One Law of Race?

I.     Introduction

When interpreting constitutional and statutory race equality law, the Supreme Court has oscillated between two contradictory interpretive strategies. Sometimes the Court has pursued a strategy of divergence, enforcing differences in constitutional and statutory law expressed by the text of particular provisions or revealed by consulting the purposes for which those provisions were proposed or enacted. At other times, the Court has pursued a strategy of convergence, aligning the substance and scope of legal protections derived from constitutional and statutory sources notwithstanding their apparent differences. The Court has never declared one strategy superior to the other, nor has it examined its own practices to diagnose the competing considerations used to justify its choice of strategy. This Article offers such a diagnosis by identifying the primary rationales on which the Court has relied to justify one choice of strategy over another.

Convergence holds considerable intuitive appeal, particularly when predicated on the empirical assumption that the nature of race discrimination is consistent across the wide range of social circumstances in which the law seeks to promote equality. This Article argues, however, that the temptation to organize race equality law around a single set of assumptions prejudices judicial interpretation and threatens to strip political institutions of the ability to develop and to implement a diversity of legal approaches to the problem of racial inequality.

The argument may seem superfluous. Some prominent cases in the Court’s race equality jurisprudence suggest that divergence is the norm. Not the least of these is Washington v. Davis, in which the Court famously concluded that disparate-impact doctrine is not a feature of the Constitution’s equality guarantee,1 even though it had previously recognized the doctrine under Title VII of the Civil Rights Act of 1964.2Davis reflects a view of constitutional governance in which Congress may enact laws requiring a “more probing judicial review of” potentially discriminatory behavior than what the Constitution requires, provided that it does so pursuant to a legitimate grant of legislative authority.3 Although the decision’s substance and rationale have been understandably controversial,4 the Davis Court justified divergence by invoking its institutional duty to honor differences in the substance and scope of constitutional and statutory law. In the area of race equality law more generally, divergence conveys the additional benefit of preserving for Congress and the states discretion to experiment with different approaches to promoting equality. Moreover, if the Court is generally mistrustful of legislative standards that diverge from constitutional norms, it will have difficulty discharging its duty to faithfully interpret statutory law.

Divergence, however, can be counterintuitive. Constitutional and statutory laws at times subscribe to a common agenda to defeat race discrimination in parallel, or even identical, circumstances; and, in situations where both constitutional and statutory standards apply, divergence may be “most troubling.”5 It contradicts basic expectations that like cases ought to be treated alike and that laws regulating the same conduct ought to be guided by a common, objective understanding of that conduct. In race equality law, this would mean a common understanding of the nature of race discrimination and its relationship to inequality. Finally, as a matter of moral intuition, it may appear that the prohibition against race discrimination ought to be absolute, and that a society that genuinely condemns such discrimination ought to make no allowances for any distinctions based on race.6 The Supreme Court’s current colorblindness approach to equal protection frequently appears guided by this intuition.7

Perhaps not surprisingly then, the Court has increasingly favored convergence. In Ricci v. DeStefano, for example, the Court resolved a purely statutory question regarding how to manage a conflict between the disparate treatment and disparate-impact provisions of Title VII by interposing a solution taken from equal protection doctrine.8 The Court adapted the constitutional test of Wygant v. Jackson Board of Education9 to hold that an employer may discard the results of an employment test that produces a racially disparate impact only if it has a strong basis in evidence to conclude it would incur disparate impact liability if it allowed the results to stand.10 Justice Scalia famously threatened a more aggressive form of convergence in his concurring opinion, warning that the Court’s decision “merely postpone[d] the evil day on which” it would be forced to determine whether the disparate impact provisions of the statute are “consistent with the Constitution’s guarantee of equal protection.”11 He also cautioned that whether disparate impact could survive a constitutional challenge might turn on the extent to which it may be defended “as simply an evidentiary tool used to identify genuine, intentional discrimination,”12 suggesting that race discrimination does not have many forms and therefore cannot justify divergent legal standards. Similarly, recent decisions regarding the Voting Rights Act of 1965 show the Court interpreting the Act’s provisions restrictively either by conforming their substantive protections to doctrinal frameworks developed in the Court’s constitutional race decisions13 or by holding unconstitutional provisions of the Act that could not be justified by the Court’s own assessment of the current breadth and nature of race discrimination in voting.14

Scholars have described the dynamics of convergence in one of two ways: as an attempt to assert control over the scope of the Constitution’s grant of legislative authority or as an attempt to conform the meaning of statutory provisions to constitutional values. Epitomizing the first approach, constitutional scholars Robert Post and Reva Siegel have argued that the Court has followed an “enforcement model” for applying the separation of powers in its decisions involving congressional lawmaking authority under the Reconstruction Amendments, limiting that authority to the enforcement of judicially-articulated constitutional rights and denying Congress the power to enforce alternative constructions of those rights.15 Post and Siegel point out that “it is the very object of the enforcement model to regulate the scope of Section 5 [of the Fourteenth Amendment] power in a world in which legislative and judicial enforcement of the Fourteenth Amendment diverge,” because Section 5 does not require that Congress simply transcribe constitutional provisions into a statutory format.16 Nevertheless, the enforcement model limits such divergence by permitting the Court to establish its “symbolic and practical control over the articulation of constitutional rights.”17

This Article describes the types of restrictions on congressional authority practiced under the enforcement model as “structural limitations,” because they restrict the scope of the structural constitutional device invoked by Congress to enact legislation. Contemporary examples of such practices are discussed here in connection with the Court’s most recent voting rights decisions.18 The Article demonstrates, however, that in race equality law—although structural limitations are certainly important for the reasons that Post and Siegel describe—they are only one source of convergence between constitutional and statutory norms. Convergence may also occur when the Court concedes congressional lawmaking authority, including under constitutional provisions broader than the enforcement provisions of the Reconstruction Amendments. The Court may interpret constitutional and statutory standards to converge not because it seeks to enforce a particular interpretation of legislative authority, but because it is guided by certain fact-based assumptions about the nature of race discrimination as a social practice that bias and restrict its interpretations of race equality law.

In addition, scholars sometimes describe convergence as a consequence of the migration of public values across the divide between constitutional and statutory domains. William Eskridge’s seminal theory of dynamic statutory interpretation suggests that we understand convergence as a product of “the gravitational pull” of constitutional principles or values.19 Certainly the dominant, but not the sole, direction of convergence is that constitutional frameworks are assimilated to statutory contexts.20 According to Eskridge, public values, identified through constitutional adjudication, serve as a set of “judicially created background principles” that operate “as driving forces in statutory interpretation.”21 The consultation of public values may assist courts to avoid the “counter-majoritarian difficulty,”22 and it may even enhance democratic responsiveness by aligning legal substance with popular sentiment.23 The trouble, however, it seems with the public values thesis is that, in the physics of statutory interpretation, gravity is not a constant, meaning that one cannot be certain when, or indeed whether, such values will influence interpretive outcomes.24 Bertrall Ross has referred to the Court’s practice of consulting constitutional values in statutory cases as “constitutional mainstreaming,” arguing that in “hard cases” the Court often turns to constitutional values to resolve statutory ambiguity.25 Ross warns that these values will not always be forward-looking, public values but may reflect the Court’s entrenched view of a subject based on its reading of prior cases.26

Both accounts add great insight to the scholarly literature on statutory interpretation. Values do indeed migrate across constitutional and statutory domains. The Court, however, often rationalizes convergence by appealing to something other than abstract values. Indeed, scholars sometimes represent the Court’s justifications for convergence as expressions of constitutional principle when they are nothing more than fact-based assertions made first in one context and then repeated in another.27 Even in those circumstances when ascendant constitutional values do appear to influence statutory interpretation, the Court rarely speaks in the register of values, perhaps out of concern that the naked imposition of judicially articulated values on statutory meaning would contradict legislative supremacy and expose “the judicial role in norm selection.”28

This Article proposes to take the Court’s explanations as they are offered and to explore the dynamics that they produce. Often they are not statements of value or commitment to principle; they are assertions about the social world in which the law is enacted and the institutional mechanisms by which it is applied, offered to constrain its interpretation. Certainly legal interpretation has normative consequences, including when it purports to be based on social facts, and discussions of those facts or other jurisprudential constraints may conceal such normative work from public view and even from the Court itself. Indeed, what is too often critically missing from decisions in which the Court negotiates the relationship between constitutional and statutory race equality law is an explicit discussion of the public values embodied therein.

The Court relies upon two different types of rationales to explain its choice to pursue either convergence or divergence in race equality cases: fact-based assumptions about the nature of discrimination as a social phenomenon and jurisprudential principles understood to constrain either lawmaking or judicial enforcement of law. In the sense used in this Article, the Supreme Court’s assumptions are empirical, or fact-based, when they purport to rely on observable facts about race and race discrimination. This does not mean that the Court’s assumptions result from the application of proper empirical methods or that these assumptions do not at times serve conceptual or ideological aims.29 As Ian Haney López has observed, a definition of race may be “empirical” because it relies on observable evidence, such as rules of racial identification that turn on traits or ancestry.30 The same definition may also be conceptual in that it provides an organizing principle that may be used to rationalize racial inequality, to identify certain groups as proper recipients of constitutional protection due to their history of disadvantage, or to expose the arbitrariness of race-based distinctions regardless of the race of the claimant.

Empirical assumptions support convergence between constitutional and statutory law when the Court believes that both bodies of law aim to address the same social phenomenon and therefore ought to do so in the same way. According to this approach, race discrimination is a social problem originating in particular social or psychological dynamics and characterized by certain common features. This approach views the nature of race discrimination as independent of legal rules prohibiting discrimination. The same approach, however, assigns to courts the responsibility to conform the substance of antidiscrimination law to the specific challenges posed by discrimination as a definable social fact.

In race equality law, the Supreme Court invokes jurisprudential principles for a variety of purposes, which are themselves sometimes in conflict. Jurisprudential principles include fidelity to differences in the letter and purpose of the law, the policing of the government’s legitimate exercise of lawmaking authority, and assumptions about the institutional capacity of courts to enforce certain types of legal rules. The Court invokes such principles sometimes in order to support the relative autonomy of political institutions to pursue a diversity of legal strategies to address racial inequality and at other times to impose limitations on political action, including by denying either the government’s authority to enact particular legislation or the judiciary’s ability to enforce the law as enacted.

This Article uses examples from employment discrimination, affirmative action, and voting rights cases to illustrate how the Court justifies its pursuit of either convergence or divergence. The Article argues against the impulse to organize race equality law around a common set of fact-based assumptions regarding the nature of discrimination or jurisprudential assumptions regarding the limitations of judicial power. Divergence reflects the diversity of democratic choices permitted within our constitutional order which authorizes political institutions to imagine and to act upon different equality commitments than what the Constitution itself embodies. Divergence may be the result of contradictory empirical assumptions about discrimination that future legislators may be called upon to reconcile, but it may also reflect a political effort to pursue different objectives and to imagine different futures for the significance of race in American society. In other words, the aim of the law rather than what it is aiming at may justify divergence. The objectives of various laws may differ even though the object of their focus is the same, and the latter difference may justify a difference of regulatory approaches.

What constitutes race discrimination is never purely a question of fact. It is determined as much from the normative commitments reflected in particular legal prohibitions as from assumptions about the nature of discrimination or the sociohistorical context in which particular examples of public and private conduct come to be repudiated as discriminatory.31 The same may be said about racial categories themselves and the significance that we attach to them. Race and race discrimination are not static concepts.32 Fundamental similarities in the social manifestation of race discrimination observed across private and public contexts, however these similarities may be described, do not require that we settle upon one law of race. Indeed, to do so would forfeit the ability of political institutions to pursue different approaches to the problem of racial inequality, to conceptualize the problem differently in different contexts, to pursue multiple regulatory objectives, and to reevaluate and evolve their equality commitments over time. What is more, it risks blinding the judiciary to important differences in the choices that political actors have made at different moments in our history and thereby signaling to future lawmakers the futility of imagining different pathways to racial equality.

Part II of the Article will discuss two celebrated cases in employment discrimination law in order to illustrate the Court’s alternating strategies of convergence and divergence in race equality law. These cases provide examples of the Court’s reliance on jurisprudential principles and fact-based assumptions to justify one choice of strategy over another. Part III will examine several of the Court’s recent decisions in the areas of employment discrimination, affirmative action, and voting rights in order to demonstrate that the Court has been aggressively pursuing convergence across a range of substantive legal areas. In Part IV, the Article will argue against the Court’s imposition of convergence without due consideration of differences in the language and purposes of laws which suggest that interpretations of their substance should diverge. Convergence ought to be a consequence of either a common political will to resolve like questions of racial inequality in a like manner or clear structural constitutional constraints on lawmaking.33 From time to time, the Court may be convinced that its own observations of social facts could set the law aright, but this does not authorize the Court to impose those observations as restraints on lawmaking.

II.     The Dynamics of Convergence and Divergence in Race Equality Law

Categorically, constitutional and statutory race equality laws address the same social problem: inequality between persons of different races attributable to public or private discrimination, or to social structure and practices that perpetuate the effects of past discrimination. When viewed strictly in terms of the values that animate race equality law, remarkable consistency exists across constitutional and statutory domains. Anticlassification and antisubordination norms extend to both areas of the law, and scholars sometimes look to developments in one in order to anticipate possible changes in the other.34 We might expect therefore that constitutional and statutory law in this area would show little variation. In fact, when they rely on different approaches to address similar instances of discrimination, we might object: (1) that they have failed to treat like cases alike; or (2) that one approach relies on a definition of discrimination that does not square with social facts. The matter, however, is not so simple.

Regarding the first objection, the opportunity for political institutions to experiment with different regulatory strategies on matters of public policy is embedded in our constitutional structure. Federalism encourages diversity of substantive law by permitting laws to be enacted at either the state or federal level and by restricting the federal government’s authority to limit the states’ lawmaking abilities.35 Scholars have recognized that states have an interest in providing equality rights that exceed federal constitutional protections,36 and states have sometimes valued their political autonomy to such an extent that, even when a state adopts identical language from a federal source, state courts can and do interpret state law to have a more capacious meaning.37 Within the federal government, the separation of powers enables similar legislative flexibility. When Congress acts within the bounds of its constitutionally delegated authority, it may confer civil rights and liberties that exceed what the Constitution itself provides.38 Indeed, divergence between constitutional and statutory law will sometimes be less a symptom of divergent understandings of the social dynamics of race discrimination and more a function of different normative choices made regarding the result that lawmakers intended.

Regarding the second objection, the notion that there is one “true” form of race discrimination is an illusion. In reality, when the government targets racial inequality, it has a number of choices to make. What, after all, is race or, for that matter, race discrimination? Discredited as a “biological essence,”39 “race” has been called a social construction,40 an identity performance,41 and even an “illusion.”42 It is a conceptual and aspirational term, signifying a desire to fit persons or traits into a particular social order rather than reflecting an existing biological order.43 As our views about the elasticity and significance of that social order evolve, so too do our uses of the term “race.”

The concept of race discrimination shows similar variability. The nature of such discrimination has been called “elusive,” particularly regarding the question whether its causes are individual or institutional.44 Race discrimination may be viewed as an interactive system spread across social contexts, according to which inequality in one area, such as healthcare or education, fuels discrimination in another, such as employment, housing or voting.45 Finally, the government may choose to prohibit certain behaviors as discrimination, and to excuse others, because of its desire to produce certain social effects. For example, Title VII prohibits employers from using race when making employment decisions, but it does not consider voluntary affirmative action programs to be unlawful discrimination when they can be justified by their remedial ends.46 Legal definitions of discrimination always serve a normative function and reflect a set of choices that look beyond the world as we see it and toward the world as we wish it to be.

The Supreme Court’s decisions in this area must navigate a difficult set of issues, and the Court itself has a critical role to play in designing doctrinal frameworks that realize the goals of constitutional and legislative provisions. This Part will expose fundamental and common features of the Court’s explanations of these decisions.

A.     Two Illustrative Cases

1.     Divergence: Washington v. Davis

In Washington v. Davis, African American plaintiffs brought an equal protection challenge under the Due Process Clause of the Fifth Amendment against the District of Columbia Metropolitan Police Department’s use of a facially neutral personnel test (“Test 21”) in connection with hiring and promotion.47 The test had originally been designed for general use by the Civil Service Commission to screen for verbal and reading ability, and the plaintiffs argued that it “bore no relationship to job performance and ha[d] a highly discriminatory impact in screening out black candidates.”48 Prior to a 1972 amendment, Title VII did not apply to claims against the federal government, and the plaintiffs raised no claim under the statute. The circuit court observed, however, that “[t]he many decisions disposing of employment discrimination claims on constitutional grounds have made no distinction between the constitutional standard and the statutory standard under Title VII.”49 The Supreme Court concluded that the circuit court’s application of Title VII’s disparate-impact standard to resolve the constitutional claim was error so “plain” that its judgment should be reversed, even though this argument was not presented in the government’s petition.50

Before Davis, the Supreme Court had held under Title VII that, if plaintiffs demonstrated that a facially neutral employment test resulted in a racially disproportionate impact, the burden then shifted to the employer to prove that the test was job related and consistent with business necessity.51 In Griggs v. Duke Power Co., African American employees challenged the employer’s use of a high school diploma requirement and a requirement of qualifying performance on two standardized tests to determine eligibility for assignment into any non-labor department within its organization.52 These criteria had the effect of disqualifying a disproportionate number of black employees from non-labor department jobs.53 The Court held that Duke Power’s reliance on these criteria violated the statute because neither requirement was ever shown to predict job performance for any position within the company.54 Applying Griggs to the constitutional claim in Davis, the circuit court granted partial summary judgment for the plaintiffs on the grounds that Test 21 had a disproportionate impact against black applicants and had not been validated as a predictor of job performance.55 The Supreme Court cautioned: “We have never held that the constitutional standard for adjudicating claims of invidious racial discrimination is identical to the standards applicable under Title VII, and we decline to do so today.”56 The Court reversed, holding that, unlike Title VII, the Constitution required proof of the government’s discriminatory purpose.57

From a factual perspective, the result in Davis is counterintuitive. The case appeared factually indistinguishable from Griggs. Both Griggs and Davis concerned claims of race discrimination based on the employer’s use of facially neutral employment tests that resulted in racially disproportionate impacts. According to the Court’s depiction of race discrimination in Griggs, an employer discriminates when it uses “employment procedures or testing mechanisms that operate as ‘built-in headwinds’ for minority groups and are unrelated to measuring job capability.”58 The practical operation and impact of a facially neutral practice, not the intent behind the practice, defines that practice as discriminatory.59 In Davis, however, the Court never took seriously the idea that Griggs represents a judgment about the nature of discrimination. True, the Griggs Court had explained its theory of discrimination as an elucidation of congressional purposes.60 However, it also spoke in terms that seemed to affirm what it construed to be Congress’s determination that discrimination has structural, as well as individual, causes and may occur due to the effects of an otherwise facially neutral employment practice.61 By contrast, the Davis Court spoke in terms that seemed to reject that determination. The Court expressed its “difficulty understanding how a law establishing a racially neutral qualification for employment” could be “nevertheless racially discriminatory,” and it suggested that the guarantee of equal protection was satisfied when the government met the standard of formally equal treatment.62 Thus, even if the Griggs disparate-impact standard represented a contemporary understanding of the subtle nature of race discrimination, one codified by Congress and acknowledged by the Court, that understanding was not used by the Court in Davis to evaluate the substance of the Constitution’s equality guarantee.

When the Davis Court concluded that the Constitution does not prohibit facially neutral conduct having a racially disproportionate impact absent proof of a discriminatory purpose, it made a judgment about what the Constitution requires and not about what, in some objective sense, race discrimination is. The Court acknowledged, but did not try to resolve, the potential disagreement over how discrimination ought to be defined. The constitutional and statutory standards diverge, according to Davis, not because of actual differences in the phenomena that they address but because of substantive differences between the two bodies of law and constitutional restrictions on judicial review. The Court affirmed the Griggs standard as a proper interpretation of congressional intent under Title VII,63 and it recognized that under the 1972 amendments that standard applied to public employers.64 Nevertheless, the Court concluded that the statute required “a more probing judicial review of, and less deference to, the seemingly reasonable acts of administrators and executives than is appropriate under the Constitution where special racial impact, without discriminatory purpose, is claimed.”65 Acknowledging that Congress could grant such authority in limited circumstances such as employment, the Court evaluated the propriety of disparate-impact liability as a constitutional standard by looking beyond the employment context. It then considered the possible impairment of legislative functions were they subjected to the same standard, arguing that:

A rule that a statute designed to serve neutral ends is nevertheless invalid, absent compelling justification, if in practice it benefits or burdens one race more than another would be far-reaching and would raise serious questions about, and perhaps invalidate, a whole range of tax, welfare, public service, regulatory, and licensing statutes that may be more burdensome to the poor and to the average black than to the more affluent white.66

The Davis Court held that, because application of the statutory test would contradict the value of judicial deference to the political branches, it could not be the constitutional test. Of course, this is not entirely satisfying. What the Constitution requires with respect to facially neutral, racially disproportionate measures may have been ambiguous absent the assumption that whatever the Constitution requires of the government as an employer it will also impose on the government when it exercises its legislative and executive functions. That assumption strengthens the precedential value of the many non-employment cases that the Court cites throughout Davis as instances when—in circumstances as wide-ranging as school desegregation,67 jury selection,68 voting rights,69 and social security70—the Court had required a showing of discriminatory purpose to determine a constitutional violation. It also makes the Court’s conclusions regarding the judiciary’s limited institutional capacity salient to the question of what the Constitution substantively requires, and it may not have been so had the Court thought that it could resolve Davis simply as an employment case and nothing more.

