding a victoryto UT in an opinion that fails to address the important issues inthe case—is simply wrong.IOver the past 20 years, UT has frequentlymodified its admissions policies, and it has generally employedrace and ethnicity in the most aggressive manner permitted undercontrolling precedent.Before 1997, race was considered directly aspart of the general admissions process, and it was frequently acontrolling factor. Admissions were based on two criteria:(1)the applicant’s Academic Index (AI), which was computedfrom standardized test scores and high school class rank, and (2)the applicant’s race. In 1996, the last year this race-conscioussystem was in place, 4.1% of enrolled freshmen wereAfrican-American, 14.7% were Asian-American, and 14.5% wereHispanic. Supp. App. 43a.The Fifth Circuit’s decision in Hopwoodv. Texas, 78 F.3d 932 (1996), prohibited UT from usingrace in admissions. In response to Hopwood, beginning withthe 1997 admissions cycle, UT instituted a “holistic review”process in which it considered an applicant’s AI as well as aPersonal Achievement Index (PAI) that was intended, among otherthings, to increase minority enrollment. The race-neutral PAI was acomposite of scores from two essays and a personal achievementscore, which in turn was based on a holistic review of anapplicant’s leadership qualities, extracurricular activities,honors and awards, work experience, community service, and specialcirc*mstances. Special consideration was given to applicants frompoor families, applicants from homes in which a language other thanEnglish was customarily spoken, and applicants from single-parenthouseholds. Because this race-neutral plan gave a preference todisadvantaged students, it had the effect of “disproportionately”benefiting minority candidates. 645 F.Supp. 2d 587, 592 (WDTex. 2009).The Texas Legislature also responded toHopwood. In 1997, it enacted the Top Ten Percent Plan, whichman-dated that UT admit all Texas seniors who rank in the top 10%of their high school classes. This facially race-neutral law servedto equalize competition between students who live in relativelyaffluent areas with superior schools and students in poorer areasserved by schools offering fewer opportunities for academicexcellence. And by benefiting the students in the latter group,this plan, like the race-neutral holistic plan already adopted byUT, tended to benefit African-American and Hispanic students, whoare often trapped in inferior public schools. 758 F.3d, at650–653.Starting in 1998, when the Top Ten Percent Plantook effect, UT’s holistic, race-neutral AI/PAI system continued tobe used to fill the seats in the entering class that were not takenby Top Ten Percent students. The AI/PAI system was also used todetermine program placement forall incoming students, including theTop Ten Percent students.“The University’s revised admissions process,coupled with the operation of the Top Ten Percent Law, resulted ina more racially diverse environment at the University.” FisherI, 570 U.S., at ___ (slip op., at 3). In 2000, UTannounced that its “enrollment levels for African American andHispanic freshmen have returned to those of 1996, the year beforethe Hopwood decision prohibited the consideration of race inadmissions policies.” App. 393a; see also Supp. App. 23a–24a(pre-Hopwood diversity levels were “restored” in 1999); App.392a–393a (“The ‘Top 10 Percent Law’ is Working for Texas” and “hasenabled us to diversify enrollment at UT Austin with talentedstudents who succeed”). And in 2003, UT proclaimed that it had“effectively compensated for the loss of affirmative action.”Id., at 396a; see also id., at 398a (“Diversityefforts at The University of Texas at Austin have brought a highernumber of freshman minority students—African Americans, Hispanicsand Asian-Americans—to the campus than were enrolled in 1996, theyear a court ruling ended the use of affirmative action in theuniversity’s enrollment process”). By 2004—the last year under theholistic, race-neutral AI/PAI system—UT’s entering class was 4.5%African-American, 17.9% Asian-American, and 16.9% Hispanic. Supp.App. 156a. The 2004 entering class thus had a higher percentage ofAfrican-Americans, Asian-Americans, and Hispanics than the classthat entered in 1996, when UT had last employed racialpreferences.Notwithstanding these lauded results, UT leaptat the opportunity to reinsert race into the process. On June 23,2003, this Court decided Grutter v. Bollinger, 539U.S. 306 (2003) , which upheld the University of Michigan LawSchool’s race-conscious admissions system. In Grutter, theCourt warned that a university contemplating the consideration ofrace as part of its admissions process must engage in “serious,good faith consideration of workable race-neutral alternatives thatwill achieve the diversity the university seeks.” Id., at339. Nevertheless, on the very day Grutter was handed down,UT’s president announced that “[t]he University of Texas at Austinwill modify its admissions procedures” in light ofGrutter, including by “implementing procedures at theundergraduate level that combine the benefits of the Top 10 PercentLaw with affirmative action programs.” App. 406a–407a (emphasisadded).[1] UT purports to havelater engaged in “almost a year of deliberations,” id., at482a, but there is no evidence that the reintroduction of race intothe admissions process was anything other than a foregoneconclusion following the president’s announcement.“The University’s plan to resume race-consciousadmissions was given formal expression in June 2004 in an internaldocument entitled Proposal to Consider Race and Ethnicity inAdmissions” (Proposal). Fisher I, supra, at ___ (slipop., at 4). The Proposal stated that UT needed race-consciousadmissions because it had not yet achieved a “critical mass ofracial diversity.” Supp. App. 25a. In support of this claim, UTcited two pieces of evidence. First, it noted that there were“significant differences between the racial and ethnic makeup ofthe University’s undergraduate population and the state’spopulation.” Id., at 24a. Second, the Proposal “relied insubstantial part,” Fisher I, supra, at ___ (slip op.,at 4), on a study of a subset of undergraduate classes containingat least five students, see Supp. App. 26a. The study showed thatamong select classes with five or more students, 52% had noAfrican-Americans, 16% had no Asian-Americans, and 12% had noHispanics. Ibid. Moreover, the study showed, only 21% ofthese classes had two or more African-Americans, 67% had two ormore Asian-Americans, and 70% had two or more Hispanics. Seeibid. Based on this study, the Proposal concluded that UT“has not reached a critical mass at the classroom level.”Id., at 24a. The Proposal did not analyze the backgrounds,life experiences, leadership qualities, awards, extracurricularactivities, community service, personal attributes, or othercharacteristics of the minority students who were already beingadmitted to UT under the holistic, race-neutral process.“To implement the Proposal the Universityincluded a student’s race as a component of the PAI score,beginning with applicants in the fall of 2004.” Fisher I,570 U.S., at ___ (slip op., at 4). “The University asksstudents to classify themselves from among five predefined racialcategories on the application.” Ibid. “Race is not assignedan explicit numerical value, but it is undisputed that race is ameaningful factor.” Ibid. UT decided to use racialpreferences to benefit African-American and Hispanic studentsbecause it considers those groups “underrepresented minorities.”Supp. App. 25a; see also App. 445a–446a (defining “underrepresentedminorities” as “Hispanic[s] and African Americans”). Even thoughUT’s classroom study showed that more classes lacked Asian-Americanstudents than lacked Hispanic students, Supp. App. 26a, UT deemedAsian-Americans “overrepresented” based on statedemographics, 645 F.Supp. 2d, at 606; see also ibid.(“It is undisputed that UT considers African-Americans andHispanics to be underrepresented but does not con-siderAsian-Americans to be underrepresented”).Although UT claims that race is but a “factor ofa factor of a factor of a factor,” id., at 608, UTacknowledges that “race is the only one of [its] holistic factorsthat appears on the cover of every application,” Tr. of Oral Arg.54 (Oct. 10, 2012). “Because an applicant’s race is identified atthe front of the admissions file, reviewers are aware of itthroughout the evaluation.” 645 F.Supp. 2d, at 597; see alsoid., at 598 (“[A] candidate’s race is known throughout theapplication process”). Consideration of race therefore pervadesevery aspect of UT’s admissions process. See App. 219a (“We arecertainly aware of the applicant’s race. It’s on the front page ofthe application that’s being read [and] is used in context witheverything else that’s part of the applicant’s file”). This is bydesign, as UT considers its use of racial classifications to be abenign form of “social engineering.” Powers, Why Schools Still NeedAffirmative Action, National L. J., Aug. 4, 2014, p. 22 (editorialby Bill Powers, President of UT from 2006–2015) (“Opponents accusedefenders of race-conscious admissions of being in favor of ‘socialengineering,’ to which I believe we should reply, ‘Guilty ascharged’”).Notwithstanding the omnipresence of racialclassifications, UT claims that it keeps no record of how thoseclassifications affect its process. “The university doesn’t keepany statistics on how many students are affected by theconsideration of race in admissions decisions,” and it “does notknow how many minority students are affected in a positive mannerby the consideration of race.” App. 337a. According to UT, it hasno way of making these determinations. See id., at320a–322a. UT says that it does not tell its admissions officershow much weight to give to race. See Deposition of Gary Lavergne43–45, Record in No. 1:08–CV–00263 (WD Tex.), Doc. 94–9 (LavergneDeposition). And because the influence of race is always“contextual,” UT claims, it cannot provide even a single example ofan instance in which race impacted a student’s odds of admission.See App. 220a (“Q. Could you give me an example where race wouldhave some impact on an applicant’s personal achievement score? A.To be honest, not really .... [I]t’s impossibleto say—to give you an example of a particular student because it’sall context-ual”). Accordingly, UT asserts that it has no ideawhich students were admitted as a result of its race-conscioussystem and which students would have been admitted under arace-neutral process. UT thus makes no effort to assess how theindividual characteristics of students admitted as the result ofracial preferences differ (or do not differ) from those of studentswho would have been admitted without them.IIUT’s race-conscious admissions program cannotsatisfy strict scrutiny. UT says that the program furthers itsinterest in the educational benefits of diversity, but it hasfailed to define that interest with any clarity or to demonstratethat its program is narrowly tailored to achieve that or any otherparticular interest. By accepting UT’s rationales as sufficient tomeet its burden, the majority licenses UT’s perverse assumptionsabout different groups of minority students—the precise assumptionsstrict scrutiny is supposed to stamp out.A“The moral imperative of racial neutrality isthe driving force of the Equal Protection Clause.” Richmondv. J. A. Croson Co., 488 U.S. 469, 518 (1989)(Kennedy, J., concurring in part and concurring in judgment). “Atthe heart of the Constitution’s guarantee of equal protection liesthe simple command that the Government must treat citizens asindividuals, not as simply components of a racial, religious,sexual or national class.” Miller v. Johnson, 515U.S. 900, 911 (1995) (internal quotation marks omitted).“Race-based assignments embody stereotypes that treat individualsas the product of their race, evaluating their thoughts andefforts—their very worth as citizens—according to a criterionbarred to the Government by history and the Constitution.”Id., at 912 (internal quotation marks omitted). Given ourconstitutional commitment to “the doctrine of equality,”“‘[d]istinctions between citizens solely because of theirancestry are by their very nature odious to a free people.’”Rice v. Cayetano, 528 U.S. 495, 517 (2000)(quoting Hirabayashi v. United States, 320 U.S.81, 100 (1943) ).