Appeal from the United States District Courtfor the Central District of CaliforniaDavid V. Kenyon, District Judge, PresidingArgued and SubmittedFebruary 1, 1999--Pasadena, CaliforniaFiled September 9, 1999Before: Harry Pregerson, Robert R. Beezer, andMichael Daly Hawkins, Circuit Judges.Opinion by Judge Pregerson;Dissent by Judge Beezer
_____________________________COUNSEL James K.T. Hunter, Los Angeles, California, and Mark T.Gallagher, Sacramento, California, for the plaintiff-appellant.Dennis M. Perluss, Los Angeles, California, for thedefendants-appellees.
_____________________________OPINION PREGERSON, Circuit Judge:I.This case is about an elementary school operated as aresearch laboratory by UCLA's Graduate School of Educationand Information Studies. The Corinne A. Seeds UniversityElementary School ("UES"), and its research and trainingmission is to help the State of California meet the needs of adramatically changing public school population. To this end,UES identifies issues relevant to the education and socialdevelopment of children in multicultural, urban communities,conducts research on these issues, and develops innovationsin teaching based on this research. UES shares its researchresults with public school teachers throughout the State ofCalifornia through seminars, workshops, teacher training pro-grams, and published articles.Each year, UES's Admissions Committee, under the direc-tion of the Dean of the Graduate School of Education andInformation Studies and the Director of UES, determineswhat characteristics are needed in UES's 460-student popula-tion to fulfill its research and training mission. UES considersgender, race/ethnicity, and family income in its admissionsprocess to obtain the desired student population. In selectingstudents, UES also considers other factors that might affect achild's suitability as a research subject, e.g., dominant lan-guage, permanence of residence, and parents' willingness tocomply with UES's mandatory involvement requirement. Par-ents of students applying to UES are informed of UES's con-sideration of race/ethnicity, gender, and family income inadmissions.Richard Hunter and Gina Brandt's older daughter Cia wasadmitted into UES through its admissions process. Apparentlypleased with Cia's experience at UES, in 1995, the year afterCia graduated, they sought to enroll their younger daughter,Keeley. Keeley was not selected for admission.1 When noti-fied that their daughter had not been admitted, Keeley's par-ents sued the Regents of the University of California("Regents") under Title VI of the Civil Rights Act of 1964,42 U.S.C. section 2000(d), and Dr. Theodore Mitchell, Deanof the Graduate School of Education and Information Studies,under 42 U.S.C. section 1983.2 The suit challenges the consti-tutionality of UES's admissions process.The district court conducted extensive hearings on theschool's purpose, its research, and its admission process, andultimately ruled in its favor. The district court found that (1)California had a compelling state interest in operating aresearch-oriented elementary school dedicated to improvingthe quality of education in urban public schools, and (2)UES's consideration of race/ethnicity in its admissions pro-cess was narrowly tailored to further that interest. We affirm.II.[1] The Equal Protection Clause of the Fourteenth Amend-ment prohibits a state from "deny[ing] to any person withinits jurisdiction the equal protection of the laws. " U.S. Const.amend. XIV, S 1. The Supreme Court has said that "any per-son, of whatever race, has the right to demand that any gov-ernmental actor subject to the Constitution justify any racialclassification subjecting that person to unequal treatmentunder the strictest judicial scrutiny." Adarand Constructors,Inc., v. Pena,
515 U.S. 200, 224
(1995). To meet the strictscrutiny test, the Regents must demonstrate that UES's con-sideration of race/ethnicity is narrowly tailored to serve acompelling governmental interest. See id. at 227.The district court's conclusion that UES's admissions pro-cedures meet the strict scrutiny test is based on extensive find-ings of fact which we review for clear error. See NationalAss'n of Radiation Survivors v. Derwinski, 994 F.2d 583, 587(9th Cir. 1993). After reviewing the entire record, we con-clude that Judge Kenyon's findings of fact were not clearlyerroneous.The district court's conclusions regarding the sufficiency ofthose facts in meeting strict scrutiny is a mixed question oflaw and fact which we review de novo. See id. We conclude,as did the district court, that the facts demonstrate that thedefendants have met the strict scrutiny test.III.[2] In applying the strict scrutiny test to UES's use of race/ethnicity as a factor in its admissions process, we first con-sider whether California's interest in the operation of aresearch-oriented elementary school dedicated to improvingthe quality of education in urban public schools is acompelling state interest.3[3] "[E]ducation is perhaps the most important function ofstate and local governments." Brown v. Board of Education,
347 U.S. 483, 493
(1954). The Supreme Court has recognized" `the public schools as a most vital civic institution for thepreservation of a democratic system of government.' " Plylerv. Doe,
457 U.S. 202, 221
(1982) (quoting School Dist. ofAbington Township v. Schempp,
374 U.S. 203, 230
(1963)(Brennan, J., concurring)); see also Wisconsin v. Yoder, 406U.S. 205, 213 (1972) ("Providing public schools ranks at thevery apex of the function of a State.").[4] The district court heard extensive expert testimony oncurrent problems in public urban education. The challengesposed by California's increasingly diverse population inten-sify the state's interest in improving urban public schools.Cultural and economic differences in the classroom pose spe-cial problems for public school teachers. In his decision,Judge Kenyon noted that defendants presented "an unexhaus-tive list of such issues and challenges [that ] includes limitedlanguage proficiency, different learning styles, involvement ofparents from diverse cultures with different expectations andvalues, and racial and ethnic conflict among families andchildren." Dr. Mitchell,4 who testified as an expert witness,stated that "[t]here is no more pressing problem facing Cali-fornia, or indeed the nation, than urban education; for it is inthe urban school system that the majority of California'sfuture citizens will be educated (either well or poorly), creat-ing the basic fabric for the society of the future."[5] UES is dedicated to providing more useful and moreaccurate information to educators facing these challenges. Dr.Deborah Stipek,5 director of UES, testified, "[t]he currentmission of UES is to do research relevant to . . . urban educa-tion and to disseminate that research to promote more effec-tive education for children in urban schools." As part of itsresearch mission, UES exchanges information with the StateDepartment of Education as well as other California educa-tional policy groups; trains teachers; develops and tests inno-vative teaching strategies; and disseminates study resultsnationwide.Dr. Mitchell also testified that "[t]he dynamic interplay of. . . research, dissemination, professional development, andthe training of an ever-expanding cadre of researchers dedi-cated to find[ing] the answers to the perplexing problems fac-ing urban schools . . . makes UES a unique and powerfulinstrument in meeting the State's fundamental obligations tothe children of its cities."[6] Given this record, the district court concluded, and weagree, that "the defendants' interest in operating a research-oriented elementary school is compelling."6The dissent expresses some concern that, as a result of ourdecision, "every stratum of a state's public education system(whether formally designated a `laboratory school' or not)may now in the name of `research on effective educationalstrategies' implement a racially classified admission system."Infra at 11093-94. We do not share this concern.It is not UES's designation as a laboratory school that justi-fies its admission process. UES's status as a laboratory schoolwith a research mission is not a designation without sub-stance. UES's research is funded in part through federal andprivate grants and its students are protected by all federal,state, and university guidelines, rules, and policies pertainingto research involving human subjects. Research results areshared through "a variety of publications, the television andfilm industries, computer technologies, and other media," aswell as through "seminars, workshops, observation opportuni-ties, and conferences" offered to teachers, administrators,researchers, and educational policy makers. Its research mis-sion and its dissemination of information makes UES "a cen-ter for the education and training of teachers and educationalleaders." Through UES, "nationally recognized scholars worktogether with educators and administrators to foster a betterschooling system for California children."Nor does UES's stated mission of "educational research"justify its admissions process. A mere statement from a gov-ernmental entity that it is committed to research, withoutmore, would not be sufficient to establish a compelling inter-est. But research is fundamental to the UES's charter. Theresearch mission affects the day-to day experience of its stu-dents and requires more resources than those available tomost, if not all, other elementary schools. In 1995, UES's ele-mentary school, with its population of 460 students, had a fac-ulty of twenty-seven professors with doctorates in fieldsincluding psychology, education, and medicine. In addition,twenty-one graduate, doctoral, and post-doctoral students,three medical students, thirty-two nursing students, andseventy-five undergraduate student teachers were involvedwith the elementary school, observing, working with students,and conducting research.All of these characteristics make UES an exceptionalschool and a valuable resource to California's public educa-tion system. Consequently, we do not share the dissent's con-cerns that this decision will lead to racial classification in"every stratum of a state's public education system." Infra at11093.IV.