Davis demonstrates the Court’s capacity to shape substantive law by denying its own institutional capacity to engage in a particular form of inquiry. It may be that the Court is sometimes motivated to deny its institutional capacity because it rejects a law on substantive grounds—the fear of what Justice Brennan called “too much justice.”71Davis is vulnerable on this score and on its contention that discriminatory purpose had always been the constitutional test.72 The Court’s explanation of its decision articulates a constitutional separation of powers rationale. Institutional capacity arguments, however, are themselves malleable, particularly when the Court raises the issue of its capacity rather than its authority.

Consider again Davis and Griggs. The Court might have concluded that equal protection and Title VII converged by relying on considerations of its own institutional capacity—the very same basis that did in fact lead the Court to conclude that constitutional equality claims could not rest on disparate impact. Prior to Davis, in the same term in which it decided Griggs, the Court held that a city’s decision to close public swimming pools did not violate equal protection, regardless whether the decision was “motivated by a desire to avoid integration of the races” following a lower court ruling that their segregation was unconstitutional.73 The Court reached this conclusion in Palmer v. Thompson by examining “the hazards of declaring a law unconstitutional because of the motivations of its sponsors.”74 The Court confessed the limitations of the judiciary’s capacity to perform the “extremely difficult” task of “ascertain[ing] the motivation, or collection of different motivations, that lie behind a legislative enactment.”75 Next, the Court warned that “there is an element of futility in a judicial attempt to invalidate a law because of the bad motives of its supporters” because a law “struck down for this reason, rather than because of its facial content or effect” would be rendered valid if passed “for different reasons.”76 It concluded that prior constitutional cases emphasizing the government’s discriminatory purpose in fact turned “on the actual effect of the enactments, not upon the motivation.”77 The Court’s reasoning in Palmer could easily be read as a precursor to Griggs. The Court did in Griggs under Title VII what in Palmer it had instructed to be done under the Constitution: it relied on an effects-based test that defined discrimination by the racial impact of the challenged practice, thus avoiding the “futility” of a motive-based test and the evidentiary problem of having to assess the intent of a collective decision-maker.78

If Palmer is a precursor to Griggs, then the history of equal protection’s relationship to Title VII is one of convergence in Griggs, which fulfilled the logic of Palmer, and then divergence in Davis, which rejected that same logic. In Palmer, the Court was encouraged to reject motivation-based tests because of concerns about its own institutional capacity to wield such tests accurately and effectively. Davis slipped free of Palmer’s precedential grasp by representing its holding to be “that the legitimate purposes of the [pool closing] ordinance—to preserve peace and avoid deficits—were not open to impeachment by evidence that the councilmen were actually motivated by racial considerations.”79 On the contrary, Palmer determined that such impeachment would have been irrelevant because the fact that blacks and whites both lost the benefit of public pools meant that the effect of the ordinance was not discriminatory.80 In short, had Davis faithfully followed Palmer, the Court’s concerns regarding its institutional capacity would have counseled in favor of a constitutional effects test, not against it.

Against the weight of Palmer, Wright v. Council of Emporia,81 and the multitude of court of appeals decisions in which proof of a racially disproportionate impact had been sufficient to sustain a constitutional violation,82Davis asserted the “prevailing rule” was that a constitutional violation would not be shown absent a discriminatory purpose.83 The Court said Title VII expresses a different, statutory rule, and Congress was authorized to enact a different rule if it wished. What is unsatisfying about Davis is not that the case looks at two parallel instances of employment discrimination and explains, unconvincingly, why the public values expressed by the passage of Title VII should not be interpreted to influence constitutional adjudication. Rather, what is unsatisfying is that Davis does not enter into a discussion of public values at all, whether related to equal protection or statutory disparate impact. The conflict, according to Davis, is between two rules, not two normative visions. Thus, if we decry the outcome in Davis and demand greater justification of its ultimate conclusion, we should recognize that the Court put its discussion of jurisprudential principles in place of any express consideration of competing values, and it is that missing discussion of values that should have explained why the Constitution requires proof of discriminatory purpose even when evidence of a substantially racially disproportionate impact exists. The Davis Court attempt to answer this why question by arguing that the invalidation of facially neutral legislation based solely on a finding of disparate impact would exceed judicial competence and therefore the legitimate exercise of judicial authority.

2.     Convergence: General Building Contractors Ass’n v. Pennsylvania

In Davis, we found a powerful illustration of how jurisprudential rules may be asserted to support divergence between the substance of laws that otherwise regulate the same social behavior and may be applied in identical circumstances. In General Building Contractors Ass’n v. Pennsylvania, we find the converse: a decision in which the Supreme Court concludes that two laws with formally distinguishable provisions are nevertheless substantively coextensive because they were both intended to prohibit the same sort of behavior.84 The latter case, therefore, represents an assertion of social fact above apparent differences in legal substance.

General Building Contractors required the Court to address the question whether a claim of race discrimination brought under 42 U.S.C. § 1981 requires proof of discriminatory purpose. The statute provides that “[a]ll persons within the jurisdiction of the United States shall have the same right in every State and Territory to make and enforce contracts . . . as is enjoyed by white citizens.”85 The State of Pennsylvania and a class of minority workers brought claims of race discrimination against a union, trade associations, contractors, and a class of construction industry employers arising out of the operation of an exclusive hiring hall and an apprenticeship program.86 Although the plaintiffs asserted a Title VII claim against the union, no such claim was raised against the employers or trade associations because the plaintiffs had failed to exhaust their administrative remedies by filing timely charges with the Equal Employment Opportunity Commission.87 The district court found that the union and the committee administering the apprenticeship program “had violated Title VII, both because they intentionally discriminated and because they enforced practices that resulted in a disparate racial impact.”88 It also interpreted § 1981 liability to rest “on roughly the same basis as a Title VII claim,” and, following that logic, held that all defendants had violated § 1981.89 The contractors and trade associations, however, were not found to have engaged in intentional discrimination. Instead, the district court concluded that the employers’ delegation of authority to the union, which in turn discriminated against minority workers, violated § 1981 because the statute “requires no proof of purposeful conduct.”90 The Supreme Court reversed, holding that the two statutes, Title VII and § 1981, were not guided by the same legislative objective and that § 1981, like the Constitution, requires a showing of purposeful discrimination.

The Court’s primary justification for reaching this conclusion involves an assertion about the nature of the conduct targeted by § 1981. The Court engaged in an “imaginative reconstruction” of § 1981 by examining the statute’s history in order to discover congressional assumptions and motivations which it then used to determine the statute’s meaning.91 The Court argued that § 1981 should be interpreted with attention to the “‘events and passions of the time’ in which the law was forged.”92 Congress originally codified language currently contained in § 1981 as section 1 of the 1866 Civil Rights Act, shortly after the ratification of the Thirteenth Amendment.93 The Court found that “[t]he principal object of the [1866 Act] was to eradicate the Black Codes, laws enacted by Southern legislatures imposing a range of civil disabilities on freedmen.”94 These laws passed to subordinate recently freed slaves were purposeful and often facially explicit.95 According to the Court, unlike during the passage of Title VII, “[t]he immediate evils with which . . . Congress was concerned [when it passed the provision] simply did not include practices that were ‘neutral on their face, and even neutral in terms of intent,’ but that had the incidental effect of disadvantaging blacks to a greater degree than whites.”96

In General Building Contractors, the Court concluded that the law’s prohibition against discrimination is limited by the nature of the discrimination prevalent at the time of its enactment. The shared substantive limitations of § 1981 and the Fourteenth Amendment are functions of the period in which the laws were forged and the circumstances they were intended to address. In General Building Contractors, purposeful discrimination is not judged to be the only “true” discrimination, but it is judged to be the particular form of discrimination targeted by Congress during the Reconstruction Era through both the Civil Rights Acts and the Fourteenth Amendment. A review of the relevant legislative history is the doorway through which the Court’s fact-based assumptions entered into its majority opinion.97 This is empirical reasoning practiced in a subordinate mode: elucidating lawmakers’ specific intent in order to conform the law’s substance and application to that intent, and to impose the Court’s present conception of discrimination.

The Court repeated this approach in Saint Francis College v. Al-Khazraji, when it concluded that § 1981’s definition of race is not informed by the “common popular understanding that there are three major races” or by biological and anthropological accounts that problematize the very concept of race.98 Instead, the Court held that Congress “intended to protect from discrimination identifiable classes of persons who are subjected to intentional discrimination solely because of their ancestry or ethnic characteristics” and that “[s]uch discrimination is racial discrimination . . . whether or not it would be classified as racial in terms of modern scientific theory.”99 As in General Building Contractors, this interpretation of § 1981 aligned it substantively with the Court’s prior interpretations of the Equal Protection Clause,100 but not because the Court imposed a single “true” account of race or race discrimination upon both the statute and the Constitution. Instead, the Court concluded that the statute and the constitutional guarantee were guided by the same historically contingent conception of race.

This reasoning leads to convergence in General Building Contractors and Al-Khazraji because the Court concluded that § 1981 and the Fourteenth Amendment originated in the same understanding of the racial wrongs that they were intended to address. The same mode supports divergence in Davis, on the ground that Title VII and the Constitution were conceived in very different eras and intended to address the different forms of discrimination associated with those times. General Building Contractors and Al-Khazraji show the Court using fact-based assumptions about the nature of race discrimination and popular conceptions of race found immediately following the Civil War to elucidate legislative intent. In this sense, the Court observes legislative supremacy by using empirical reasoning in a subordinate mode to discover political will. Though the Court’s reasoning in General Building Contractors has other deficiencies, this is its greatest virtue.

More problematically, the Court further supported its decision by arguing that it was bound by its previous interpretations of other provisions of the 1866 Act and the Fourteenth Amendment. The Court concluded that the 1866 Act and the Enforcement Act of 1870 are but “legislative cousins” of the amendment itself101 and express the “same general congressional policy” held by the 39th Congress.102 This jurisprudential argument is sadly unfocused. It appears to be partially structural in pointing out that the 1870 Act was passed relying on the authority granted to Congress by Section 5 of the Fourteenth Amendment to enforce the amendment’s provisions.103 It also appears to be partially an invocation of stare decisis: The Court uses General Building Contractors as an opportunity to solidify its interpretation of several provisions originating in the 1866 Act as requiring a showing of animus or discriminatory purpose.104 It also relied on Davis and its progeny to support the conclusion that neither the amendment nor the statutes with which it shared such a close history were intended to confer liability for race discrimination based on evidence of a racially disparate impact without a finding that “the impact can be traced to a discriminatory purpose.”105 The Court construed the 1866 Act as “an initial blueprint of the Fourteenth Amendment, which” itself was intended to “incorporat[e] the guaranties of the Civil Rights Act of 1866 in the organic law of the land.”106 It concluded that, since Davis had restricted equal protection liability to purposeful discrimination, the statute must be bound by the same limitation.

Among the authorities cited in General Building Contractors, only Davis answered directly the question whether a claim of discrimination under any law proposed or enacted by the 39th Congress could be sustained solely on the basis of a racially disproportionate impact. And, as in Davis, the Court followed a prevailing rule purportedly established by prior cases to reach its conclusion without examination of the public values that established the propriety of that rule. In General Building Contractors, the Court substitutes for a discussion of the equality values that might have been shared by the Equal Protection Clause and § 1981 an example of the type of practice at which the clause was aimed in order to constrain interpretation of the statute. This limitation in the Court’s analysis simply provokes other questions, such as: why does the Court interpret the example of the Black Codes in the way that it does? Pulling on the thread that this question presents unravels other aspects of the Court’s rationale.

The Court assumes that what is relevant about the Black Codes is their purposeful, rather than their subordinating, nature, and this assumption shapes how the Court puts their example to use. Linking the statute to the Fourteenth Amendment implicated the Davis rule, and interpreting the Black Codes as purposeful discrimination helped to solidify Davis.107 The Court might have read the Black Codes as a form of structural racial subordination—an attempt to perpetuate a racial caste system following the end of the Civil War through the denial of civil rights to former slaves and their descendants. If it had, then interpreting § 1981 to contain an effects-based test would have been consistent with such an antisubordination agenda.108 The statute’s origins in the Thirteenth Amendment would have further supported such an interpretation.109 Differences between the language of the Constitution and of the statute—the latter of which evidences an attempt to elevate the civil status of blacks to that of whites110—could have supported an interpretation that the statute reached effects-based claims though the Constitution did not, especially since the Court’s prior interpretation that the statute reached private activity had already granted the statute broader application than the Constitution.111 Finally, had the Court recognized the original text of § 1981 as a “blueprint” for the Fourteenth Amendment in Davis, it might have arrived at the opposite conclusion in that case and cited the antisubordination reading of the Black Codes in support. This would have meant that the Reconstruction Amendments and civil rights laws were intended to disestablish not simply what the Black Codes were, but the result—a racial caste system—that they were intended to bring about.

This alternative interpretation would establish convergence between the Constitution, § 1981, and Title VII standards, but not because the Court itself judged a particular set of values or assumptions to best reflect the most desirable vision of racial equality, the nature of race discrimination, or the judiciary’s limited institutional capacity. Instead, its decision would have been based on the conclusion that two congresses acting at different times had, as evidenced by the text and history of their enactments, converged around a common ideal of equality. The point is that even the Court’s subordinate use of empirical reasoning is subject to manipulation and conceals the work done by values orienting the Court’s proffered example of discrimination. In other words, the Black Codes example illustrates the statute’s scope only if we have a basis to identify what aspects of the example are relevant to the question, but the Court never articulates that normative framework.

B.     Comparing Convergence and Divergence Strategies

The virtues of convergence may seem obvious. Convergence promises to promote predictability and uniformity in the interpretation and application of race equality law by streamlining the doctrinal tests and substantive liability standards that apply to discrimination claims. Furthermore, reliance on empirical reasoning gives courts the opportunity “to get the facts right,” so to speak, with respect to the nature of the problem addressed by the law and will lead to convergence when the nature of the problem is considered stable across the social contexts in which the law is applied. It also grants courts the opportunity to adapt the law to social change—to pivot from overt “first generation” discrimination to more subtle forms of “second generation” discrimination.112

But convergence strategies are not always so appealing. Convergence based on empirical assumptions divides into two types. The first type describes decisions that treat the nature of race discrimination as static, freestanding, and independent of the conduct of democratic institutions which may for political or policy reasons seek to articulate different definitions of discrimination at particular moments in time. The second type describes decisions that attempt to identify the particular form of discrimination that the legislature sought to prohibit and to interpret the law in such a way as to enhance its effectiveness to combat those specific behaviors. The first approach pits the Court’s aspirations of regulatory potency and efficiency against fidelity to the letter and purpose of the law. Fact-based assumptions about the nature of race discrimination have the potential to be tenacious, uncompromising, and unresponsive to the concerns that motivated the passage of particular laws. They often are based on nothing more than the Court’s view of common sense.

The second approach uses empirical reasoning in order to understand legislative intent. General Building Contractors is an example of this approach. It shows that empirical reasoning practiced in this subordinate mode affirms the value of legislative supremacy and allows Congress considerable latitude to choose the factual assumptions on which interpretations of legislation should rest. This judicial approach also has the benefit of relying on a basis for restricting the statute’s interpretation that is itself within democratic control, in that Congress may override the Court’s interpretation. It has the additional advantage of promoting finality because the Court’s ascription of legislative intent is capable of resolving an issue of legal interpretation once and for all, regardless whether popular or scientific understandings of the issue change. However, General Building Contractors also shows that, even in the subordinate mode, empirical reasoning is disturbingly malleable because the attempt to resolve an interpretive question with a factual example masks an important normative choice regarding what aspects of the example are relevant to the question posed.

Attention to jurisprudential principles drives convergence when the letter or purpose of constitutional and statutory laws demonstrates that they were intended to share the same standard, when courts are presumed to lack the institutional capacity to perform a function which both laws require, or when the Constitution is understood to impose either structural or enforcement limitations on the substance or application of statutory law. When political institutions intend that two bodies of law substantively converge and design them to accomplish that end, convergence poses no counter-majoritarian difficulty. Structural limitations on the substance of statutory law force convergence when the Court determines that the constitutional provision authorizing congressional lawmaking itself contains a substantive limitation restricting the reach of laws enacted under its authority.

Consider as an example Title VII’s disparate-impact theory. To enact civil rights legislation under the authority of the Commerce Clause, Congress must target conduct that “‘substantially affects’ interstate commerce”;113 when Congress enacts legislation pursuant to its enforcement powers under Section 5 of the Fourteenth Amendment, “there must be . . . congruence and proportionality between the injury to be prevented or remedied.”114 As Post and Siegel have warned, if Title VII’s disparate-impact provisions were held to require authorization under Section 5 (e.g., to reach facially neutral conduct by the states and their subdivisions), those provisions may need to be reinterpreted as doing no more than providing an effects-based test for “smok[ing] out” intentional discrimination because—according to the constitutional vision articulated by Davis and its progeny—race discrimination is motive-based.115

Enforcement limitations on the application of statutory law occur when statutory law requires state action that does not survive constitutional challenge. For example, in Shaw v. Reno, the Court concluded that the Constitution places an enforcement limitation on the form of majority-minority districts drawn in order to comply with the Voting Rights Act of 1965 (“VRA”).116 Specifically, the Court held that districts that failed to observe “traditional districting principles [of] compactness, contiguity, and respect for political subdivisions” may violate equal protection.117 The Court extended this argument in Miller v. Johnson, when it ruled that majority-minority districts drawn to comply with the Act will violate equal protection if the government considers race as a “predominant factor.”118 These cases forbid the Act to be enforced in a manner that runs afoul of the Constitution. Justice Scalia articulated a kind of enforcement limitation argument against disparate-impact theory when he forewarned of a conflict between statutory disparate impact and equal protection. His assertion is not that disparate impact is unconstitutional because it cannot be supported under Section 5, but that disparate-impact liability cannot be enforced without violating equal protection because the standard forces employers to do what the government may not—base their decisions upon considerations of race.119 The implication here is that, in the context of employment, disparate-impact theory may be authorized by the Constitution (e.g., under the commerce power), and yet it may also be unconstitutional because, through its enforcement, the government violates equal protection.

The Court’s decisions in this area are also informed by the view that courts have limited institutional capacity. Lack of institutional capacity may support legal interpretations that relieve courts of the responsibility to perform functions for which they purportedly are not well suited; and, to the extent that these functions may be implicated by either constitutional or statutory law, restricting them in the same way would force convergence. For example, if Palmer v. Thompson had prevailed and the Court had continued to presume that courts generally lack the ability to judge the motives of collective actors, the Court might have upheld disparate impact under the Constitution in Davis or upheld a method for determining intent modeled after disparate-impact theory, perhaps following the statute’s burden-shifting approach.120 Conversely, following Justice Scalia’s suggestion in Ricci,121 if the Court were to determine that disparate-impact theory is defensible only as a device for smoking out discriminatory purposes, it may conclude that due to limitations in its own institutional capacity it is simply not up to the task. This is not a likely outcome given the history of the Court’s endorsement of statistical proof as circumstantial evidence of discriminatory purposes in constitutional cases.122 It is, however, a theoretically plausible one demonstrating that arguments based on the limits of the judiciary’s institutional capacity may cut in more than one direction.

* * * * *

The virtues of divergence may not be immediately appreciable from Davis. The case is best known as an example of civil rights retrenchment. General Building Contractors shows that its effects reach beyond constitutional claims. Davis establishes that the substantive equality protections of the Constitution are less broad than those of an act of Congress, even so far as claims against the government are concerned. This result is counterintuitive. Disparate-impact theory distributes liability on a structural basis, examining the interaction between facially neutral conduct and racially identifiable external factors, including the effects of discriminatory public policy, that produce an adverse impact regardless of the defendant’s intent.123 Governmental actors would appear to be the most appropriate targets for such a theory of liability, but, according to Davis, the Constitution does not hold the government to such a standard. An interpretation of the Constitution viewing the latter as indifferent to the structural sources of racial inequality may seem difficult to justify, but, even as illustrated by Davis, divergence has its virtues.

In Davis, supporting divergence through reliance on jurisprudential rules requires a double movement. First, the Court determined that, absent a suspect facial classification, the Constitution’s equality guarantee requires proof of a discriminatory purpose. Second, the Court affirmed congressional authority to enact legislation exceeding the scope of the constitutional guarantee by prohibiting employers from using facially neutral employment practices that produce racially disproportionate impacts even absent proof of a discriminatory purpose. Without the second movement there is no opportunity for divergence. It shows the Court conceding to Congress the authority to pursue legal strategies that depart from the substantive requirements of the Constitution. Read in this way, Davis is a case about enabling substantive diversity in race equality law by respecting differences in the letter and purpose of the law. As a jurisprudential matter, attentiveness to substantive distinctions between statutory and constitutional law may be understood to express fidelity to the law as written, to avoid judicial usurpation of political functions, to promote the value of legislative supremacy, and to admit limitations of the judicial power that issue from the limited ability of courts to effectively review certain types of actions or to second-guess certain types of actors.

