Concept
Race and Strict Scrutiny
Classifications subject to heightened scrutiny
Race and Strict Scrutiny
Race is the paradigm suspect classification, and the Court’s statement in Bolling v. Sharpe still frames the subject: classifications based solely upon race must be scrutinized with particular care, since they are contrary to our traditions and hence constitutionally suspect.1 Two reasons explain the suspicion. The clear and central purpose of the Fourteenth Amendment was to eliminate all official state sources of invidious racial discrimination.2 And the long history of racial discrimination makes it likely that racial classifications rest on stereotype and prejudice. In Chief Justice Burger’s words, “[c]lassifying persons according to their race is more likely to reflect racial prejudice than legitimate public concerns.”3 The Court has also faulted classifications that rest on “an accident of birth,” a phrase from a gender case that fits race at least as well.4
Strict scrutiny attaches once a law is shown to classify by race, and there are two ways to show it. The text of the law may draw the line on its face. Or a facially neutral law may be shown to classify through discriminatory administration or impact, but only with proof of a discriminatory purpose.5 The rule reaches even a prison policy that separates inmates by race, as Johnson v. California illustrates.
The modern rule displaced an older view that a racial distinction, unlike a racial discrimination, was harmless if it merely recognized difference. Plessy v. Ferguson gave that view its classic statement. Brown v. Board of Education rejected the doctrine of “separate but equal” in public education, holding that separate educational facilities are inherently unequal and that the segregated plaintiffs were deprived of the equal protection of the laws.6 Chief Justice Warren explained that segregating children solely because of their race “generates a feeling of inferiority as to their status in the community that may affect their hearts and minds in a way unlikely ever to be undone.”7 The consequence is that every racial classification is now constitutionally suspect and merits strict scrutiny, whether it bluntly disadvantages persons of color or is offered as a mere recognition of race distinctions.8 Equal application does not rescue a classification. Loving v. Virginia refused to treat the “mere equal application” of a statute containing racial classifications as enough to remove it from the Fourteenth Amendment’s proscription of invidious racial discrimination, and it rejected the argument that such statutes survive whenever any rational purpose could support them.9 A rule forbidding a white student to date a student of color is therefore unconstitutional even if it forbids every student to date across racial lines.10
The subsections below follow the order in which a problem presents itself. The first fixes the standard that governs every governmental actor. The second and third ask what a government must prove, and how closely it must tailor its means, to defend a remedial program in public contracting. The fourth through sixth turn to diversity in university admissions: the interest the Court accepted in Grutter, the line between holistic review and point systems, and the 2023 decision that reshaped the field.
Adarand and Uniform Strict Scrutiny
Adarand Constructors, Inc. v. Pena holds that every racial classification imposed by a federal, state, or local government, whether it confers a burden or a benefit, must be reviewed under strict scrutiny: the classification is valid only if it is a narrowly tailored measure that furthers a compelling governmental interest.11 The preceding section used the case only to show that the Fifth Amendment binds the federal government to the same equality norm as the Fourteenth. The full decision does three more things. It grounds the uniform standard in three principles, it overrules the decision that had given Congress an easier test, and it makes clear that strict scrutiny does not doom every racial classification.