[7] To complete our strict scrutiny analysis, we nextaddress the question whether the district court correctly con-cluded that "[t]he defendants have successfully proven thatthe use of racial and ethnic identity criteria in UES's admis-sion policy is narrowly tailored to serve the purpose of acompelling state interest." (Emphasis added.) In support ofthis conclusion, the district court pointed to "a parade ofexperts [who testified] about the necessity of a race-consciousadmission policy at UES. Each expert's testimony was under-scored by the belief that the State must `continue to conductresearch on issues involving how children learn and how wecan do a better job of teaching them.' " (quoting Dr. HarryHandler's7 testimony).Dr. Carollee Howes8 testified that "[t]here is a simple ruleabout being a researcher. . . . If you're trying to find a samplethat has some [particular] distribution of race, you use race asthe variable to make that. You don't use an approximation orsome variable of it." Dr. Stipek further testified that "even ifthe applicant pool in the aggregate [was] sufficiently diverse,an entirely random selection would not yield a population thatbalances ethnicity with other factors, such as age, gender andfamily income." Dr. Handler also testified that"[b]ecause ofthe small sample size, it is highly unlikely that such a smallgroup, if selected without some explicit consideration of race/ethnicity, would be representative of Los Angeles' or theState's urban school population."9[8] The district court commented on the testimony pres-ented: The Court simply cannot hope to recount each of the particular innovative educational techniques devel- oped at UES, or each of the specific studies con- ducted at UES, which rely on the diversity of the laboratory school's student population. Having examined the testimony of the defendants' wit- nesses, the Court is convinced that without a racially and ethnically diverse student population, the bene- fits to be gained by these innovations and studies would be lost.Accordingly, the district court concluded that "it would not bepossible, nor would it be reasonable, to require the defendantsto attempt to obtain an ethnically diverse representative sam-ple of students without the use of specific racial targets andclassifications."The dissent suggests a number of alternatives to UES's cur-rent admissions process. These alternatives range from locat-ing laboratory schools elsewhere to mandating laboratoryconditions in public schools throughout California. See infraat 11098. But both Dr. Stipek and Dr. Handler testified thatit was necessary to explicitly consider race/ethnicity in UES'sadmissions process to achieve the precise student populationrequired for UES's research.10 Therefore, even if Californiawere to establish one or more other lab schools elsewhere,this would not address UES's need to maintain the representa-tive sample of students UES needs to fulfill its research mis-sion.Finally, in evaluating whether UES's use of race/ethnicityin its admissions process is narrowly tailored, we recognize,as did the district court, that courts should defer to research-ers' decisions about what they need for their research.11 TheSupreme Court has stated: "Courts have stressed the impor-tance of avoiding second-guessing of legitimate academicjudgments. This Court itself has cautioned that`judges . . .asked to review the substance of a genuinely academic deci-sion . . . should show great respect for the faculty's profes-sional judgment.' " University of Penn v. EEOC, 493 U.S.182, 199 (1990) (internal quotation marks and citation omit-ted).[9] Based on the evidence in the record, we agree with thedistrict court's determination that UES's use of race/ethnicityin its admissions process is narrowly tailored to achieve thenecessary laboratory environment.V.In short, UES is a research-oriented institution dedicated todeveloping effective techniques for use in urban publicschools--a project that benefits public school childrenthroughout the state. California has a compelling interest inproviding effective education to its diverse, multi-ethnic, pub-lic school population. UES's use of race/ethnicity in itsadmissions process is narrowly tailored to achieve the neces-sary laboratory environment to produce research results whichcan be used to improve the education of California's ethni-cally diverse urban public school population.AFFIRMED.
_____________________________BEEZER, Circuit Judge, Dissenting:Keeley Tatsuyo Hunter appeals the district court's determi-nation that the use of a racially classified admissions proce-dure at the Corrine A. Seeds University Elementary School("UES") does not violate the Equal Protection Clause of theFourteenth Amendment.I write separately to express my fundamental disagreementwith the court's opinion. It contravenes the central purpose ofthe Equal Protection Clause: to purge racial classificationsfrom public life. The opinion strays from our precedent andfails to take heed of the Supreme Court's repeated warningsagainst allowing the use of racial classifications in non-remedial contexts. More generally, the opinion reflects a dis-quieting renewed tolerance for the use of race in governmen-tal decisionmaking.IUES is a laboratory elementary school operated by theGraduate School of Education and Information Studies("GSE&IS") at the University of California, Los Angeles("UCLA"). Students at UES are research subjects.1 UES'sstated mission is to conduct research relevant to an urban edu-cational experience, to work with teachers, communities andschools to disseminate that research and to foster a moreeffective education system primarily for urban elementarystudents. More specifically, as described in appellees' brief,UES is devoted to "[s]tudying how children learn and howtheir backgrounds and family experiences -- including theirracial and ethnic identities -- impact on their educationalexperience." UES has a student population of approximately460 children, ages 4 to 12, who attend pre-kindergartenthrough sixth grade classes.The UES admissions committee is comprised of three UESteachers and two GSE&IS faculty members. In selecting stu-dents for each year's incoming class, the UES admissionscommittee does not use pre-selection interviews,achievement/ability testing or any other type of competitivecriteria. However, the UES admissions committee explicitlyconsiders, inter alia, every applicant's racial/ethnic identity.2It does so in an attempt to obtain what it considers an ade-quate cross-sample of the general student population, thusmaintaining the scientific credibility of its educational studies.3In December 1994 Hunter, who is now eight years old,applied for admission into UES's 1995-1996 entering classfor four-year-olds (the so-called Early Childhood, or "EC-1,"Program). Six racial/ethnic categories were used for the 1995-1996 admissions cycle: African-American, Asian-American,Native American, Latino(a), Caucasian and Multi-Ethnic. Inthe section of the admissions application entitled "Child'sEthnic Identity," Hunter identified herself as "Asian-American (specify): Japanese" and "Caucasian."4 Hunter wasone of 215 applicants for admission to UES's EC-1 program,46 of whom were admitted. On March 11, 1995, Hunter wasnotified that she had been denied admission.The UES admissions procedure for the 1995-1996 schoolyear operated as follows:First, the 20 siblings of current UES students who appliedfor admission were identified and admitted;Second, the UES admissions committee determined a spe-cific number of each racial/ethnic group to be admitted;Third, dominant Spanish-speaking applicants were identi-fied and a number of them were admitted;Fourth, applicants from each self-identified racial/ethnicgroup were chosen at random to create a shorter list of poten-tial admitees;Fifth, from the children randomly selected within eachracial/ethnic category, further sorting was done to ensure areasonable gender balance and to distribute the admitted chil-dren among income groups;Finally, a number of children were specially selected by theDean of the GSE&IS (so-called "Dean's Admits") and substi-tuted for randomly-selected children of the same gender,racial/ethnic identity and income group.5 Hunter brought suit on May 17, 1995 against the Regentsof the University of California ("Regents") under Title VI ofthe Civil Rights Act of 1964, 42 U.S.C. S 2000(d), and againstTheodore R. Mitchell, Dean of the GSE&IS and Vice Chan-cellor of Academic Planning and Budget at UCLA, under 42U.S.C. S 1983. Hunter claimed that UES's utilization of racialquotas in denying her admission to UES's 1995-1996 EC-1entering class violated the Equal Protection Clause of theFourteenth Amendment. Following a two-day bench trial, thedistrict court held that appellees' use of a racially classifiedadmissions procedure is constitutional. In particular, the dis-trict court held that the use of racial and ethnic identitycriteria in UES's admission policy is narrowly tailored toserve a compelling governmental interest. This timely appealfollowed.IIThe Equal Protection Clause of the Fourteenth Amendmentprovides that "[n]o State shall . . . deny to any person withinits jurisdiction the equal protection of the laws. " U.S. Const.amend. XIV, S1. It is grounded on the principle that"[d]istinctions between citizens solely because of their ances-try are by their very nature odious to a free people whoseinstitutions are founded upon the doctrine of equality."Hirabayashi v. United States,
320 U.S. 81, 100