As a matter of legal policy, divergence provides opportunities for flexibility—for lawmakers to explore different approaches to a particular legal problem—which is an advantage when facing a problem as vexatious as racial inequality. Although the outcome in Davis effectively placed out of the reach of equal protection doctrine facially neutral governmental practices for which the plaintiff lacks evidence of discriminatory purpose, the rule of Davis permits the interpretation of congressional statutes to be independent of constitutional equality doctrine. This allows the rights and remedies available under these statutes to serve different objectives than the Constitution, such as substantive equality measured in terms of proportional representation or achievement. The Equal Protection Clause and the Fifteenth Amendment each require proof of a discriminatory purpose,124 while Title VII and the VRA are violated by facially neutral practices that produce racially disproportionate effects.125 This difference has allowed Title VII to incentivize employers to utilize formal employment procedures that have helped to diversify the workforce, while increasing fairness to all workers.126 It has also helped the VRA to ban voting practices that would otherwise have had the effect of holding minorities in a position of second-class citizenship.127 The constitutional discriminatory purpose doctrine requires evidence of malice, or animus,128 a standard that is ill-suited to address “second generation discrimination” that frequently results from unconscious stereotyping.129 Title VII’s prohibitions against race discrimination in employment have no animus requirement.130 The Constitution permits race-based affirmative action only when the government can establish that its use of race was narrowly tailored to fulfill a compelling interest in either remedying past discrimination or obtaining diversity.131 Title VII permits race to be considered in employment without a showing that the employer engaged in past discrimination and for the purpose of correcting a manifest racial imbalance even if to do so would exceed what would be required to obtain a critical mass of racial minorities necessary to obtain the benefits of diversity.132

Were the Supreme Court to deny congressional authority to enact laws that deviate substantively from the Court’s interpretations of the Constitution, this would hobble race equality legislation and undermine our constitutional order by effectively outlawing—depending on one’s point of view—desirable or, at the very least, legitimate policy objectives. Thus, read in the reverse, Davis itself is not a case about restricting civil rights protections under the Constitution, but a case about preserving for Congress the option to exceed constitutional protections in those areas where the Constitution grants Congress the legislative authority to do so.

III.     Shifting Strategies of Convergence and Divergence: Race Equality Law Today

Part I discussed two decisions in which the Supreme Court faced a choice between enforcing convergence or divergence of constitutional and statutory race equality law. These examples come from the period in which the Court laid the foundations for our contemporary understanding of the relationship between these two sources of law. This Part will show that the Court appears to be in the process of charting a new course by pursuing more aggressive convergence strategies than it had in earlier cases.

A.     Employment Discrimination and Affirmative Action

In Ricci, the Court discussed two kinds of convergence between constitutional and statutory law. The first concerns the Court’s holding that, if an employer commits race-based disparate treatment in order to avoid potential liability under Title VII’s disparate-impact test, it may avoid disparate treatment liability only if it is able to show that it had “a strong basis in evidence” to conclude that it would have been liable for disparate impact had it not taken the challenged action.133 The Court imported the “strong basis in evidence” test from Wygant v. Jackson Board of Education, a constitutional affirmative action decision which instructs that the government may employ affirmative action to remedy past discrimination only if it has “a strong basis in evidence [to conclude] that the remedial action [is] necessary.”134

The second is reflected in Justice Scalia’s concurring opinion in which he prophesied that the Ricci decision “merely postpones the evil day on which the Court will have to confront the question” of the constitutionality of the statutory disparate-impact test.135 Justice Scalia did not argue that the Constitution structurally limits the substance of Title VII, as it would if Title VII depends, in some measure, on authority from Section 5 of the Fourteenth Amendment. Instead, he argued that by establishing disparate-impact liability, “Congress pressured employers to give preferences to minority employees in ways that violate” equal protection.136

In essence, Justice Scalia argued that the Constitution will not permit the government (including the states to which the same principle would apply) to enact laws that define discrimination as disparate impact because such laws cannot be enforced without violating equal protection.137 This argument flatly rejects the basis for divergence articulated in Davis: that Congress possesses authority to enact a disparate-impact standard even though the standard is not available under the Constitution.138 But Justice Scalia goes even further, suggesting that the government cannot prohibit disparate impact because it is not “genuine” discrimination.139 This Section will examine Justice Scalia’s opinion as a call for further convergence.

1.     Disparate Impact and Discriminatory Purpose After Ricci

The Ricci case originally concerned claims under Title VII and the Equal Protection Clause. White and Latino firefighters who had taken, and sought certification of the results of, a test determining eligibility for promotion to the rank of lieutenant or captain within the City of New Haven’s fire department, brought these claims. The city invalidated the results of the test when it observed that the test had a racially disproportionate impact and was threatened with suit by black firefighters.140 On cross-motions for summary judgment, the district court ruled for the defendants, and the Second Circuit summarily affirmed. The district court found that the defendants’ desire to comply with Title VII’s disparate-impact provisions did not prove disparate treatment under the statute or a discriminatory purpose violating equal protection. Relying on prior circuit precedent, the court concluded that the “[d]efendants’ motivation to avoid making promotions based on a test with a racially disparate impact . . . does not, as a matter of law, constitute discriminatory intent”141 and that therefore nothing in the record raised a genuine issue of fact as to whether the defendants had acted with a discriminatory purpose.142

The Supreme Court decided only the statutory issue and reversed, entering summary judgment for the plaintiffs, because it rejected the district court’s conclusion that the city’s motivation to comply with the statute did not constitute discriminatory intent as a matter of law. Rather, the Court began its analysis from the “premise [that] [t]he City’s actions would violate the disparate-treatment prohibition of Title VII absent some valid defense.”143 The Court further concluded that the statute’s disparate treatment and disparate-impact provisions were “in conflict absent a rule to reconcile them.”144 The Court then looked to its constitutional affirmative action decisions to provide such a rule, concluding that the city needed to demonstrate “a strong basis in evidence to believe” that certification of the test results would have exposed it to disparate-impact liability in order to mount a proper disparate treatment liability defense.145 Writing for the majority, Justice Kennedy was careful to point out that the case “does not call on us to consider whether the statutory constraints under Title VII must be parallel in all respects to those under the Constitution.”146 He reasoned, however, that the constitutional cases could “provide helpful guidance in” resolving the intra-statutory conflict between disparate treatment and disparate impact.147 Justice Kennedy argued that the “strong basis in evidence” test was the right mechanism to address this conflict because requiring an employer to be certain that its affirmative action measure remedied an actual violation of law “would bring compliance efforts to a near standstill,” while “[a] minimal standard could cause employers to discard the results of lawful and beneficial promotional examinations” and set in place “a de facto quota system.”148

Whether or not the Court would hold that a strong basis in evidence to avoid disparate-impact liability would provide a defense against an equal protection claim,149 the Court’s opinion in Ricci does much to unify the constitutional and statutory standards: it restricts the employer’s ability to anticipate and to avoid discriminatory actions by appropriating the same doctrinal device that it had previously used to limit the government’s authority to implement voluntary racially remedial measures. The logic here is at once conceptual and jurisprudential. Conceptually, it holds that race-conscious action is discrimination unless the defendant has a basis to conclude that it is directly remedial. Jurisprudentially, it forbids the judiciary from sitting in absolute judgment of the defendant’s calculation. The question is not whether the defendant truly violated the law, but whether the defendant had a strong basis in evidence to conclude that it had.

In Ricci, the Court chose convergence over some very obvious obstacles in the case. Refusal to certify the results of a test producing a disparate impact is distinguishable from affirmative action because it involves no race-based preference. In addition, if the Court wanted to obtain guidance by analogizing the certification decision to affirmative action, it should more appropriately have looked to statutory precedents which reviewed employers’ voluntary affirmative action programs under a more lenient standard.150 And so, the form of convergence practiced in Ricci is aggressive, pushing past intra-statutory continuity and instead promoting continuity between constitutional and statutory law.151

Ricci undermines disparate-impact theory’s capacity to shape employer behavior because it gives employers a disincentive to respond voluntarily and proactively to evidence of disparate impact. The “strong basis in evidence” standard defines the narrow safe haven in which the employer is permitted to act. Ricci’s implications for constitutional discriminatory purpose doctrine are far less clear. Under Personnel Administrator of Massachusetts v. Feeney, the Court would have been required to find that the city acted “‘because of,’ not merely ‘in spite of,’ its adverse effects” on the white and Latino firefighters.152 The Court avoided a constitutional decision, and yet tied the Constitution and the statute together by repurposing a constitutional rule to mend what it represented to be a defect in the statute’s design. In finding that the City’s actions constituted disparate treatment, the Court did not require evidence that the city acted with any animus.153 Nor did it require evidence that the City treated plaintiffs differently because of their race—that, differently than if the racial impact of the test had been reversed.154 The Court’s indifference to the government’s aims begs the question whether, if Ricci foreshadows revision of the constitutional standard, the Court may conclude that compliance with antidiscrimination law, diversity, and integration are discriminatory purposes because satisfaction of such purposes requires attention to racial outcomes.155

In his concurrence, Justice Scalia presumed that the two standards were the same and that, having found disparate treatment based on the city’s consideration of a disparate impact, the Court must conclude that the same motivation would violate equal protection.156 Were the Court to adopt that view, it would negate the animus requirement of Feeney and “raise[] constitutional questions about other race-conscious, facially neutral efforts to rectify bias, to increase diversity, or to integrate.”157 This outcome is, of course, not accomplished in Ricci, though the decision suggests its possibility. Assimilating the statutory rule to equal protection would require more than a rejection of Feeney. It would also require a rejection of the rationale of General Building Contractors which found that, contrary to Title VII, the Reconstruction statutes and amendments restricted liability to actions taken for a discriminatory purpose in response to the practices of that era.158Ricci does not command this result, but Justice Scalia’s concurrence makes it a plausible prediction of the Court’s future direction.

The Court’s already aggressive pursuit of convergence makes Justice Scalia’s concurrence in Ricci all the more earthshaking. The decision imposes a constitutional standard on the statute for which the Court and Congress arguably have most strongly favored divergence. In doing so, it throws into question the continuing legitimacy of the Court’s statutory affirmative action cases.159 According to Justice Scalia, more work is still to be done, because the Court’s decision leaves unanswered “the question: Whether, or to what extent, . . . the disparate-impact provisions of Title VII . . . [are] consistent with the Constitution’s guarantee of equal protection.”160 The difficulty, according to Justice Scalia, comes from his observation “that Title VII not only permits but affirmatively requires [remedial race-based] actions when a disparate-impact violation would otherwise result.”161 This view makes it seem unlikely that Title VII could be justified “as simply an evidentiary tool used to identify genuine, intentional discrimination.”162 The disparate-impact provisions instead appear to “place a racial thumb on the scales” by requiring employers to evaluate and to make decisions based on “the racial outcomes of their” employment practices.163 This amounts to “[g]overnment compulsion of” employers to implement race-conscious practices—in short, to do what the government itself could not without violating equal protection.164

Justice Scalia’s argument is not that Congress lacked the structural authority to enact the disparate-impact provisions of Title VII, but that those provisions cannot be enforced without compelling employers to adopt race-conscious practices in order to avoid disparate-impact liability and that the compulsion of such race-conscious action violates equal protection.165 This is an enforcement limitation argument. Justice Scalia does not develop the argument by resolving whether or not disparate-impact provisions requiring employers to evaluate the racial outcome of their practices themselves classify by race, thus authorizing the Court to apply strict scrutiny, or whether they appear driven by a discriminatory purpose.166 The specific test by which disparate impact could be found unconstitutional is unclear from his opinion. Nevertheless, what is clear is his conclusion that Congress cannot compel employers to do what it cannot do itself: attempt to remedy, or to avoid, racial impacts that are not themselves caused by committing genuine discrimination through the direct consideration of race.

2.     Affirmative Action After Ricci and Fisher

Legal scholars George Rutherglen and Daniel Ortiz wrote a quarter century ago that, in the area of affirmative action, “the divergence between the Supreme Court’s treatment of the statutory and constitutional issues has largely escaped notice” and that “the constitutional and statutory standards not only diverge but diverge differently in different contexts”167 and also “converge at several points.”168 In the intervening years since Rutherglen’s and Ortiz’s article, little has been said that illuminates how and why these patterns of divergence and convergence occur in affirmative action and across race equality law. Of particular concern is the Court’s treatment of voluntary affirmative action programs under Title VII and equal protection.169 Since the publication of their article, the two bodies of law have continued to diverge on this issue, but, following Ricci, the future direction of the law is uncertain.

One year after Rutherglen and Ortiz published their article, the Court decided City of Richmond v. J.A. Croson Co.170 Under Croson and its progeny, equal protection doctrine purports to subject all racial classifications, including public, race-based affirmative action programs, to strict scrutiny regardless whether the government’s purpose is “benign.”171 The doctrine precludes the government from justifying its use of race on the grounds that it seeks to remedy “societal discrimination”172 or to achieve a proportional representation of minority group members, which the Court calls “racial balancing.”173 While the Court has recognized diversity as a compelling interest in education,174 it has never done so in employment.

Instead, under Wygant, an employer’s voluntary affirmative action program will survive strict scrutiny only if the government has “a strong basis in evidence for its conclusion that the remedial action was necessary.”175 In contrast, Title VII permits public and private employers to consider race with the goal of producing a racially-balanced workforce. The Court held in United Steelworkers of America v. Weber that race-conscious decision making executed pursuant to a voluntary affirmative action plan complies with the statute if the plan meets certain criteria: that it is designed to correct a “manifest racial imbalance,” “does not unnecessarily trammel the interests of the white employees” or “create an absolute bar to [their] advancement,” and is a “temporary measure” that will not extend beyond the correction of the imbalance.176 No admission or evidence of past discrimination is required.177 The doctrine permits the employer to defend its action as a response to societal, or structural, discrimination. To a certain extent, identifying a traditionally segregated job category requires awareness of societal racial subordination patterns,178 and the employer is authorized to correct an imbalance by seeking proportional representation among racial groups provided it does not attempt to maintain such representation through continuing consideration of race once it achieves its goal.

According to Eskridge, the Court’s decision in Regents of the University of California v. Bakke crystallized certain public and constitutional values regarding the propriety of affirmative action as a remedy for discrimination.179 Those values, in turn, Eskridge argues, influenced Weber and its progeny.180 Some evidence of this can be seen in Johnson v. Transportation Agency of Santa Clara, when the Court upheld a plan that used sex as one factor among many, “set[] aside no positions for women,” and “did not authorize . . . blind hiring.”181 It therefore resembled the sort of “plus” factor plan that could pass constitutional muster,182 and the Court indeed cited Justice Powell’s approving discussion of the “Harvard Plan” for university admissions in support of its conclusion that the agency’s affirmative action plan was lawful under Title VII.183Johnson, however, applied Weber, and Weber involved precisely the kind of racial quota that Justice Powell rejected in his Bakke opinion.

The “strong basis in evidence” test is squarely at odds with the statutory Weber-Johnson rationale, and so the Court’s bypass of the statutory test by adopting the constitutional standard in Ricci is aggressive and confusing. Ricci is not an affirmative action case, but it intersects affirmative action doctrine on two levels: on its facts and as a matter of law. It intersects on its facts by reviewing an employer’s attempt to remedy what the employer perceived to be discrimination by adopting a race conscious practice. As a matter of law, it intersects by borrowing from constitutional affirmative action doctrine to create a new statutory rule. However, that rule sharply contradicts established statutory affirmative action doctrine. In short, if an employer uses a racial preference, it enjoys a lenient standard that preserves its discretion and does not require proof of past discrimination. If an employer voids the results of a facially neutral test because the test produced a disparate racial impact, strong evidence of prior discrimination is required. This result is extremely difficult to justify, and both equal protection jurisprudence and disparate treatment doctrine before and after Ricci suggest that the heavier burden should fall on the use of racial preferences. It also begs the question of how to distinguish between affirmative action programs and other mechanisms by which the employer considers race in order to remedy either a racial imbalance or past discrimination.

The Court’s statement that the test should be applied “to resolve any conflict between the disparate-treatment and disparate-impact provisions” is also aggressive.184 The timing of the city’s decision not to certify the test results is important in that the Court noted the plaintiffs’ reliance investment.185 However, the Court had no occasion to consider whether designing a test to preclude a particular racial distribution was permissible—in his concurrence, Justice Scalia stated his belief that it is not.186 If Ricci compels that conclusion because the statute makes no distinction between racial considerations made before or after the application of a test, then Ricci will have implications for cases in which facially neutral practices (e.g., designing a test) are alleged to have racial purposes (e.g., promoting racial diversity), as well as for affirmative action cases.

Already, the decision has impacted affirmative action jurisprudence in lower federal courts. The Second Circuit, which affirmed the decision to which the Supreme Court granted certiorari in Ricci, has read Ricci to “indicate[] that not all voluntary race- or gender-conscious employer action is properly analyzed under Weber and Johnson.”187 In United States v. Brennan, the circuit court acknowledged that, prior to Ricci, it had interpreted Weber and Johnson to mean that if an employer undertook voluntary racial remedies when faced with a prima facie case of disparate-impact discrimination, it would not face Title VII liability for disparate treatment.188Ricci, conversely, demonstrated that some voluntary race-conscious measures are not affirmative action, but “individualized,” “make-whole relief,” and are therefore not subject to the Weber–Johnson rationale.189 The benefits provided under an affirmative action plan are forward-looking and class-wide; the relief provided by the city in Ricci was backward-looking and addressed only the situation of persons who would have been negatively affected by certifying the results of the promotion test. The Second Circuit concluded that the defendants’ voluntary implementation of a settlement agreement providing retroactive seniority benefits to minority workers who were subjected to a discriminatory employment test should be reviewed under Ricci’s “strong basis in evidence” test because it constituted make-whole relief, even though the grant of seniority did not automatically appoint workers to the positions they sought.

The Brennan decision purports to delineate distinct, non-overlapping areas of employment discrimination law in which Ricci and the Court’s prior affirmative action precedents may coexist by avoiding direct collision. Brennan, however, does not tell us why less-broad, less-invasive, facially neutral individualized relief that may result in no actual change of employment status ought to command a higher standard than class-wide racial preferences that may entitle beneficiaries to employment positions. For all its efforts, the Second Circuit’s opinion only underscores the aggressiveness of the convergence strategy implemented in Ricci by showing the radical adjustment required to render Weber–Johnson compatible with Ricci in practical terms. If the Second Circuit is correct, Ricci has curtailed the scope of the Weber–Johnson rationale and that, in and of itself, accomplishes significant convergence simply by restricting the reach of the statutory doctrine that conflicts with the constitutional rule. The appropriation of the constitutional affirmative action test impacted the statutory affirmative action test, even though Ricci did not present the Court with an affirmative action question. Given Justice Scalia’s warnings about future conflict between disparate impact and equal protection, we ought also to consider whether greater convergence in affirmative action doctrine is on the horizon.

To address this question, we should consider not only how Ricci has altered Title VII doctrine, but also how changes in the Court’s equal protection doctrine may influence future interpretations of the statute. The analytical dissonance of applying a higher standard in Ricci than would be applied in a true affirmative action case under Weber and Johnson is itself sufficient reason to suspect that the Court will be compelled to revisit the statutory affirmative action doctrine. Furthermore, although the Court has yet to recognize diversity as a basis for workplace affirmative action programs, some scholars have suggested that it should.190 Doing so would allow affirmative action programs to receive statutory protection absent evidence of a manifest racial imbalance in a traditionally segregated job category. However, the potential effect of diversity on statutory affirmative action doctrine is uncertain. It is difficult to imagine how diversity would be imported into the statutory framework—for example, whether it would come with something akin to the constitutional test of strict scrutiny, and whether it would require demonstration of some downstream instrumental benefit to the employer or to society just as diversity has been held to provide in education.191

Under the Court’s equal protection affirmative action decisions, to sustain a race-based affirmative action plan based on a university’s pursuit of diversity, a court must “verify that it is ‘necessary’ for the university to use race to achieve the educational benefits of diversity” and that “court[s] must ultimately be satisfied that no workable race-neutral alternatives would produce the educational benefits of diversity.”192 Under equal protection, the concept of “critical mass” constrains the pursuit of diversity. This means that the defendant’s pursuit of diversity may permit it to justify policies necessary to obtain a critical mass, or “meaningful numbers,” of members of a particular racial group.193 This is, in rough terms, the number necessary to obtain the benefits of diversity—in Grutter and its progeny, the benefits of diversity necessary to advance the university’s educational mission.194 The Court has never held that critical mass could be synonymous with proportional representation. To the contrary, in Grutter, the University of Michigan Law School’s pursuit of a critical mass of minority students was praised because it did not result in the achievement of a fixed, or even a consistent, benchmark,195 and the Court has otherwise rejected the objective of racial balancing in its equal protection jurisprudence.196

Weber and Johnson do not require the exhaustion of race-neutral alternatives to affirmative action in order to justify the employer’s use of racial preferences. They permit the employer to set an employment target, unrestricted by the concept of critical mass, to correct manifest imbalances in its workforce. Were these doctrines of exhaustion and critical mass to be incorporated under Title VII to amend the Weber–Johnson framework, they would likely restrict employers’ efforts to make their workplaces more representative of their specific labor pools. Under equal protection doctrine, these requirements have attached to the concept of diversity, but they could be severed from diversity and applied to the statutory concept of manifest imbalance. Furthermore, even if diversity itself came with the doctrinal components intended to limit the use of diversity-motivated programs, there may be little payoff for employers because a situation in which an employer can credibly claim that it needs to pursue a racial critical mass is also very likely to be a situation in which the employer’s workforce exhibits a manifest racial imbalance, unless the labor pool itself is virtually devoid of minority workers. Moreover, if an employer can establish a manifest imbalance, correcting the imbalance justifies greater affirmative measures than obtaining a critical mass because a critical mass will only be a number large enough to produce some performance benefit for the employer rather than large enough to mirror the proportion of minority workers in the relevant labor pool.