“[B]ecause racial characteristics so seldomprovide a relevant basis for disparate treatment, the EqualProtection Clause demands that racial classifications... be subjected to the most rigid scrutiny.” FisherI, 570 U.S., at ___ (slip op., at 8) (internal quotationmarks and citations omitted). “[J]udicial review must begin fromthe position that ‘any official action that treats a persondifferently on account of his race or ethnic origin is inher-entlysuspect.’” Ibid.; see also Grutter, 539U.S., at 388 (Kennedy, J., dissenting) (“‘Racial andethnic distinctions of any sort are inherently suspect and thuscall for the most exacting judicial examination’”). Understrict scrutiny, the use of race must be “necessary to further acompelling governmental interest,” and the means employed must be“‘specifically and narrowly’” tailored to accomplishthe compelling interest. Id., at 327, 333 (O’Connor, J., forthe Court).The “higher education dynamic does not change”this standard. Fisher I, supra, at ___ (slip op., at12). “Racial discrimination [is] invidious in all contexts,”Edmonson v. Leesville Concrete Co., 500 U.S.614, 619 (1991) , and “‘[t]he analysis and level of scrutinyapplied to determine the validity of [a racial] classification donot vary simply because the objective appears acceptable,’”Fisher I, supra, at ___ (slip op., at 12).Nor does the standard of review “‘depen[d]on the race of those burdened or benefited by a particularclassification.’” Gratz v. Bollinger, 539U.S. 244, 270 (2003) (quoting Adarand Constructors,Inc. v. Peña, 515 U.S. 200, 224 (1995) ); see alsoMiller, supra, at 904 (“This rule obtains with equalforce regardless of ‘the race of those burdened or benefited by aparticular classification’” (quoting Croson,supra, at 494 (plurality opinion of O’Connor, J.)). “Thus,‘any person, of whatever race, has the right to demand that anygovernmental actor subject to the Constitution justify any racialclassification subjecting that person to unequal treatment underthe strictest of judicial scrutiny.’” Gratz,supra, at 270 (quoting Adarand, supra, at224).In short, in “all contexts,” Edmonson,supra, at 619, racial classifications are permitted only “asa last resort,” when all else has failed, Croson,supra, at 519 (opinion of Kennedy, J.). “Strict scrutiny isa searching examination, and it is the government that bears theburden” of proof. Fisher I, 570 U.S., at ___ (slipop., at 8). To meet this burden, the government must “demonstratewith clarity that its ‘purpose or interest is bothconstitutionally permissible and substantial, and that its use ofthe classification is necessary ... to theaccomplishment of its purpose.’” Id., at ___ (slipop., at 7) (emphasis added).BHere, UT has failed to define its interest inusing racial preferences with clarity. As a result, the narrowtailoring inquiry is impossible, and UT cannot satisfy strictscrutiny.When UT adopted its challenged policy, itcharacterized its compelling interest as obtaining a“‘critical mass’” of underrepresented minorities.Id., at ___ (slip op., at 1). The 2004 Proposal claimed that“[t]he use of race-neutral policies and programs has not beensuccessful in achieving a critical mass of racial diversity.” Supp.App. 25a; see Fisher v. University of Tex. at Austin,631 F.3d 213, 226 (CA5 2011) (“[T]he 2004 Proposalexplained that UT had not yet achieved the critical mass ofunderrepresented minority students needed to obtain the fulleducational benefits of diversity”). But to this day, UT has notexplained in anything other than the vaguest terms what it means by“critical mass.” In fact, UT argues that it need not identifyany interest more specific than “securing the educationalbenefits of diversity.” Brief for Respondents 15.UT has insisted that critical mass is not anabsolute number. See Tr. of Oral Arg. 39 (Oct. 10, 2012) (declaringthat UT is not working toward any particular number ofAfrican-American or Hispanic students); App. 315a (confirming thatUT has not defined critical mass as a number and has not projectedwhen it will attain critical mass). Instead, UT prefers adeliberately malleable “we’ll know it when we see it” notion ofcritical mass. It defines “critical mass” as “an adequaterepresentation of minority students so that the ...educational benefits that can be derived from diversity canactually happen,” and it declares that it “will ...know [that] it has reached critical mass” when it “see[s] theeducational benefits happening.” Id., at 314a–315a. In otherwords: Trust us.This intentionally imprecise interest isdesigned to insulate UT’s program from meaningful judicial review.As Judge Garza explained:“[T]o meet its narrow tailoring burden,the University must explain its goal to us in some meaningful way.We cannot undertake a rigorous ends-to-means narrow tailoringanalysis when the University will not define the ends. We cannottell whether the admissions program closely ‘fits’ the University’sgoal when it fails to objectively articulate its goal. Nor can wedetermine whether considering race is necessary for the Universityto achieve ‘critical mass,’ or whether there are effectiverace-neutral alternatives, when it has not described what ‘criticalmass’ requires.” 758 F.3d, at 667 (dissenting opinion).Indeed, without knowing in reasonably specificterms what critical mass is or how it can be measured, a reviewingcourt cannot conduct the requisite “careful judicial inquiry” intowhether the use of race was “‘necessary.’” FisherI, supra, at ___ (slip op., at 10).To be sure, I agree with the majority that ourprecedents do not require UT to pinpoint “an interest in enrollinga certain number of minority students.” Ante, at 11. But inorder for us to assess whether UT’s program is narrowly tailored,the University must identify some sort of concrete interest.“Classifying and assigning” students according to race “requiresmore than ... an amorphous end to justify it.”Parents Involved in Community Schools v. Seattle SchoolDist. No. 1, 551 U.S. 701, 735 (2007) . Because UT hasfailed to explain “with clarity,” Fisher I, supra, at___ (slip op., at 7), why it needs a race-conscious policy and howit will know when its goals have been met, the narrow tailoringanalysis cannot be meaningfully conducted. UT therefore cannotsatisfy strict scrutiny.The majority acknowledges that “asserting aninterest in the educational benefits of diversity writ large isinsufficient,” and that “[a] university’s goals cannot be elusoryor amorphous—they must be sufficiently measurable to permitjudicial scrutiny of the policies adopted to reach them.”Ante, at 12. According to the majority, however, UT hasarticulated the following “concrete and precise goals”: “thedestruction of stereotypes, the promot[ion of] cross-racialunderstanding, the preparation of a student body for anincreasingly diverse workforce and society, and the cultivat[ionof] a set of leaders with legitimacy in the eyes of the citizenry.”Ibid. (internal quotation marks omitted).These are laudable goals, but they are notconcrete or precise, and they offer no limiting principle for theuse of racial preferences. For instance, how will a court ever beable to determine whether stereotypes have been adequatelydestroyed? Or whether cross-racial understanding has beenadequately achieved? If a university can justify racialdiscrimination simply by having a few employees opine that racialpreferences are necessary to accomplish these nebulous goals, seeante, at 12–13 (citing only self-serving statementsfrom UT officials), then the narrow tailoring inquiry ismeaningless. Courts will be required to defer to the judgment ofuniversity administrators, and affirmative-action policies will becompletely insulated from judicial review.By accepting these amorphous goals as sufficientfor UT to carry its burden, the majority violates decades ofprecedent rejecting blind deference to government officialsdefending “‘inherently suspect’” classifications.Miller, 515 U.S., at 904 (citing Regents of Univ.of Cal. v. Bakke, 438 U.S. 265, 291 (1978)(opinion of Powell, J.)); see also, e.g., Miller,supra, at 922 (“Our presumptive skepticism of all racialclassifications ... prohibits us ... fromaccepting on its face the Justice Department’s conclusion”(citation omitted)); Croson, 488 U.S., at 500 (“[T]hemere recitation of a ‘benign’ or legitimate purpose for a racialclassification is entitled to little or no weight”); id., at501 (“The history of racial classifications in this countrysuggests that blind judicial deference to legislative or executivepronouncements of necessity has no place in equal protectionanalysis”). Most troublingly, the majority’s uncritical deferenceto UT’s self-serving claims blatantly contradicts our decision inthe prior iteration of this very case, in which we faulted theFifth Circuit for improperly “deferring to the University’s goodfaith in its use of racial classifications.” Fisher I, 570U.S., at ___ (slip op., at 12). As we emphasized just threeyears ago, our precedent “ma[kes] clear that it is for the courts,not for university administrators, to ensure that” an admissionsprocess is narrowly tailored. Id., at ___ (slip op., at10).A court cannot ensure that an admissions processis narrowly tailored if it cannot pin down the goals that theprocess is designed to achieve. UT’s vague policy goals are “sobroad and imprecise that they cannot withstand strict scrutiny.”Parents Involved, supra, at 785 (Kennedy, J.,concurring in part and concurring in judgment).CAlthough UT’s primary argument is that it neednot point to any interest more specific than “the educationalbenefits of diversity,” Brief for Respondents 15, it has—at variouspoints in this litigation—identified four more specific goals:demographic parity, classroom diversity, intraracial diversity, andavoiding racial isolation. Neither UT nor the majority hasdemonstrated that any of these four goals provides a sufficientbasis for satisfying strict scrutiny. And UT’s arguments to thecontrary depend on a series of invidious assumptions.1First, both UT and the majority citedemographic data as evidence that African-American and Hispanicstudents are “underrepresented” at UT and that racial preferencesare necessary to compensate for this underrepresentation. See,e.g., Supp. App. 24a; ante, at 14. But neither UT northe majority is clear about the relationship between Texasdemographics and UT’s interest in obtaining a critical mass.Does critical mass depend on the relative sizeof a particular group in the population of a State? For example, isthe critical mass of African-Americans and Hispanics in Texas,where African-Americans are about 11.8% of the population andHispanics are about 37.6%, different from the critical mass inneighboring New Mexico, where the African-American population ismuch smaller (about 2.1%) and the Hispanic population constitutes ahigher percentage of the State’s total (about 46.3%)? See UnitedStates Census Bureau, QuickFacts, online athttps://www.census.gov/quickfacts/table/PST045215/35,48 (allInter-net materials as last visited June 21, 2016).UT’s answer to this question has veered back andforth. At oral argument in Fisher I, UT’s lawyer indicatedthat critical mass “could” vary “from group to group” and from“state to state.” See Tr. of Oral Arg. 40 (Oct. 10, 2012). And UTinitially justified its race-conscious plan at least in part on theground that “significant differences between the racial and ethnicmakeup of the University’s undergraduate population and the state’spopulation prevent the University from fully achieving itsmission.” Supp. App. 24a; see also id., at 16a (“[A]critical mass in Texas is necessarily larger than a critical massin Michigan,” because “[a] majority of the college-age populationin Texas is African American or Hispanic”); Fisher, 631F.3d, at 225–226, 236 (concluding that UT’s reliance on Texasdemographics reflects “measured attention to the community itserves”); Brief for Respondents in No. 11–345, at41 (notingthat critical mass may hinge, in part, on “the communities thatuniversities serve”). UT’s extensive reliance on state demographicsis also revealed by its substantial focus on increasing therepresentation of Hispanics, but not Asian-Americans, see,e.g., 645 F.Supp. 2d, at 606; Supp. App. 25a; App.445a–446a, because Hispanics, but not Asian-Americans, areunderrepre-sented at UT when compared to the demographics of theState.