(1943); seealso Shaw v. Reno,
509 U.S. 630, 643
(1993) ("Shaw I")(racial distinctions "threaten to stigmatize individuals by rea-son of their membership in a racial group and to incite racialhostility"). The central purpose of the Equal Protection Clauseis "to prevent the States from purposefully discriminatingbetween individuals on the basis of race." Shaw I, 509 U.S.at 642. It is meant to ensure that all persons will be treated "asindividuals, not simply as components of a racial . . . class."Miller v. Johnson,
515 U.S. 900, 911
(1995) (internal quota-tion marks omitted). "[W]henever the government treats anyperson unequally because of his or her race, that person hassuffered an injury that falls squarely within the language andspirit of the Constitution's guarantee of equal protection."Adarand Constructors, Inc. v. Pena,
515 U.S. 200
, 229-30(1995)."[R]egardless of [the] purported motivation," Coalition forEconomic Equity v. Wilson, 122 F.3d 692, 702 (9th Cir.1997), "all racial classifications, imposed by whatever federal,state, or local governmental actor, must be analyzed by areviewing court under strict scrutiny."6 Adarand, 515 U.S. at227; see Coalition for Economic Equity, 122 F.3d at 702("[a]ny governmental action that classifies persons by race ispresumptively unconstitutional and subject to the most exact-ing judicial scrutiny"). "[S]uch classifications are constitu-tional only if they are narrowly tailored measures that furthercompelling governmental interests." Adarand , 515 U.S. at227. It is the government's burden to satisfy the demands ofthis "extraordinary justification," Coalition for EconomicEquity, 122 F.3d at 702. See Adarand,
515 U.S. at 224
("anyperson, of whatever race, has the right to demand that anygovernmental actor subject to the Constitution justify anyracial classification subjecting that person to unequal treat-ment under the strictest judicial scrutiny").IIIIt is uncontroverted that the racially classified admissionsprocedure at UES treats individuals unequally. Cf. CoalitionFor Economic Equality v. Wilson, 122 F.3d 692, 707 (9th Cir.1997). The only issues on appeal are whether appellees haveasserted a sufficiently compelling interest to do so, andwhether their chosen means are narrowly tailored to serve thatinterest.AAppellees assert two separate, but related, compelling gov-ernmental interests: (1) California's "interest in research oneffective urban educational strategies and dissemination ofnew knowledge about educational practices"7 and (2) Califor-nia's "interest in promoting freedom of inquiry at the Univer-sity of California." Appellees' Brief at 18, 41. Neitherinterest, whether considered singly or together, is sufficientlycompelling to withstand strict scrutiny.1As the First Circuit recently noted in Wessmann v. Gittens,160 F.3d 790, 795 (1st Cir. 1998), "[t]he question of preciselywhat interests government may legitimately invoke to justifyrace-based classifications is largely unsettled. " A few things,however, are certain. "A State's interest in remedying theeffects of past or present racial discrimination may in theproper case justify a government's use of racial distinctions."Shaw v. Hunt,
517 U.S. 899, 909
(1996) ("Shaw II"). For thatinterest to constitute a compelling governmental interest, itmust satisfy two conditions. First, the discrimination must bespecific, identified discrimination. See id. Second, "the insti-tution that makes the racial distinction must have had a`strong basis in evidence' to conclude that remedial actionwas necessary . . . ." Id. at 910.It is also certain that "a generalized assertion of past dis-crimination in a particular industry or region is not adequate"to justify a race-conscious remedial scheme. See Shaw II, 517U.S. at 909. Such a broadbrush justification "provides noguidance for a legislative body to determine the precise scopeof the injury it seeks to remedy." Id. (internal quotation marksomitted). For like reasons, the "role model" theory8 is unac-ceptable as a compelling governmental interest. Not only doessuch a theory lack any connection to "the kind of prior dis-crimination . . . that would justify race-based relief," but alsoit "could be used to justify race-based decisionmaking essen-tially limitless in scope and duration." Richmond v. J.A. Cro-son Co.,
488 U.S. 469, 497
-98 (1989) (plurality opinion)(internal quotation marks omitted). Unlike a race-based rem-edy that is specifically tied to the eradication of identified pastdiscrimination, the role model theory "has no logical stoppingpoint." Wygant,
476 U.S. at 275
(plurality opinion). It wouldallow the government "to engage in discriminatory. . . prac-tices long past the point required by any legitimate remedialpurpose." Id.Beyond those few certainties, the caselaw is more opaque.9In Croson, a plurality of the Supreme Court held that racialclassifications are justified only when used to remedy theeffects of racial discrimination. See Croson,
488 U.S. at 493
(O'Connor, J., joined by Rehnquist, C.J., and White and Ken-nedy, JJ.) ("Classifications based on race carry a danger ofstigmatic harm. Unless they are strictly reserved for remedialsettings, they may in fact promote notions of racial inferiorityand lead to a politics of racial hostility."). One year later, fourmembers of the current Supreme Court reiterated this view:"We subject even racial classifications claimed to be remedialto strict scrutiny . . . to ensure that the Government in factemploys any race-conscious measures to further this remedialinterest [in redressing the effects of identified race discrimina-tion] and employs them only when, and no more broadly than,the interest demands." Metro Broadcasting, Inc. v. FCC, 497U.S. 547, 611 (1990) (O'Connor, J., joined by Rehnquist,C.J., and Scalia and Kennedy, JJ., dissenting) (citing Croson
488 U.S. at 493
-95, 498-502 and Wygant,
476 U.S. 267
(plu-rality opinion)).Six of our sister circuits have adopted this view and havedefinitively held that racial classifications may only be usedfor the purpose of remedying racial discrimination. SeeHopwood v. State of Texas, 78 F.3d 932, 944 (5th Cir. 1996);Contractors Ass'n v. City of Philadelphia, 91 F.3d 586, 596(3d Cir. 1996); Aiken v. City of Memphis, 37 F.3d 1155, 1162-63 (6th Cir. 1994); In re Birmingham Reverse DiscriminationEmployment Litig., 20 F.3d 1525, 1544 (11th Cir. 1994);O'Donnell Construction Co. v. District of Columbia , 963 F.2d420, 424 (D.C. Cir. 1992); Podberesky v. Kirwan , 956 F.2d52, 56 (4th Cir. 1992). But see Wittmer v. Peters, 87 F.3d 916,919 (7th Cir. 1996) (holding that effective operation of prisonboot camps is compelling governmental interest). We havealso clearly held that "[r]ace-based classifications must bereserved strictly for remedial settings." Coral Constr. Co. v.King County, 941 F.2d 910, 920 (9th Cir. 1991) (citingCroson,
488 U.S. at 493
, 524); see also Monterey MechanicalCo. v. Wilson, 125 F.3d 702, 714 (9th Cir. 1997). But cf.Coalition for Economic Equity v. Wilson, 122 F.3d 692, 708(9th Cir. 1997) ("The Equal Protection Clause . .. prohibits[racial classifications] in all but the most compelling circum-stances. [Racial classifications are] in most circumstancesirrelevant and therefore prohibited [and] should be subject todetailed judicial inquiry to ensure that the personal right toequal protection of the laws has not been infringed.").A majority of the Supreme Court has never accepted a non-remedial justification for a racial classification. 10 See Metro,
497 U.S. at 612