Finally, Ricci shares certain qualities with equal protection cases generally and with Fisher v. University of Texas at Austin, the Court’s most recent equal protection affirmative action decision, in particular that may increase the likelihood of further convergence between constitutional and statutory law. First, in both Ricci and Fisher, the Court demonstrated a willingness to defend particular notions of merit and to second-guess the defendants’ decisions with regard to what constitutes merit. Title VII historically has afforded employers broader discretion to determine the qualifications for work and the mechanisms for identifying such qualifications than is recognized under Ricci.197 The statute’s voluntary affirmative action doctrine reflects this respect for employer discretion. Both Ricci and Fisher open with a discussion of the underlying activity at issue in the case (i.e., firefighter promotion, university admission) as the pursuit of a merit-based benefit that is “prized and competitive.”198 Both show the Court asserting its authority to review merit systems and to defend those systems in concert with guaranteeing an individual’s right to equal protection. Fisher does so through its “no workable race-neutral alternatives” test.199Ricci does so by denying the city the discretion to revise its conditions for promotion when it finds that it has reasons to doubt the accuracy and legality of its existing measures, and by treating the city’s refusal to certify the test results as itself an injury. The Court, in both cases, appears to be operating under the fact-based assumption that measures directed at relieving racial subordination, such as affirmative action and disparate-impact theory, undermine merit systems to the detriment of deserving whites. This is an assumption serving both empirical and conceptual ends in which whites and minorities are considered equally vulnerable. “Genuine, intentional discrimination,” whether as disparate treatment or affirmative action, and efforts to alleviate structural subordination and not those subordinating structures themselves are viewed as discrimination.

Second, Ricci shares a peculiar notion of injury with equal protection jurisprudence. The Court’s equal protection decisions make clear that “being forced to compete in a race-based system that may prejudice the plaintiff” is a form of cognizable constitutional injury.200 No material loss or denial of benefit or opportunity is necessary. The harm suffered by the plaintiff is expressive and dignitary.201 By contrast, Title VII recognizes only “adverse employment actions” as giving rise to a claim of disparate treatment, and not all slights or inconveniences experienced in the workplace will satisfy that criterion, even if they are the result of prejudice.202 Insults and purely dignitary harms will not suffice. Ricci, however, does not follow the Title VII paradigm. The plaintiffs in Ricci were not denied a position or benefit; the promotions for which performance on the city’s exam were relevant were simply suspended when the city decided not to certify the results of its test. As constitutional scholar Richard Primus has recognized, the Ricci plaintiffs having been subjected to “a decisionmaking process infected by a state actor’s illicit consideration of race” would adequately support an equal protection claim.203 The Court may be willing to carry the constitutional concept of injury to Title VII’s affirmative action doctrine. Rather than invalidating only those voluntary affirmative action plans that raise an “absolute bar” to the success, or that “unnecessarily trammel” the interests of whites,204 the Court may conclude that, as in the equal protection context, consideration of race is an injury requiring greater justification.

B.     Voting Rights

Since the 1990s, the Supreme Court has interpreted the Equal Protection Clause to place strong enforcement limitations on the VRA.205 States and their political subdivisions are not permitted to comply with the Act by considering race as a predominant factor in the drawing of district lines or by ignoring traditional districting principles to draw district lines that trace the racial composition of included communities.206 These cases, however, dealt with constitutional challenges to districts drawn to satisfy the VRA’s requirements, and the enforcement limitations that they establish stop short of restricting the substance of the VRA itself. In its recent cases, however, the Court has done just that.

In Bartlett v. Strickland, the Court did not consider a constitutional claim challenging districts drawn to comply with the VRA as it had in Miller and Shaw.207 Instead, it considered whether section 2 of the VRA itself should be construed to require state officials to draw district lines so as to allow a racial minority to join with “crossover majority voters” to “elect its candidate of choice.”208 Section 2 provides for a claim of vote dilution if a voting qualification or procedure is “imposed or applied . . . in a manner which results in a denial or abridgment of the right of any citizen . . . to vote on account of race” and if members of a racial class “have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.”209 The Court concluded that section 2 liability requires proof that the minority population in the potential district is greater than 50%. In his opinion on behalf of the Court, Justice Kennedy argued that as a statistical minority, “African Americans standing alone have no better or worse opportunity to elect a candidate than does any other group of voters with the same relative voting strength” and therefore do not “have less opportunity than other members of the electorate to” participate or to elect the candidate of their choice.210

In a sense, this interpretation imposes the constitutional norm of colorblindness on the VRA, because it “limit[s] the contexts in which government can rely on racial classifications in implementing the Act” and denies the government potentially effective race-conscious measures, such as cross-over districts.211 This is particularly significant because Congress added the effects-based language of section 2 in an effort to differentiate between the statutory test for vote dilution and the constitutional discriminatory purpose standard following City of Mobile v. Bolden, in which the Court interpreted the statutory test to incorporate the constitutional standard.212 Nevertheless, although the Bartlett decision may have the practical effect of restricting the application of section 2 to accord with mainstream constitutional values, Justice Kennedy did not argue that the statute’s interpretation must be congruent with the substance of the Constitution. He invoked constitutional cases establishing the colorblindness norm only when he discussed the “serious constitutional concerns” that would be raised were section 2 to provide a basis for liability unenforceable under the Equal Protection Clause.213

In other words, Justice Kennedy’s opinion does not transfer colorblindness to the interpretation of section 2 without identifying a discrete, jurisprudential basis for doing so. First, Justice Kennedy argued that the statute would face a serious enforcement limitation under equal protection if it were construed to require the creation of crossover districts. Justice Kennedy cited constitutional precedent to establish that “the ‘moral imperative’ of racial neutrality is the driving force of the Equal Protection Clause,”214 and he explicitly warned of the danger that a cross-over district requirement would run afoul of Miller v. Johnson.215 Second, even assuming that crossover districts could be designed to avoid violating equal protection, determining when a crossover district would be required under the states “would place courts in the untenable position of predicting many political variables and tying them to race-based assumptions.”216 The majority-minority requirement, by contrast, “draws clear lines” and creates “workable standards” for “sound judicial and legislative administration.”217 Justice Souter argued that this position overestimated the ease of determining that a racial majority is sufficiently large to elect the candidate of its choice, because the threshold population is “elastic” and courts have, in the past, required significantly higher percentages than 50% to obtain assurance that a district was indeed “safe.”218 This argument demonstrates that questions of the courts’ capacity are relative. Claims of limited capacity may turn on overestimations of the difficulty of tasks purported to be outside of judicial competence or on underestimations of the difficulty of tasks otherwise characterized as administrable and within ordinary competence.

Justice Kennedy also considered the fact-based assertion that “racially polarized voting is waning,” which could be taken to provide a basis imposing a requirement of crossover districts on the ground that healthy crossover margins should often be expected.219 He rejected this argument, countering that because “racial discrimination and racially polarized voting are not ancient history” the statute should not be “interpreted to entrench racial differences by expanding” its basis for liability.220 He explained that “[c]rossover districts are, by definition, the result of white voters joining forces with minority voters” and that the Court had “decline[d] now to expand the reaches of [section] 2 to require, by force of law, the voluntary cooperation our society has achieved.”221 The conclusion that the expansion of liability to require in some instances the creation of crossover districts would “entrench racial differences” is odd given that the purpose of crossover districts is to facilitate interracial voting solidarity. It seems, however, to turn on an assumption familiar in the Court’s constitutional cases that the government’s consideration of race itself causes injury and leads to racial balkanization. Bartlett strikes an aggressive step toward constitutional and statutory convergence almost three decades after Congress signaled, in response to City of Mobile, that it wanted the standards to remain separate. Justice Kennedy’s opinion elaborates on a series of jurisprudential and empirical justifications for the Court’s interpretation of section 2 and does not simply rely on the moral force of colorblindness as a value adopted from constitutional jurisprudence. The effect in either case would have been convergence, but the justificatory modality of the Court’s convergence-divergence choices, the subject of this Article, remains dependent on recurrent and malleable jurisprudential and empirical rationales.

The Court’s most recent VRA decision charts an even more aggressive path toward convergence, one in which the Court relies upon fact-based assumptions to impose structural, constitutional limitations on congressional authority. In Shelby County v. Holder, the Court held unconstitutional section 4 of the VRA, which established the formula used to determine which jurisdictions are covered by the preclearance requirement of section 5.222 Preclearance requires covered jurisdictions to demonstrate that changes in voting procedures “had neither ‘the purpose [nor] the effect of denying or abridging the right to vote on account of race or color.’”223 Both the preclearance requirement and its effects-based approach to determining what voting procedures may be blocked from enforcement strongly distinguish the statute from the Fifteenth Amendment’s purpose and classification-based approach to ending race discrimination in voting.224 Though the Court made clear that its opinion does not address the constitutionality of section 5,225 the mechanism of preclearance is effectively suspended awaiting Congress’s reenactment of a constitutionally permissible coverage formula. The ultimate expression of convergence regarding the VRA would indeed be a declaration that preclearance is unconstitutional, and that day has not yet come. Nevertheless, the effects-based approach of section 5 is, for the time being, suspended, rendering yet another provision of the VRA designed to pursue racial equality in a manner more ambitious than the Constitution crippled, if not fatally wounded.

Shelby County concerns a question of the scope of congressional lawmaking authority under Section 2 of the Fifteenth Amendment, which grants Congress the authority to “to enforce . . . ‘by appropriate legislation’” the constitutional right to vote free from race-based discrimination.226 The question was whether section 4 of the VRA was a proper exercise of this power. The Court pinned its answer squarely on an empirical rationale: that the government’s “‘current need[]’ for a preclearance system that treats States differently from one another” was contradicted by fact-based judgments about the present conditions found in the jurisdictions identified for preclearance by section 4’s coverage formula.227 The provision required preclearance of changes made by jurisdictions which, at the time of the VRA’s passage in 1965, used blatantly discriminatory voting measures and had low voter turnout. However, when Congress reauthorized the VRA in 2006, conditions in the covered jurisdictions had evolved.228 The Court had warned prior to reauthorization, in Northwest Austin Municipal Utility District No. One v. Holder, that the “current burdens” imposed by the VRA “must be justified by current needs’” and that disparities in the VRA’s “geographic coverage [must be] sufficiently related to the problem that it targets.”229 The Court quoted Congress’s own 2006 statement that “[s]ignificant progress has been made in eliminating first generation barriers experienced by minority voters,”230 and argued that Congress had failed to update the Act accordingly:

[H]istory did not end in 1965. By the time the Act was reauthorized in 2006, there had been 40 more years of it. In assessing the “current need[]” for a preclearance system that treats States differently from one another today, that history cannot be ignored. During that time, largely because of the Voting Rights Act, voting tests were abolished, disparities in voter registration and turnout due to race were erased, and African-Americans attained political office in record numbers. And yet the coverage formula that Congress reauthorized in 2006 ignores these developments, keeping the focus on decades-old data relevant to decades-old problems, rather than current data reflecting current needs.231

The Court thus objected to a continuation of preclearance “as if nothing had changed” regarding the voting conditions found across the nation.232

The Court’s factual assertions serve several purposes. As discussed above, they undermine the rational basis for the 2006 reauthorization of section 4 and purport to demonstrate that, in its current form, section 4 cannot fulfill the Northwest Austin requirement of “justifi[cation] by current needs.”233 In doing so, the Court appears to set aside any dispute it may have with Congress over constitutional meaning and to turn instead to a question of objective fact. Post and Siegel have shown that in cases involving Section 5 of the Fourteenth Amendment the modern Court has repeatedly applied what they call an “enforcement model” to Section 5 litigation. According to that model, the Court asserts its authority to interpret constitutional meaning by denying Congress the authority to act upon alternative constructions of constitutional rights.234Shelby County, by contrast, does not address the question of what the substantive protections of the Fifteenth Amendment mean but of whether present circumstances can justify the conclusion that those protections are in need of strict regulatory enforcement. There is of course an underlying normative question—whether Section 2 of the Fifteenth Amendment ought to require Congress to justify the “current burdens” imposed by the VRA on the basis of “current needs”—but the Court treated that question as settled by Northwest Austin, leaving only the matter of how to apply the latter’s test in Shelby County.

The Court’s empirical claims also interrupt the ordinary force of stare decisis in two respects. The first concerns the Court’s basis for upholding the constitutionality of the VRA in South Carolina v. Katzenbach.235 There, the Court stated—in overarching terms—the VRA’s constitutionality “must be judged with reference to the historical experience which it reflects.”236 That experience included “the blight of racial discrimination in voting, which ha[d] infected the electoral process in parts of our country for nearly a century”237 and appeared in 1965 as “an insidious and pervasive evil which had been perpetuated . . . through unremitting and ingenious defiance of the Constitution,” justifying “sterner and more elaborate measures.”238 The Court also noted that, in passing more extraordinary and comprehensive measures to combat voting discrimination, Congress was acting against the trend of public opinion in the cause of civil rights.

As the dissent pointed out in Shelby County, with the Court’s suspension of preclearance, “history repeats itself” and the Act will no longer be able to “prevent backsliding” by precluding the proliferation of “more subtle second-generation barriers” to voting equality.239 In short, the very backsliding that the Court authorized Congress to attempt to prevent in Katzenbach was held in Shelby County to be an insufficient basis to support reauthorization of section 4.

The second disruption of stare decisis concerns the proper standard of review and the doctrine of equal sovereignty. Katzenbach held that Congress was empowered under the Fifteenth Amendment to “use any rational means to effectuate the constitutional prohibition of racial discrimination in voting.”240 The Shelby County Court purported to follow Katzenbach in requiring “that ‘the coverage formula [be] rational in both practice and theory.’”241 In Katzenbach, however, the Court concluded that the principle of “equal sovereignty” effected no “bar on differential treatment outside [the] context [of the admission of new States].’”242 By contrast, the Shelby County Court found in Northwest Austin support for the conclusion that the principle of equal sovereignty “remains highly pertinent in assessing subsequent disparate treatment of States.”243 The Court’s concern for equal sovereignty, in effect, permitted it to deploy the rational basis test as if it were required to subject the coverage formula to a higher standard than rationality. Justice Ginsburg recognized this argument as a silent rejection of stare decisis protection for the holding of Katzenbach.244 Interestingly, the Court raises no “proportionality and congruence” argument based on City of Boerne in order to resolve the question whether section 4 appropriately invokes Congress’s Fifteenth Amendment enforcement power. Were it to address head-on the constitutionality of the VRA’s section 5, it may be required to do so.245 In Shelby County, however, the Court’s assessment of “current needs” through empirical reasoning about the staleness of the coverage formula both supports the claim of disparate treatment and seems to take the place of any City of Boerne analysis.

The Court also objected to, but did not hold unconstitutional, Congress’s amendments to section 5, which extend its prohibition to redistricting plans that might have favored minority groups but failed to do so due to a discriminatory purpose and to voting measures “that ha[ve] the purpose of or will have the effect of diminishing the ability of any citizens . . . to elect their preferred candidates of choice” because of race.246 These amendments further distinguish section 5 from the Fifteenth Amendment’s standard, and the Court concluded that they raise the bar for covered jurisdictions “even as the conditions justifying [preclearance] have dramatically improved.”247 The Court left to Congress the task of “draft[ing] another formula based on current conditions.”248 Doing so, however, would not necessarily address the Court’s obvious concern that the very mechanism of preclearance, particularly as amended in 2006, may not be able to be justified as an enforcement of constitutional norms. Like Justice Scalia’s concurrence in Ricci, Justice Kennedy’s dictum regarding the amendments to section 5 appear to signal a looming battle over further convergence.

In addition, Shelby County’s empirical reasoning, that the effectiveness of the Act is itself an evolving social fact that supported striking down the coverage formula,249 echoes Justice O’Connor’s controversial dictum in Grutter. Writing on behalf of the Grutter Court, Justice O’Conner opined “that 25 years from now, the use of racial preferences will no longer be necessary to further the interest” of educational diversity.250 Justice O’Connor observed that, in the prior 25 years since Justice Powell’s opinion in Bakke, “the number of minority applicants with high grades and test scores ha[d] indeed increased.”251 She then hypothesized that changing societal conditions in which university admission would be conducted in the not-so-distant future would ensure that affirmative action could no longer be justified. Both decisions view race discrimination as a time-bounded phenomenon, already disappearing from social and political life, with the consequence that the government’s affirmative efforts to promote race equality must also be time-bounded. Shelby County not only mobilizes an argument parallel to Justice O’Connor’s but pushes the argument further, holding that the VRA itself had already created conditions that caused section 4’s constitutional clock to run out.

Like Ricci and Bartlett, Shelby County represents an especially aggressive form of constitutional and statutory convergence, rewriting the substance of statutory law (or requiring that Congress do so) based on empirical arguments about the nature of the regulatory object and jurisprudential arguments about limitations on either judicial or congressional power. The decision also demonstrates a certain continuity in the recent Court’s empirical assumptions regarding the continued need for strong antidiscrimination measures to address race discrimination. Although the Court relies on Congress’s own statistical data to support its factual assertions in Shelby County, Justice O’Connor’s dictum in Grutter shows that similar conclusions can be, and sometimes are, derived from little more than the Justices’ own ruminations about the evolution of social facts. As the Court continues to follow Shelby County’s logic that the time for robust civil rights enforcement has ended, we might expect other antidiscrimination statutes to be reinterpreted through the prism of what the Court considers to be the current, dwindling nature of race discrimination.

IV.     Resisting Convergence in Race Equality Law

Convergence is appealing. Whether because of the “gravitational pull” of constitutional values or the judicial economy of maintaining a single set of legal standards that organize constitutional and statutory law, convergence seems a conceptually and administratively advantageous choice. Add assumptions about the nature of race and race discrimination that presume that the sources of racial inequality are fundamentally consistent, whether encountered in one social context or another, and convergence may appear to be the only reasonable option. To rationalize and to promote the effectiveness of race equality law is, it seems, to pursue convergence. One could hardly fault the Court then for pursuing convergence in Ricci, Bartlett, and Shelby County.

This view, however, is mistaken. When the Supreme Court imposes convergence on the basis of its own empirical assumptions regarding the nature of race discrimination, it undermines the rule of law and sounds the alarm of counter-majoritarianism. When the Court imposes convergence based on the limitations of its own institutional capacity, it restricts the power of political institutions by denying its own ability to complete the task delivered by statute. If convergence is a potential remedy to inconsistency and arbitrariness in the application of race equality law, it is one that should be applied sparingly because the side effects are likely to be worse than the cure. Revision and reconsideration of policy objectives and of the effectiveness of the means chosen to achieve them are fundamental features of democratic order, and the task of revisiting these concerns is committed to political institutions. Judicial review is a tool for determining whether those institutions have exceeded their constitutional authority, not for aligning the substance of legislation to repeat constitutional choices.

A.     The Appeal of Convergence

Convergence strategies, in general, have a special normative appeal in race equality law because of the moral opprobrium directed at race discrimination. Convergence offers the promise of a unifying vision of racial equality, one that enhances the administrability, predictability, and consistency of race equality law. It offers the judiciary the opportunity to ground this vision within an interpretation of constitutional authority, such as by recognizing structural and enforcement limitations that restrict the scope of congressional lawmaking. It also offers the judiciary the opportunity to project its understanding of what constitutes race discrimination as an observable social phenomenon onto the law and to adapt the law to changing social circumstances, including popular views about what constitutes discrimination or what forms of discrimination require the law’s attention. To withhold from the judiciary the power to effect convergence through its awareness of social facts is to limit the judiciary’s participation in defining and addressing the moral wrongs of racism and race discrimination. This would deny the judiciary a significant opportunity to make a democratic contribution to regulatory order, and it would commit such matters strictly to politics, where bias against particular groups or interests may distort legislative outcomes. It would contradict what Strauss has called in the constitutional context the judiciary’s capacity for modernization, and it would deny courts the opportunity to echo or to anticipate popular will.252

The most significant normative objection to divergence may be that like cases are not treated alike. If two bodies of law are meant to regulate the same social behavior, why should they not define the regulated behavior in the same way and set out to determine its lawfulness in the same way? This is the Davis problem. To that end, the most significant practical problem with relying on jurisprudential rules to support the divergence of constitutional and statutory standards would appear to be inefficient. In addition, the failure to conform the law to an empirically accurate account of the social phenomenon of race discrimination may undermine regulatory objectives. By contrast, if convergence is a consequence of jurisprudential rules, such as structural or enforcement limitations on legislation, then consistency and administrability may be consequences of convergence—but not accuracy or efficiency.

As discussed above, when the Supreme Court upholds convergence, it typically does not do so by relying openly on a discussion of the superiority of particular equality values. The Court seems to appreciate that to declare a particular set of values either normatively superior or the victor in a public contest is to assume the position of a political actor and potentially to exceed the judicial role.253 The Court may seek to avoid this problem by relying on assertions of social fact or invoking constitutional constraints, thereby reinforcing a particular set of equality values while denying that it is participating in a contest of values at all. Empirical considerations are especially attractive in this light because they appear to externalize the reason for convergence, taking it out of the political, and even the judicial, realm. Having now identified these justifications for convergence we can begin to reevaluate their appeal and to make the case against convergence, except where authorized as an expression of political will.

B.     The Pitfalls of Convergence

Convergence should occur when constitutional and statutory law demonstrate that they emanated from a common popular will, as evidenced by the substance of their provisions or the purposes that inspired their proposal or enactment. To an extent, this approach describes the Court’s decision in General Building Contractors.254

There, the Court engaged in fact-based reasoning, consulting the historical circumstances in which § 1981 was enacted, in order to ascertain Congress’s intent, rather than imposing on the statute a static definition of race discrimination that merely reflects the Court’s present understanding. The deficiency of General Building Contractors, however, lies in its failure to articulate a governing value, or set of values, that could explain why evidence that Congress sought to respond to the Black Codes should be construed in one way and not another.255 Whether one agrees with the Court’s interpretation or not, General Building Contractors represents an effort by the Court to recognize the empirical understanding of discrimination that framed legislators’ intentions. In this way, the case strikes the type of balance recommended by this Article. To the fullest extent possible, convergence should not be a judicial choice; it should be compelled by the Court’s fidelity to the will of political actors, unless those political actors operated beyond the bounds of their legal authority. Without such constraints, convergence comes with many costs.