[2]On the other hand, UT’s counsel asserted thatthe critical mass for the University is “not at all” dependent onthe demographics of Texas, and that UT’s “concept [of] criticalmass isn’t tied to demographic[s].” Tr. of Oral Arg. 40, 49 (Oct.10, 2012). And UT’s Fisher I brief expressly agreed that “auniversity cannot look to racial demographics—and then workbackward in its admissions process to meet a target tied to suchdemographics.” Brief for Respondents in No. 11–345, at31; seealso Brief for Respondents 26–27 (disclaiming any interest indemographic parity).To the extent that UT is pursuing parity withTexas demographics, that is nothing more than “outright racialbalancing,” which this Court has time and again held “patentlyunconstitutional.” Fisher I, 570 U.S., at ___ (slipop., at 9); see Grutter, 539 U.S., at 330 (“[O]utrightracial balancing ... is patently unconstitutional”);Freeman v. Pitts, 503 U.S. 467, 494 (1992)(“Racial balance is not to be achieved for its own sake”);Croson, 488 U.S., at 507 (rejecting goal of “outrightracial balancing”); Bakke, 438 U.S., at 307 (opinionof Powell, J.) (“If petitioner’s purpose is to assure within itsstudent body some specified percentage of a particular group merelybecause of its race or ethnic origin, such a preferential purposemust be rejected ... as facially invalid”). An interest“linked to nothing other than proportional representation ofvarious races ... would support indefinite use ofracial classifications, employed first to obtain the appropriatemixture of racial views and then to ensure that the [program]continues to reflect that mixture.” Metro Broadcasting, Inc.v. FCC, 497 U.S. 547, 614 (1990) (O’Connor, J.,dissenting). And as we held in Fisher I, “‘[r]acialbalancing is not transformed from “patently unconstitutional” to acompelling state interest simply by relabeling it “racialdiver-sity.”’” 570 U.S., at ___ (slip op., at 9)(quoting Parents Involved, 551 U.S., at 732).The record here demonstrates the pitfallsinherent in racial balancing. Although UT claims an interest in theeducational benefits of diversity, it appears to have paid littleattention to anything other than the number of minority students onits campus and in its classrooms. UT’s 2004 Proposal illustratesthis approach by repeatedly citing numerical assessments of theracial makeup of the student body and various classes as thejustification for adopting a race-conscious plan. See, e.g.,Supp. App. 24a–26a, 30a. Instead of focusing on the benefits ofdiversity, UT seems to have resorted to a simple racial census.The majority, for its part, claims that“[a]lthough demographics alone are by no means dispositive, they dohave some value as a gauge of the University’s ability to enrollstudents who can offer underrepresented perspectives.” Ante,at 14. But even if UT merely “view[s] the demographic disparity ascause for concern,” Brief for United States as Amicus Curiae29, and is seeking only to reduce—rather than eliminate—thedisparity, that undefined goal cannot be properly subjected tostrict scrutiny. In that case, there is simply no way for a courtto know what specific demographic interest UT is pursuing, why arace-neutral alternative could not achieve that interest, and whenthat demographic goal would be satisfied. If a demographicdiscrepancy can serve as “a gauge” that justifies the use of racialdiscrimination, ante, at 14, then racial discrimination canbe justified on that basis until demographic parity is reached.There is no logical stopping point short of patentlyunconstitutional racial balancing. Demographic disparities thuscannot be used to satisfy strict scrutiny here. See Croson,supra, at 498 (rejecting a municipality’s assertion that itsracial set-aside program was justified in light of pastdiscrimination because that assertion had “‘no logicalstopping point’” and could continue until the percentage ofgovernment contracts awarded to minorities “mirrored the percentageof minorities in the population as a whole”); Wygant v.Jackson Bd. of Ed., 476 U.S. 267, 275 (1986)(plurality opinion) (rejecting the government’s asserted interestbecause it had “no logical stopping point”).2The other major explanation UT offered in theProposal was its desire to promote classroom diversity. TheProposal stressed that UT “has not reached a critical mass at theclassroom level.” Supp. App. 24a (emphasis added); see alsoid., at 1a, 25a, 39a; App. 316a. In support of thisproposition, UT relied on a study of select classes containing fiveor more students. As noted above, the study indicated that 52% ofthese classes had no African-Americans, 16% had no Asian-Americans,and 12% had no Hispanics. Supp. App. 26a. The study furthersuggested that only 21% of these classes had two or moreAfrican-Americans, 67% had two or more Asian-Americans, and 70% hadtwo or more Hispanics. See ibid. Based on this study, UTconcluded that it had a “compelling educational interest” inemploying racial preferences to ensure that it did not “have largenumbers of classes in which there are no students—or only a singlestudent—of a given underrepresented race or ethnicity.” Id.,at 25a.UT now equivocates, disclaiming any discreteinterest in classroom diversity. See Brief for Respondents 26–27.Instead, UT has taken the position that the lack of classroomdiversity was merely a “red flag that UT had not yet fullyrealized” “the constitutionally permissible educational benefits ofdiversity.” Brief for Respondents in No. 11–345, at 43. But UT hasfailed to identify the level of classroom diversity it deemssufficient, again making it im-possible to apply strictscrutiny.[3] A reviewing courtcannot determine whether UT’s race-conscious programwas necessaryto remove the so-called “red flag” without understanding theprecise nature of that goal or know-ing when the “red flag” will beconsidered to havedisappeared.Putting aside UT’s effective abandonment of itsinterest in classroom diversity, the evidence cited in support ofthat interest is woefully insufficient to show that UT’srace-conscious plan was necessary to achieve the educationalbenefits of a diverse student body. As far as the record shows, UTfailed to even scratch the surface of the available data beforereflexively resorting to racial preferences. For instance, becauseUT knows which students were admitted through the Top Ten PercentPlan and which were not, as well as which students enrolled inwhich classes, it would seem relatively easy to determine whetherTop Ten Percent students were more or less likely than holisticadmittees to enroll in the types of classes where diversity waslacking. But UT never bothered to figure this out. See ante,at 9 (acknowledging that UT submitted no evidence regarding “howstudents admitted solely based on their class rank differ in theircontribution to diversity from students admitted through holisticreview”). Nor is there any indication that UT instructed admissionsofficers to search for African-American and Hispanic applicants whowould fill particular gaps at the classroom level. Given UT’sfailure to present such evidence, it has not demonstrated that itsrace-conscious policy would promote classroom diversity any betterthan race-neutral options, such as expanding the Top Ten PercentPlan or using race-neutral holistic admissions.Moreover, if UT is truly seeking to expose itsstudents to a diversity of ideas and perspectives, its policy ispoorly tailored to serve that end. UT’s own study—which themajority touts as the best “nuanced quantitative data” supportingUT’s position, ante, at 15—demonstrated that classroomdiversity was more lacking for students classified asAsian-American than for those classified as Hispanic. Supp. App.26a. But the UT plan discriminates against Asian-Americanstudents.[4] UT is apparentlyunconcerned that Asian-Americans “may be made to feel isolated ormay be seen as ... ‘spokesperson[s]’ of their race orethnicity.” Id., at 69a; see id., at 25a. And unlessthe University is engaged in unconstitutional racial balancingbased on Texas demographics (where Hispanics outnumberAsian-Americans), see Part II–C–1, supra, it seemingly viewsthe classroom contributions of Asian-American students as lessvaluable than those of Hispanic students. In UT’s view, apparently,“Asian Americans are not worth as much as Hispanics in promoting‘cross-racial understanding,’ breaking down ‘racial stereotypes,’and enabling students to ‘better understand persons of differentraces.’” Brief for Asian American Legal Foundationetal. as Amici Curiae 11 (representing 117Asian-American organizations). The majority opinion effectivelyendorses this view, crediting UT’s reliance on the classroom studyas proof that the University assessed its need for racialdiscrimination (including racial discrimination that undeniablyharms Asian-Americans) “with care.” Ante, at 15.While both the majority and the Fifth Circuitrely on UT’s classroom study, see ante, at 15; 758 F. 3d, at658–659, they completely ignore its finding that Hispanics arebetter represented than Asian-Americans in UT classrooms. In fact,they act almost as if Asian-American students do not exist. Seeante, at 14 (mentioning Asian-Americans only a single timeoutside of parentheticals, and not in the context of the classroomstudy); 758 F. 3d, at 658 (mentioning Asian-Americans only a singletime).[5] Only the DistrictCourt acknowledged the impact of UT’s policy on Asian-Americanstudents. But it brushed aside this impact,concluding—astoundingly—that UT can pick and choose which racialand ethnic groups it would like to favor. According to the DistrictCourt, “nothing in Grutter requires a university to giveequal preference to every minority group,” and UT is allowed “toexercise its discretion in determining which minority groups shouldbenefit from the consideration of race.” 645 F. Supp. 2d, at606.This reasoning, which the majority implicitlyaccepts by blessing UT’s reliance on the classroom study, placesthe Court on the “tortuous” path of “decid[ing] which races tofavor.” Metro Broadcasting, 497 U.S., at 632 (Kennedy,J., dissenting). And the Court’s willingness to allow this“discrimination against individuals of Asian descent in UTadmissions is particularly troubling, in light of the long historyof discrimination against Asian Americans, especially ineducation.” Brief for Asian American Legal Foundation etal.as Amici Curiae 6; see also, e.g., id., at 16–17(discussing the placement of Chinese-Americans in “‘separatebut equal’” public schools); Gong Lum v. Rice,275 U.S. 78 –82 (1927) (holding that a9-year-oldChinese-American girl could be denied entry to a “white”school because she was “a member of the Mongolian or yellow race”).In sum, “[w]hile the Court repeatedly refers to the preferences asfavoring ‘minorities,’ ... it must be emphasized thatthe discriminatory policies upheld today operate to exclude”Asian-American students, who “have not made [UT’s] list” of favoredgroups. Metro Broadcasting, supra, at 632 (Kennedy,J., dissenting).Perhaps the majority finds discriminationagainst Asian-American students benign, since Asian-Americans are“overrepresented” at UT. 645 F.Supp. 2d, at 606. But“[h]istory should teach greater humility.” MetroBroadcasting, 497 U.S., at 609 (O’Connor, J.,dissenting). “‘[B]enign’ carries with it no independentmeaning, but reflects only acceptance of the current generation’sconclusion that a politically acceptable burden, imposed onparticular citizens on the basis of race, is reasonable.”Id., at 610. Where, as here, the government has providedlittle explanation for why it needs to discriminate based on race,“‘there is simply no way of determining what classificationsare “benign” ... and what classifications are in factmotivated by illegitimate notions of racial inferiority or simpleracial politics.’” Parents Involved, 551 U. S., at 783(opinion of Kennedy, J.) (quoting Croson, 488 U.S., at493 (plurality opinion of O’Connor, J.)). By accepting theclassroom study as proof that UT satisfied strict scrutiny, themajority “move[s] us from ‘separate but equal’ to ‘un-equal butbenign.’” Metro Broadcasting, supra, at 638(Kennedy, J., dissenting).In addition to demonstrating that UTdiscriminates against Asian-American students, the classroom studyalso exhibits UT’s use of a few crude, overly simplistic racial andethnic categories. Under the UT plan, both the favored and thedisfavored groups are broad and consist of students from enormouslydiverse backgrounds. See Supp. App. 30a; see also Fisher I,570 U.S., at ___ (slip op., at 4) (“five predefined racialcategories”). Because “[c]rude measures of this sort threaten toreduce [students] to racial chits,” Parents Involved, 551U.S., at 798 (opinion of Kennedy, J.), UT’s reliance on suchmeasures further undermines any claim based on classroom diver-sitystatistics, see id., at 723 (majority opinion) (criticizingschool policies that viewed race in rough “white/nonwhite” or“black/‘other’” terms); id., at 786 (opinion ofKennedy, J.) (faulting government for relying on “crude racialcat-egories”); Metro Broadcasting, supra, at 633, n.1 (Kennedy, J., dissenting) (concluding that “‘the veryattempt to define with precision a beneficiary’s qualifying racialcharacteristics is repugnant to our constitutional ideals,’”and noting that if the government “‘is to make a seriouseffort to define racial classes by criteria that can beadministered objectively, it must study precedents such as theFirst Regulation to the Reichs Citizenship Law of November 14,1935’”).For example, students labeled “Asian American,”Supp. App. 26a, seemingly include “individuals of Chinese,Japanese, Korean, Vietnamese, Cambodian, Hmong, Indian and otherbackgrounds comprising roughly 60% of the world’s population,”Brief for Asian American Legal Foundation etal. as AmiciCuriae, O. T. 2012, No. 11–345, p.28.