(O'Connor, J., joined by Rehnquist, C.J., andScalia and Kennedy, JJ., dissenting) ("Modern equal protec-tion doctrine has recognized only one [compelling govern-mental] interest [in using racial classifications]: remedyingthe effects of racial discrimination."). In fact, four membersof the Court, dissenting in Metro, squarely rejected a non-remedial compelling governmental interest in "diversity ofbroadcast viewpoints."11 See id. Two circuit courts have alsospecifically rejected the "diversity" justification. SeeHopwood, 78 F.3d at 948 ("the use of race to achieve adiverse student body . . . simply cannot be a state interestcompelling enough to meet the steep standard of strictscrutiny"); Lutheran Church-Mo. Synod v. FCC , 141 F.3d344, 354 (D.C. Cir. 1998) (noting, in the employment context,that "[w]e do not think diversity can be elevated to the `com-pelling' level, particularly when the Court has given everyindication of wanting to cut back Metro Broadcasting").Based on its review of the relevant Supreme Court caselaw,the Hopwood court even went so far as to state that "non-remedial state interests will never justify racial classifica-tions." Hopwood, 78 F.3d at 944.Justice O'Connor's dissenting opinion in Metro representedthe view of four Justices. The majority opinion in Metroupheld the federal government's non-remedial interest in"diversity of broadcast viewpoints" only by applying interme-diate scrutiny (which the Court believed appropriate for fed-eral racial classifications). Metro has since been overruled,see Adarand,
515 U.S. at 225
-27, in part because its applica-tion of intermediate scrutiny to federal racial classificationswas inconsistent with the strict scrutiny applied to state racialclassifications. In overruling Metro, Adarand did not specifi-cally address the question whether "diversity of broadcastviewpoints" could count as a compelling governmental inter-est under strict scrutiny. See id. at 258-59 (Stevens, J., dis-senting) ("the question is not remotely presented in thiscase"). Thus, the Metro dissent's thorough and thoughtfulrejection of an asserted non-remedial governmental interest in"diversity" under strict scrutiny provides significant guidanceon the question whether UES's asserted interests in educa-tional research and academic freedom constitute compellinggovernmental interests.12The leitmotif of Justice O'Connor's dissent in Metro is bestcaptured by the following: "Social scientists may debate howpeoples' thoughts and behavior reflect their background, butthe Constitution provides that the Government may not allo-cate benefits and burdens among individuals based on theassumption that race or ethnicity determines how they act orthink." Metro,
497 U.S. at 602
(O'Connor, J., joined by Rehn-quist, C.J., and Scalia and Kennedy, JJ., dissenting). Racialclassifications, "whether providing benefits to or burdeningparticular racial or ethnic groups," are inherently perniciousbecause "[t]hey endorse race-based reasoning and the concep-tion of a Nation divided into racial blocs, thus contributing toan escalation of racial hostility and conflict." Id. at 603-04.Such policies "may stigmatize those groups singled out fordifferent treatment" and "may embody stereotypes that treatindividuals as the product of their race . . . . " Id. at 604.The majority in Metro found solace in their argument that"diversity" would only justify "benign" uses of race-conscious measures. See Metro,
497 U.S. at 563
-65 & n.12.The dissenters found this cold comfort, noting that"[t]heCourt's emphasis on `benign racial classifications' suggestsconfidence in its ability to distinguish good from harmfulgovernmental uses of racial criteria. History should teachgreater humility." Id. at 609 (O'Connor, J., joined by Rehn-quist, C.J., and Scalia and Kennedy, JJ., dissenting)."Divorced from any remedial purpose and otherwise unde-fined, `benign' means only what shifting fashions and chang-ing politics deem acceptable." Id. at 615. Thus, "racialdistinctions might be directed expressly or in practice at anyracial or ethnic group" depending on "the preference of themoment." Id. at 610.It is precisely because racial classifications are "potentiallyso harmful to the entire body politic," id. at 604 (internal quo-tation marks omitted), that the Supreme Court has only toler-ated them in carefully defined remedial contexts. See id. at612; Wygant,
476 U.S. at 275
(plurality opinion) (noting "theCourt's focus on prior discrimination as the justification for,and the limitation on, a State's adoption of race-basedremedies"). A non-remedial "interest in increasing the diver-sity of broadcast viewpoints is clearly not a compelling inter-est. It is simply too amorphous, too insubstantial, and toounrelated to any legitimate basis for employing racialclassifications." Metro,
497 U.S. at 612
(O'Connor, J., joinedby Rehnquist, C.J., and Scalia and Kennedy, JJ., dissenting).Justice O'Connor's dissenting opinion excoriated themajority for "too casually extend[ing] the justifications thatmight support racial classifications, beyond that of remedyingpast discrimination." Id. at 613. Justice O'Connor's opinioncharacterized "diversity" as "certainly insufficiently weightyto justify tolerance" of government-sponsored racial distinc-tions; indeed, to accept "diversity" as a justification would beto "trivialize[ ] the constitutional command to guard againstsuch discrimination." Id. Of particular concern was the possi-bility that, "[l]ike the vague assertion of societal discrimina-tion, a claim of insufficiently diverse broadcasting viewpointsmight be used to justify . . . unconstrained racial preferences. . . [and] would support indefinite use of racialclassifications." Id. at 614. Justice O'Connor's discussion of"diversity" as a possible compelling governmental interestconcluded with a stern admonition: We should not accept as adequate for equal protec- tion purposes an interest unrelated to race, yet capa- ble of supporting measures so difficult to distinguish from proscribed discrimination. The remedial inter- est may support race classifications because that interest is necessarily related to past racial discrimi- nation; yet the interest in diversity of viewpoints pro- vides no legitimate, much less important, reason to employ race classifications apart from generaliza- tions impermissibly equating race with thoughts and behavior.Id. at 615.The same fear expressed in Justice O'Connor's opinion that"diversity" could be used to justify the indefinite use of racialclassifications explains the Supreme Court's requirement thateven remedial race-conscious measures be supported by a"specific and verifiable," id. at 613, interest in eradicatingracial discrimination. See Shaw II,
517 U.S. at 909
. The Courtrequires such showings to ensure that racial classificationswill have "only limited and carefully defined uses." Metro,
497 U.S. at 613
(O'Connor, J., joined by Rehnquist, C.J., andScalia and Kennedy, JJ., dissenting). For example, discussingJustice Powell's opinion in Bakke, the Court in Croson high-lighted the contrast between the "focused" goal of remedying"specific instances of racial discrimination" and the compara-tively "amorphous concept of injury" inherent in "societal dis-crimination" that "may be ageless in its reach into the past."Croson,
488 U.S. at 496
-97 (plurality opinion) (internal quo-tation marks omitted). The idea of "societal discrimination""does little to define the scope of any injury . . . [and] couldjustify a preference of any size or duration." Id. at 505. Like-wise in Wygant, the Court rejected the asserted interest in pro-viding minority role models to redress societal discriminationbecause such a rationale would allow remedies "timeless intheir ability to affect the future." Wygant ,
476 U.S. at 276
(plurality opinion).To prevent such an unbridled use of racial classifications,the Court has imposed rigorous evidentiary safeguards: anygovernmental entity endeavoring to classify by race mustpoint to specific, identified instances of past or present dis-crimination, Shaw II,
517 U.S. at 909
, for which that govern-mental entity has been either actively or passivelyresponsible, Croson,
488 U.S. at 482
-84, 490-91; and thatgovernmental entity must come forward with "a strong basisin evidence for its conclusion that remedial action wasnecessary." Wygant,
476 U.S. at 277
(plurality opinion);Bakke,
438 U.S. at 308
-09 (Powell, J.) ("findings of constitu-tional or statutory violations" necessary to justify a racialclassification). Proper findings in this regard are necessary to define both the scope of the injury and the extent of the remedy necessary to cure its effects. Such findings also serve to assure all citizens that the deviation from the norm of equal treatment of all racial and ethnic groups is a temporary matter, a measure taken in the service of the goal of equality itself. Absent such findings, there is a danger that a racial classifi- cation is merely the product of unthinking stereo- types or a form of racial politics.Croson,
488 U.S. at 510
(plurality opinion).2I recognize that the great majority of equal protection juris-prudence has been formulated in the context of challenges toaffirmative action programs of one stripe or another. The caseat bar is not an affirmative action program; indeed, as appel-lees put it, the admissions procedures at UES "are not reme-dial in purpose and are not intended either to enhance theeducational opportunities of disadvantaged children admittedto UES or to improve the general educational experience atUES by promoting classroom diversity." Students are selectedin fixed numbers according to race or ethnicity only to serveas subjects for educational research.Let us not forget, however, the fundamental purpose of theEqual Protection Clause of the Fourteenth Amendment:"toprevent the States from purposefully discriminating betweenindividuals on the basis of race." Shaw I,
509 U.S. at 642
. Inthe face of such a powerful constitutional proscription, UES'snovel justifications quickly wither. So "noxious, " Adarand,