The challenge targeted a federal program that gave prime contractors a financial incentive to hire subcontractors certified as disadvantaged, under rules that presumed members of certain racial groups to qualify. A subcontractor argued that the presumption violated the equal protection component of the Fifth Amendment’s Due Process Clause. The district court granted summary judgment for the Government, and the Tenth Circuit affirmed under the lenient standard that Fullilove and Metro Broadcasting had supported.12 The Supreme Court held that race-based measures of this kind call for strict scrutiny. It remanded because the court of appeals had applied intermediate scrutiny and had never decided whether the interests served were compelling or the programs narrowly tailored.13
The Court rested its holding on three propositions from earlier cases: skepticism, consistency, and congruence.14 Skepticism requires the most searching examination of any race-based preference, and the need for it is a problem of sorting. A court cannot always tell whether a classification is truly “benign,” so a lesser standard risks permitting illegitimate uses of race.15 Strict scrutiny supplies the searching look through its two requirements. The compelling-interest requirement confirms that the goal is important enough to warrant so suspect a tool, and the tailoring requirement confirms that the means fit the goal so closely that little or no possibility remains that the real motive was racial prejudice or stereotype.16
Consistency holds that the standard of review does not depend on the race of those burdened or benefited by a classification.17 The right is personal. The Equal Protection Clause protects “any person,” so a contractor shut out of a bid by a racial preference has been treated unequally because of race just as surely as one shut out by a racial exclusion, and the guarantee cannot mean one thing for one individual and something else for a person of another color.18 Consistency fixes the test and leaves the outcome open. The unequal treatment is the injury that triggers review, and the principle says nothing about whether a particular law is valid; that determination belongs to the court applying strict scrutiny.19
Congruence holds that equal protection analysis under the Fifth Amendment is the same as under the Fourteenth.20 The principle mattered because the Court had earlier let Congress proceed under a lesser standard. Metro Broadcasting upheld two FCC policies designed to increase minority ownership of radio and television stations, policies Congress required the FCC to maintain.21 When Justice Thomas joined the four Justices who had dissented in Metro Broadcasting, the Court overruled it.22 The Court found it inconsistent with the principle that the Fifth and Fourteenth Amendments protect persons rather than groups and that government action based on race receives the most exacting judicial inquiry.23 Fullilove, which had upheld a federal minority set-aside, is no longer controlling to the extent it applied a less rigorous standard, although the Court left open whether that program would survive strict scrutiny.24
Three separate views show the range of positions around the majority’s. Justice Stevens, dissenting, argued that no moral or constitutional equivalence exists between a policy designed to perpetuate a caste system and one that seeks to eradicate racial subordination, a difference he compared to the one between a “No Trespassing” sign and a welcome mat.25 Other dissenters would have applied Fullilove under stare decisis, or would have left the statutes in effect out of deference to the political branches.26 Justice Scalia, concurring in part, moved the other way: the Government can never have a compelling interest in discriminating by race to “make up” for past discrimination in the opposite direction, because the Constitution recognizes no “creditor” or “debtor” race.27 The majority’s holding rests on the personal character of the right. The injury falls on the individual contractor in the same way whichever government drew the line.
Strict scrutiny under Adarand is demanding without being automatic invalidation. Federal racial classifications, like those of a State, must serve a compelling governmental interest and be narrowly tailored to further it,28 and a classification that meets that test is valid. The Court recalled that every Justice had agreed that Alabama’s “pervasive, systematic, and obstinate discriminatory conduct” in its Department of Public Safety justified a narrowly tailored race-based remedy.29 Adarand’s own program received no verdict, because the Court sent the compelling-interest and tailoring questions back to the lower courts.30
Strict scrutiny attaches only to a racial classification, so race-neutral tools stand outside the rule. Croson itself observed that States and their local subdivisions have many legislative weapons to punish and prevent present discrimination and to remove arbitrary barriers to minority advancement.31 Simplified bidding procedures, relaxed bonding requirements, and training and financial aid for entrepreneurs of all races are examples.32
A stem that describes a federal program meant to help a historically disadvantaged group tests whether you reach for a gentler standard because Congress acted or because the purpose is remedial. After Adarand neither fact lowers the standard, so the analysis goes straight to compelling interest and narrow tailoring. The common error is to stop once strict scrutiny is named, as if the label settled the result. The program survives if the government proves a compelling interest and a tight fit.
Proving a Compelling Remedial Interest
A state or local government that adopts a race-based remedy must first prove that a remedy is needed, and City of Richmond v. J.A. Croson Co. shows what that proof demands: a strong evidentiary basis of specific past discrimination, in the relevant industry and jurisdiction, that the program is designed to correct.34 Strict scrutiny reaches every racial classification, so the first question about a remedial program is whether its interest is compelling, and a remedial interest consists in correcting discrimination that actually occurred. Richmond’s plan set a 30 percent quota for minority business enterprises in city contracting.35 The council adopted it after a public hearing at which no one presented direct evidence that the city or any of its prime contractors had discriminated in letting subcontracts.36 The Court held that the plan rested on no compelling interest, and its reasons define the record a government must build before it sorts people by race.37
Two requirements emerge. The first concerns the target of the remedy. A majority of the Justices held that a state or local government has a compelling interest in remedying past discrimination only if it can identify discrimination against a racial minority within its own jurisdiction, whether governmental or private, and that it may not justify a plan as correcting “societal discrimination.”38 The second concerns the evidence. Before adopting the remedial scheme, the government must have gathered evidence that factually demonstrates a prima facie case of a constitutional or statutory violation that needs redressing.39 The standard is a strong basis in evidence that remedial action was necessary.40 A prima facie case is evidence of a violation, not a judgment, so a city need not wait for a court to find one.