[6] It would be ludicrous to suggest that all ofthese students have similar backgrounds and similar ideas andexperiences to share. So why has UT lumped them together andconcluded that it is appropriate to discriminate againstAsian-American students because they are “overrepresented” in theUT student body? UT has no good answer. And UT makes no effort toensure that it has a critical mass of, say, “Filipino Americans” or“Cambodian Americans.” Tr. of Oral Arg. 52 (Oct. 10, 2012). As longas there are a sufficient number of “Asian Americans,” UT isapparently satisfied.UT’s failure to provide any definition of thevarious racial and ethnic groups is also revealing. UT does notspecify what it means to be “African-American,” “His-panic,” “AsianAmerican,” “Native American,” or “White.” Supp. App. 30a. And UTevidently labels each student as falling into only a single racialor ethnic group, see, e.g., id., at 10a–13a, 30a,43a–44a, 71a, 156a–157a, 169a–170a, without explaining howindividuals with ancestors from different groups are to becharacterized. As racial and ethnic prejudice recedes, more andmore students will have parents (or grandparents) who fall intomore than one of UT’s five groups. According to census figures,individuals describing themselves as members of multiple races grewby 32% from 2000 to 2010.[7] Arecent survey reported that 26% of Hispanics and 28% ofAsian-Americans marry a spouse of a different race orethnicity.[8] UT’s crudeclassification system is ill suited for the more integrated countrythat we are rapidly becoming. UT assumes that if an applicantdescribes himself or herself as a member of a particular race orethnicity, that applicant will have a perspective that differs fromthat of applicants who describe themselves as members of differentgroups. But is this necessarily so? If an applicant hasonegrandparent, great-grandparent, or great-great-grandparent who wasa member of a favored group, is that enough to permit UT to inferthat this student’s classroom contribution will reflect adistinctive perspective or set of experiences associated with thatgroup? UT does not say. It instead relies on applicants to“classify themselves.” Fisher I, 570 U.S., at ___(slip op., at 4). This is an invitation for applicants to game thesystem.Finally, it seems clear that the lack ofclassroom diver-sity is attributable in good part to factors otherthan the representation of the favored groups in the UT studentpopulation. UT offers an enormous number of classes in a wide rangeof subjects, and it gives undergraduates a very large measure offreedom to choose their classes. UT also offers courses in subjectsthat are likely to have special appeal to members of the minoritygroups given preferential treatment under its challenged plan, andthis of course diminishes the number of other courses in whichthese students can enroll. See, e.g., Supp. App. 72a–73a(indicating that the representation of African-Americans andHispanics in UT classrooms varies substantially from major tomajor). Having designed an undergraduate program that virtuallyensures a lack of classroom diver-sity, UT is poorly positioned toargue that this very result provides a justification for racial andethnic discrimination, which the Constitution rarely allows.3UT’s purported interest in intraracialdiversity, or “diversity within diversity,” Brief for Respondents34, also falls short. At bottom, this argument relies on theunsupported assumption that there is something deficient or atleast radically different about the African-American and Hispanicstudents admitted through the Top Ten Percent Plan.Throughout this litigation, UT has repeatedlyshifted its position on the need for intraracial diversity.Initially, in the 2004 Proposal, UT did not rely on this allegedneed at all. Rather, the Proposal “examined two metrics—classroomdiversity and demographic disparities—that it concluded wererelevant to its ability to provide [the] benefits of diversity.”Brief for United States as Amicus Curiae 27–28. Thosemetrics looked only to the numbers of African-Americans andHispanics, not to diversity within each group.On appeal to the Fifth Circuit and in FisherI, however, UT began to emphasize its intraracial diversityargument. UT complained that the Top Ten Percent Law hinders itsefforts to assemble a broadly diverse class because the minoritiesadmitted under that law are drawn largely from certain areas ofTexas where there are majority-minority schools. These students, UTargued, tend to come from poor, disadvantaged families, and theUniver-sity would prefer a system that gives it substantial leewayto seek broad diversity within groups of underrepresentedminorities. In particular, UT asserted a need for moreAfrican-American and Hispanic students from privileged backgrounds.See, e.g., Brief for Respondents in No. 11–345, at 34(explaining that UT needs race-conscious admissions in order toadmit “[t]he African-American or Hispanic child of successfulprofessionals in Dallas”); ibid. (claiming that privilegedminorities “have great potential for serving as a ‘bridge’ inpromoting cross-racial understanding, as well as in breaking downracial stereotypes”); ibid. (intimating that theunderprivileged minority students admitted under the Top TenPercent Plan “reinforc[e]” “stereotypical assumptions”); Tr.of Oral Arg. 43–45 (Oct. 10, 2012) (“[A]lthough the percentage plancertainly helps with minority admissions, by and large, the—theminorities who are admitted tend to come from seg-regated,racially-identifiable schools,” and “we wantminorities fromdifferent backgrounds”). Thus, the Top Ten Percent Law is faultedfor admitting the wrong kind of African-American and Hispanicstudents.The Fifth Circuit embraced this argument onremand, endorsing UT’s claimed need to enroll minorities from“high-performing,” “majority-white” high schools. 758 F.3d,at 653. According to the Fifth Circuit, these more privilegedminorities “bring a perspective not captured by” students admittedunder the Top Ten Percent Law, who often come “from highlysegregated, underfunded, and underperforming schools.” Ibid.For instance, the court determined, privileged minorities “canenrich the diversity of the student body in distinct ways” becausesuch students have “higher levels of preparation and betterprospects for admission to UT Austin’s more demanding colleges”than underprivileged minorities. Id., at 654; see alsoFisher, 631 F.3d, at 240, n. 149 (concluding thattheTop Ten Percent Plan “widens the ‘credentials gap’ between minorityand non-minority students at the University, which risks drivingaway matriculating minor-ity students from difficult majors likebusiness or the sciences”).Remarkably, UT now contends that petitioner has“fabricat[ed]” the argument that it is seeking affluent minorities.Brief for Respondents 2. That claim is impossible to square withUT’s prior statements to this Court in the briefing and oralargument in Fisher I.[9]Moreover, although UT reframes its argument, it continues to assertthat it needs affirmative action to admit privileged minorities.For instance, UT’s brief highlights its interest in admitting“[t]he black student with high grades from Andover.” Brief forRespondents 33. Similarly, at oral argument, UT claimed that its“interests in the educa-tional benefits of diversity would not bemet if all of [the] minority students were ... comingfrom depressed socioeconomic backgrounds.” Tr. of Oral Arg. 53(Dec. 9, 2015); see also id., at 43, 45.Ultimately, UT’s intraracial diversity rationalerelies on the baseless assumption that there is something wrongwith African-American and Hispanic students admitted through theTop Ten Percent Plan, because they are “from the lower-performing,racially identifiable schools.” Id., at 43; see id.,at 42–43 (explaining that “the basis” for UT’s conclusion that itwas “not getting a variety of perspectives among African-Americansor Hispanics” was the fact that the Top Ten Percent Plan admitsunderprivileged minorities from highly segregated schools). Ineffect, UT asks the Court “to assume”—without anyevidence—“that minorities admitted under the Top Ten Percent Law... are somehow more hom*ogenous, less dynamic, and moreundesirably stereotypical than those admitted under holisticreview.” 758 F.3d, at 669–670 (Garza, J., dissenting). AndUT’s assumptions appear to be based on the pernicious stereotypethat the African-Americans and Hispanics admitted through the TopTen Percent Plan only got in because they did not have to competeagainst very many whites and Asian-Americans. See Tr. of Oral Arg.42–43 (Dec. 9, 2015). These are “the very stereotypical assumptions[that] the Equal Protection Clause forbids.” Miller, 515U.S., at 914. UT cannot satisfy its burden by attempting to“substitute racial stereotype for evidence, and racial prejudicefor reason.” Calhoun v. United States, 568 U.S.___, ___ (2013) (slip op., at 4) (Sotomayor, J., respecting denialof certiorari).In addition to relying on stereotypes, UT’sargument that it needs racial preferences to admit privilegedminorities turns the concept of affirmative action on its head.When affirmative action programs were first adopted, it was for thepurpose of helping the disadvantaged. See, e.g., Bakke, 438U.S., at 272–275 (opinion of Powell, J.) (explaining that theschool’s affirmative action program was designed “to increase therepresentation” of “‘economically and/or educationallydisadvantaged’ applicants”). Now we are told that a programthat tends to admit poor and disadvantaged minority students isinadequate because it does not work to the advantage of those whoare more fortunate. This is affirmative action gone wild.It is also far from clear that UT’s assumptionsabout the socioeconomic status of minorities admitted through theTop Ten Percent Plan are even remotely accurate. Take, for example,parental education. In 2008, when petitioner applied to UT,approximately 79% of Texans aged 25 years or older had a highschool diploma, 17% had a bachelor’s degree, and 8% had a graduateor professional degree. Dept. of Educ., Nat. Center for Educ.Statistics, T. Snyder & S. Dillow,Digest of EducationStatistics 2010, p.29 (2011). In contrast, 96% ofAfrican-Americans admitted through the Top Ten Percent Plan had aparent with a high school diploma, 59% had a parent with abachelor’s degree, and 26% had a parent with a graduate orprofessional degree. See UT, Office of Admissions, Student Profile,Admitted Freshman Class of 2008, p.8 (rev. Aug. 1,2012) (2008Student Profile), online athttps://uteas.app.box.com/s/twqozsbm2vb9lhm14o0v0czvqs1ygzqr/1/7732448553/23476747441/1.Similarly, 83% of Hispanics admitted through the Top Ten PercentPlan had a parent with a high school diploma, 42% had a parent witha bachelor’s degree, and 21% had a parent with a graduate orprofessional degree. Ibid. As these statistics make plain,the minorities that UT characterizes as “coming from depressedsocioeconomic backgrounds,” Tr. of Oral Arg. 53 (Dec. 9, 2015),generally come from households with education levels exceeding thenorm in Texas.Or consider income levels. In 2008, the medianannual household income in Texas was $49,453. United States CensusBureau, A. Noss, Household Income for States: 2008 and 2009,p.4 (2010), online athttps://www.census.gov/prod/2010pubs/acsbr09-2.pdf. The householdincome levels for Top Ten Percent African-American and Hispanicadmittees were on par: Roughly half of such admittees came fromhouseholds below the Texas median, and half came from householdsabove the median. See 2008 Student Profile 6. And a large portionof these admittees are from households with income levels farexceeding the Texas median. Specifically, 25% of African-Americansand 27% of Hispanics admitted through the Top Ten Percent Plan in2008 were raised in households with incomes exceeding $80,000.Ibid. In light of this evidence, UT’s actual argument is notthat it needs affirmative action to ensure that its minorityadmittees are representative of the State of Texas. Rather, UT isasserting that it needs affirmative action to ensure that itsminority students disproportionally come from families that arewealthier and better educated than the average Texas