515 U.S. at 241
(Thomas, J., concurring), so "odious,"Hirabayashi,
320 U.S. at 100
, are racial classifications in ourconstitutional democracy, that four members of the currentSupreme Court have held that they have absolutely no placeexcept in the most narrowly defined remedial settings. SeeCroson,
488 U.S. at 493
(O'Connor, J., joined by Rehnquist,C.J., and White and Kennedy, JJ.); id. at 524 (Scalia, J., con-curring). Another member of the Court, Justice Thomas, haswritten that "government-sponsored racial discriminationbased on benign prejudice is just as noxious as discriminationinspired by malicious prejudice."13 Adarand,
515 U.S. at 241
(Thomas, J., concurring).14 I find these opinions instructive, ifnot controlling.Even if UES's asserted interests are examined as possiblecompelling governmental interests, it is evident that they suf-fer from the same defects that doomed the "role model" the-ory in Wygant, the "societal discrimination" justification inCroson and the "diversity" rationale in Metro. As a prelimi-nary matter, I would emphasize that academic freedom, stand-ing alone, is clearly too flimsy an interest to justify racialdiscrimination. See Bakke,
438 U.S. at 313
-14 (Powell, J.)(although academic freedom constitutes a "countervailingconstitutional interest, . . . constitutional limitations protectingindividual rights may not be disregarded"). If "academicfreedom" could justify a racial classification, Plessy v.Ferguson,
163 U.S. 537
(1896), would still be the law of theland. Thus, "academic freedom" will be considered only as aninterest that might buttress UES's alternative claimed interestin educational research.The district court held that "the defendants' interest inoperating a research-oriented elementary school iscompelling." I cannot agree. The "Supreme Court's decisionsin Croson and Adarand indicate quite plainly that a majorityof the Justices are highly skeptical of racial preferences andbelieve that the Constitution imposes a heavy burden of justi-fication on their use." Wessmann, 160 F.3d at 808. UES'sasserted interest in "research on effective urban educationalstrategies and dissemination of new knowledge about educa-tional practices," Appellees' Brief at 18, 41, cannot bear thatburden.Just like the "role model" theory in Wygant, an"educational research" rationale is "amorphous" and admitsof "no logical stopping point." Wygant,
476 U.S. at 275
-76(plurality opinion). Just like "the vague assertion of societaldiscrimination [in Croson, or the] claim of insufficientlydiverse broadcasting viewpoints [in Metro]," Metro, 497 U.S.at 614 (O'Connor, J., joined by Rehnquist, C.J., and Scaliaand Kennedy, JJ., dissenting), a governmental interest in edu-cational research "might be used to justify . . . unconstrainedracial preferences," id., that are "timeless in their ability toaffect the future," Wygant,
476 U.S. at 276
(plurality opinion).Its inherent "indefiniteness," id., could easily "justify race-based decisionmaking essentially limitless in scope andduration." Croson,
488 U.S. at 498
(plurality opinion) (inter-nal quotation marks omitted).Because an "educational research" justification contains"no viable limiting principle," it "may be expanded beyondany reasonable limits." Hopwood, 78 F.3d at 950-51.Although this case arises in the unique setting of California'sonly publicly supported elementary laboratory school, oneshudders at the uses to which an "educational research" justi-fication might be put. The holding articulated in the opinionfiled today provides no principled basis for limiting the use ofracial classifications in the service of "educational research,"nor even for restricting the type of state actors who may con-duct such research. Every stratum of a state's public educationsystem (whether formally designated a "laboratory school" ornot) may now, in the name of "research on effective educa-tional strategies,"15 implement a racially classified admissionssystem. The sure result would be "a mosaic of shifting prefer-ences based on inherently unmeasurable claims." 16 Croson,
488 U.S. at 506
.For example, the University of California at Davis MedicalSchool might decide that its school would provide a valuable"research laboratory" site to examine whether a class made upof fixed percentages of members of various races and ethnici-ties would result in improved educational outcomes for thosemembers. A local school board may determine that all of theelementary and secondary schools within its jurisdiction arenow to be educational "laboratories" and that racial quotas inadmissions will be utilized in order to guarantee the "researchrelevance" of its student population. The court's opinionbetrays a disturbing tolerance for racial classifications, and ahistorically unjustified confidence in the ability of govern-ment to employ them for "benign" purposes. See Plessy, 163U.S. 537; see generally Fullilove v. Klutznick ,
448 U.S. 448
,486-87 (1980) (Burger, C.J.) (plurality) ("The history of gov-ernmental tolerance of practices using racial or ethnic criteriafor the purpose or with the effect of imposing an invidiousdiscrimination must alert us to the deleterious effects of evenbenign racial or ethnic classifications when they stray fromnarrow remedial justifications.").Appellees proferred governmental interest in educationalresearch, even when considered along with its claimed inter-est in academic freedom, is "certainly insufficiently weighty,"Metro,
497 U.S. at 613
(O'Connor, J., joined by Rehnquist,C.J., and Scalia and Kennedy, JJ., dissenting), to justify adeparture from the constitutional mandate of equal protection.It is "amorphous" and "indefinite[ ]" and, consequently,"overexpansive." Wygant,
476 U.S. at 275
-76 (plurality opin-ion). It also lacks any connection to the kind of past or presentracial discrimination that might, under appropriate circum-stances, justify a government's use of racial distinctions. SeeShaw II,
517 U.S. at 909
; Monterey Mechanical, 125 F.3d at714; cf. Metro,
497 U.S. at 613
(O'Connor, J., joined byRehnquist, C.J., and Scalia and Kennedy, JJ., dissenting)(accepting non-remedial interest in "diversity " would "toocasually extend[ ] the justifications that might support racialclassifications, beyond that of remedying pastdiscrimination"). I would reverse the district court's holdingthat appellees claimed "interest in operating a research-oriented elementary school is compelling," Hunter v. Regentsof the Univ. of Calif., 971 F. Supp. 1316, 1324 (C.D. Cal.1997).IV"Under strict scrutiny the means chosen to accomplish theState's asserted purpose must be specifically and narrowlyframed to accomplish that purpose." Wygant ,
476 U.S. at 280
(plurality opinion). This narrow tailoring requirementdemands "the most exact connection between justification andclassification." Adarand,
515 U.S. at 229
(internal quotationmarks omitted). In part, this serves to ensure "that there is lit-tle or no possibility that the motive for the classification wasillegitimate racial prejudice or stereotype." Croson, 488 U.S.493 (plurality opinion).We look to a number of factors to gauge whether a racialclassification is narrowly tailored: "whether there was anyconsideration of the use of race-neutral means," Adarand, 515U.S. at 237-38, and the "efficacy of [those ] alternative[s],"United States v. Paradise,
480 U.S. 149, 171
(1987) (pluralityopinion); whether the racial classification was adopted for thesake of "administrative convenience," Croson, 488 U.S. at508; and whether the chosen means are underinclusive oroverinclusive, Metro,
497 U.S. at 621
(O'Connor, J., joinedby Rehnquist, C.J., and Scalia and Kennedy, JJ., dissenting).17UES's racially classified admissions procedure is not nar-rowly tailored.AUES's race-based admissions policy lacks the "exact con-nection between justification and classification, " Adarand,
515 U.S. at 229
(internal quotation marks omitted), that nar-row tailoring requires. Appellees appear to believe that achild's race and ethnicity is somehow linked to a distinct"learning style,"18 and that these supposed cognitive differ-ences between the races call for study. However, none ofappellees' evidence concretely demonstrates the existence ofan ineluctable connection between any particular race/ethnicity and any particular "learning style."19 Cf. Wessmann,160 F.3d at 799 ("The School Committee has provided abso-lutely no competent evidence that the proportional representa-tion promoted by the Policy is in any way tied to the vigorousexchange of ideas . . . ."). At best, record evidence shows onlya possible connection. For example, Professor Deborah Stipektestified that "children with different ethnic backgrounds oftenhave different learning styles." (emphasis supplied). ProfessorGeoffrey Saxe testified that "ethnic diversity of targeted pop-ulations is particularly important for studies involving inter-vention and assessment, where we can expect that there maybe important differences in the way children make sense ofinterventions, assessment procedures, and in the nature ofmathematics individuals use in their everyday practices."(emphasis supplied).But strict scrutiny "requires a direct rather than approxi-mate fit of means to ends." Metro,
497 U.S. at 620