Each requirement does separate work. Identified discrimination forces the government to define the injury before it prescribes the cure. Justice Powell contrasted the “focused” goal of remedying wrongs worked by specific instances of discrimination with remedying “societal discrimination,” which he called “an amorphous concept of injury that may be ageless in its reach into the past.”41 An injury without limits in time or scope gives a court nothing against which to size a remedy. Croson made the same point about speculation. How many minority firms Richmond would have had absent societal discrimination was sheer speculation, and calling that speculation identified discrimination would let any local government assemble a patchwork of racial preferences from statistical generalizations about its field.42 The strong-basis requirement tests whether the stated remedial purpose is the real one. A legislature’s recital of a benign purpose carries little or no weight when it uses a racial classification, because racial classifications are suspect and simple assurances of good intention cannot suffice.43
Richmond’s record shows both requirements failing at once. The city and the district court relied on five things, and none of them, singly or together, gave the city a strong basis in evidence that remedial action was necessary.44 Each fell short for its own reason, and the reasons show what probative evidence would look like:
- The ordinance’s own declaration that it was “remedial” counted for little or nothing, because a bare recital of a benign purpose cannot stand in for evidence when the classification is suspect.45
- Proponents’ statements that discrimination existed in construction “in this area, and the State, and around the nation” were conclusory and of little probative value in establishing identified discrimination in Richmond’s own industry.46
- The city’s general population was 50 percent Black while minority-owned businesses received 0.67 percent of its prime construction contracts,47 but where special qualifications are necessary, the relevant pool is the number of minorities qualified to do the particular task.48
- Low minority membership in contractors’ associations proved little unless the city linked it to the number of local minority firms eligible for membership.49
- Congress’s findings of societal discrimination had extremely limited value for Richmond, because the scope of the problem varies from market to market.50
The population comparison is the most tempting item, because a gap between 50 percent and 0.67 percent looks like proof of exclusion. The comparison treats every resident as a potential contractor, so a shortage of qualified minority firms could produce the same gap without anyone having discriminated. The other four items share a common defect: each is general in geography, subject, or the group it covers, and none identifies discrimination in the market the program regulates.
The same insistence appears in Wygant v. Jackson Board of Education, where the lower courts upheld a racial layoff preference on the theory that minority students needed “role models” to alleviate the effects of prior societal discrimination. The Court reversed, and a plurality of four Justices reiterated Justice Powell’s view that “[s]ocietal discrimination, without more, is too amorphous a basis for imposing a racially classified remedy.”51
Identified discrimination need not be the city’s own. A city with legislative authority over its procurement may use its spending power to remedy private discrimination if it identifies that discrimination with the particularity the Fourteenth Amendment requires.52 The discrimination must still lie within the city’s own jurisdiction, because the Court has never approved extrapolating discrimination in one jurisdiction from the experience of another.53 A state program likewise needs the State to identify governmental or private discrimination within the State.54
The Court also said what would have been enough. Had Richmond possessed evidence that nonminority contractors were systematically excluding minority businesses from subcontracting, it could have acted to end the exclusion, and a significant disparity between the number of qualified minority contractors willing and able to perform a service and the number actually engaged by the locality or its prime contractors would permit an inference of discriminatory exclusion.55 In the extreme case, some form of narrowly tailored racial preference might be necessary to break down patterns of deliberate exclusion.56
A city cannot fill the gaps in its own record by borrowing from Congress. Richmond’s plan was modeled on the federal set-aside upheld in Fullilove v. Klutznick.57 If the mere recitation of a benign purpose entitled a State to exercise Congress’s full power under section 5 of the Fourteenth Amendment, section 1 would no longer constrain it, and control over the content of equal protection would pass to the fifty state legislatures and their subdivisions. The Framers wanted clear limits on the States’ use of race as a criterion for legislative action, enforced by the federal courts.58 The same logic governs the evidence: a local government that needed only a congressional report on societal discrimination to enact a set-aside would reduce the constraints of the Equal Protection Clause to a nullity.59 The contrast between Congress and a city has narrowed since. Adarand subjected federal racial classifications to strict scrutiny, and a majority of the Justices now appears to reject the idea that section 5 lets Congress take race-based action that would violate section 1 if a State took it.60
Separate opinions mark the edges of the debate. Justice Scalia would allow a state or local government to use race for remedial ends only in a social emergency rising to the level of imminent danger to life and limb, such as a prison race riot requiring temporary segregation of inmates.61 Justice Kennedy saw structural virtues in a rule striking down every preference that is not a necessary remedy for victims of unlawful discrimination, yet he stopped short of adopting a rule of automatic invalidity.62 The governing rule sits between these positions: identified discrimination can justify a race-conscious remedy, and society-wide discrimination cannot.