First, when predicated on static assumptions of social fact, convergence undermines rule of law values and invites the counter-majoritarian difficulty. Values associated with convergence, such as efficiency and administrability, are orthogonal to respect for the letter and purpose of the law. If a statute’s provisions differ in meaningful ways, or were intended to serve objectives distinguishable from a constitutional guarantee, to interpret them to have congruent meaning is to steal from legislators the authority to respond to changes in popular will regarding how to define race discrimination or to address the problem of racial inequality. This version of the counter-majoritarian difficulty is particular to statutory interpretation because it concerns the Court’s use of fact-based assumptions to overshadow legislative objectives.

Though the dynamics of counter-majoritarianism differ in statutory interpretation from the more familiar constitutional context, they are nevertheless real. Scholars have warned generally against overestimating the corrective effect of legislative overrides because, in practice, overrides are rare, difficult to obtain, and have other negative consequences such as disrupting current legislative agendas.256 Legislators face additional challenges when they seek to respond to the Court’s limiting construction of a statute due to the Court’s factual assumptions about the statute’s regulatory object. Here, in race equality law, the Court has not simply invited them to provide a clear statement of their objectives or to rectify textual ambiguities. Instead, the Court has effectively indicated to legislators what assumptions about the nature of race discrimination, including its prevalence and its current manifestations, the Court will use to interpret future statutes.

For example, in Ricci, mere consideration of the racial impact of an employment practice is discrimination because the employer’s consideration of race frustrated the expectation of white workers that promotions would be assigned according to a predetermined merit system. This outcome cannot be consistent with legislative intent or with the understanding of the nature of race discrimination that motivated Title VII’s enactment or its subsequent amendment in 1991 to codify the Griggs disparate-impact test. After all, under that test, the employer’s consideration of racial impacts was necessary in order for it to assess its own compliance with the statute. Similarly, in Bartlett, racial minorities were held not to be entitled to crossover districts in which their political interests could be satisfied with the aid of nonminority voters because, in the Court’s view, a statistical minority’s electoral losses are not evidence of discrimination. In Shelby County, the Court charged Congress to return a coverage formula appropriate for the present era in which it views voting discrimination to be largely a thing of the past and substantially rectified by the VRA. Each of these rationales provides a basis for restrictive interpretations of future statutes, even if Congress were to make explicit its intention to achieve broad antidiscrimination coverage.

The same types of assumptions may also be used to withhold stare decisis protection from prior statutory interpretations. As discussed in Shelby County, assertions of social fact may be used to justify the Court’s rejection of prior cases’ rationales.257 Stare decisis does permit the Court to retire precedents that have been “tested by experience” yet failed to demonstrate their continued consistency “with the sense of justice or with the social welfare.”258 But in Shelby County the Court spared itself the unenviable task of arguing that Katzenbach had been proved unjust259 by instead arguing that the VRA’s coverage formula could no longer be justified on the same empirical basis on which Katzenbach had relied. Once again, by arguing that the coverage formula must be justified in relation to “current conditions,” the Court articulated a basis to constrain future legislation and not just to overrule the decades-old formula that it faced in that case. By design, the “current conditions” test licenses the Court to recast its own fact-based assumptions as constraints on legislative authority. Even when, as in Ricci, the doctrine of stare decisis is not directly threatened, convergence may still lead to the disruption of statutory coherence as the Court ignores statutory precedents and engages its constitutional decisions as sources of statutory meaning.260 Finally, as the looming threat to disparate impact sounded in Ricci and the decision’s potential implications for statutory affirmative action doctrine make clear, once the Court goes down the path of aligning constitutional and statutory law, it is not clear where that path will end.

Second, convergence restricts the government’s options when formulating regulatory approaches to the problem of race discrimination. Indeed, to represent race discrimination as a single “problem” is already to fall into the trap of commonplace, but unexamined, empirical assumptions supporting convergence. Race discrimination manifests itself in multiple areas of public and private life, and in many different ways. It depends on factors such as the social context in which it occurs (e.g., employment, housing, voting, etc.), whether the discrimination occurs on an individual or class-wide basis, whether it is the result of animus, unconscious bias, or formally race neutral structural factors that perpetuate historical patterns of racial subordination.261 For some groups, race discrimination has been described to operate as a system in which persistent patterns of inequality in some areas contribute to and reinforce discrimination in others.262 For other groups, instances of discrimination may seem discrete or disconnected and what the law defines as discrimination may be limited to particular manifestations or contexts, such as reverse discrimination against whites through the use of racial preferences in education and employment. In any event, what discrimination is and how the law defines it are not easily separated, and neither the courts nor lawmakers can free themselves from making normative choices by explaining their decisions on the basis of social facts.

Consistency and administrability may be enhanced by convergence, but they should not be confused with coordination. To say that race discrimination spans across areas of social life that we often consider discrete does not mean that it manifests itself in just the same way regardless where we find it—that each manifestation is, in effect, the mirror image of another. A coordinated effort to combat discrimination across social contexts may well choose to apply different standards in different contexts, provided that the coordinating system of standards disrupts historical patterns of discrimination. Consistency and administrability also come at the price of flexibility and adaptability. Congress provided for effects-based liability in race discrimination employment, voting and, arguably, housing. Congress has twice reaffirmed its commitment to provide such claims and to ensure their integrity under both Title VII and the VRA.263 In each instance, Congress judged protection against disparate impact caused by facially neutral employment practices to be a fundamental feature of the law’s normative commitment to equal opportunity. The Supreme Court recently demonstrated its willingness to consider this issue yet again when it granted certiorari on the question whether the Fair Housing Act includes claims for disparate impact.264 It has also voiced concerns about the constitutionality of the express effects-based tests of Title VII and section 5 of the VRA.265 These recent developments beg the question, once the Court starts down the path of convergence: how far should it go? Indeed, it is difficult to go down this path at all without crossing the line between interpreting and revising the law.

Third, one may easily overestimate the clarity, permanence, and infallibility of empirical support for convergence. Even if the Court were rigorously consulting social science to support its assumptions, measurements are often inexact or context-specific, and social facts are subject to change. Scientists, philosophers, and legal scholars alike have written about the pitfalls of defining race from a biological perspective, and yet the biological definitions still have their proponents.266 Some have cautioned that the dimensions of race that are relevant to race equality law exceed biological and cultural understandings.267 Even within a single discipline, such as social psychology, the definition of discrimination and descriptions of its social dynamics may change dramatically within the span of a few decades, a relatively short period of time so far as the law is concerned.268 What race is, and what constitutes race discrimination, are questions the answers to which are always in a state of reexamination and revision. More critically, the Court’s assertions about the nature of race discrimination are typically made without reference to any scientifically vetted data, as if it were merely restating common sense; and its opinions on the subject are themselves subject to change. Indeed, as discussed above, we should expect greater reliance on empirical reasoning to further undermine stare decisis, as prior interpretations give way to developments in areas of research that formerly supported their governing assumptions.

In the end, no matter how reliable the source or methodology, empirical reasoning cannot tell us what questions about a regulatory object are worth asking or what features of the object are important to the law. The task of determining the legal salience of social facts is assigned to legal norms. As in Shelby County, where the Court relied upon congressional data to criticize the justification for reauthorization of the VRA, the Court may select the data that serve its normative agenda or, to give it a more diplomatic reading, the Court may neglect to explain why it has privileged one type of data in place of another.269

Fourth, and finally, sometimes convergence is not the result of fact-based assumptions at all. Instead, it may arise through reliance on jurisprudential rules that appear to compel convergence regardless whether there is a reason to believe that more than one account of race discrimination is objectively rational or that the accuracy of an account depends on context. This form of convergence may itself promote judicial economy and administrability of the law, and it has the virtue of being subject to some democratic control. At times, however, it will be no less volatile than constructions of social fact. Like the Court’s views about the nature of race discrimination, its views regarding the judiciary’s institutional capacity to enforce antidiscrimination norms can change within a very short period of time.270

The structural and enforcement limitations that the Constitution imposes on certain antidiscrimination statutes represent, in a sense, a contest of popular wills. The Court’s interpretations of the substance of the Fourteenth and Fifteenth Amendments themselves reflect expressions of popular will. The fact that these interpretations are then used to restrict the substance of statutes enacted pursuant to the amendments’ enforcement powers, or to limit the remedies that may be available when enforcing particular statutes, does not therefore represent a rejection of popular will, but merely a respect for constitutional superiority. The problem, as Post and Siegel have described it, is that the Court must answer why the Constitution requires that civil rights enforcement legislation receive “the kind of stringent judicial supervision” that City of Boerne v. Flores and its progeny require.271 The cases discussed in Part III show that the question deserves renewed attention, for it now must be posed of the Court’s Fifteenth Amendment doctrine under Northwest Austin and Shelby County.

Enforcement limitations, such as are found in Shaw and threatened by Justice Scalia’s concurrence in Ricci, may be just as effective at reducing congressional authority to remedy discrimination as are the structural limitations directly interpreting Congress’s grant of legislative authority. Indeed, they apply more broadly with far reaching implications for public policy. Post and Siegel describe Congress a “vital resource for the Court to consider” when interpreting constitutional equality guarantees because Congress’s status “as a popular legislative body” provides it with a superior vantage point from which to discern “evolving cultural norms.”272 In this sense, the convergence that is imposed by Bartlett and Shelby County, and threatened by the Court in Ricci, strips Congress of the ability to adapt statutory law to anticipate evolving forms of discrimination and forecloses what might otherwise be an important conversation between Congress and the Court concerning the nature of race discrimination. If the Court is truly interested in adapting race equality law to respond either to evolving social norms or to current empirical understandings of the nature of race discrimination, then it ought to seek out this conversation and acknowledge appropriate areas of judicial deference.

C.     Divergence Is Only What We Make of It

We end where we began, but with a challenge for the future. Davis demonstrated nicely the virtues of divergence: It observes legislative supremacy, respects the authority of political institutions to exceed constitutional equality guarantees, and preserves possibilities for experimentation and adaptation to evolving social problems. When the Court interprets constitutional and statutory law to express a single legal approach to race discrimination, even though the text or the purpose of particular laws may indicate otherwise, the Court nullifies political choices and typically does so without assessing the values represented by those choices. Divergence enhances democratic responsiveness, permitting political institutions to tailor laws to meet the challenges of particular eras and to pursue the objectives of an ever-changing electorate. However, the opportunity to adapt and to experiment with new approaches to racial inequality means nothing if it is not taken. Divergence is only what we make of it.

Race equality law at present divides principally into two types of claims: intentional and effects-based claims. Rather than asking why we have accepted this limitation, instead we seem to be asking which theory paints the truer picture of discrimination. A “war” on the horizon has been prophesied between equal protection’s discriminatory purpose doctrine and statutory disparate-impact tests, and skirmishes have continued over whether, and under what conditions, the law ought to permit race-based affirmative action as a permissible form of disparate treatment. What lawmakers choose to identify as race discrimination may be a function of their judgment regarding how discrimination presents itself at a particular moment in history—or, as both a normative and practical matter, how specific prohibitions against discrimination will promote racial equality. Even when fact-based assumptions about race discrimination remain constant, political institutions may choose to pursue different regulatory approaches in order to promote different agendas.

For example, theories of disparate impact and purposeful discrimination may differ in terms of how they define discrimination. By the standard account, disparate impact defines discrimination in terms of racially disproportionate impact that the employer fails to justify, and disparate treatment, like the constitutional purpose test, requires an illicit motive.273 Disparate impact seems to assume that discrimination is structural and systemic, resulting from institutional arrangements that transcend individual choices, and disparate treatment seems to assume that discrimination results from bias at the level of individual decision-making. However, as Primus has observed, disparate-impact theory may be justified as an evidentiary tool to “ferret[] out” disparate treatment.274 The two, therefore, may coincide—and yet, even if they do, the more important difference between the two theories may be how they imagine the future. Disparate treatment imagines a future that is much like the past: bad actors are motivated by prejudice to discriminate, and the law intervenes to punish those responsible; then the cycle continues. Disparate impact is predicated on the assumption that, by focusing its attention on proper decision-making criteria, an institution can remove “built-in headwinds” that sustained patterns of social subordination.275 Disparate-impact theory is, in other words, designed to render a particular outcome by forcing employers either to eliminate or to justify practices that trend against that outcome. It imagines a future in which formal, well-tailored decision-making methods discourage discrimination and diminish inequality. Whether disparate-impact theory’s assumption is sound is, of course, an empirical question. However, perfecting its understanding of the social dynamics of discrimination and inequality is not the law’s objective. Imagining a future of equal opportunity through more rigorous and accurate procedural formalism and pursuing that future through discriminatory effects tests, Congress has made, and courts are guided by, a normative choice.276

Race equality law is, therefore, not just about the inequality that we see before us, but the futures that we imagine. Divergence between constitutional and statutory law may occur because we see, and separately focus upon, different types of racial injustice, or because we imagine and attempt to pursue alternative possibilities for a future in which discrimination and inequality are less prevalent. A third reason arises when more than one law is directed at achieving a common goal and, to coordinate their implementation in distinct social fields, the two must differ and each be molded to fit the other—a lock and key approach.277 For example, race equality law could take a less tolerant view of racial inequality in education, authorizing effects-based tests and liability based on de facto segregation or granting public universities authorization to use affirmative action to achieve some form of proportional representation, while applying in employment a more deferential view targeting only intentional discrimination. The rationale would be that, because employment is a downstream beneficiary of public education, if schools enforce an absolute policy of equal access, employers could maintain high qualification standards for labor with confidence that all members of the workforce have had equal educational opportunities regardless of race. Constitutional law might also defer to governmental actors by limiting equal protection violations in education to de jure segregation and purposeful discrimination, and then force employers to internalize the costs of unequal educational opportunity by imposing a more exacting combination of intent- and effects-based standards in employment. The latter resembles the law as it is today. If it is a coordinated system, we may choose to sustain divergence to preserve coordination and we may even choose to coordinate these bodies of law differently. The objective here is not to advocate for one policy or another but to highlight the nature of the choices before us.

As discussed in Part II, there is considerable room for variation in terms of how disparate treatment, disparate impact, and affirmative action doctrines may be conceptualized and implemented. Meaningful divergence exists, and it is only because it is meaningful that it stirs such controversy. Nevertheless, the debate about whether and to what extent the Supreme Court should seek convergence across constitutional and statutory race equality law does not turn alone on whether to maintain the status quo. It also concerns the law’s openness to future innovation.

We live in a time when much seems to be changing about how our society views race and discrimination. The election of President Barack Obama has been said to herald an era of “post-racialism.” The latter term is difficult to define. It may express the view that, notwithstanding the persistence of inequality, “racial discrimination is [now] rare and aberrant behavior.”278 It may also refer to changes in the significance of racial affiliation as a matter of personal identity and institutional value—that we now live in a time in which racial identity is “fluid,”279 allowing race to be negotiated and wielded in order to signal group affiliation or, under the model of “racial capitalism,”280 to exploit its value as a commodity. This perspective could also be described as “racial realism,”281 or the recognition that we already exploit race to achieve institutional and individual gains, notwithstanding general mandates against discrimination. The first tends to support convergence around a particular, conservative view of racial equality—one that forbids purposeful discrimination and would likely find race-based affirmative action and race-conscious legislation to lack present-day justification. The second asks us to re-imagine the relationship between law and race if we wish to permit individuals to determine the salience of race for their own sense of self while carefully controlling the extent to which institutions may exploit racial difference for their self-interest.282 These are some of the choices on the horizon. They and others will be moot, however, if Supreme Court equality jurisprudence refuses to recognize legal diversity and instead conforms all choices to fulfill a single law of racial equality.

Not all of the Court’s recent decisions in race equality law have followed the trend toward convergence, and the present moment offers some hope that the Court can be turned away from this trend. The Court’s recent decision in Schuette v. BAMN suggests that the Court continues to recognize the value of preserving for political institutions the authority to pursue a diversity of paths toward racial equality.283 In BAMN, the Court held that Michigan’s constitutional amendment banning public use of race-based affirmative action did not violate equal protection. Much like Davis, the decision may be read in more than one direction. On the one hand, the decision offers a new example of civil rights retrenchment: BAMN rejected the circuit court’s reliance on decades-old precedents in which the Court had previously held that alterations of the political process exhibiting a “racial focus” were constitutionally suspect and demanded strict scrutiny.284 Indeed, conservative members of the Court may have been motivated to uphold the state constitutional amendment because its language well-encapsulated the color-blindness approach that they believe best explains how equal protection itself ought to apply to race-based affirmative action. On the other hand, BAMN is a clear example of a state acting as a laboratory for legal experimentation. By vindicating the people of Michigan’s right to explore an alternative approach to promoting racial equality from what the Court has constructed as its equal protection jurisprudence, the Court has left open the possibility that other states and indeed various political institutions may pursue racial equality by applying a variety of legal standards to combat race discrimination and its effects. We live therefore in the middle of an ever-unfolding story, the end of which should not be presumed; it simply has not yet been written.

V.     Conclusion

Supreme Court doctrine establishes no definite or predictable relationship between constitutional and statutory race equality law. Instead, the Court oscillates between strategies of divergence and convergence, sometimes permitting constitutional and statutory law to differ substantially and other times requiring that they converge around a single rule. Convergence has great intuitive appeal both as a matter of judicial economy and because it reflects a common assumption that the nature of race discrimination is fairly static across social fields and fulfills a moral intuition of great significance in antidiscrimination law that like cases should be treated alike. In recent years, the Court has pursued convergence aggressively. But convergence has its price, in the form of restrictions on innovative, legal adaptation and disruptions of democratic responsiveness. Convergence should not be achieved as a consequence of the Supreme Court’s counter-majoritarian imposition of its own fact-based assumptions about race discrimination, and the Court should take great care when interpreting statutory law to be restrained by a lack of judicial competence or by the Constitution’s structural or enforcement limitations. Instead, the Court should respect substantive differences in existing constitutional and statutory law, and it should seek to preserve opportunities for divergence as a fundamental feature of constitutional order such that political institutions may be encouraged to consider ways in which the law may be adapted to meet the racial challenges of the present and future.




  1. [1]. Washington v. Davis, 426 U.S. 229, 242 (1976).

  2. [2]. See Griggs v. Duke Power Co., 401 U.S. 424, 431 (1971) (interpreting Title VII to prohibit practices that “operate[] to exclude [African Americans]”).

  3. [3]. Davis, 426 U.S. at 247.

  4. [4]. See, e.g., Kenneth L. Karst, Foreword: Equal Citizenship Under the Fourteenth Amendment, 91 Harv. L. Rev. 1, 48–51 (1977) (criticizing Davis on the ground that its discriminatory purpose theory of race discrimination failed to take into account the historical formation of a racial caste system); Michael J. Perry, The Disproportionate Impact Theory of Racial Discrimination, 125 U. Pa. L. Rev. 540, 544 (1977) (criticizing Davis for its failure to adequately explain why disparate-impact theory is not a proper theory of race discrimination under the Constitution); see also Owen M. Fiss, Groups and the Equal Protection Clause, 5 Phil. & Pub. Aff. 107, 157–59 (1976) (arguing that a “differential impact” test should apply to equal protection claims because the Constitution should not tolerate the perpetuation of social underclasses).

  5. [5]. George Rutherglen & Daniel R. Ortiz, Affirmative Action Under the Constitution and Title VII: From Confusion to Convergence, 35 UCLA L. Rev. 467, 469–70 (1988) (arguing that the difference in “coverage” under constitutional and statutory law leading to the application of different legal standards to public and private employment is less troubling than the substantive divergence between those standards when both apply to the same class of public employers).

  6. [6]. Shaw v. Reno, 509 U.S. 630, 643 (1993) (“Classifications of citizens solely on the basis of race ‘are by their very nature odious to a free people whose institutions are founded upon the doctrine of equality.’” (quoting Hirabayashi v. United States, 320 U.S. 81, 100 (1943)).

  7. [7]. For example, in Parents Involved in Community Schools v. Seattle School District No. 1, writing on behalf of himself and three other Justices, Chief Justice Roberts famously explained his rejection of the school district’s voluntary integration efforts by stating that “[t]he way to stop discrimination on the basis of race is to stop discriminating on the basis of race.” Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 748 (2007) (plurality opinion); see also Gratz v. Bollinger, 539 U.S. 244, 270 (2003) (“[R]acial classifications are simply too pernicious to permit any but the most exact connection between justification and classification....” (quoting Fullilove v. Klutznick, 448 U.S. 448, 537 (1980) (Stevens, J., dissenting)) (internal quotation marks omitted)); Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 230 (1995) (“[A]ny individual suffers an injury when he or she is disadvantaged by the government because of his or her race....”); Reva B. Siegel, Foreword: Equality Divided, 127 Harv. L. Rev. 1, 42 (2013) (“Over time, it has become more common for the Court to explain not only the benefits of strict scrutiny but also the harms of racial classification in universal terms.”).

  8. [8]. See Ricci v. DeStefano, 557 U.S. 557, 582 (2009).

  9. [9]. Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 277 (1986) (plurality opinion) (stating that, in a case involving voluntary affirmative action, “the trial court must make a factual determination that the employer had a strong basis in evidence for its conclusion that remedial action was necessary”).

  10. [10]. Ricci, 557 U.S. at 563.

  11. [11]. Id. at 594 (Scalia, J., concurring).

  12. [12]. Id. at 595 (Scalia, J., concurring) (emphasis added).

  13. [13]. See Bartlett v. Strickland, 556 U.S. 1, 16 (2009).

  14. [14]. See Shelby Cnty. v. Holder, 133 S. Ct. 2612, 2629 (2013).