(O'Connor, J., joined by Rehnquist, C.J., and Scalia and Ken-nedy, JJ., dissenting). For example, the four dissenting Jus-tices in Metro heard, and condemned, the argument that anindividual's "race will likely indicate that[that individual]possesses a distinct perspective."20 Id. at 619. They warnedthat "even if the . . . equation of race and . . . viewpoint hassome empirical basis, equal protection principles prohibit theGovernment from relying upon that basis to employ racialclassifications." Id. at 620 (emphasis supplied); cf. Los Ange-les Dept. of Water and Power v. Manhart,
435 U.S. 702
, 709(1978) ("Practices that classify employees in terms of reli-gion, race, or sex tend to preserve traditional assumptionsabout groups rather than thoughtful scrutiny of individuals.").The dissenting opinion in Metro applies with full force inthe instant case. Even if "some empirical basis" might supportthe existence of a causal relationship between race/ethnicityand a distinct "learning style," the Equal Protection Clauseforbids relying on that basis to classify by race."[E]ssentialequal protection principles . . . prohibit racial generaliza-tions." Id. at 619; see Shaw I,
509 U.S. at 647
(striking downracial gerrymander because "[i]t reinforces the perception thatmembers of the same racial group -- regardless of their age,education, economic status, or the community in which theylive -- think alike, share the same political interests, and willprefer the same candidates at the polls").B"Among the various narrow tailoring requirements, there isno doubt that consideration of race-neutral alternatives isamong the most important." Coral Constr., 941 F.2d at 922.Appellees fail to satisfy this requirement. Specifically, appel-lees neglected to undertake any consideration -- let alone"serious, good faith consideration," id. at 923 -- of one obvi-ous race-neutral alternative: the establishment of one or moreadditional laboratory elementary schools in areas of Califor-nia where the demographic diversity would naturally produceapplicant pools with any desired racial/ethnic mix. UES iscurrently California's only state-supported laboratory schooland is necessarily limited to an applicant pool that is drawnfrom Westwood and its environs. But nothing limits Califor-nia to only one laboratory school.21Appellees only response is to cite Coral Construction forthe proposition that they are not required to "exhaust everyalternative, however irrational, costly, unreasonable, andunlikely to succeed." Id. at 923 (alteration omitted). However,Coral Construction teaches that a state must "exhaust race-neutral measures that the state is authorized to enact, and thathave a reasonable possibility of being effective. " Id. Califor-nia certainly has the authority to authorize the establishmentof new laboratory schools, and appellees have failed to carrytheir burden of showing that such an alternative would not"have a reasonable possibility of being effective."Another reasonable race-neutral option would be for UESto conduct its research on "the ethnically diverse student pop-ulation now present in the urban school community. " Hunter,971 F. Supp. at 1328. The district court held that appelleeshad demonstrated that "the race and ethnicity-orientedresearch conducted at UES could not otherwise be performedin the actual urban elementary schools" in California.22 Id. at1332. The district court erred in discounting "the efficacy of[this] alternative," Paradise,
480 U.S. at 171
(plurality opin-ion). This race-neutral alternative clearly has a "reasonablepossibility of being effective." Coral Constr., 941 F.2d at 923.The district court found that existing public elementaryschools would not allow teachers and researchers to controlsuch things as class size, age groupings or student-teacherratios. See Hunter, 971 F. Supp. at 1332. This is hardly aninsuperable barrier. Any public school policies or proceduresthat present a potential impediment to conducting research inthe public schools could be modified or eliminated, as appro-priate, by the Board of Education and/or the Californialegislature.23The district court also found that ensuring necessary levelsof parental involvement in, and cooperation with, researchcarried out in public schools would be difficult. See id. Thedistrict court cited Professor Harry Handler's testimony that"depending on voluntary participation by parents producesadditional bias in the study being undertaken through the self-selection involved in volunteering. At UES, in contrast, par-ents . . . agree to participate in all of the types of activities."24Appellees produce no evidence, however, that the putative"bias" caused by self-selection in a public school contextcould not be substantially or totally eliminated by simplyoffering suitable incentives (e.g., tax credits or cash pay-ments) to ensure that parents participate in any research asrequired.The district court, referring to Professor Ronald Galli-more's testimony, found that "attrition of teachers and schoolsupport"25 presented a significant obstacle to educationalresearch in public schools. Hunter, 971 F. Supp. at 1332. Attrial, Professor Stipek expressed a similar view, stating that"[a] laboratory school . . . needs teachers who are experiencedworking with researchers and who understand and accept thattheir job responsibilities require them to collaborate withresearchers, to try, experimentally, innovative practices, andto evaluate those new methods and practices." These practicalconcerns are not, however, insurmountable.Appellees fail to demonstrate why public school teacherscould not be provided with any requisite training and theninduced, or even required through contract, to collaboratewith UES researchers. Even if public school teachers couldnot be utilized, appellees offered no evidence showing thatUES teachers and research teams, given appropriate supportand authorization from the California legislature and/or theBoard of Education, could not simply be placed at certainpublic schools as their research requires. Furthermore, thelack of public school "support" for educational research couldeasily be addressed through appropriate state legislation and/or changes in public school policy and operating procedure.Finally, even if appellees' theory of racial reductionismwere valid, there is no reason for UES not to eschew the useof race in its admissions procedure and instead to choose stu-dents based on their particular "learning style. " A properlyconstructed test, or series of tests, could allow UES to identifychildren who display whatever "learning style " it wished tostudy "without resorting to stigmatizing and fractionalizingracial classifications," Coral Constr., 941 F.2d at 923. Giventhis possibility of an individualized admissions method, theonly plausible explanation for appellees' use of a race-basedsystem "would seem to be simple administrative conve-nience," Croson,
488 U.S. at 508
. That defense is unaccept-able. Appellees' interest in avoiding the bureaucratic effortnecessary to tailor UES's admissions procedure to race-neutral factors "cannot justify a rigid line drawn on the basisof a suspect classification." Id.CAppellees' offer essentially six rationales to justify theirreliance on race in the UES admissions process, namely: toensure funding;26 to study the race-specific "learning styles"of children; to publish the results of research; to disseminatethe results of research; to train future researchers and teach-ers; and to meet the requirements of specific studies. Theserationales cannot justify the racial quotas that UES employed.27The "funding" rationale is legally untenable, as "a Statemay not protect the public fisc by drawing an invidious dis-tinction between classes of its citizens." Memorial Hospital v.Maricopa County,
415 U.S. 250, 263
(1974). Further, as bestone can discern, the rationales based on the race-based"learning styles" of children, the desire to publish and dissem-inate research results,28 and the need to train future researchersand teachers, all appear to require only some general level ofracial/ethnic diversity. Appellees' expert witnesses testified,for example, that: UES requires "a diversity comparable inrough terms to urban schools" (Professor Saxe) (emphasissupplied); "it is imperative that UES have a student popula-tion that is perceived and accepted as being representative ofthe population in the public schools" (Professor Handler)(emphasis supplied); UES needs "the kind of diverse popula-tions that challenge . . . schools all over the state" (ProfessorGallimore) (emphasis supplied). Professor Stipek, director ofUES, testified that the required "diversity" at UES wouldneed to include some unspecified "critical mass" of each sig-nificant ethnic group represented in the urban school popula-tion, but conceded that she could not "give specificpercentages." Plainly, this need for some undefined level ofracial/ethnic diversity at UES bears scant relation, let alonethe constitutionally required "exact connection, " Adarand,
515 U.S. at 229