Narrow Tailoring of Set-Asides
A government that proves a compelling remedial interest has finished only half of strict scrutiny. Croson holds that a set-aside for minority contractors must also be narrowly tailored to identified discrimination against minorities in the relevant industry and jurisdiction.65 Tailoring is the means half of the test. It asks whether the classification fits the goal closely, which includes whether race-neutral means would have done the job and whether the classification reaches the people the goal concerns and few others. Richmond’s 30 percent quota failed this half too. The Court observed that assessing the plan’s tailoring was almost impossible because the plan was not linked to identified discrimination in any way, and it then limited itself to two observations.66
The first observation concerns race-neutral alternatives. The record showed no consideration of race-neutral means of increasing minority participation in city contracting, and many of the barriers the city relied on appeared to be race neutral. If minority firms lack capital or cannot meet bonding requirements, a program of city financing for small firms would lead to greater minority participation without sorting any bidder by race.67 That reasoning follows from what necessity means. A racial classification burdens every bidder it disfavors, so a government cannot call the classification necessary while a measure that imposes no such burden remains untried.
Fullilove shows what a sufficient record looks like. Congress had carefully examined and rejected race-neutral alternatives before enacting its set-aside, whereas nothing showed that Richmond’s council had considered any alternative to a race-based quota.68 Even without proof of discrimination, a city has a whole array of race-neutral devices at its disposal, among them simplified bidding procedures, relaxed bonding requirements, and training and financial aid for disadvantaged entrepreneurs of all races, which would open public contracting to all those who have suffered the effects of past societal discrimination or neglect.69 A government that wants to use race must therefore show that it considered the neutral options and explain why each fell short.
The second observation concerns the number itself. The 30 percent quota could not be said to be narrowly tailored to any goal except perhaps outright racial balancing, because it rested on the “completely unrealistic” assumption that minorities will choose a particular trade in lockstep proportion to their share of the local population.70 The defect lies in what the figure measures. A population share tells how many contractors there would be if every resident were equally likely to enter the construction trade, but contracting demands skills, capital, licenses, and experience, which are unevenly spread even where no one has discriminated. Where a job requires special qualifications, the relevant pool is the group qualified to do the work.71 A number keyed to population therefore follows demographics and says nothing about how much injury remains to be cured. Lowering the percentage would not answer the objection, which runs to a fixed figure untethered from a measured injury.
The third defect lies in who receives the benefit. Under Richmond’s scheme a successful Black, Hispanic, or Oriental entrepreneur from anywhere in the country enjoyed an absolute preference over other citizens based solely on race, and the Court found it obvious that such a program is not narrowly tailored to remedy the effects of prior discrimination.72 A remedy is justified by an injury identified in the relevant industry and jurisdiction, so a preference that reaches beneficiaries outside that injury is overinclusive.73 Justice Stevens, concurring, described the same defect as stereotyping on the burdened side. The class of disadvantaged white contractors includes some who were guilty of unlawful discrimination, some who practiced discrimination before it was forbidden, and some who never discriminated against anyone, and imposing a common burden on so mixed a group merely because each member shares a race rests on stereotype rather than fact or reason.74
The fourth defect lies in the waiver. Unlike the program upheld in Fullilove, Richmond’s waiver system focused solely on the availability of minority firms; no inquiry asked whether the particular firm seeking the preference had suffered the effects of past discrimination by the city or its prime contractors.75 An availability test can show whether the quota can be filled, but it cannot show whether the firm that receives the preference was ever harmed, and that is the question that ties the remedy to the injury one firm at a time.