  15. [15]. Robert C. Post & Reva B. Siegel, Legislative Constitutionalism and Section Five Power: Policentric Interpretation of the Family and Medical Leave Act, 112 Yale L.J. 1943, 1946 (2003) [hereinafter Post & Siegel, Legislative Constitutionalism] (“Condemning Section 5 legislation that might establish Congress as an independent interpreter of the Constitution, the Court has announced that ‘Congress’ power under § 5... extends only to enforcing the provisions of the Fourteenth Amendment,’ and that ‘Congress does not enforce a constitutional right by changing what the right is.’” (alteration in original) (quoting City of Boerne v. Flores, 521 U.S. 507, 519 (1997))); see also Robert C. Post & Reva B. Siegel, Protecting the Constitution from the People: Juricentric Restrictions on Section Five Power, 78 Ind. L.J. 1, 2 (2003) (defining the “juricentric Constitution” as a vision of constitutional interpretation that “allows the Court’s coordinate branches to enforce the Constitution only insofar as they enforce judicial interpretations of constitutional meaning”).

  16. [16]. Post & Siegel, Legislative Constitutionalism, supra note 15, at 1954.

  17. [17]. Id. at 1960. Post and Siegel recommend instead a polycentric approach to constitutional interpretation that would respect the authority of coordinate branches to interpret constitutional meaning in a manner reflective of their unique institutional roles. Id. at 2033–34.

  18. [18]. See infra Part III.B.

  19. [19]. William N. Eskridge, Jr., Public Values in Statutory Interpretation, 137 U. Pa. L. Rev. 1007, 1013 (1989).

  20. [20]. William Eskridge and John Ferejohn have also identified certain statutes as “super-statutes” which reverse the “gravitational pull” hypothesis by “establish[ing] a new normative or institutional framework” for legal policy and popular opinion. Therefore, according to Eskridge and Ferejohn, a super-statute may become “one of the baselines against which other sources of law—sometimes including the Constitution itself—are read.” William N. Eskridge, Jr. & John Ferejohn, Super-Statutes, 50 Duke L.J. 1215, 1216 (2001). Compare Eskridge, supra note 19, at 1013 (accusing Hart and Sacks of having “scrupulously avoided detailed analysis of constitutional values and the gravitational pull these values have on statutory interpretation”), with Eskridge & Ferejohn, supra, at 1237 (identifying Title VII as a super-statute because it “has pervasively affected federal statutes and constitutional law”).

  21. [21]. Eskridge, supra note 19, at 1013.

  22. [22]. Alexander M. Bickel, The Least Dangerous Branch: The Supreme Court at the Bar of Politics 16–21 (1962) (criticizing judicial review as “counter-majoritarian” because it authorizes the judiciary to sit in judgment of the popular will as expressed through representative politics with the consequence that judicial review may undermine the democratic process). In constitutional interpretation, David Strauss has described this phenomenon as “modernization,” an approach by which the Court’s attention to changes in public values may reconcile the exercise of judicial review with democratic order. David A. Strauss, The Modernizing Mission of Judicial Review, 76 U. Chi. L. Rev. 859, 860 (2009) (describing modernization as “an approach that, more or less consciously, looks to the future, not the past; that tries to bring laws up to date, rather than deferring to tradition; and that anticipates and accommodates, rather than limits, developments in popular opinion”). The version of the counter-majoritarian difficulty at issue here is related primarily to statutory interpretation, because it concerns the Court’s pursuit of convergence even when to do so undermines legislative objectives. Although some scholars have argued that the counter-majoritarian difficulty is muted in statutory contexts due to the availability of legislative override. See Christopher S. Elmendorf, Refining the Democracy Canon, 95 Cornell L. Rev. 1051, 1065 (2010) (“Though statutory interpretations are, in principle, subject to a majoritarian legislative override, and therefore less vulnerable to the countermajoritarian critique often levied at controversial constitutional rulings, the costs of legislative action sometimes make this possibility of override more theoretical than real.”). Such overrides are rare. Abbe R. Gluck & Lisa Schultz Bressman, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part I, 65 Stan. L. Rev. 901, 912 (2013). And such overrides disrupt current legislative agendas, and may be at special disadvantage when the Court relies on fact-based assumptions that, once established, bind lawmakers or when the Court relies on structural constitutional restraints on legislative authority. See Elmendorf, supra, at 1066.

  23. [23]. Eskridge, supra note 19, at 1016; see also Strauss, supra note 22.

  24. [24]. Eskridge himself has complained that “[t]here is a randomness to the [Supreme] Court’s invocation of public values which is quite troubling and which prevents public values from contributing as much to [the] law’s integrity as they theoretically could.” Eskridge, supra note 19, at 1073.

  25. [25]. Bertrall L. Ross, Against Constitutional Mainstreaming, 78 U. Chi. L. Rev. 1203, 1206 (2011).

  26. [26]. Id. (defining “constitutional mainstreaming” as when “the Court interprets an ambiguous statute in unforeseen contexts to accord with the evolving values that it has emphasized in its decisions interpreting the Constitution but in a manner that conflicts with the values reflected in subsequent legislative enactments”).

  27. [27]. For example, Eskridge asserts that “[t]he [Supreme] Court, or at least some of the Justices, adverted to the constitutional principles in several [Title VII] cases” involving affirmative action “for the proposition that ‘affirmative race-conscious relief may provide an effective means of remedying the effects of past discrimination.’” Eskridge, supra note 19, at 1034 (quoting Local 28, Sheet Metal Workers’ Int’l Ass’n v. EEOC, 478 U.S. 421, 450 n.27 (1986)). The Court’s quoted language is an empirical assertion about the efficacy of affirmative action remedies, not a statement of principle. It may be right to say that certain values, and individual Justices’ commitments to those values, are at play in these decisions. But, in terms of how the Court explains itself—its articulation of the considerations that led to its decision—statements of principle are sometimes, as they are in the quotation, displaced by statements of fact.

  28. [28]. Suzanne B. Goldberg, Constitutional Tipping Points: Civil Rights, Social Change, and Fact-Based Adjudication, 106 Colum. L. Rev. 1955, 1958–59 (2006) (remarking at the Court’s “practice of obfuscating” its role in constitutional norm selection when required by constitutional challenge to “intervene in conflicts regarding popular views of social groups and, at times, [to] ‘tip’ from one understanding of a social group and its constitutional claims to another”).

  29. [29]. Throughout the Article, I use the term “empirical” interchangeably with “fact-based,” and I mean it in the broadest sense—that is, as “originating in or based on observation or experience.” The Merriam-Webster Dictionary 409 (11th ed. 2004). This Article raises no question and offers no argument regarding what methods of observation, if any, are sufficiently reliable to justify the influence of their conclusions on the crafting of legal rules. Challenging the role of empirical assumptions in the shaping of legal rules must be distinguished from challenging empirical methods, even if such a challenge implicitly rejects the view that adherence to proper methods in and of itself establishes the authoritative nature of empirical evidence in legal discourse. In addition, it can be difficult to disentangle the extent to which the Court’s assumptions are conceptual rather than empirical, and certainly fact-based claims are sometimes wielded in order to serve a conceptual function, such as to resolve ambiguity in the meaning of a legal term. However, the empirical nature of these assertions is revealed by their attempt to articulate some truth—not a moral truth, but an observable truth about our social world.

  30. [30]. See Ian F. Haney López, The Social Construction of Race: Some Observations on Illusion, Fabrication, and Choice, 29 Harv. C.R.-C.L. L. Rev. 1, 4–5 (1994).

  31. [31]. Stephen M. Rich, Against Prejudice, 80 Geo. Wash. L. Rev. 1, 6 (2011) (“Antidiscrimination law requires us to make normative choices regarding what sorts of conduct we ought to hold unlawful, what sorts we may excuse, and why.”).

  32. [32]. See infra notes 39–46 and accompanying text; see also Ariela J. Gross, What Blood Won’t Tell: A History of Race on Trial in America 306 (2008) (“Race did not just happen, and [the] law was not imposed on us. We made race, through legal institutions, and we continue to make it every day. ...”).

  33. [33]. See Post & Siegel, Legislative Constitutionalism, supra note 15, at 1946–49 (arguing in favor of a “policentric model” to constitutional interpretation that would not permit the Court to restrict congressional lawmaking on the ground of divergence between Congress’ and the Court’s interpretations of the constitutional right that Congress sought to enforce); see also infra Part III.B. (arguing against the Court’s aggressive interpretation of structural constitutional limitations on congressional authority in its recent voting rights cases).

  34. [34]. See, e.g., Siegel, supra note 7, at 51–58 (considering implications of the Supreme Court’s statutory ruling in Ricci v. DeStefano for future applications of discriminatory purpose doctrine in equal protection jurisprudence); see also Reva B. Siegel, Discrimination in the Eyes of the Law: How “Color Blindness” Discourse Disrupts and Rationalizes Social Stratification, 88 Calif. L. Rev. 77, 107–14 (2000) (describing how anticlassification and antisubordination principles shape antidiscrimination discourse).

  35. [35]. See Deborah Jones Merritt, The Guarantee Clause and State Autonomy: Federalism for a Third Century, 88 Colum. L. Rev. 1, 8–9 (1988). See generally David L. Shapiro, Federalism: A Dialogue (1995).

  36. [36]. See, e.g., Scott Titshaw, The Reactionary Road to Free Love: How DOMA, State Marriage Amendments, and Social Conservatives Undermine Traditional Marriage, 115 W. Va. L. Rev. 205, 295 n.459 (2012) (“Most American state constitutions have equality provisions overlapping, and often exceeding, the protections provided by the federal Equal Protection Clause.” (citing Robert F. Williams, Equality Guarantees in State Constitutional Law, 63 Tex. L. Rev. 1195, 1222 (1985))).

  37. [37]. See, e.g., Jensen ex rel. Jensen v. Cunningham, 250 P.3d 465, 478 (Utah 2011) (upholding more generous interpretation of constitutional rights under state law, because “we owe federal law no more deference in that regard than we do sister state interpretation of identical state language” (internal quotation marks omitted)). Indeed, some scholars have warned that “uncritical ‘reception’ of federal equal protection doctrine... has drained the state equality provisions of much of their vitality,” suggesting that some state courts too may be under the influence of empirical and jurisprudential assumptions that drive the interpretation of state law toward convergence with federal constitutional doctrine. Williams, supra note 36, at 1222; see also Robert F. Williams, State Courts Adopting Federal Constitutional Doctrine: Case-by-Case Adoptionism or Prospective Lockstepping?, 46 Wm. & Mary L. Rev. 1499, 1504–06 (2005) (articulating criticism of state courts that unreflectively adopt Supreme Court interpretations of federal constitutional law when deciding matters of state law).

  38. [38]. To that end, some have criticized the Supreme Court for interpreting structural limitations on congressional lawmaking too narrowly to permit Congress to respond to inequality and to reflect evolving equality norms. See, e.g., Robert C. Post & Reva B. Siegel, Equal Protection by Law: Federal Antidiscrimination Legislation After Morrison and Kimel, 110 Yale L.J. 441, 442–43 (2000).

  39. [39]. Audrey Smedley, “Race” and the Construction of Human Identity, 100 Am. Anthropologist 690, 696 (1998) (arguing that “the assumption at the heart of ‘race’” is a “powerful social lie... that a presumed biological essence is the basis of one’s true identity”). Geneticists and social constructionists have now criticized the biological concept of race to the point where “[t]he rejection of race in science is now almost complete.” Haney López, supra note 30, at 16; see also Matt Cartmill, The Status of the Race Concept in Physical Anthropology, 100 Am. Anthropologist 651, 651 (1998) (“If races are defined as geographically delimited conspecific populations characterized by distinctive regional phenotypes, then human races do not exist now and have not exited for centuries.”). But see Neven Sesardic, Race: A Social Destruction of a Biological Concept, 25 Biology & Phil. 143, 160 (2010) (“[T]ypical attempts to disconnect the concept of race from genetics have too quickly and too uncritically been accepted by many ‘race critics.’”). See generally Neven Sesardic, Confusions About Race: A New Installment, 44 Stud. Hist. & Phil. Biological & Biomedical Sci. 287 (2013) (responding to criticisms of his 2010 article).

  40. [40]. Social constructionists embrace a definition of race as an ongoing social process, influenced by macro-level structural forces and micro-level individual behaviors. Haney López, supra note 30, at 7 (“Race is neither an essence nor an illusion, but rather an ongoing, contradictory, self-reinforcing process subject to the macro forces of social and political struggle and the micro effects of daily decisions.”). See generally Michael Omi & Howard Winant, Racial Formation in the United States: From the 1960s to the 1990s, at 4 (2d ed. 1994) (proposing to set forth a racial theory framework through which to analyze racial politics). Theories of race as social construction depict race as a process of social formation of which the law is an integral part.

  41. [41]. Some have argued that race is a matter of personal performance and, as such, is integral to a person’s sense of self and social identity. See, e.g., Devon W. Carbado & Mitu Gulati, Acting White? Rethinking Race in Post-Racial America 23–24 (2013); Camille Gear Rich, Affirmative Action in the Era of Elective Race: Racial Commodification and the Promise of New Functionalism, 102 Geo. L.J. 179, 189–90 (2013); Camille Gear Rich, Performing Racial and Ethnic Identity: Discrimination by Proxy and the Future of Title VII, 79 N.Y.U. L. Rev. 1134, 1145–46 (2004).

  42. [42]. See, e.g., Anthony Appiah, The Uncompleted Argument: Du Bois and the Illusion of Race, 12 Critical Inquiry 21 (1985) (evaluating the truth about race through the life and writings of W.E.B. DuBois).

  43. [43]. See Julian S. Huxley & A. C. Haddon, We Europeans: A Survey of “Racial” Problems 112 (1936) (“A true ‘race’... is thus a hypothetical group inferred to have existed in the past.”).

  44. [44]. See generally Rachel F. Moran, The Elusive Nature of Race Discrimination, 55 Stan. L. Rev. 2365 (2003) (reviewing Ian Ayres, Pervasive Prejudice? Unconventional Evidence of Race and Gender Discrimination (2001); Crossroads, Directions, and a New Critical Race Theory (Francisco Valdes et al., eds., 2002)); Barbara F. Reskin, The Proximate Causes of Employment Discrimination, 29 Contemp. Soc. 319, 319, 326 (2000) (distinguishing between “original” causes of discrimination, understood to be individual “intrapsychic processes” that bias decisionmaking and “proximate” causes of discrimination, understood most often to be institutional practices that fail to constrain such bias).

  45. [45]. See generally Barbara Reskin, The Race Discrimination System, 38 Ann. Rev. Soc. 17 (2012).

  46. [46]. Compare 42 U.S.C. § 2000e-2(a)(1) (2012) (forbidding “employer[s] to fail or refuse to hire or to discharge... or otherwise to discriminate against any individual... because of such individual’s race...”), with United Steelworkers v. Weber, 443 U.S. 193, 197 (1979) (holding that the employer’s use of affirmative action to correct a manifest racial imbalance in its workforce was not unlawful discrimination).

  47. [47]. Washington v. Davis, 426 U.S. 229, 233–34 (1976).

  48. [48]. Id. at 234–35 (internal quotation marks omitted). Four times as many blacks failed the test as whites. Id. at 237.

  49. [49]. Id. at 236–37 n.6 (quoting Davis v. Washington, 512 F.2d 956, 958 n.2 (D.C. Cir. 1975)) (internal quotation marks omitted). The Constitution’s equality guarantees are derived from the Fourteenth Amendment (“No state shall... deny to any person within its jurisdiction the equal protection of the laws.”), and Fifth Amendment (“No person shall... be deprived of life, liberty, or property, without due process of law....”). See Davis, 426 U.S. at 239 (recognizing plaintiffs’ equal protection claim under the Fifth Amendment’s Due Process Clause). The operative language on which disparate-impact liability was sustained under Griggs v. Duke Power Co. provides that “[i]t shall be an unlawful employment practice for an employer... to limit, segregate, or classify his employees... in any way which would deprive or tend to deprive any individual of employment opportunities or otherwise adversely affect his status as an employee, because of such individual’s race....” 42 U.S.C. § 2000e-2(a)(2).

  50. [50]. Davis, 426 U.S. at 238–39.

  51. [51]. Griggs v. Duke Power Co., 401 U.S. 424, 431 (1971).

  52. [52]. Id. at 427.

  53. [53]. Id. at 429.

  54. [54]. Id. at 431 (concluding that neither requirement was “shown to bear a demonstrable relationship to successful job performance of the jobs for which it was used” even though the Company claimed it had “instituted [the requirement] on the Company’s judgment that they generally would improve the overall quality of the work force”).

  55. [55]. Davis, 426 U.S. at 236–37.

  56. [56]. Id. at 239.

  57. [57]. Id. at 245–48.

  58. [58]. Griggs, 401 U.S. at 432.

  59. [59]. Id. at 431–32.

  60. [60]. See, e.g., id. at 429–30 (stating that “[t]he objective of Congress... was to... remove barriers that have operated in the past to favor an identifiable group of white employees over other employees”); id. at 431 (“What is required by Congress is the removal of artificial, arbitrary, and unnecessary barriers to employment when the barriers operate invidiously to discriminate on the basis of racial or other impermissible classification.”); id. (“Congress has now required that the posture and condition of the job-seeker be taken into account.”); id. at 432 (“Congress directed the thrust of the Act to the consequences of employment practices, not simply the motivation.”).

  61. [61]. See, e.g., id. at 430 (explaining that the circuit court’s conclusion that whites faired “far better” on the testing requirements than blacks as a “consequence... directly traceable to race” because blacks had “long received inferior education in segregated schools”); id. at 433 (“The facts of this case demonstrate the inadequacy of broad and general testing devices.... History is filled with examples of men and women who rendered highly effective performance without conventional badges of accomplishment....”).

  62. [62]. Davis, 426 U.S. at 245 (“As an initial matter, we have difficulty understanding how a law establishing a racially neutral qualification for employment is nevertheless racially discriminatory... simply because a greater proportion of Negroes fail to qualify than members of any other racial or ethnic groups.”); id. at 246 (“Respondents, as Negroes, could not more successfully claim that the test denied them equal protection than could white applicants who also failed.... That other Negroes also failed to score well would, alone, not demonstrate that respondents individually were being denied equal protection of the laws by the application of an otherwise valid qualifying test....”).

  63. [63]. Id. at 246−47 (“Under Title VII, Congress provided that when hiring and promotion practices disqualifying substantially disproportionate numbers of blacks are challenged, discriminatory purpose need not be proved, and that it is an insufficient response to demonstrate some rational basis for the challenged practices.”); see also Connecticut v. Teal, 457 U.S. 440, 448 (1982) (explaining that disparate impact arises out of section 703(a)(2) of the statute, which prohibits practices that adversely affect a plaintiff’s employment status because of his race).

  64. [64]. Davis, 426 U.S. at 248.

  65. [65]. Id. at 247.

  66. [66]. Id. at 248.

  67. [67]. Id. at 240 (citing Keyes v. Sch. Dist. No. 1, 413 U.S. 189, 205 (1973)).

  68. [68]. See, e.g., id. at 239 (citing Alexander v. Louisiana, 405 U.S. 625, 628−29 (1972); Akins v. Texas, 325 U.S. 398, 403−04 (1945)).

  69. [69]. Id. at 240 (citing Wright v. Rockefeller, 376 U.S. 52 (1964)).

  70. [70]. Id. at 240–41 (citing Jefferson v. Hackney, 406 U.S. 535, 548 (1972)).

  71. [71]. McCleskey v. Kemp, 481 U.S. 279, 339 (1987) (Brennan, J., dissenting) (describing the majority’s “fear that [the] recognition of [the petitioner’s] claim would open the door to widespread challenges to all aspects of criminal sentencing” as “a fear of too much justice”).

  72. [72]. See Siegel, supra note 7, at 14−15 (demonstrating that, prior to Davis, the law was unsettled and lower courts seeking guidance had applied the Griggs test to constitutional claims); see also infra notes 79−83 and accompanying text.

  73. [73]. Palmer v. Thompson, 403 U.S. 217, 224 (1971).

  74. [74]. Id.

  75. [75]. Id.; see also id. at 225 (“It is difficult or impossible for any court to determine the ‘sole’ or ‘dominant’ motivation behind the choices of a group of legislators.”).

  76. [76]. Id. (emphasis added).

  77. [77]. Id.

  78. [78]. Id.

  79. [79]. Washington v. Davis, 426 U.S. 229, 243 (1976).

  80. [80]. Palmer, 403 U.S. at 220 (“[T]his is not a case where whites are permitted to use public facilities while blacks are denied access.... Unless, therefore, as petitioners urge, certain past cases require us to hold that closing the pools to all denied equal protection to Negroes, we must agree with the courts below and affirm.” (emphasis added)).

  81. [81]. Wright v. Council of Emporia, 407 U.S. 451, 462 (1972) (upholding invalidation of a school districting decision that had the effect of violating a federal court’s desegregation order); see also Davis, 426 U.S. at 243 (acknowledging Wright’s reliance on Palmer).

  82. [82]. Davis, 426 U.S. at 244−45 & n.12 (citing cases and acknowledging that “[b]oth before and after Palmer... various Courts of Appeals ha[d] held in several contexts, including public employment, that the substantially disproportionate racial impact of a statute or official practice standing alone and without regard to discriminatory purpose, suffices to prove racial discrimination violating the Equal Protection Clause”).

  83. [83]. Id. at 243−44.

  84. [84]. Gen. Bldg. Contractors Ass’n v. Pennsylvania, 458 U.S. 375, 391 (1982).

  85. [85]. 42 U.S.C. § 1981(a) (2012).

  86. [86]. Gen. Bldg. Contractors, 458 U.S. at 378.

  87. [87]. Id. at 380 & n.4.

  88. [88]. Id. at 381.

  89. [89]. Id. (internal quotation marks omitted). In its discussion of § 1981 liability, the district court expressly referenced Griggs. Pennsylvania v. Local Union 542, Int’l Union of Operating Eng’rs, 469 F. Supp. 329, 400 (E.D. Pa. 1978).