(internal quotation marks omitted), to the useof a racial/ethnic quota system.The rationale premised on the need for racial/ethnic diver-sity in specific studies conducted at UES also lacks any kindof connection to the racial/ethnic quotas used. The onlystudies mentioned by appellees' expert witnesses were a bilin-gual research project involving native Spanish and Englishspeaking students, and an instructional program being devel-oped on the Harlem Renaissance. The bilingual study onlyrequires the selection of children based on their native lan-guage, and is therefore irrelevant to any purported need forracial/ethnic diversity or racial quotas. The Harlem Renais-sance project is geared solely toward African-American stu-dents and it does not, even by its own terms, require anyspecific number of those students. Thus, it, too, lacks any con-nection to UES's asserted need for a set number of studentsfrom different racial/ethnic groups.DFinally, UES's utilization of a mixed-race category is a"red flag[ ]" signaling that its admissions procedure "is not, asthe Equal Protection Clause requires, narrowly tailored."Monterey Mech., 125 F.3d at 714. Indeed, UES's raciallyclassified admissions system is overinclusive. The theoryunderlying UES's research mission is, in important part, thata child's race/ethnicity is specifically tied to a distinct"learning style." But if this is true, the targeted selection ofmixed-race children is incomprehensible -- how could sucha theory possibly require the selection of mixed-race students?Use of this puzzling category is compelling evidence thatUES's admissions method is not narrowly tailored. Cf.Croson,
488 U.S. at 506
("random inclusion of racial groupsthat . . . may never have suffered from discrimination . . .strongly impugns the city's claim of remedial motivation");Wygant,
476 U.S. at 284
n.13 (plurality opinion) ("TheBoard's definition of minority to include blacks, Orientals,American Indians, and persons of Spanish descent furtherillustrates the undifferentiated nature of the plan.") (citationomitted).Appellees only attempted justification for the anomalouspresence of a mixed-race category in UES's admission proce-dures came from Professor Stipek. She claimed that "manychildren identified in California statistics in a single ethniccategory are in fact of mixed ethnicities," and noted that"researchers . . . as well as government organizations . . . areworking on alternative strategies to collect ethnicity data thatdo not force individuals into single categories. " Accordingly,Professor Stipek explained that it was "inappropriate" torequire multi-ethnic applicants to choose just one racial/ethniccategory. This explanation manifestly fails to explain how thetheory of race-specific "learning styles" justifies the selectionof mixed-race students. The use of a mixed-race category inUES's admissions procedure is simply irreconcilable with the"onerous `narrowly tailored' requirement," Wygant, 476 U.S.at 294 (O'Connor, J., concurring).VI would reverse the judgment of the district court. UES'sracially classified admissions procedure does not comportwith the Equal Protection Clause of the Fourteenth Amend-ment. In my judgment, Hunter is "entitled to reapply under anadmissions system that invokes none of the[ ] serious consti-tutional infirmities," Hopwood, 78 F.3d at 962, discussed inthis opinion.I respectfully dissent. the end
___________________________FOOTNOTES 1 There is no suggestion that any different admission procedures wereemployed in the process resulting in Keeley's rejection than in Cia's ear-lier acceptance.2 Keeley was four years old when the suit was initiated. Gina Brandtbrought the suit as Keeley's "mother and next friend," and Keeley is rep-resented by her father, Richard Hunter, attorney at law.3 The parties agree that UES's admission's process is not part of a reme-dial program.4 The district court accepted each of the defendants' witnesses as anexpert in their field. Dr. Mitchell received a Ph.D. from Stanford Univer-sity in 1983. He served as chair to Dartmouth's Department of Education,and Deputy to the President of Stanford University before becoming Deanof UCLA's Graduate School of Education in 1992. He has publishednumerous scholarly articles on education and school reform.5 Dr. Stipek also testified as an expert witness. She received a Ph.D. indevelopmental psychology from Yale University. She has publishednumerous articles on early childhood and elementary education. She hasbeen with the Graduate School of Education and Information Studies since1977.6 The appellant argues that only an interest in remedying past discrimi-nation can justify UES use of race/ethnicity as one of a number of factorsin its admissions process. We disagree. The Supreme Court has never heldthat only a state's interest in remedial action can meet strict scrutiny. Infact, in Miller v. Johnson,
515 U.S. 900
(1995), the Court expressly leftopen the question whether "compliance with the[Voting Rights] Act,standing alone, can provide a compelling interest independent of anyinterest in remedying past discrimination." Id. at 921 (emphasis added).The Court in Shaw v. Hunt followed Miller and left this question open: "InMiller, we expressly left open the question whether under the proper cir-cumstances compliance with the Voting Rights Act, on its own, could bea compelling interest. Here once again we do not reach that question . . . ."Shaw v. Hunt,
517 U.S. 899, 911
(1996) (internal citation and parentheti-cal omitted).In addition, contrary to the dissent's assertion, this court has not "heldthat `[r]ace based classifications must be reserved strictly for remedialsettings.' " Infra at 11085 (quoting Coral Construction v. King County,941 F.2d 910, 920 (1991)). Rather, this court held that where the assertedstate interest is remedying past discrimination, this remedial interest mustbe supported by concrete evidence of discrimination. See id; see alsoMonterey Mechanical v. Wilson, 125 F.3d 702, 714 (9th Cir. 1997)(same). In Coral and Monterey, this court simply followed the SupremeCourt's decision in City of Richmond v. J.A. Croson Co.,
488 U.S. 469
(1989), which required a "strong basis in evidence" for race-based reme-dial action. These holdings have no bearing on the question whether anon-remedial interest, such as the operation of a research-oriented elemen-tary school dedicated to improving the quality of education in urban publicschools, can serve as a compelling interest sufficient to survive strict scru-tiny.7 Dr. Harry Handler has a Ph.D. in Educational Psychology from theUniversity of Southern California, and, at the time of his testimony, wasAdjunct Professor and Special Assistant to the Graduate School of Educa-tion and Information Studies. Dr. Handler was Superintendent of the LosAngeles Unified School District for seven years.8 Dr. Howes has a master's degree in child study from Tufts University,a Ph.D. in developmental psychology from Boston University, and post-doctoral training in social psychiatry at Harvard University. Dr. Howeshas published books, treatises and studies in her field including a series onchild care policy studies.9 In 1995, there were 46 students admitted to the Early Childhood Pro-gram. Consequently, each child made up 2.2% of the class.10 "We cannot have a subject sample that does not have meaningful dis-tribution of ethnicity and still meets the scientific standards that we areheld to . . . . Otherwise, you can't do research there." Dr. Stipek's TrialTestimony for April 9, 1997.11 No one would challenge a decision of UCLA medical school toexplicitly consider ethnicity in selecting study participants for research onGauchers disease or Tay-Sachs--diseases that occur predominantly in theJewish population. Nor would anyone have a problem with a study on theeffects of nutrition in the prevention of sickle-cell anemia that limitedstudy participants to Black children. Nor would anyone object to a similarstudy of pernicious anemia that limited participants to older persons ofNorthern European descent. The National Institute for Health is currentlycalling for grant applications for research investigating why prostate can-cer occurs with greater frequency in white and black men than in Hispanicand Asian men.1 Appellees characterize UES as "a controlled laboratory for educationalresearchers, akin to the laboratories used by physical scientists."2 An information brochure prepared by UES states: "As a school that isseeking and disseminating ways to improve the children's learning andsocial development, Seeds UES must have a population of children that isrelevant to the student population in California. To achieve this,[GSE&IS] sets gender, ethnicity and family income population goals forthe school."3 A representative cross-sample is necessary, according to appellees,because "findings cannot be generalized beyond the population on whichthe research was conducted."4 Hunter is one-quarter Japanese and three-quarters Caucasian.5 The demographic racial breakdown, as self-reported, of the 215 chil-dren who applied for admission to UES's EC-1 program was as follows:110 Caucasians (51.2%); 49 mixed-race (22.8%); 23 African-Americans(10.7%); 19 Asians (8.8%); and 14 Latinos (6.5%). The demographicracial breakdown, as self-reported, of the 46 admitted applicants was asfollows: 18 Caucasians (39.1%); 8 mixed race (17.4%); 6 African-Americans (13%); 4 Asians (8.7%); and 10 Latinos (21.7%).6 Although appellants bring suit under Title VI, a racial classification ispermissible under Title VI only where it satisfies the strict scrutiny stan-dard applied under the Equal Protection Clause. See Regents of the Uni-versity of California v. Bakke,