The city’s remaining reason for a quota over individualized review would have been simple administrative convenience. The Court held that the interest in avoiding the bureaucratic effort needed to tailor relief to those who truly suffered the effects of prior discrimination cannot justify a rigid line drawn on the basis of a suspect classification, and it cited Frontiero for the point.76 The reasoning allocates the cost of the investigation. If the government investigates, the effort falls on the government; if it skips the investigation, the cost falls on the bidders excluded by race, some of whom were never connected to any wrong.
The burden a remedy places on third parties belongs to the same inquiry. In Wygant, a layoff preference for minority teachers was not sufficiently narrowly tailored even if a compelling interest in remedying prior employment discrimination had been shown, because hiring goals impose a diffuse burden while layoffs impose the entire burden of achieving racial equality on particular individuals.77 A fixed contracting quota concentrates its burden in a similar way on the subcontractor who loses a reserved share of work.
Proof of discrimination does not end the analysis, and a rigid quota will rarely be the necessary means. The Court left room for a narrowly tailored racial preference in the extreme case, where one might be needed to break down patterns of deliberate exclusion.78
Diversity as Compelling Interest
In Grutter v. Bollinger the Court held that diversity in the student body is a compelling interest in public higher education, and that a law school’s race-conscious policy, using race as a plus factor within individualized review, is narrowly tailored to the educational benefits that diversity yields.80 The two preceding subsections showed how much a government must prove to justify a remedial classification. Grutter accepted a different kind of interest, one that depends on what a classroom teaches and requires no proof of anyone’s past wrongdoing. Language in earlier decisions, Croson among them, could be read to make remedying past discrimination the only permissible justification, and the Court answered that it had never so held.81 This subsection takes up the compelling interest. The next two take up how closely a university must tailor its means to it.
The interest came from Regents of the University of California v. Bakke, which reviewed a medical school set-aside reserving 16 of 100 seats in each entering class for members of certain minority groups.82 The Court splintered, and Justice Powell’s opinion announcing the judgment became the touchstone for analyzing race-conscious admissions policies. Grutter endorsed his view that student body diversity is a compelling interest that can justify the use of race in university admissions.83
What Powell rejected explains as much as what he accepted. He found three of the medical school’s four asserted purposes invalid and the fourth, obtaining the educational benefits of a diverse student body, compelling.84 He contrasted the “focused” goal of remedying wrongs worked by specific instances of discrimination with remedying “societal discrimination,” an “amorphous concept of injury that may be ageless in its reach into the past.”85 An injury with no edges gives a court nothing against which to measure a remedy, which is the problem Croson addressed when it demanded findings of identified discrimination in the relevant market.
The interest Powell accepted has a different structure. It rests on a university’s academic freedom to choose a student body that ensures “wide exposure” to the ideas and mores of students as diverse as the Nation of many peoples,86 which lets the university fulfill its mission of selecting those students who will contribute the most to the “robust exchange of ideas.”87 It encompasses a broad array of qualifications and characteristics of which race and ethnicity are a single factor among many.88 Because the goal is defined by what happens in the classroom, a school can state it and a court can test it without measuring any historical injury. A program aimed at ethnic diversity alone could not claim the interest, since a focus solely on ethnic diversity would hinder genuine diversity.89
In Grutter the Law School asserted one interest, the educational benefits that flow from a diverse student body, and the Court held that interest compelling.90 The Court gave a degree of deference to the school’s academic judgment. Public education has an important purpose, universities occupy a special niche in the constitutional tradition because of the expansive freedoms of speech and thought associated with the university environment, and attaining a diverse student body is at the heart of the Law School’s institutional mission.91 A judge has no neutral yardstick for deciding how much exposure to classmates of different backgrounds a legal education requires, while the faculty that designs the program does. The Court nonetheless insisted that its scrutiny of the interest was no less strict for taking those educational judgments into account.92
That deference reaches the importance of the goal and stops there. In Fisher v. University of Texas at Austin the Court held that strict scrutiny requires a judicial inquiry into whether the university could achieve sufficient diversity without racial classifications. A court must examine, not defer to, the university’s serious, good-faith consideration of workable race-neutral alternatives, and the court of appeals erred in extending deference to the narrow-tailoring inquiry.93 A faculty is well placed to say what its mission requires. Whether a neutral alternative would have worked is a question of fit, which courts decide in every strict scrutiny case.