family.In addition to using socioeconomic status tofalsely denigrate the minority students admitted through the TopTen Percent Plan, UT also argues that such students areacademically inferior. See, e.g., Brief for Respondents inNo. 11–345, at 33 (“[T]he top 10% law systematically hinders UT’sefforts to assemble a class that is ... academicallyexcellent”). “On average,” UT claims, “African-American andHispanic holistic admits have higher SAT scores than their Top 10%counterparts.” Brief for Respondents 43, n.8. As a result, UTargues that it needs race-conscious admissions to enrollacademically superior minority students with higher SAT scores.Regrettably, the majority seems to embrace this argument as well.See ante, at 16 (“[T]he Equal Protection Clause does notforce universities to choose between a diverse student body and areputation for academic excellence”).This argument fails for a number of reasons.First, it is simply not true that Top Ten Percent minorityadmittees are academically inferior to holistic admittees. In fact,as UT’s president explained in 2000, “top 10 percent high schoolstudents make much higher grades in college than non-top 10 percentstudents,” and “[s]trong academic performance in high school is aneven better predictor of success in college than standardized testscores.” App. 393a–394a; see also Lavergne Deposition 41–42(agreeing that “it’s generally true that students admitted pursuantto HB 588 [the Top Ten Percent Law] have a higher level of academicperformance at the University than students admitted outside of HB588”). Indeed, the statistics in the record reveal that, for eachyear between 2003 and 2007, African-American in-state freshmen whowere admitted under the Top Ten Percent Law earned a higher meangrade point average than those admitted outside of the Top TenPercent Law. Supp. App. 164a. The same is true for Hispanicstudents. Id., at 165a. These conclusions correspond to theresults of nationwide studies showing that high school grades are abetter predictor of success in college than SAT scores.[10]It is also more than a little ironic that UTuses the SAT, which has often been accused of reflecting racial andcultural bias,[11] as areason for dissatisfaction with poor and disadvantagedAfrican-American and Hispanic students who excel both in highschool and in college. Even if the SAT does not reflect such bias(and I am ill equipped to express a view on that subject), SATscores clearly correlate with wealth.[12]UT certainly has a compelling interest inadmitting students who will achieve academic success, but it doesnot follow that it has a compelling interest in maximizingadmittees’ SAT scores. Approximately 850 4-year-degree institutionsdo not require the SAT or ACT as part of the admissions process.See J. Soares, SAT Wars: The Case for Test-Optional CollegeAdmissions 2 (2012). This includes many excellent schools.[13]To the extent that intraracial diversity refersto something other than admitting privileged minorities andminorities with higher SAT scores, UT has failed to define thatinterest with any clarity. UT “has not provided any concretetargets for admitting more minority students possessing [the]unique qualitative-diversity characteristics” it desires. 758F.3d, at 669 (Garza, J., dissenting). Nor has UT specifiedwhich characteristics, viewpoints, and life experiences aresupposedly lacking in the African-Americans and Hispanics admittedthrough the Top Ten Percent Plan. In fact, because UTadministrators make no collective, qualitative assessment of theminorities admitted automatically, they have no way of knowingwhich attributes are missing. See ante, at 9 (admitting thatthere is no way of knowing “how students admitted solely based ontheir class rank differ in their contribution to diversity fromstudents admitted through holistic review”); 758 F.3d, at 669(Garza, J., dissenting) (“The University does not assess whetherTop Ten Percent Law admittees exhibit sufficient diversity withindiversity, whether the requisite ‘change agents’ are among them,and whether these admittees are able, collectively or individually,to combat pernicious stereotypes”). Furthermore, UT has notidentified “when, if ever, its goal (which remains undefined) forqualitative diversity will be reached.” Id., at 671. UT’sintraracial diversity rationale is thus too imprecise to permitstrict scrutiny analysis.Finally, UT’s shifting positions on intraracialdiversity, and the fact that intraracial diversity was notemphasized in the Proposal, suggest that it was not “the actualpurpose underlying the discriminatory classification.”Mississippi Univ. for Women v. Hogan, 458 U.S.718, 730 (1982) . Instead, it appears to be a post hocrationalization.4UT also alleges—and the majority embraces—aninterest in avoiding “feelings of loneliness and isolation” amongminority students. Ante, at 14–15; see Brief for Respondents7–8, 38–39. In support of this argument, they cite only demographicdata and anecdotal statements by UT officials that some students(we are not told how many) feel “isolated.” This vague interestcannot possibly satisfy strict scrutiny.If UT is seeking demographic parity to avoidisolation, that is impermissible racial balancing. See Part II–C–1,supra. And linking racial loneliness and isolation to statedemographics is illogical. Imagine, for example, that anAfrican-American student attends a university that is 20%African-American. If racial isolation depends on a comparison tostate demographics, then that student is more likely to feelisolated if the school is located in Mississippi (which is 37.0%African-American) than if it is located in Montana (which is 0.4%African-American). See United States Census Bureau, QuickFacts,online at https://www.census.gov/quickfacts/table/PST045215/28,30.In reality, however, the student may feel—if anything—lessisolated in Mississippi, where African-Americans are more prevalentin the population at large.If, on the other hand, state demographics arenot driving UT’s interest in avoiding racial isolation, then itstreatment of Asian-American students is hard to understand. As theDistrict Court noted, “the gross number of Hispanic studentsattending UT exceeds the gross number of Asian-American students.”645 F.Supp. 2d, at 606. In 2008, for example, UT enrolled1,338 Hispanic freshmen and 1,249 Asian-American freshmen. Supp.App. 156a. UT never explains why the Hispanic students—but not theAsian-American students—are isolated and lonely enough to receivean admissions boost, notwithstanding the fact that there are moreHispanics than Asian-Americans in the student population. Theanecdotal statements from UT officials certainly do not indicatethat Hispanics are somehow lonelier than Asian-Americans.Ultimately, UT has failed to articulate itsinterest in preventing racial isolation with any clarity, and ithas provided no clear indication of how it will know when suchisolation no longer exists. Like UT’s purported interests indemographic parity, classroom diversity, and intraracial diversity,its interest in avoiding racial isolation cannot justify the use ofracial preferences.DEven assuming UT is correct that, underGrutter, it need only cite a generic interest in theeducational benefits of diversity, its plan still fails strictscrutiny because it is not narrowly tailored. Narrow tailoringrequires “a careful judicial inquiry into whether a universitycould achieve sufficient diversity without using racialclassifications.” Fisher I, 570 U.S., at ___ (slipop., at 10). “If a ‘“nonracial approach ... couldpromote the substantial interest about as well and at tolerableadministrative expense,”’ then the university may notconsider race.” Id., at ___ (slip op., at 11) (citationsomitted). Here, there is no evidence that race-blind, holisticreview would not achieve UT’s goals at least “about as well” asUT’s race-based policy. In addition, UT could have adopted otherapproaches to further its goals, such as intensifying its outreachefforts, uncapping the Top Ten Percent Law, or placing greaterweight on socioeconomic factors.The majority argues that none of thesealternatives is “a workable means for the University to attain thebenefits of diversity it sought.” Ante, at 16. Tellingly,however, the majority devotes only a single, conclusory sentence tothe most obvious race-neutral alternative: race-blind, holisticreview that considers the applicant’s unique characteristics andpersonal circ*mstances. See ibid.[14] Under a system that combines the Top Ten PercentPlan with race-blind, holistic review, UT could still admit “thestar athlete or musician whose grades suffered because of dailypractices and training,” the “talented young biologist whostruggled to maintain above-average grades in humanities classes,”and the “student whose freshman-year grades were poor because of afamily crisis but who got herself back on track in her last threeyears of school.” Ante, at 17. All of these uniquecirc*mstances can be considered without injecting race into theprocess. Because UT has failed to provide any evidence whatsoeverthat race-conscious holistic review will achieve its diversityobjectives more effectively than race-blind holistic review, itcannot satisfy the heavy burden imposed by the strict scrutinystandard.The fact that UT’s racial preferences areunnecessary to achieve its stated goals is further demonstrated bytheir minimal effect on UT’s diversity. In 2004, when race was nota factor, 3.6% of non-Top Ten Percent Texas enrollees wereAfrican-American and 11.6% were Hispanic. See Supp. App. 157a. Itwould stand to reason that at least the same percentages ofAfrican-American and Hispanic students would have been admittedthrough holistic review in 2008 even if race were not a factor. Ifthat assumption is correct, then race was determinative for only 15African-American students and 18 Hispanic students in 2008(representing 0.2% and 0.3%, respectively, of the total enrolledfirst-time freshmen from Texas high schools). Seeibid.[15]The majority contends that “[t]he fact that raceconsciousness played a role in only a small portion of admissionsdecisions should be a hallmark of narrow tailoring, not evidence ofunconstitutionality.” Ante, at 15. This argument directlycontradicts this Court’s precedent. Because racial classificationsare “‘a highly suspect tool,’” Grutter, 539U.S, at 326, they should be employed only “as a last resort,”Croson, 488 U.S., at 519 (opinion of Kennedy, J.); seealso Grutter, supra, at 342 (“[R]acialclassifications, however compelling their goals, are potentially sodangerous that they may be employed no more broadly than theinterest demands”). Where, as here, racial preferences have only aslight impact on minority enrollment, a race-neutral alternativelikely could have reached the same result. See ParentsInvolved, 551 U.S., at 733–734 (holding that the “minimaleffect” of school districts’ racial classifications “casts doubt onthe necessity of using [such] classifications” and “suggests thatother means [of achieving their objectives] would be effective”).As Justice Kennedy once aptly put it, “the small number of[students] affected suggests that the schoo[l] could have achieved[its] stated ends through different means.” Id., at 790(opinion concurring in part and concurring in judgment). And inthis case, a race-neutral alternative could accomplish UT’sobjectives without gratuitously branding the covers of tens ofthousands of applications with a bare racial stamp and “tell[ing]each student he or she is to be defined by race.” Id., at789.IIIThe majority purports to agree with much ofthe above analysis. The Court acknowledges that “‘becauseracial characteristics so seldom provide a relevant basis fordisparate treatment,’” “‘[r]ace may not be considered[by a university] unless the admissions process can withstandstrict scrutiny.’” Ante, at 6–7. The Court admits thatthe burden of proof is on UT, ante, at 7, and that “auniversity bears a heavy burden in showing that it had not obtainedthe educational benefits of diversity before it turned to arace-conscious plan,” ante, at 13–14. And the Courtrecognizes that the record here is “almost devoid of informationabout the students who secured admission to the Univer-sity throughthe Plan,” and that “[t]he Court thus cannot know how studentsadmitted solely based on their class rank differ in theircontribution to diversity from students admitted through holisticreview.” Ante, at 9. This should be the end of the case:Without identifying what was missing from the African-American andHispanic students it was already admitting through its race-neutralprocess, and without showing how the use of race-based admissionscould rectify the deficiency, UT