438 U.S. 265, 287
(1978) (Powell, J.); id.at 328 (Brennan, White, Marshall, and Blackmun, JJ.).7 Throughout this opinion, I will refer to this asserted interest with theshorthand "educational research."8 First seen in Wygant v. Jackson Bd. of Ed.,
476 U.S. 267, 274
(1986),this theory holds that minority students are in need of "role models" toalleviate the effects of prior societal discrimination.9 As Judge Wiener, concurring in Hopwood v. State of Texas, 78 F.3d932 (5th Cir. 1996), observed, "the Supreme Court[has] declined to definecompelling interest or to tell us how to apply that term." Id. at 965 n.19(Wiener, J., concurring); see also Susan M. Maxwell, Note, Racial Classi-fications Under Strict Scrutiny: Policy Considerations and the Remedial-Plus Approach, 77 Tex. L. Rev. 259, 296 (1998) ("the Court's view on theconstitutionality of nonremedial racial classifications remainsambiguous").10 Justice Scalia has, however, offered one: "a social emergency risingto the level of imminent danger to life and limb -- for example, a prisonrace riot, requiring temporary segregation of inmates." Croson, 488 U.S.at 521 (Scalia, J., concurring).11 It bears noting that Justice Powell's oft-cited opinion in Bakke, whichapproved the consideration of race as "one element in a range of factorsa university may properly consider" in achieving diversity among its stu-dent body, Bakke,
438 U.S. at 314
, "garnered only his own vote and hasnever represented the view of a majority of the Court," Hopwood, 78 F.3dat 944.12 Like UES in the case at bar, the FCC in Metro "concede[d] that itspolicies embodied no remedial purpose." Metro,
497 U.S. at 611
(O'Connor, J., joined by Rehnquist, C.J., and Scalia and Kennedy, JJ., dis-senting).13 Justice Thomas intolerance for racial classifications would presum-ably only yield in a situation where such classifications are unavoidably"necessary to eliminate [a State's] own maintenance of a system of unlaw-ful racial classification," Croson,
488 U.S. at 526
(Scalia, J., concurring).14 As noted above, seven circuit courts of appeal (including this circuit)have agreed that only the remediation of identified past discrimination canjustify government-sponsored racial classifications.15 I deliberately omit the word "urban" from the articulation of UES'sinterest, as nothing in today's holding would prevent research into effec-tive strategies for the education of any particular racial mix of childrenfound in any particular geographical setting.16 The Court in Croson expressed an unwillingness to accept a justifica-tion that lacked a specific link to identified discrimination because that"would be to open the door to competing claims for `remedial relief' forevery disadvantaged group . . . . Courts would be asked to evaluate theextent of the prejudice and consequent harm suffered by various minoritygroups. Those whose societal injury is thought to exceed some arbitrarylevel of tolerability then would be entitled to preferential classifications.We think such a result would be contrary to both the letter and spirit ofa constitutional provision whose central command is equality." Croson,
488 U.S. at 505
-06 (internal quotation marks and citation omitted). Like-wise under the holding in the opinion filed today. Courts will now bethrust into the unseemly position of having to determine which "research"projects are sufficiently genuine and worthy to justify the use of racialclassifications. A look back at our history belies this court's apparent faithin the judiciary's capacity to "distinguish good from harmful governmen-tal uses of racial criteria." Metro,
497 U.S. at 609
(O'Connor, J., joinedby Rehnquist, C.J., and Scalia and Kennedy, JJ., dissenting); see alsoWittmer, 87 F.3d at 919 ("common sense undergirded the pernicious dis-crimination against blacks now universally regretted"); cf. Korematsu v.United States,
323 U.S. 214
(1944); Plessy ,
163 U.S. 537
.17 Another important consideration of narrow tailoring, noted inAdarand in the context of a challenge to a remedial race-conscious pro-gram, is "whether the program [is] appropriately limited" in scope andduration "such that it will not last longer than the discriminatory effectsit is designed to eliminate." Adarand,
515 U.S. at 238
(internal quotationmarks omitted). In this case, because appellees asserted interest in educa-tional research is vague and amorphous, UES's admissions procedure, afortiori, cannot be "appropriately limited" in scope and duration. Cf.Croson,
488 U.S. at 507
("it is almost impossible to assess whether theRichmond Plan is narrowly tailored to remedy prior discrimination sinceit is not linked to identified discrimination in any way").18 According to Professor Stipek, one of appellees' expert witnesses attrial, "[e]thnicity is an important variable in research related to educationbecause there is evidence that children with different ethnic backgroundsoften have different learning styles, respond differently to the sameinstructional approaches and have different styles of interaction with theteacher."19 The district court did not make a specific finding on this discrete ques-tion. The closest it came was to make the undifferentiated statement thatan ethnically diverse student population presents "issues and challenges[that] include[ ] limited language proficiency, different learning styles,involvement of parents from diverse cultures with different expectationsand values and racial and ethnic conflict among families and children."Hunter, 971 F. Supp. at 1328-29. To the extent that this statement couldbe construed as containing a finding of fact that any particular race/ethnicity is inextricably linked to any particular "learning style," such afinding is, on this record, clearly erroneous. More broadly, the "premisethat differences in race . . . reflect real differences . . . is utterly irrationaland repugnant to the principles of a free and democratic society." Metro,
497 U.S. at 618
(O'Connor, J., joined by Rehnquist, C.J., and Scalia andKennedy, JJ., dissenting) (internal quotation marks omitted).20 It bears emphasis that, although the majority opinion in Metroaccepted the approximate fit of means (the FCC's race-based preferentiallicensing policy) to ends (diversity of broadcast viewpoints), it did so onlythrough the use of intermediate scrutiny. Strict scrutiny may well have ledto a different majority opinion.21 According to Mitchell, three other University of California campuses(i.e., San Diego, Santa Cruz and Irvine) are presently discussing plans tobuild their own research-oriented laboratory schools. He testified, how-ever, that he "ha[d] no idea" whether the naturally occurring diversityaround those campuses would produce a sufficiently diverse applicantpool to obviate the need for racial quotas. Mitchell also testified that he"ha[d] not undertaken to ask the question[whether] any of the other ninecampuses of the University of California" outside of UCLA are suffi-ciently diverse to eliminate the need for racial quotas.22 The district court's conclusion that this race-neutral alternative wasunavailable presents a mixed question of law and fact that we review denovo. See National Ass'n of Radiation Survivors v. Derwinski, 994 F.2d583, 587 (9th Cir. 1993).23 The California legislature has shown itself willing to support theestablishment of nontraditional schools. For example, as appellant notes,the California legislature enacted the Charter Schools Act of 1992, Cal.Ed. Code SS 47600, et seq., which authorized the creation of up to 100charter schools to, inter alia, "[e]ncourage the use of different and innova-tive teaching methods," Cal. Ed. Code S 47601(c).24 I note in passing that neither Professor Handler, nor any other ofappellees' witnesses, explained how the "bias " resulting from the "self-selection" of public school parents differs from the bias resulting from the"self-selection" of UES parents who, as a condition of their childrens'enrollment, agree to participate in research projects as needed.25 Professor Gallimore described this lack of "school support" as fol-lows: "researchers can be asked to leave with no notice by a principal orsuperintendent even after years of investment and effort."26 Professor Stipek testified, for example, that racial/ethnic diversity isnecessary to secure funding from the National Institutes of Health, a"primary outside funding source for childhood education research."27 UES admitted exactly 18 Caucasian (39.1%), 10 Latino (21.7%), 8mixed-race (17.4%), 6 African-American (13%), and 4 Asian (8.7%) stu-dents. Although appellees' witnesses preferred to characterize these num-bers as "general targets" or "general goals," rather than outright quotas,UES "appeared to be especially adept at meeting its yearly `goals.' "Hopwood, 78 F.3d at 948 n.36. UES did so, as Professor Stipek explained,by translating any "goal" into a "specific number" in the course of makingadmissions decisions.28 The publication and dissemination rationales are based on the need tomaintain credibility among the academic and teaching communities.According to Professor Stipek, members of these communities would"question the relevance "of UES's research unless it had "a population ofchildren and families that reflects the diversity of California's urban publicschools."