Critics read the Court’s review as strict scrutiny applied with an unusually forgiving hand. The Court deferred both to the university’s judgment that diversity was essential to its mission and to its judgment that it needed a “critical mass” of underrepresented minority students, a form of “strict-scrutiny-light.”94 The complaint that a Court names one standard and applies another is a familiar one. Justice Marshall made it about rational basis review in Cleburne, a case taken up in the last part of this chapter.95
The Court found the educational benefits substantial and defined the school’s “critical mass” by reference to them and not to a percentage. An interest in assuring some specified percentage of a group merely because of its race or ethnic origin would be outright racial balancing, which is patently unconstitutional.96 The Court added a point about leadership. A handful of highly selective law schools supply a large share of the Nation’s senators and judges, and the path to leadership must be visibly open to talented and qualified individuals of every race and ethnicity if leaders are to hold legitimacy in the eyes of the citizenry.97 The school’s own concept of diversity was broad. It did not define diversity solely by race and ethnicity, and applicants could present their own version of diversity through personal statements and other application materials.98
A stem that gives a public university two justifications for a race-conscious admissions policy asks you to sort them with the distinction Powell drew and Grutter adopted.
The boundary of the interest appears at the edge of higher education. In Parents Involved in Community Schools v. Seattle School District No. 1 two districts assigned students to elementary and secondary schools partly by race, and the Court refused to extend Grutter to them.99 Grutter had relied on considerations unique to higher education, where universities occupy a special niche, and had stressed that “[c]ontext matters” in applying strict scrutiny.100 It also expressly limited its holding by defining a specific type of broad-based diversity and by noting the unique context of higher education. The lower courts largely disregarded those limits in extending Grutter to uphold race-based assignments in elementary and secondary schools.101 The school setting raises its own questions, because all students are included and school assignments are a zero-sum game, unlike a selective admissions process.102 A plan that aims to match each school’s racial makeup to a target also runs into Grutter’s own statement that securing a specified percentage of a group merely because of its race is outright racial balancing.103 In the school context, the Court recognized two interests as compelling. The first is remedying the effects of past intentional discrimination, which Seattle could not claim, because its schools had never been segregated by law or subject to court-ordered desegregation decrees.104 The second is the interest in diversity in higher education that Grutter upheld.
Grutter also expected its interest to be temporary. The Court said that race-conscious admissions programs must have a logical end point and surmised that in 25 years such policies would no longer be needed. Justice Ginsburg’s concurrence hoped, without being able to know, that progress toward nondiscrimination and genuinely equal opportunity would make it safe to sunset affirmative action within a generation.105
Grutter therefore marks the high point of the diversity rationale: the interest is compelling, it is defined by educational benefits, and it earns deference from the court. Whether a particular admissions process is narrowly tailored to it is the harder question, and the next two subsections show how that requirement limited what universities could do. A 2023 decision rejecting two college admissions plans that used race changed this area, as the last subsection explains.106
Holistic Review Versus Point Systems
On a single day in 2003 the Court decided two challenges to the University of Michigan’s use of race in admissions, upholding one policy and striking down the other.108 In Gratz v. Bollinger the Court held that an undergraduate policy that automatically awarded substantial points to applicants solely because of race was not narrowly tailored to the interest in diversity, while Grutter upheld a law school’s individualized, holistic process that treated race as one factor among many.109 The interest half of strict scrutiny did not divide the two cases. The comparison shows what the tailoring requirement demands: race may enter an admissions decision through a judgment about the particular file, and it may not enter through a value fixed before the file is read.