cannot demonstrate that itsprocedure is narrowly tailored.Yet, somehow, the majority concludes thatpetitioner must lose as a result of UT’s failure to provideevidence justifying its decision to employ racial discrimination.Tellingly, the Court frames its analysis as if petitioner bears theburden of proof here. See ante, at 11–19. But it is not thepetitioner’s burden to show that the consideration of race isunconstitutional. To the extent the record is inadequate, theresponsibility lies with UT. For “[w]hen a court subjectsgovernmental action to strict scrutiny, it cannot construeambiguities in favor of the State,” Parents Involved,supra, at 786 (opinion of Kennedy, J.), particularly where,as here, the summary judgment posture obligates the Court to viewthe facts in the light most favorable to petitioner, seeMatsush*ta Elec. Industrial Co. v. Zenith RadioCorp., 475 U.S. 574, 587 (1986) .Given that the University bears the burden ofproof, it is not surprising that UT never made the argument that itshould win based on the lack of evidence. UT instead assertsthat “if the Court believes there are any deficiencies in [the]record that cast doubt on the constitutionality of UT’s policy, theanswer is to order a trial, not to grant summary judgment.” Brieffor Respondents 51; see also id., at 52–53 (“[I]f this Courthas any doubts about how the Top 10% Law works, or how UT’sholistic plan offsets the tradeoffs of the Top 10% Law, the answeris to remand for a trial”). Nevertheless, the majority cites threereasons for breaking from the normal strict scrutiny standard. Noneof these is convincing.AFirst, the Court states that, while “th[e]evidentiary gap perhaps could be filled by a remand to the districtcourt for further factfinding” in “an ordinary case,” that will notwork here because “[w]hen petitioner’s application was rejected,... the University’s combinedpercentage-plan/holistic-review approach to admission had been ineffect for just three years,” so “further factfinding” “might yieldlittle insight.” Ante, at 9. This reasoning is dangerouslyincorrect. The Equal Protection Clause does not provide a 3-yeargrace period for racial discrimination. Under strict scrutiny, UTwas required to identify evidence that race-based admissions werenecessary to achieve a compelling interest before it putthem in place—not three or more years after. See ante, at13–14 (“Petitioner is correct that a university bears a heavyburden in showing that it had not obtained the educational benefitsof diversity before it turned to a race-conscious plan”(emphasis added)); Fisher I, 570 U.S., at ___ (slipop., at 11) (“[S]trict scrutiny imposes on the university theultimate burden of demonstrating, before turning to racialclassifications, that available, workable race-neutral alternativesdo not suffice” (emphasis added)). UT’s failure to obtain actualevidence that racial preferences were necessary before resolving touse them only confirms that its decision to inject race intoadmissions was a reflexive response to Grutter,[16] and that UT did not seriouslyconsider whether race-neutral means would serve its goals as wellas a race-based process.BSecond, in an effort to excuse UT’s lack ofevidence, the Court argues that because “the University lacks anyauthority to alter the role of the Top Ten Percent Plan,” “itsimilarly had no reason to keep extensive data on the Plan or thestudents admitted under it—particularly in the years beforeFisher I clarified the stringency of the strict-scrutinyburden for a school that employs race-conscious review.”Ante, at 9–10. But UT has long been aware that it bears theburden of justifying its racial discrimination under strictscrutiny. See, e.g., Brief for Respondents in No. 11–345, at22 (“It is undisputed that UT’s consideration of race in itsholistic admissions process triggers strict scrutiny,” and “thatinquiry is undeniably rigorous”).[17] In light of this burden, UT had every reasonto keep data on the students admitted through the Top Ten PercentPlan. Without such data, how could UT have possibly identified anycharacteristics that were lacking in Top Ten Percent admittees andthat could be obtained via race-conscious admissions? How could UTdetermine that employing a race-based process would serve its goalsbetter than, for instance, expanding the Top Ten Percent Plan? UTcould not possibly make such determinations without studying thestudents admitted under the Top Ten Percent Plan. Its failure to doso demonstrates that UT unthinkingly employed a race-based processwithout examining whether the use of race was actually necessary.This is not—as the Court claims—a “good-faith effor[t] to complywith the law.” Ante, at 10.The majority’s willingness to cite UT’s “goodfaith” as the basis for excusing its failure to adduce evidence isparticularly inappropriate in light of UT’s well-documented absenceof good faith. Since UT described its admissions policy to thisCourt in Fisher I, it has been revealed that thisdescription was incomplete. As explained in an independentinvestigation into UT admissions, UT maintained a clandestineadmissions system that evaded public scrutiny until a formeradmissions officer blew the whistle in 2014. See Kroll, Inc.,Univer-sity of Texas at Austin—Investigation of AdmissionsPrac-tices and Allegations of Undue Influence 4 (Feb. 6, 2015)(Kroll Report). Under this longstanding, secret process, universityofficials regularly overrode normal holistic review to allowpolitically connected individuals—such as donors, alumni,legislators, members of the Board of Regents, and UT officials andfaculty—to get family members and other friends admitted to UT,despite having grades and standardized test scores substantiallybelow the median for admitted students. Id., at 12–14; seealso Blanchard & Hoppe, Influential Texans HelpedUnderqualified Students Get Into UT, Dallas Morning News, July 20,2015, online at http://www.dallasnews.com/news/education/headlines/20150720-influential-texans-helped-underqualified-students-get-into-ut.ece(“Dozens of highly influential Texans—including lawmakers,millionaire donors and university regents—helped underqualifiedstudents get into the University of Texas, often by writing to UTofficials, records show”).UT officials involved in this covert processintentionally kept few records and destroyed those that did exist.See, e.g., Kroll Report 43 (“Efforts were made to minimizepaper trails and written lists during this end-of-cycle process. Atone meeting, the administrative assistants tried not keeping anynotes, but this proved difficult, so they took notes and latershredded them. One administrative assistant usually brought tothese meetings a stack of index cards that were subsequentlydestroyed”); see also id., at 13 (finding that “writtenrecords or notes” of the secret admissions meetings “are notmaintained and are typically shredded”). And in the course of thislitigation, UT has been less than forthright concerning itstreatment of well-connected applicants. Compare, e.g., Tr.of Oral Arg. 51 (Dec. 9, 2015) (“University of Texas does not dolegacy, Your Honor”), and App. 281a (“[O]ur legacy policy is suchthat we don’t consider legacy”), with Kroll Report 29 (discussingevidence that “alumni/legacy influence” “results each year incertain applicants receiving a competitive boost or specialconsideration in the admissions process,” and noting that this is“an aspect of the admissions process that does not appear in thepublic representations of UT-Austin’s admissions process”). DespiteUT’s apparent readiness to mislead the public and the Court, themajority is “willing to be satisfied by [UT’s] profession of itsown good faith.” Grutter, 539 U.S., at 394 (Kennedy,J., dissenting).[18]Notwithstanding the majority’s claims to thecontrary, UT should have access to plenty of information about “howstudents admitted solely based on their class rank differ in theircontribution to diversity from students admitted through holisticreview.” Ante, at 9. UT undoubtedly knows which studentswere admitted through the Top Ten Percent Plan and which wereadmitted through holistic review. See, e.g., Supp. App.157a. And it undoubtedly has a record of all of the classes inwhich these students enrolled. See, e.g., UT, Office of theReg-istrar, Transcript—Official, online athttps://registrar.utexas.edu/students/transcripts-official(instructing graduates on how to obtain a transcript listing a“comprehensive record” of classes taken). UT could use thisinformation to demonstrate whether the Top Ten Percent minorityadmittees were more or less likely than the holistic minorityadmittees to choose to enroll in the courses lacking diversity.In addition, UT assigns PAI scores to allstudents—including those admitted through the Top Ten PercentPlan—for purposes of admission to individual majors. Accordingly,all students must submit a full application containing essays,letters of recommendation, a resume, a list of courses taken inhigh school, and a description of any extracurricular activities,leadership experience, or special circ*mstances. See App.212a–214a; 235a–236a; 758 F.3d, at 669, n.14 (Garza,J., dissenting). Unless UT has destroyed these files,[19] it could use them to compare theunique personal characteristics of Top Ten minority admittees withthose of holistic minority admittees, and to determine whether theTop Ten admittees are, in fact, less desirable than the holisticadmittees. This may require UT to expend some resources, but thatis an appropriate burden in light of the strict scrutiny standardand the fact that all of the relevant information is in UT’spossession. The cost of factfinding is a strange basis for awardinga victory to UT, which has a huge budget, and a loss to petitioner,who does not.Finally, while I agree with the majority and theFifth Circuit that Fisher I significantly changed thegoverning law by clarifying the stringency of the strict scrutinystandard,[20] that does notexcuse UT from meeting that heavy burden. In Adarand, forinstance, another case in which the Court clarified the rigor ofthe strict scrutiny standard, the Court acknowledged that itsdecision “alter[ed] the playing field in some important respects.”515 U.S., at 237. As a result, it “remand[ed] the case to thelower courts for further consideration in light of theprinciples [it had] announced.” Ibid. (emphasis added).In other words, the Court made clear that—notwithstanding the shiftin the law—the government had to meet the clarified burden it wasannouncing. The Court did not embrace the notion that its decisionto alter the stringency of the strict scrutiny standard somehowallowed the government to automatically prevail.CThird, the majority notes that this litigationhas persisted for many years, that petitioner has already graduatedfrom another college, that UT’s policy may have changed over time,and that this case may offer little prospective guidance. At most,these considerations counsel in favor of dismissing this case asimprovidently granted. But see, e.g., Gratz, 539 U.S.,at 251, and n.1, 260–262 (rejecting the dissent’s argumentthat, because the case had already persisted long enough for thepetitioners to graduate from other schools, the case should bedismissed); id., at 282 (Stevens, J., dissenting). None ofthese considerations has any bearing whatsoever on the merits ofthis suit. The majority cannot side with UT simply because it istired of this case.IVIt is important to understand what is and whatis not at stake in this case. What is not at stake iswhether UT or any other university may adopt an admissions planthat results in a student body with a broad representation ofstudents from all racial and ethnic groups. UT previously had arace-neutral plan that it claimed had “effectively compensated forthe loss of affirmative action,” App. 396a, and UT could have takenother steps that would have increased the diversity of its admittedstudents without taking race or ethnic background into account.What is at stake is whether universityadministrators may justify systematic racial discrimination simplyby asserting that such discrimination is necessary to achieve “theeducational benefits of diversity,” without explaining—much lessproving—why the discrimination is needed or how the discriminatoryplan is well crafted to serve its objectives. Even though UT hasnever provided any coherent explanation for its asserted need todiscriminate on the basis of race, and even though UT’s positionrelies on a series of unsupported and noxious racial assumptions,the majority concludes that UT has met its heavy burden. Thisconclusion is remarkable—and remarkably wrong.Because UT has failed to satisfy strictscrutiny, I respectfully dissent.