  90. [90]. Gen. Bldg. Contractors, 458 U.S. at 382 (internal quotation marks omitted).

  91. [91]. William N. Eskridge, Jr., The New Textualism, 37 UCLA L. Rev. 621, 630 (1990) (describing “imaginative reconstruction” as an interpretive “mode” in which “the Court will trace the evolution of the statute and its debating history” in order “to recreate the general assumptions, goals, and limitations of the enacting Congress”); see also id. at 630 n.34 (citing General Building Contractors as an example of “imaginative reconstruction”).

  92. [92]. Gen. Bldg. Contractors, 458 U.S. at 386 (quoting United States v. Price, 383 U.S. 787, 803 (1966)).

  93. [93]. Id. at 384.

  94. [94]. Id. at 386.

  95. [95]. Id. at 386–87.

  96. [96]. Id. at 388 (citation omitted) (quoting Griggs v. Duke Power Co., 401 U.S. 424, 430 (1971)).

  97. [97]. See id. at 386 n.13 (noting that discussion of the Black Codes “occupied a central place in the congressional debates leading to the enactment of the 1866 Act”); see also id. at 387.

  98. [98]. Saint Francis Coll. v. Al-Khazraji, 481 U.S. 604, 610 n.4 (1987). Justice White’s opinion for the Court may be construed to favor the view “that racial classifications are for the most part sociopolitical, rather than biological, in nature.” Id. (citing a variety of scientific publications); see also Gross, supra note 32, at 304–05.

  99. [99]. See Al-Khazraji, 481 U.S. at 613; see also Shaare Tefila Congregation v. Cobb, 481 U.S. 615, 617–18 (1987) (holding that § 1982 permitted a Jewish plaintiff to bring a claim of race discrimination in connection with the desecration of a synagogue because, following Al-Khazraji, “Jews and Arabs were among the peoples [during the nineteenth century] considered to be distinct races and hence within the protection of the statute”).

  100. [100]. See Al-Khazraji, 481 U.S. at 613 n.5 (listing “prior cases” in which the Court had held that “discrimination by States on the basis of ancestry violates the Equal Protection Clause”).

  101. [101]. Gen. Bldg. Contractors, 458 U.S. at 389.

  102. [102]. Id. at 385 (quoting Hurd v. Hodge, 334 U.S. 24, 32 (1948)) (internal quotation marks omitted).

  103. [103]. Id. at 389. The structural assumption here is that the statute should not be construed to prohibit conduct exceeding the constitutional principles that it was intended to enforce under Section 5 as those principles have been identified through judicial interpretation of the protections and prohibitions of Section 1. See United States v. Morrison, 529 U.S. 598, 619 (2000) (“In fact,... several limitations inherent in § 5’s text and constitutional context have been recognized since the Fourteenth Amendment was adopted.”); Kimel v. Fla. Bd. of Regents, 528 U.S. 62, 81 (2000) (“The ultimate interpretation and determination of the Fourteenth Amendment’s substantive meaning remains the province of the Judicial Branch.”); City of Boerne v. Flores, 521 U.S. 507, 519 (1997) (holding that Congress cannot “decree the substance of the Fourteenth Amendment’s restrictions on the states” because it “has been given the power ‘to enforce,’ not the power to determine what constitutes a constitutional violation”).

  104. [104]. See Runyon v. McCrary, 427 U.S. 160, 170–71 (1976) (stating that § 1981 will be violated if a private actor refuses to contract on equal terms with a black counterparty “solely because he [is] a Negro”); Griffin v. Breckenridge, 403 U.S. 88, 102 (1971) (stating that a violation of §1985(3) requires “some racial, or perhaps otherwise class-based, invidiously discriminatory animus behind the conspirators’ action”); Jones v. Alfred H. Mayer Co., 392 U.S. 409, 426 (1968) (stating that § 1982 “was meant to prohibit all racially motivated deprivations of the rights enumerated in the statute”).

  105. [105]. Gen. Bldg. Contractors, 458 U.S. at 390 (quoting Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 272 (1979)) (internal quotation marks omitted).

  106. [106]. Id. (quoting Hurd, 334 U.S. at 32) (internal quotation marks omitted).

  107. [107]. Indeed, General Building Contractors may be a cunning example of dynamic statutory interpretation in which the Court adopted the ascendant constitutional values of Davis when interpreting the Black Codes. See William N. Eskridge, Jr., Dynamic Statutory Interpretation, 135 U. Pa. L. Rev. 1479, 1479–80 (1987) (arguing that when faced with statutory ambiguity courts should consider contemporary values, which may be expressed by Congress or in the Supreme Court’s constitutional rulings). Yet, to underscore the focus of this Article, the Court explains its decision in terms of legislative intent and not in terms of ascendant public values.

  108. [108]. Similarly, in the absence of Davis, Al-Khazraji might have relied upon the example of the Black Codes to neutralize the evidence of nineteenth century racial theory on which the Court relied and to conclude that what race means under § 1981 is not a function of the legislators’ fidelity to that racial theory but rather a function of what types of practices they intended the law to disestablish. Race discrimination would then have been understood based on a theory of social construction to which a history of social subordination would be highly relevant. See supra note 40.

  109. [109]. Indeed, the text of § 1981, “derived in part from the 1866 Act, has roots in the Thirteenth as well as the Fourteenth Amendment.” Gen. Bldg. Contractors, 458 U.S. at 390 n.17 (emphasis added).

  110. [110]. Section 1981 expressly provides that “[a]ll persons within the jurisdiction of the United States shall have the same right . . . to make and enforce contracts... as is enjoyed by white citizens.” 42 U.S.C. § 1981 (2012) (emphasis added). Whether to interpret the statute to apply to whites represents an important interpretive choice, already foreclosed by the time General Building Contractors was decided. See McDonald v. Sante Fe Trail Transp. Co., 427 U.S. 273, 280 (1976) (holding that § 1981 applies to claims brought by whites); see also Saint Francis Coll. v. Al-Khazraji, 481 U.S. 604, 609–10 (1987) (“Concededly, McDonald... held that white persons could maintain a § 1981 suit....”). But that interpretation did not foreclose a reading of the statute that emphasized its effort to lift non-whites to equal civil status.

  111. [111]. Runyon v. McCrary, 427 U.S. 160, 172 (1976) (holding that § 1981 prohibits private schools from excluding otherwise qualified students on the basis of race); see also Johnson v. Ry. Express Agency, Inc. 421 U.S. 454, 460–61 (1975) (concluding that § 1981 provides a remedy against discrimination by private employers); Jones v. Alfred H. Mayer Co., 392 U.S. 409, 413 (1968) (interpreting § 1982 to apply to private conduct).

  112. [112]. Susan Sturm, Second Generation Employment Discrimination: A Structural Approach, 101 Colum. L. Rev. 458, 468 (2001) (describing “second generation discrimination” as originating in “structural,” “subtle” and “complex” forms of bias in contrast to the overtly segregationist practices that characterized “first generation discrimination”); see also Linda Hamilton Krieger, The Content of Our Categories: A Cognitive Bias Approach to Discrimination and Equal Employment Opportunity, 47 Stan. L. Rev. 1161, 1241–44 (1995) (arguing that the “motivating factor” approach to defining discrimination is critical to addressing discrimination caused by cognitive bias).

  113. [113]. United States v. Lopez, 514 U.S. 549, 559 (1995). As Robert Post and Reva Siegel have observed, divergences between constitutional and statutory antidiscrimination law “were not generally understood by the Court or others as constitutionally problematic [because] they could always be accommodated by the broad authority of the Commerce Clause.” Post & Siegel, supra note 38, at 448–49. This interpretation explains the endorsement of divergence in Davis. Nevertheless, structural limitations on congressional lawmaking may pose very serious challenges for antidiscrimination law. See, e.g., United States v. Morrison, 529 U.S. 598, 617 (2000) (holding that Congress cannot enact a private right of action for gender-motivated violence under the commerce power because such violence does not “substantially affect” interstate commerce).

  114. [114]. City of Boerne v. Flores, 521 U.S. 507, 520 (1997); see also Morrison, 529 U.S. at 626–27 (finding that a remedy directed at individuals for gender bias exceeded Congress’s Section 5 power); Kimel v. Fla. Bd. of Regents, 528 U.S. 62, 82–83 (2000) (“Applying the same ‘congruence and proportionality’ test... , we conclude that the [Age Discrimination in Employment Act of 1967] is not ‘appropriate legislation’ under § 5 of the Fourteenth Amendment.”).

  115. [115]. Ricci v. DeStefano, 557 U.S. 557, 595 (2009) (Scalia, J., concurring) (internal quotation marks omitted); see also Post & Siegel, supra note 38, at 452 (describing this as “a fundamental reworking of an important area of Title VII jurisprudence” and warning that it may cause us to “imagine an incremental judicial reworking of the body of Title VII law so as to bring it into line with the constricted set of standards constitutionally applicable to states”).

  116. [116]. Shaw v. Reno, 509 U.S. 630, 658 (1993).

  117. [117]. Id. at 647; see also Stephen M. Rich, Inferred Classifications, 99 Va. L. Rev. 1525, 1547–58 (2013) (discussing the application of equal protection to the state’s redistricting plan in Shaw as a constraint on race-conscious state action).

  118. [118]. Miller v. Johnson, 515 U.S. 900, 915–17 (1995).

  119. [119]. Ricci v. DeStefano, 557 U.S. 557, 594 (2009) (Scalia, J., concurring) (“[I]f the Federal Government is prohibited from discriminating on the basis of race, then surely it is also prohibited from enacting laws mandating that third parties—e.g., employers, whether private, State, or municipal—discriminate on the basis of race.” (internal citation omitted)).

  120. [120]. For example, the Court might have held that evidence of a racial disparate impact raises a presumption of purposeful discrimination and shifts the burden to the government to justify the challenged action. Cf. Griggs v. Duke Power Co., 401 U.S. 424, 432 (1971) (holding that evidence that a facially neutral employment practice results in a racially disparate impact shifts the burden to the employer to justify the practice on the basis of job-relatedness and business necessity).

  121. [121]. See supra note 12 and accompanying text.

  122. [122]. See Washington v. Davis, 426 U.S. 229, 241 (1976) (disclaiming “that the necessary discriminatory racial purpose must be express or appear on the face of the statute, or that a law’s disproportionate impact is irrelevant in cases involving Constitution-based claims of racial discrimination”); see also Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 266 (1977) (listing as a potential “starting point” for the consideration of evidence of discriminatory purpose “[t]he impact of the official action—whether it ‘bears more heavily on one race than another’” (quoting Davis, 426 U.S. at 242)).

  123. [123]. See, e.g., Griggs, 401 U.S. at 430 (justifying the application of disparate-impact theory to facially neutral employment practices based on the government’s history of providing African Americans “inferior education in segregated schools”).

  124. [124]. See Davis, 426 U.S. at 238–39; Gomillion v. Lightfoot, 364 U.S. 339, 346–47 (1960) (granting relief for a racial discrimination claim under the Fifteenth Amendment).

  125. [125]. See Thornburg v. Gingles, 478 U.S. 30, 60 (1986) (applying an effects test to a racially polarized voting claim under the VRA); Griggs, 401 U.S. at 431.

  126. [126]. See Kevin Stainback & Donald Tomaskovic-Devey, Documenting Desegregation: Racial and Gender Segregation in Private-Sector Employment Since the Civil Rights Act 15, 150 (2012).

  127. [127]. See James A. Gardner, Liberty, Community and the Constitutional Structure of Political Influence: A Reconsideration of the Right to Vote, 145 U. Pa. L. Rev. 893, 940 (1997).

  128. [128]. See Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 272 (1979).

  129. [129]. See supra note 112 and accompanying text.

  130. [130]. Ricci v. DeStefano, 557 U.S. 557, 579 (2009) (finding disparate treatment regardless how “well intentioned or benevolent” the employer’s reasons may have been); UAW v. Johnson Controls, Inc., 499 U.S. 187, 200 (1991) (finding liability regardless of “[t]he beneficence of [the] employer’s purpose”); see also Rich, supra note 31, at 65–69 (discussing the absence of a prejudice or animus requirement under Title VII).

  131. [131]. See Grutter v. Bollinger, 539 U.S. 306, 328 (2003) (upholding the University of Michigan Law School’s affirmative action admissions policy).

  132. [132]. Compare United Steelworkers of Am. v. Weber, 443 U.S. 193, 197 (1979) (holding that an employer may use race in a voluntary affirmative action program in order to correct a “manifest racial imbalance” in the employer’s workforce), with Grutter, 539 U.S. at 330 (permitting the law school to seek a “critical mass” of minority students, “defined by reference to the educational benefits that diversity is designed to produce”).

  133. [133]. Ricci, 557 U.S. at 584.

  134. [134]. Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 277 (1986); accord City of Richmond v. J.A. Croson Co., 488 U.S. 469, 500 (1989).

  135. [135]. Ricci, 557 U.S. at 594 (Scalia, J., concurring).

  136. [136]. Siegel, supra note 7, at 55; see also supra note 119 and accompanying text.

  137. [137]. Siegel, supra note 7, at 56 (“Justice Scalia asserted, without argument, that government can only remedy employers’ intentional discrimination and has no constitutional prerogative to combat the forms of unconscious and structural bias disparate impact is most commonly invoked to correct.”).

  138. [138]. See supra Part II.A.1.

  139. [139]. Ricci, 557 U.S. at 595 (Scalia, J., concurring) (opining that “[i]t might be possible to defend [disparate impact] by framing it as simply an evidentiary tool used to identify genuine, intentional discrimination”).

  140. [140]. Id. at 566 (majority opinion).

  141. [141]. Ricci v. DeStefano, 554 F. Supp. 2d 142, 160 (D. Conn. 2006); see also Hayden v. Cnty. of Nassau, 180 F.3d 42, 51 (2d Cir. 1999) (“[T]he intent to remedy the disparate impact of the prior exams is not equivalent to an intent to discriminate against non-minority applicants.”).

  142. [142]. Ricci, 544 F. Supp. 2d at 162 (denying defendants’ motion for summary judgment).

  143. [143]. Ricci, 557 U.S. at 579.

  144. [144]. Id. at 580.

  145. [145]. Id. at 585; see also id. at 584 (“[W]e adopt the strong-basis-in-evidence standard... to resolve any conflict between the disparate-treatment and disparate-impact provisions of Title VII.”).

  146. [146]. Id. at 582.

  147. [147]. Id.

  148. [148]. Id. at 581.

  149. [149]. Id. at 584 (“We... do not hold that meeting the strong-basis-in-evidence standard would satisfy the Equal Protection Clause in a future case.... [W]e need not decide whether a legitimate fear of disparate impact is ever sufficient to justify discriminatory treatment under the Constitution.”).

  150. [150]. See infra Part III.A.2.

  151. [151]. This is true also in other ways. For example, Ricci does not require the plaintiffs to demonstrate that they suffered an adverse employment action, a requirement of “orthodox Title VII doctrine.” Richard Primus, The Future of Disparate Impact, 108 Mich. L. Rev. 1341, 1356 (2010). In this way, too, Ricci makes the statutory standard look more like the constitutional standard by situating the plaintiffs cognizable harm at the instance at which race is considered and not at the consequences that flow from its consideration.

  152. [152]. Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 279 (1979).

  153. [153]. Ricci, 557 U.S. at 579–80 (rejecting the government’s argument that intent to comply with the statute cannot support a claim of discrimination and explaining that “[w]hatever the City’s ultimate aim—however well intentioned or benevolent it might have seemed—the City made its employment decision because of race”).

  154. [154]. Rich, supra note 31, at 47–48.

  155. [155]. See Siegel, supra note 7, at 56 (describing this view as a “rewriting [of] Feeney” that would “raise[] serious constitutional questions about other race-conscious facially neutral efforts to rectify bias, to increase diversity, or to integrate”); id. at 58 (“It could be that the Court is preparing to eliminate Feeney’s required showing of specific intent to harm....”). See generally Michelle Adams, Is Integration a Discriminatory Purpose?, 96 Iowa L. Rev. 837 (2011).

  156. [156]. See Ricci, 557 U.S. at 594 (Scalia, J., concurring) (arguing that disparate impact’s requirement that employers take into account racial outcomes violates Feeney); see also id. at 595 (Scalia, J., concurring) (arguing that the disparate-impact provision’s “purportedly benign motive... cannot save [it]”).

  157. [157]. Siegel, supra note 7, at 56.

  158. [158]. See supra notes 94–97 and accompanying text. But see Goodman v. Lukens Steel Co., 482 U.S. 656, 668–69 (1981) (finding a basis for liability under Title VII and § 1981 in the absence of racial animus by the union, because the union had “in effect, categorized racial grievances as unworthy of pursuit and... ignored racial discrimination claims on behalf of blacks”).

  159. [159]. See infra Part III.A.2.

  160. [160]. Ricci, 557 U.S. at 594 (Scalia, J., concurring).

  161. [161]. Id. (Scalia, J., concurring).

  162. [162]. Id. at 595 (Scalia, J., concurring). Constitutional scholar Richard Primus has made a similar argument in defense of disparate impact as it might be authorized under Section 5 of the Fourteenth Amendment. See Richard A. Primus, Equal Protection and Disparate Impact: Round Three, 117 Harv. L. Rev. 494, 520–21 (2003). Justice Scalia, however, expressed doubt about this argument, stating that “arguably the disparate-impact provisions sweep too broadly to be fairly characterized in such a fashion.” Ricci, 557 U.S. at 595 (Scalia, J., concurring).

  163. [163]. Id. at 594 (Scalia, J., concurring).

  164. [164]. Id. at 595 (Scalia, J., concurring).

  165. [165]. See id. at 595 (Scalia, J., concurring).

  166. [166]. See id. (Scalia, J., concurring). Justice Scalia’s citation to Feeney suggests that he believes that the disparate-impact provisions may be unconstitutional because they compel employers to act with a discriminatory purpose. His citations to Miller v. Johnson, 515 U.S. 900 (1995) and Adarand v. Pena, 515 U.S. 200 (1995), suggest that he believes the disparate-impact provisions contain a racial classification or that such a classification can be discerned from the form and practical effect of those provisions. See Ricci, 557 U.S. at 595 (Scalia, J., concurring); see also Rich, supra note 117, at 1533–61 (discussing circumstances in which the Court has inferred racial classifications from the form and practical effect of a challenged policy).

  167. [167]. Rutherglen & Ortiz, supra note 5, at 470.

  168. [168]. Id. at 503.

  169. [169]. This is an area in which Rutherglen and Ortiz previously noted divergence and where there continues to be divergence, unlike in the area of court-ordered racial remedies. Id. at 490–503.

  170. [170]. See City of Richmond v. J.A. Croson, Co., 488 U.S. 469, 511 (1989) (holding the City violated the Equal Protection Clause by awarding construction contracts on the basis of race).

  171. [171]. See, e.g., Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411, 2417 (2013) (stating that “[i]t is... irrelevant that a system of racial preferences in admissions may seem benign” because “[a]ny racial classification must meet strict scrutiny....”); Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 227 (1995) (“[W]e hold today that all racial classifications, imposed by whatever federal, state, or local governmental actor, must be analyzed by a reviewing court under strict scrutiny.”); Croson, 488 U.S. at 508 (applying strict scrutiny to the City’s race-based program). Strict scrutiny is an exacting standard, requiring the government to show that the challenged classification is narrowly tailored to serve a compelling government interest.

  172. [172]. Shaw v. Hunt, 517 U.S. 899, 909–10 (1996) (plurality opinion) (“[A]n effort to alleviate the effects of societal discrimination is not a compelling interest.”); Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 276 (1986) (“Societal discrimination, without more, is too amorphous a basis for imposing a racially classified remedy.”).

  173. [173]. Grutter v. Bollinger, 539 U.S. 306, 330 (2003) (describing “outright racial balancing” as “patently unconstitutional”); see also Regents of Univ. of Cal. v. Bakke, 438 U.S. 265, 307 (rejecting as “facially invalid” a purpose “to assure... some specified percentage of a particular group merely because of its race or ethnic origin”).

  174. [174]. See Grutter, 539 U.S at 343.

  175. [175]. Wygant, 476 U.S. at 277 (plurality opinion); accord Croson, 488 U.S. at 500 (affirming the “strong basis in evidence” test as the constitutional rule for remedial measures).

  176. [176]. United Steelworkers of Am. v. Weber, 443 U.S. 193, 208 (1979).

  177. [177]. Justice Blackmun suggested that the employer’s consideration of race could be sustained only by evidence of an “arguable violation” of law. Id. at 211 (Blackmun, J., concurring). Later, Justice O’Connor suggested that the standard should be that “the employer must have had a firm basis for believing that remedial action was required.” Johnson v. Transp. Agency of Santa Clara, 480 U.S. 616, 649 (1987) (O’Connor, J., concurring). In both cases, the Court rejected these arguments.

  178. [178]. See Weber, 443 U.S. at 212 (Blackmun, J., concurring) (“The sources cited [in the majority opinion] suggest that the Court considers a job category to be ‘traditionally segregated’ when there has been a societal history of purposeful exclusion of blacks from the job category, resulting in a persistent [racial] disparity....”).

  179. [179]. See Eskridge, supra note 19, at 1034.

  180. [180]. See id. (“The Court, or at least some of the Justices, adverted to the constitutional principles [established in Bakke] in several of these cases for the proposition that ‘affirmative race-conscious relief may provide an effective means of remedying the effects of past discrimination.’”). Interestingly, Justice Powell turned to McDonald v. Santa Fe Trail Transportation Co., a statutory case in which the Court held that whites could bring race discrimination claims under Title VII and § 1981, to support the conclusion that Congress intended the Fourteenth Amendment to reflect the same substantive understanding that race discrimination is not confined to particular racial subjects. McDonald v. Santa Fe Trail Transp. Co., 427 U.S. 273, 293 (1976).