The Law School’s program shows what individualized review looks like in operation. The school engaged in a highly individualized, holistic review of each applicant’s file, giving serious consideration to all the ways an applicant might contribute to a diverse educational environment, and it afforded that consideration to applicants of all races.110 The program followed Powell’s model, in which diversity covers a broad array of qualifications and characteristics and race and ethnicity are a single factor among many.111 The school frequently accepted nonminority applicants with lower grades and test scores than some minority applicants, which showed that it weighed other diversity factors seriously.112
The undergraduate college used a different mechanism. It automatically awarded 20 points, one-fifth of those needed to guarantee admission, to every applicant from an underrepresented minority group solely because of race.113 Points for other diversity contributions, such as leadership and service, personal achievement, and geographic diversity, were capped at much lower levels. Even the most outstanding national high school leader could never receive more than five points for those accomplishments, a mere quarter of the points automatically assigned on the basis of race.114 Justice O’Connor, concurring, explained why that structure was fatal: the selection index, by setting up automatic, predetermined point allocations, ensures that the diversity contributions of applicants cannot be individually assessed.115
Grutter states the rule that explains both outcomes. A program that does not operate as a quota does not, by that fact alone, satisfy the requirement of individualized consideration. When race serves as a plus factor, the program must remain flexible enough that each applicant is evaluated as an individual and not in a way that makes race or ethnicity the defining feature of the application.116 The Gratz plan reserved no seats, yet nothing in an applicant’s file could change the 20 points, so the plan could not assess each applicant’s contribution to diversity individually. At the Law School, the officer who reads a file decides how much race matters for that applicant, so the rest of the file can move the result. In the undergraduate index, the weight was set before any file was opened, and the same weight applied to every member of the favored groups. An assumption that every member of a group brings the same contribution is the kind of generalization that strict scrutiny exists to catch.
Individualized consideration decided the Michigan cases, but the tailoring inquiry in the diversity setting looks to more than that. It “must be calibrated to fit the distinct issues raised by the use of race to achieve student body diversity,” because the very purpose of strict scrutiny is to take such relevant differences into account.117 Courts also ask whether the university seriously considered workable race-neutral alternatives and whether the use of race has a logical end point.118 The alternatives requirement is the one most often misstated. Narrow tailoring does not require a university to exhaust every conceivable race-neutral alternative, but it does require serious, good-faith consideration of workable ones, and a court examines that consideration without deferring to it.119 An exhaustion rule would force a university to try each option at the expense of its mission. A good-faith-consideration rule still obliges the university to explain why workable neutral options fell short.
The usual error is to treat the size of the bonus as the problem. The objection runs to the automatic mechanism. The reverse error is to treat the word “holistic” as a safe harbor, a mistake the next subsection explains.
The line between the two programs is narrower than the labels suggest. Three Justices would have upheld both programs and four would have struck down both, while Justices O’Connor and Breyer found a distinction between them, which explains the divergent results.120 A dissent in Gratz noted that percentage plans are just as race conscious as a point scheme but reach diverse results without saying what they are doing, adding that equal protection cannot become an exercise in which the winners are the ones who hide the ball.121 Justice Thomas joined the Gratz majority but would hold that a State’s use of racial discrimination in higher education admissions is categorically prohibited by the Equal Protection Clause.122 Justice Scalia, writing separately in Grutter, predicted that later suits would ask whether a scheme evaluates the applicant as an individual enough, and avoids separate admissions tracks enough, to fall under Grutter rather than Gratz.123
These are the rules the Michigan decisions announced. The Court’s 2023 decision rejecting two college admissions plans changed this area, and the final subsection explains how the framework above fits with it.124
After Students for Fair Admissions
In Students for Fair Admissions, Inc. v. President & Fellows of Harvard College the Court held in 2023 that the admissions programs of Harvard College and the University of North Carolina could not be reconciled with the Equal Protection Clause.126 The Court gave four reasons: both programs lacked sufficiently focused and measurable objectives warranting the use of race, unavoidably employed race in a negative manner, involved racial stereotyping, and lacked meaningful end points.127 The earlier subsections describe the framework that Grutter, Gratz, and Fisher built, and this decision reached the opposite result for programs of the Harvard and UNC type. The decision changed the landscape of affirmative action in education but is not completely clear about how it intersects with prior opinions, so the question for a student is what remains of that framework.128
The majority did not expressly overrule Bakke or Grutter,129 yet the opinion effectively, though not expressly, overruled Grutter and Fisher II.130 An answer should report what the majority held and describe its relationship to Grutter in those terms, without asserting a formal overruling the opinion does not announce.