Notes

1See also Nissimov, UT ToResume Factoring in Applicants’ Race: UT To Reintroduce Race-BasedCriteria, Houston Chronicle, June 24, 2003, p. 4A (“President LarryFaulkner said Monday his institution will quickly developrace-based admissions criteria by the fall that would be used forthe summer and fall of 2004, after being given the green light todo so by Monday’s U.S. Supreme Court ruling”); Silverstein,Hong, & Trounson, State Finds Itself Hemmed In, L.A.Times, June 24, 2003,p. A1 (explaining UT’s “intention, afterdropping race as a consideration, to move swiftly to restore itsuse in admissions” in time for “the next admissions cycle”); Hart,Texas Ponders Changes to 10% Law, Boston Globe, June 25, 2003, p.A3 (“Soon after Monday’s ruling, University of Texas PresidentLarry Faulkner said that the school will overhaul procedures” inorder to allow consideration of “[t]he race of an applicant” for“students enrolling in fall 2004”); Ambiguity Remains; High CourtLeaves Quota Questions Looming, El Paso Times, June 25, 2003, p. 6B(“The University of Texas at Austin’s president, Larry Faulkner,has already announced that new admissions policies would be draftedto include race as a factor”).

2In 2010, 3.8% of Texas’spopulation was Asian, but 18.6% of UT’s enrolled, first-timefreshmen in 2008 were Asian-American. See Supp. App. 156a; UnitedStates Census Bureau, QuickFacts (QuickFacts Texas), online athttps://www.census.gov/quickfacts/table/PST045215/48. By contrast,37.6% of Texas’s 2010 population identified as Hispanic or Latino,but a lower percentage—19.9%—of UT’s enrolled, first-time freshmenin 2008 were Hispanic. See Supp. App. 156a; QuickFactsTexas.

3If UT’s goal is to haveat least two African-Americans, two Hispanics, and twoAsian-Americans present in each of the relevant classrooms, thatgoal is literally unreachable in classes of five and practi-callyunreachable in many other small classes.

4The majority’s assertionthat UT’s race-based policy does not discriminate againstAsian-American students, see ante, at 5–6, defies the lawsof mathematics. UT’s program is clearly designed to increase thenumber of African-American and Hispanic students by giving them anadmissions boost vis-à-vis other applicants. See, e.g.,Supp. App. 25a; App. 445a–446a; cf. 645 F. Supp. 2d 587, 606 (WDTex. 2009); see also ante, at 15 (citing increases in thepresence of African-Americans and Hispanics at UT as evidence thatit* race-based program was successful). Given a “limited number ofspaces,” App. 250a, providing a boost to African-Americans andHispanics inevitably harms students who do not receive the sameboost by decreasing their odds of admission.

5In particular, the FifthCircuit’s willful blindness to Asian-American students isabsolutely shameless. For instance, one of the Fifth Circuit’sprimary contentions—which UT repeatedly highlighted in its briefand at argument—is that, given the SAT score gaps between whites onthe one hand and African-Americans and Hispanics on the other,“holistic admissions would approach an all-white enterprise” in theabsence of racial preferences. 758 F.3d, at 647. In makingthis argument, the court below failed to mention Asian-Americans.The reason for this omission is obvious: As indicated in thevery sources that the Fifth Circuit relied on for thispoint, on the very pages it cited, Asian-Americanenrollees admitted to UT through holistic review have consistentlyhigher average SAT scores than white enrollees admittedthrough holistic review. See UT, Office of Admissions,Implementation and Results of the Texas Automatic Admissions Law(HB 588) at the University of Texas at Austin, Demographic Analysisof Entering Freshmen Fall of 2006, pp.11–14 (rev. Dec. 6,2007), cited at 758 F.3d, at 647, n.71; UT, Office ofAdmissions, Implementation and Results of the Texas AutomaticAdmissions Law (HB 588) at the University of Texas at Austin,Demographic Analysis of Entering Freshmen Fall of 2008,pp.12–15 (Oct. 28, 2008), cited at 758 F. 3d, at 647,n.72. The Fifth Circuit’s intentional omission ofAsian-Americans from its analysis is also evident in the appendicesto its opinion, which either omit any reference to Asian-Americansor misleadingly label them as“other.” See id., at 661. Thereality of how UT treats Asian-American applicants apparently doesnot fit into the neat story the Fifth Circuit wanted totell.

6And it is anybody’s guesswhether this group also includes applicants “of full or partialArab, Armenian, Azerbaijani, Georgian, Kurdish, Persian, or Turkishdescent, or whether such applicants are to be considered‘White.’” Brief for Judicial Watch, Inc., etal. asAmici Curiae 16.