  181. [181]. Johnson, 480 U.S. at 637–38.

  182. [182]. The Court would ultimately find a similar “plus factor” plan constitutional in Grutter. See Grutter v. Bollinger, 539 U.S. 306, 334 (2003) (upholding a race-based plan in university admissions).

  183. [183]. Johnson, 480 U.S. at 638 (observing the “contrast” between the plan challenged in Weber and the Santa Clara plan, which “resemble[d] the ‘Harvard Plan’ approvingly noted by Justice Powell in [Bakke]” and required all applicants to compete with one another for available positions with consideration given for the sex of the applicant).

  184. [184]. Ricci v. DeStefano, 557 U.S. 557, 584 (2009) (emphasis added).

  185. [185]. See, e.g., id. at 567–68 (describing the costs and inconveniences incurred by the plaintiffs as a consequence of relying on the city’s publicized system for awarding promotions); id. at 585 (“Nor do we question an employer’s affirmative efforts to ensure that all groups have a fair opportunity to apply for promotions and to participate in the process by which promotions will be made. But once that process has been established and employers have made clear their selections criteria, they may not then invalidate the test results....”).

  186. [186]. See id. at 594 (Scalia, J., concurring) (opining that “[s]urely” an employer would be guilty of unlawful discrimination “if he refrained from establishing a racial hiring quota but intentionally designed his hiring practices to achieve the same end”).

  187. [187]. United States v. Brennan, 650 F.3d 65, 97 (2d Cir. 2011).

  188. [188]. Id. at 98 (citing Bushey v. N.Y. State Civil Serv. Comm’n, 733 F.2d 220, 228 (2d Cir. 1984)).

  189. [189]. Id. at 102.

  190. [190]. See, e.g., Cynthia L. Estlund, Putting Grutter to Work: Diversity, Integration, and Affirmative Action in the Workplace, 26 Berkeley J. Emp. & Lab. L. 1, 5 (2005) (arguing that the diversity rationale for affirmative action should be embraced in the employment context because diversity provides a “forward-looking,” non-remedial basis for affirmative action that would extend the latter’s reach beyond the circumstances permitted currently by Title VII).

  191. [191]. In Grutter, diversity is uniquely constitutionally salient in higher education because it helps “to break down racial stereotypes” and to encourage “livelier, more spirited” class discussion. Grutter v. Bollinger, 539 U.S. 306, 330 (2003). Writing as amicus curiae, certain major American corporations argued before the Court “that the skills needed in today’s increasingly global marketplace can only be developed through exposure to widely diverse people, cultures, ideas, and viewpoints.” Id. at 330. Perhaps this expression of corporate self-interest is also a rationale for workplace diversity, but it seems to run afoul of Title VII’s prohibition against the use of race strictly to benefit the employer. See, e.g., 42 U.S.C. § 2000e-2(e) (2012) (providing “a bona fide occupational qualification” defense for employment practices designed “on the basis of [the employee’s] religion, sex, or national origin” but not race).

  192. [192]. Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411, 2420 (2013).

  193. [193]. Grutter, 539 U.S. at 318.

  194. [194]. Id. at 340.

  195. [195]. Id. at 334.

  196. [196]. See supra note 174 and accompanying text.

  197. [197]. Rich, supra note 31, at 77 (“[G]enerally the employer’s latitude to define the qualifications for work trumps either external or internal theories of merit; in other words, the employer may choose selection criteria that are foolish or inefficient and may apply its own internal rules inconsistently without running afoul of Title VII so long as it does so for reasons unrelated to the plaintiff’s protected status.”).

  198. [198]. Fisher, 133 S. Ct. at 2415; see also Ricci v. DeStefano, 557 U.S. 557, 561–62 (“[F]irefighters prize their promotion to and within the officer ranks.... Aware of the intense competition for promotions, New Haven, like many cities, relies on objective examinations to identify the best qualified candidates.”).

  199. [199]. Fisher, 133 S. Ct. at 2420.

  200. [200]. Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 719 (2007) (emphasis added); see also Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 211 (1995) (“The injury in cases of this kind is that a ‘discriminatory classification prevent[s] the plaintiff from competing on an equal footing.” (alteration in original) (quoting Ne. Fla. Chapter, Associated Gen. Contractors of Am. v. Jacksonville, 508 U.S. 656, 667 (1993))); id. at 229–30 (“[W]henever the government treats any person unequally because of his or her race, that person has suffered an injury that falls squarely within the language and spirit of the Constitution’s guarantee of equal protection.”); Rich, supra note 117, at 1564–66 (discussing this conception of the constitutional injury in connection with the Court’s affirmative action and racial redistricting cases).

  201. [201]. See, e.g., Parents Involved in Cmty. Sch., 551 U.S. at 797 (Kennedy, J., concurring) (“To be forced to live under a state-mandated racial label is inconsistent with the dignity of individuals in our society.”); Rice v. Cayetano, 528 U.S. 495, 517 (2000) (“One of the principal reasons race is treated as a forbidden classification is that it demeans the dignity and worth of a person to be judged by ancestry instead of by his or her own merit and essential qualities.”); Rich, supra note 117, at 1566–67 (arguing that the expressive equality harms recognized by the Court in Shaw v. Reno take further root in the Court’s affirmative action cases which also make clear that such harms are “personal” or “individual”).

  202. [202]. See, e.g, Burlington N. & Santa Fe Ry. Co. v. White, 548 U.S. 53, 57 (2006) (holding that Title VII’s antiretaliation provision applies only to “employer actions that would have been materially adverse to a reasonable employee or job applicant”); St. Mary’s Honor Ctr. v. Hicks, 509 U.S. 502, 506–07 (1993) (discussing the presumption employers must overcome to establish “that the adverse employment actions were taken ‘for a legitimate, non-discriminatory reason’” (quoting Tex. Dep’t of Cmty. Affairs v. Burdine, 450 U.S. 248, 254 (1981))); see also Primus, supra note 151, at 1356 (“The easiest examples of adverse employment actions include dismissals, demotions, failures to hire, failures to promote, and reductions in pay. But not every undesirable thing that happens in the workplace counts as an adverse employment action.” (footnotes omitted)).

  203. [203]. Primus, supra note 151, at 1357–58.

  204. [204]. See supra note 177–78 and accompanying text.

  205. [205]. See supra notes 116–18 and accompanying text.

  206. [206]. See generally Miller v. Johnson, 515 U.S. 900 (1995); Shaw v. Reno, 509 U.S. 630 (1993).

  207. [207]. Bartlett v. Strickland, 556 U.S. 1, 6 (2009)

  208. [208]. Id.

  209. [209]. 42 U.S.C. § 1973 (2012).

  210. [210]. Bartlett, 556 U.S. at 14 (quoting 42 U.S.C. § 1973).

  211. [211]. Ross, supra note 25, at 1205.

  212. [212]. City of Mobile v. Bolden, 446 U.S. 55, 61 (1980). Indeed, the Bartlett Court recognizes upfront that, although originally enacted in a form that “intended to have an effect no different from that of the Fifteenth Amendment,’” Bartlett, 556 U.S. at 14 (quoting Mobile, 446 U.S. at 61), section 2 was amended to include an effects test that extends beyond the constitutional protection. Id.

  213. [213]. Bartlett, 556 U.S. at 21.

  214. [214]. Id. (quoting City of Richmond v. J.A. Croson Co., 488 U.S. 469, 518 (1989) (Kennedy, J., concurring)).

  215. [215]. See id. at 21−22 (arguing that such an “interpretation would result in a substantial increase in the number of mandatory districts drawn with race as ‘the predominant factor motivating the legislature’s decision’” (quoting Miller v. Johnson, 515 U.S. 900, 916 (1995)).

  216. [216]. Id. at 17.

  217. [217]. Id. (“Though courts are capable of making refined and exacting factual inquiries, they ‘are inherently ill-equipped’ to ‘make decisions based on highly political judgments’ of the sort that crossover-district claims would require.” (quoting Holder v. Hall, 512 U.S. 874, 894 (1994))).

  218. [218]. Id. at 33 (Souter, J., dissenting).

  219. [219]. Id. at 25.

  220. [220]. Id.

  221. [221]. Id. at 25−26.

  222. [222]. Shelby Cnty. v. Holder, 133 S. Ct. 2612, 2631 (2013).

  223. [223]. Id. at 2620 (alteration in original) (quoting 42 U.S.C. § 1973a(b) (2012)).

  224. [224]. See Nw. Austin Mun. Util. Dist. No. One v. Holder, 557 U.S. 193, 202 (2009) (noting that “Section 5 goes beyond the prohibition of the Fifteenth Amendment”).

  225. [225]. Shelby Cnty., 133 S. Ct. at 2631.

  226. [226]. Id. at 2632 (Ginsburg, J., dissenting).

  227. [227]. Id. at 2628 (alteration in original).

  228. [228]. See id. at 2621.

  229. [229]. Nw. Austin, 557 U.S. at 203; see also Shelby Cnty., 133 S. Ct. at 2622 (applying the Northwest Austin rule).

  230. [230]. Id. at 2625 (alteration in original) (quoting Fanie Lou Hamer, Rosa Parks, and Corretta Scott King Voting Rights Act Reauthorization and Amendment Act of 2006, Pub. L. No. 109-246, § 2(b)(1), 120 Stat. 557, 557) (internal quotation marks omitted).

  231. [231]. Id. at 2628−29.

  232. [232]. Id. at 2626. The Court’s appraisal of voting conditions nationwide is significant and shows again the malleability of its empirical reasoning. Had the Court limited its disposition of the case to resolving the constitutionality of section 4 as applied to Alabama, it would have been compelled to decide the issue based on whether conditions within the state demonstrated a current need for preclearance, a very different empirical question. See John Paul Stevens, The Court & the Right to Vote: A Dissent, N.Y. Rev. Books (Aug. 15, 2013),
    articles/archives/2013/aug/15/the-court-right-to-vote-dissent; see also Shelby Cnty., 133 S. Ct. at 2646 (Ginsburg, J., dissenting) (opining “that, at least in Alabama, the ‘current burdens’ imposed by [section] 5’s preclearance requirement are ‘justified by current needs’”).

  233. [233]. Shelby Cnty., 133 S. Ct. at 2627 (quoting Nw. Austin, 557 U.S. at 203) (internal quotation marks omitted).

  234. [234]. See Post & Siegel, Legislative Constitutionalism, supra note 15; see also supra notes 15–17 and accompanying text (discussing Post and Siegel’s interpretation of the Court’s Section 5 jurisprudence).

  235. [235]. South Carolina v. Katzenbach, 383 U.S. 301 (1966).

  236. [236]. Id. at 308.

  237. [237]. Id.

  238. [238]. Id. at 309.

  239. [239]. Shelby Cnty. v. Holder, 133 S. Ct. 2612, 2651 (2013) (Ginsburg, J., dissenting).

  240. [240]. Katzenbach, 383 U.S. at 324.

  241. [241]. Shelby Cnty., 133 S. Ct. at 2625 (quoting Katzenbach, 383 U.S. at 330).

  242. [242]. Id. at 2649 (Ginsburg, J., dissenting) (quoting Katzenbach, 383 U.S. at 330).

  243. [243]. Id. at 2624 (majority opinion) (citing Nw. Austin Mun. Util. Dis. No. One v. Holder, 557 U.S. 193, 203 (2009)).

  244. [244]. Id. at 2649 (Ginsburg, J., dissenting).

  245. [245]. This is particularly likely given that the Court’s reliance on Northwest Austin already aligns Fourteenth Amendment and Fifteenth Amendment doctrines. See Richard Hasen, The Curious Disappearance of Boerne and the Future Jurisprudence of Voting Rights and Race, SCOTUSblog (June 25, 2013, 7:10 PM), (discussing the court’s failure to resolve the “question of how to scrutinize Congress’s power to enforce the Fifteenth Amendment in voting rights cases”).

  246. [246]. Shelby Cnty., 133 S. Ct. at 2627 (quoting 42 U.S.C. § 1973c(b) (2012)) (internal quotation marks omitted).

  247. [247]. Id. at 2627.

  248. [248]. Id. at 2631.

  249. [249]. Id. at 2626 (“There is no doubt that these improvements are in large part because of the Voting Rights Act. The Act has proved immensely successful at redressing racial discrimination and integrating the voting process.”).

  250. [250]. Grutter v. Bollinger, 539 U.S. 306, 343 (2003).

  251. [251]. Id.

  252. [252]. See supra note 22.

  253. [253]. See Strauss, supra note 22, at 892 (“In a word, why shouldn’t modernization be the job of officials who must face the electorate periodically and who are periodically replaced, rather than the job of politically insulated, long-tenured judges?”); see also id. at 860 (stating that, though it sometimes practices modernization, the Supreme Court has not “fully avowed” the approach).

  254. [254]. See supra notes 97–103 and accompanying text (discussing the Court’s reasoning that §1981 and the Fourteenth Amendment were enacted and proposed by the 39th Congress to fulfill common purposes).

  255. [255]. See supra notes 107–11 and accompanying text.

  256. [256]. See supra note 22; see also Ross, supra note 25, at 1228–29.

  257. [257]. See supra notes 235–39 and accompanying text.

  258. [258]. Patterson v. McLean Credit Union, 491 U.S. 164, 174 (1989) (quoting Runyon v. McCrary, 427 U.S. 160, 191 (1976) (Stevens, J., concurring)).

  259. [259]. To so argue would hardly be consistent with the Shelby County Court’s praise for the success of the VRA in the years following Katzenbach. See Shelby Cty. v. Holder, 133 S. Ct. 2612, 2626 (2013) (“There is no doubt that these improvements [in African-American voter turnout] are in large part because of the Voting Rights Act.”).

  260. [260]. See supra notes 150–52 and accompanying text.

  261. [261]. See supra notes 44–45, 103 and accompanying text.

  262. [262]. See supra note 45 and accompanying text.

  263. [263]. For example, Congress clarified its intention to provide an effects-based test under section 2 of the VRA following the Court’s decision in City of Mobile v. Bolden, 446 U.S. 55 (1980) (holding that section 2 required a showing of discriminatory purpose). The Senate Judiciary Committee Report on the amendments to section 2 “concluded that th[e] intent test [from Bolden] places an unacceptably difficult burden on plaintiffs” and “diverts the judicial injury [sic] from the crucial question of whether minorities have equal access to the electoral process to a historical question of individual motives,” and it concluded that practices with racial impacts should be understood to violate the statute to the same practical effect as purposeful discrimination. S. Rep. No. 97-417, at 16 (1982), reprinted in 1982 U.S.C.C.A.N. 177, 192. The committee further determined that voting systems that “operate, designedly or otherwise, to minimize or cancel out the voting strength and political effectiveness of minority groups” are just as “impermissible” a denial of the right to vote “as outright denial of access to the ballot box.” Id. at 28. Similarly, Congress concluded in 1991 that the Court’s dilution of the disparate-impact standard “in Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989) ha[d] weakened the scope and effectiveness of Federal civil rights protections” by rejecting the Court’s prior interpretation of disparate-impact liability in Griggs. Civil Rights Act of 1991, Pub. L. No. 102-166, § 2, 105 Stat. 1071.

  264. [264]. See Mt. Holly Gardens Citizens in Action, Inc. v. Twp. of Mount Holly, 658 F.3d 375 (3d Cir. 2011), cert. granted, 133 S. Ct. 2824 (2013), cert. dismissed, 134 S. Ct. 636 (2013) (dismissing certiorari after settlement by the parties). Civil rights advocates had good reason to favor settlement in Mt. Holly Gardens: unlike Title VII and the VRA, the Fair Housing Act lacks express language providing for effects-based claims. Had the Court heard the case and ruled that no disparate-impact liability is available under the Act, it may have done so based on a constitutional argument that such liability exceeds congressional authority or based on a statutory argument that the text of the Act simply does not support such an interpretation. And, had it chosen the latter, Mt. Holly Gardens might have been, in a sense, a repeat of Davis.

  265. [265]. See Shelby Cnty. v. Holder, 133 S. Ct. 2612, 2627 (2013) (opining that amendments to section 5 that enhance its effects-based approach may be unconstitutional on empirical grounds similar to those on which the Court invalidated section 4, because they raised the bar on covered jurisdictions “even as the conditions justifying [preclearance] have dramatically improved”); Ricci v. DeStefano, 557 U.S. 557, 594 (2009) (Scalia, J., concurring) (questioning whether Title VII’s disparate-impact provisions are “consistent with the Constitution’s guarantee of equal protection”).

  266. [266]. See supra notes 39–42 and accompanying text.

  267. [267]. See, e.g., Neil Gotanda, A Critique of “Our Constitution is Color-Blind, 44 Stan. L. Rev. 1, 4 (1991) (proposing distinct concepts of “status-race,” “formal-race,” “culture-race,” and “historical-race” to explain the Supreme Court’s references to race in its constitutional jurisprudence).

  268. [268]. See generally Rich, supra note 31 (arguing that psychological understandings of discrimination are inconstant, evolving, and cannot be expected to reliably ground the normative commitments of antidiscrimination law).

  269. [269]. See supra notes 228–38 and accompanying text.

  270. [270]. Compare Griggs v. Duke Power Co., 401 U.S. 424, 429–31 (1971) (recognizing racially disproportionate impact resulting from facially neutral employment practices as discrimination), with Washington v. Davis, 426 U.S. 229, 245 (1976) (expressing “difficulty understanding how a law establishing a racially neutral qualification for employment is nevertheless racially discriminatory... simply because a greater proportion of Negroes fail to qualify than members of other racial or ethnic groups”); compare Palmer v. Thompson, 403 U.S. 217, 224 (1971) (describing the “hazards of declaring a law unconstitutional because of the motivations of its sponsors,” including the “difficult[y] for a court to ascertain the motivation, or collection of different motivations, that lie behind a legislative enactment”), with Davis, 426 U.S. at 238–40 (arguing that equal protection has always required courts to determine whether the government acted with a discriminatory purpose).

  271. [271]. Post & Siegel, supra note 38, at 444.

  272. [272]. Id. at 520.

  273. [273]. See, e.g., Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 335–36 n.15 (1977).

  274. [274]. Primus, supra note 162, at 498.

  275. [275]. Griggs, 401 U.S. at 432; see also id. at 436 (“Congress has made such qualifications the controlling factor, so that race, religion, nationality, and sex become irrelevant.” (internal quotation marks omitted)).

  276. [276]. One may view disparate impact in employment and voting as a success by its own standard of measure. In employment, disparate impact has influenced employers to adopt formal and competent selection techniques that have had the benefit of reducing inequality and inhibiting individual acts of discrimination. See supra note 127 and accompanying text. In voting, the Supreme Court’s own conclusions in Shelby County demonstrate its confidence that the statutory scheme of the VRA has been remarkably successful at promoting fair voting measures. See supra note 250 and accompanying text. Whether success should result in retirement is fundamentally a political choice.

  277. [277]. We might adopt this view if we thought about discrimination as an integrated system of subordinating practices. See supra note 45 and accompanying text.

  278. [278]. Mario L. Barnes et al., A Post-Race Equal Protection?, 98 Geo. L.J. 967, 968 (2010); see also Sumi Cho, Post-Racialism, 94 Iowa L. Rev. 1589, 1594 (2009) (arguing that post-racialism is an ideology based on the belief that as a result of progress in race relations, the state does not need to consider race in making policy decisions).

  279. [279]. See Aliya Saperstein & Andrew M. Penner, Racial Fluidity and Inequality in the United States, 118 Am. J. Soc. 676, 678 (2012) (observing a “perverse[]” relationship in which increased racial fluidity correlates with “more entrenched racial inequality”); see also David R. Harris & Jeremiah Joseph Sim, Who Is Multiracial? Assessing the Complexity of Lived Race, 67 Am. Soc. Rev. 614, 623 (2002) (demonstrating that racial self-identification varies with context).

  280. [280]. Nancy Leong, Racial Capitalism, 126 Harv. L. Rev. 2151, 2153 (2013) (defining “racial capitalism” as “the process of deriving social and economic value from the racial identity of another person”).

  281. [281]. See John D. Skrentny, After Civil Rights: Racial Realism in the New American Workplace 10 (2014) (defining “racial realism” as the view that “race has both significance and usefulness... and this is true irrespective of government policy or lofty concerns about equality and justice”).

  282. [282]. For an extended and provocative discussion of this dynamic, see generally Carbado & Gulati, supra note 41.

  283. [283]. Schuette v. Coal. to Defend Affirmative Action, Integration & Immigrant Rights & Fight for Equality by Any Means Necessary (BAMN), 134 S. Ct. 1623, 1638 (2014).

  284. [284]. See Washington v. Seattle Sch. Dist. No. 1, 458 U.S. 457, 474 (1982) (holding that the “racial focus” of a challenged state initiative that prohibited local school districts from engaging in racially integrative busing required strict scrutiny); Hunter v. Erickson, 393 U.S. 385, 393 (1969) (applying strict scrutiny to an amendment to the Akron city charter “making it more difficult to enact legislation” serving the interests of racial minorities); see also BAMN, 134 S. Ct. at 1631 (rejecting the circuit court’s interpretation of and reliance on Seattle); id. at 1643 (Scalia, J., concurring) (“Patently atextual, unadministrable, and contrary to our traditional equal-protection jurisprudence, Hunter and Seattle should be overruled.”).


Associate Professor of Law, Gould School of Law, University of Southern California. 

I am grateful for generous comments on early drafts of this paper by Ariela Gross, David Cruz, Neil Siegel, Vicki Schultz, and Nomi Stolzenberg. I also thank the attendees of the USC Gould School of Law Faculty Workshop and the Labor and Employment Law Colloquium for their thoughtful questions and comments.