Outside the circumstances of these cases, the Court said, its precedents have identified only two compelling interests that permit resort to race-based government action. One is remediating specific, identified instances of past discrimination that violated the Constitution or a statute. The other is avoiding imminent and serious risks to human safety in prisons, such as a race riot.131 With so few recognized interests, the universities’ educational rationale had to be capable of judicial review, and the four defects below explain why the Court found that it was not.
The first defect lies in the objectives themselves. A court asked whether race is necessary to an objective needs a target against which to measure necessity. An objective that cannot be measured leaves the court nothing to test, and review would shrink into acceptance of the university’s own description. That is why the Court demanded objectives focused and measurable enough to warrant the use of race.132
The second defect is the negative use of race. College admissions offer a limited number of seats, so a preference for some applicants ordinarily comes at the expense of others, and the Court concluded that the programs unavoidably employed race in a negative manner.133
The third defect is stereotyping. A preference tied to race alone treats the racial label as evidence of what an applicant will contribute, which applies a generalization about a group to an individual. The Court insisted that benefits be tied to the person. A student who overcame racial discrimination is credited for courage and determination, and a student whose heritage or culture motivated a leadership role is credited for a unique ability to contribute to the university. In both cases the student is treated on the basis of experience as an individual and not on the basis of race.134
The fourth defect is the absence of a meaningful end point. Grutter had said that race-conscious programs must be limited in time, repeating the requirement in several formulations, and the Court treated that limit as the reason it was willing to dispense temporarily with the Constitution’s guarantee of equal protection.135 Harvard admitted that its race-based program has no end point, and UNC said that it might use race to a greater extent in the future.136 Justice Kavanaugh’s concurrence read Grutter as imposing a 25-year time limit.137
Separate opinions went further in both directions. Justice Thomas offered an originalist defense of the colorblind Constitution and reiterated his view that all forms of racial discrimination, including so-called affirmative action, are prohibited. Justice Gorsuch, joined by Justice Thomas, argued that both programs violated Title VI of the Civil Rights Act of 1964.138 Justice Sotomayor, joined by Justices Kagan and Jackson, dissented, arguing that the limited use of race in holistic admissions has equalized educational opportunity, that race is only one factor among many in the programs under review, and that the majority refuses to recognize that racial identity can shape the viewpoints of some students.139 Justice Jackson added that the country has never been colorblind and that it fails to acknowledge the intergenerational transmission of inequality to say that anyone is victimized when a college considers the legacy of discrimination.140 The majority answered that the dissenters would permit race-based admissions to remedy past societal discrimination, an idea Justice Powell rejected in Bakke, and that they read Grutter as authorizing such programs indefinitely.141
What survives is race-based action that responds to a specific, identified violation of the Constitution or a statute.142 That is the remedial exception Croson and Adarand left open, with the demands the earlier subsections described: the government must be able to point to the identified discrimination, and the remedy must fit it. The structure of the exception explains why it survives while the diversity rationale did not. An identified violation supplies a measurable target, the injury that was proved, and a built-in stopping point, the cure of that injury. The diversity objectives at issue supplied neither.
The opinion also leaves room for an applicant’s own account of how race has shaped a life. Some applicants may gain an advantage by discussing their racial background, so race may still make a difference.143 The limit is the one stated above: credit must attach to the individual’s own courage, determination, or unique contribution, and a reader who awards the same credit for the box checked on the form has returned to the generalization the Court rejected.144
The decision also leaves neighboring questions open. It raises the question whether race can still be a factor in elementary and secondary public education and whether governments can still use racial preferences in contracting decisions.145 Programs that pursue diversity without using racial classifications, such as the Texas Top Ten Percent plan, have never been specifically held constitutional, and their validity is likely to be the next major battle over racial equality.146
Step 2 now does the work that deference once did in the earlier framework. The label “holistic review” also gives no safe harbor, since the dissent stressed that race was only one factor among many in the programs under review and the Court invalidated them anyway.147