7United States CensusBureau, 2010 Census Shows Multiple-Race Population Grew Faster ThanSingle-Race Population (Sept. 27, 2012), online athttps://www.census.gov/newsroom/releases/archives/race/cb12-182.html.

8W. Wang, Pew ResearchCenter, Interracial Marriage: Who Is “Marrying Out”? (June 12,2015), online at http://www.pewresearch.org/fact-tank/2015/06/12/interracial-marriage-who-is-marrying-out/; W.Wang, Pew Research Center, The Rise of Intermarriage (Feb. 16,2012), online athttp://www.pewsocialtrends.org/2012/02/16/the-rise-of-intermarriage/.

9Amici supportingUT certainly understood it to be arguing that it needs affirmativeaction to admit privileged minorities. See Brief for SixEducational Nonprofit Organizations 38 (citing Brief forRespondents in No. 11–345, p.34). And UT’s amicicontinue to press the full-throated version of the argument. SeeBrief for Six Educational Nonprofit Organizations 12–13(“Intraracial diversity ... explodesperceivedassociations between racial groups and particular demographiccharacteristics, such as the ‘common stereotype of Black andLatina/o students[ ] that all students from these groups come frompoor, inner-city backgrounds.’ Schools like UT combat suchstereotypes by seeking to admit African-American and Latinostudents from elevated socioeconomic and/or non-urban backgrounds”(citation omitted)); id., at 15 (arguing that UT needsracial preferences to admit minority students from “elevated”“socioeconomic backgrounds,” because “such students are on a moreequal social footing with the average nonminority student”);id., at 37–38 (“African-American and Latino students who maycome from higher socioeconomic status ... may serve as‘debiasing agent[s],’ promoting disequilibrium to disruptstereotypical associations. These students are also likely to bebetter able to promote communication and integration on campus”(citation omitted)).

10 See,e.g., Strauss, Study: High School Grades Best Predictor ofCollege Success—Not SAT/ACT Scores, Washington Post, Feb.21, 2014, online athttps://www.washingtonpost.com/news/answer-sheet/wp/2014/02/21/a-telling-study-about-act-sat-scores/.

11 See,e.g., Freedle, Correcting the SAT’s Ethnic and Social-ClassBias: A Method for Reestimating SAT Scores, 73 Harv. Ed. Rev. 1(2003) (“The SAT has been shown to be both culturally andstatistically biased against African Americans, Hispanic Americans,and Asian Americans”); Santelices & Wilson, Unfair Treatment?The Case of Freedle, the SAT, and the Standardization Approach toDifferential Item Functioning, 80 Harv. Ed. Rev. 106, 127 (2010)(questioning the validity of African-American SAT scores and,consequently, admissions decisions based on those scores); Brieffor Amherst University etal. as Amici Curiae 15–16(“[E]xperience has taught amici that SAT and ACT scores forAfrican-American students do not accurately predict achievementlater in college and beyond”); Brief for Experimental Psychologistsas Amici Curiae 7 (“A substantial body of research by socialscientists has revealed that standardized test scores and gradesoften underestimate the true academic capacity of members ofcertain minority groups”); Brief for Six Educational NonprofitOrganizations as Amici Curiae 21 (“Underrepresentation ofAfrican-American and Latino students by conventional academicmetrics was also a reflection of the racial bias in standardizedtesting”).

12 Zumbrun, SAT Scores and IncomeInequality: How Wealthier Kids Rank Higher, Wall Street Journal,Oct. 7, 2014, online athttp://blogs.wsj.com/economics/2014/10/07/sat-scores-and-income-inequality-how-wealthier-kids-rank-higher/.

13 Seee.g., Brief for California Institute of Technologyetal. as Amici Curiae 15 (“[I]n amicus GeorgeWashington University’s experience, standardized test scores areconsidered so limited in what they can reveal about an applicantthat the University recently has done away with the requirementaltogether”); see also American University, Applying Test Optional,online at http://www.american.edu/admissions/ testoptional.cfm; TheUniversity of Arizona, Office of Admissions, Frequently AskedQuestions, online at https://admissions.arizona.edu/freshmen/frequently-asked-questions; Bowdoin College, Test OptionalPolicy, online athttp://www.bowdoin.edu/admissions/apply/testing-policy.shtml;Brandeis University, Test-Optional Policy, online athttp://www.brandeis.edu/admissions/apply/testing.html; Bryn MawrCollege, Standardized Testing Policy, online athttp://www.brynmawr.edu/admissions/standardized-testing-policy;College of the Holy Cross, What We Look For, online athttp://www.holycross.edu/admissions-aid/what-we-look-for; GeorgeWashington University, Test-Optional Policy, online athttps://undergraduate.admissions.gwu.edu/test-optional-policy; NewYork University, Standardized Tests, online athttp://www.nyu.edu/admissions/undergraduate-admissions/how-to-apply/all-freshmen-applicants/instructions/standardized-tests.html; SmithCollege, For First-Year Students, online at http://www.smith.edu/admission/firstyear_apply.php; Temple University, Temple OptionFAQ, online at http://admissions.temple.edu/node/441; Wake ForestUniversity, Test Optional, online athttp://admissions.wfu.edu/apply/test-optional/.

14 TheCourt asserts that race-blind, holistic review is not a workablealternative because UT tried, and failed, to meet its goals viathat method from 1996 to 2003. See ante, at 16 (“Perhapsmore significantly, in the wake of Hopwood, the Universityspent seven years attempt-ing to achieve its compelling interestusing race-neutral holistic review”). But the Court never explainsits basis for concluding that UT’s previous system failed. We arenot told how the Court is measuring success or how it knows that arace-conscious program will satisfy UT’s goals more effectivelythan race-neutral, holistic review. And although the majorityelsewhere emphasizes “the University’s continuing obligation tosatisfy the burden of strict scrutiny in light of changingcirc*mstances,” ante, at 10, its rejection of race-blind,holistic review relies exclusively on “evidence” predatingpetitioner’s suit by five years.

15 In2008, 1,208 first-time freshmen from Texas high schools enrolled atUT after being admitted outside the Top Ten Percent Plan. Supp.App. 157a. Based on the 2004 statistics, it is reasonable to assumethat, if the University had undertaken a race-neutralholistic review in 2008, 3.6% (43) of these students would havebeen African-American and 11.6% (140) would have been Hispanic. Seeibid. Under the University’s race-conscious holisticreview, 58 African-American freshmen from Texas and 158 Hispanicfreshmen from Texas were enrolled in 2008, thus reflecting anincrease of only 15 African-American students and 18 Hispanicstudents. And if those marginal increases (of 15 and 18 students)are divided by the number of total enrolled first-time freshmenfrom Texas high schools (6,322), see ibid., the calculationyields the 0.2% and 0.3% percentages mentioned in the textabove.

16 Recall that UT’s president vowed toreinstate race-conscious admissions within hours ofGrutter’s release. See Part I, supra.

17 Seealso, e.g., Parents Involved in Community Schools v.Seattle School Dist. No. 1, 551 U.S. 701, 720 (2007)(“It is well established that when the government distributesburdens or benefits on the basis of individual racialclassifications, that action is reviewed under strict scrutiny”);Grutter v. Bollinger, 539 U.S. 306, 326 (2003)(“We have held that all racial classifications imposed bygovernment ‘must be analyzed by a reviewing court under strictscrutiny’”); Gratz v. Bollinger, 539 U.S.244, 270 (2003) (“It is by now well established that ‘all racialclassifications reviewable under the Equal Protection Clause mustbe strictly scrutinized’”); Adarand Constructors, Inc.v. Peña, 515 U.S. 200, 227 (1995) (“[W]e hold todaythat all racial classifications, imposed by whatever federal,state, or local government actor, must be analyzed by a reviewingcourt under strict scrutiny”).

18 Themajority’s claim that UT has not “had a full opportunity to respondto” the Kroll Report, ante, at 14, is simply wrong. Thereport was discussed in no less than six of the briefs filed inthis case. See Brief in Opposition 19–20, n.2; Reply to Briefin Opposition 6; Brief for Respondents 51, n.9; Brief forCato Institute as Amicus Curiae 8–12 (certiorari stage);Brief for Cato Institute as Amicus Curiae 12, and n.4(merits stage); Brief for Judicial Education Project as AmicusCuriae 5–17. Not only did UT have an “opportunity to respond”to the Kroll Report—it did in fact respond at both thecertiorari stage and the merits stage. See Brief in Opposition19–20, n.2 (explicitly discussing the “recently releasedKroll Report”); Brief for Respondents 51, n.9 (similar). Andthe Court’s purported concern about reliance on “extrarecordmaterials,” ante, at 14, rings especially hollow in light ofits willingness to affirm the decision below, which relied heavilyon the Fifth Circuit’s own extrarecord Internet research, see,e.g., 758 F.3d, at 650–653.

19 UT’scurrent records retention policy requires it to retain studentrecords, including application materials, for at least five yearsafter a student graduates. See University of Texas at Austin,Records Retention Schedule, Agency Item No. AALL358, p. 58 (Nov.14, 2014), online athttps://www.tsl.texas.gov/sites/default/files/public/tslac/slrm/state/schedules/721.pdf. If this policy was in place when UT resumedrace-conscious admissions in 2004, then it still had thesematerials when petitioner filed this suit in 2008, and likely stillhad them at the time of Fisher I in 2013. At the very least,the application materials for the 2008 freshman class appear to besubject to a litigation hold. See App. 290a–292a. To the extentthat UT failed to preserve these records, the consequences of thatdecision should fall on the University, not on petitioner. Cf.Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. ___,___ (2016) (slip op., at 12) (allowing “a representative sample tofill an evidentiary gap created by the employer’s failure to keepadequate records”).

20 Seeante, at 10 (“Fisher I clarified the stringency ofthe strict-scrutiny burden for a school that employs race-consciousreview”); 758 F.3d, at 642 (“Bringing forward JusticeKennedy’s dissent in Grutter, the Supreme Court faulted thedistrict court’s and this Court’s review of UT Austin’s means toachieve the permissible goal of diversity”); id., at 665,n.5 (Garza, J., dissenting) (“I agree with the majority thatFisher represents a decisive shift in thelaw”).

Fisher v. University of Texas at Austin, 579 U.S. ___ (2016) (2024)

References