The Yale Law Journal


An Intersectional Critique of Tiers of Scrutiny: Beyond “Either/Or” Approaches to Equal Protection

06 Nov 2019

abstract. For the past forty years, Justice Powell’s concurring opinion in University of California v. Bakke has been at the center of scholarly debates about affirmative action. Notwithstanding the enormous attention Justice Powell’s concurrence has received, scholars have paid little attention to a passage in that opinion that expressly takes up the issue of gender. Drawing on the theory of intersectionality, this Essay explains several ways in which its reasoning is flawed. The Essay also shows how interrogating Justice Powell’s “single axis” race and gender analysis raises broader questions about tiers of scrutiny for Black women. Through a hypothetical of a university’s affirmative-action plan that specifically targets Black women, the Essay considers what tier of scrutiny should apply. Because, for the most part, scholars take a race or gender approach to equal-protection law, they have not engaged that doctrinal puzzle and its implications for tiers-of-scrutiny writ large.


For the past forty years, affirmative action has been one of the most hotly contested constitutional issues. The literature on the policy is voluminous.1 Perhaps unsurprisingly, much of that literature references, if not fully engages, University of California v. Bakke.2 In that case, Justice Powell’s concurring opinion expressed three critical ideas: (1) that affirmative-action policies warrant the highest level of judicial review (strict scrutiny);3 (2) that diversity can serve as a “compelling state interest” for the policy;4 and (3) that colleges and universities may employ race as “one factor, among many” (but not as quotas) in deciding which students to admit.5 A majority of the Supreme Court would subsequently endorse Justice Powell’s constitutional approach to affirmative action.6 And that approach continues to set the stage on which scholars rehearse their arguments for and against the policy.

Notwithstanding the enormous attention Justice Powell’s concurrence has received, scholars have paid little attention to a passage in it that expressly takes up the issue of gender.7 Drawing on the theory of intersectionality, we engage that passage, explain several ways in which its reasoning is flawed, and employ that analysis to articulate a more general critique of Equal Protection’s tiers of scrutiny.

Our intersectional reading of Bakke is timely for at least four reasons. First, last year marked the case’s fortieth anniversary, and this year is the thirtieth anniversary of the theory of intersectionality. One of us introduced that theory in a 1989 article titled Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics.8 Like the literature on Bakke, the literature of intersectionality is similarly wide-ranging. Indeed, since the publication of Demarginalizing, actors across different contexts, disciplines, and organizations have employed intersectionality to inform their approaches to and critiques of knowledge production and epistemology, community organizing and political advocacy, and litigation and legal reform initiatives.9 Despite the breadth and depth of the literatures on intersectionality and affirmative action, intersectionality has not figured as an important framework in debates over affirmative action, and affirmative action has not figured prominently in discourse about intersectionality. This Essay stages a long overdue interface.

Second, a central intervention Demarginalizing performed was to demonstrate how “single-axis” race-or-gender approaches to antidiscrimination law created hidden baselines against which courts framed Black10 women’s allegations of discrimination as extraordinary, preferential, and unfeasible. This Essay employs Bakke as a point of departure to illustrate how these baseline and representational problems are present in antidiscrimination law more broadly, including in Equal Protection’s tiers-of-scrutiny regime.

Third, our intersectional analysis of tiers of scrutiny broadens the normative, analytical, and doctrinal terms on which scholars have contested that framework.11 Absent from existing critiques is the view that tiers of scrutiny constitutionally embody a “single-axis” race-or-gender logic.12 This either/or logic becomes readily apparent upon asking which tier of scrutiny is applicable to remedial projects that target Black women. Intermediate scrutiny because they are women?13 Strict scrutiny because they are Black?14 Or “strict-scrutiny-plus,” because as Black women they occupy two classifications to which heightened scrutiny typically applies?

To tiers-of-scrutiny enthusiasts, the dilemma we pose may seem to encourage the simple-enough task of generating a coherent justification for one tier or another. And to some extent, we do approach the matter as a doctrinal puzzle in that we offer a provisional argument within the tiers-of-scrutiny universe for reviewing remedial efforts that address particular intersectional dilemmas. At the same time, we share the sense of many skeptics of current equal-protection doctrine who argue that the tiers have become fossilized commitments of a bygone era.15 We should be clear to note, then, that we are neither embracing the overall tiers-of-scrutiny framework nor proposing “a tier of one’s own” for Black women. Our more fundamental goal is to draw attention to and disrupt the racialized baselines that have grounded gender-based intermediate scrutiny and have made Black women’s remedial status in equal-protection law more suspect than white women’s.

Finally, our foregrounding of Black women in our intersectional analysis of Bakke and tiers of scrutiny provides an important window into how, historically, courts have marginalized Black women’s identity and experiences, and into the presumptively neutral baselines that underwrite antidiscrimination law. Although rarely noted, the suspect status of Black women’s antidiscrimination lawsuits contrasts dramatically with the presumptive cognizability of the white male subjects of “reverse discrimination” claims. As first observed in Demarginalizing, the same questions that courts have raised about the scope of Title VII’s antidiscrimination mandate, and that we have raised about tiers of scrutiny, could be raised with reference to white men. As best we can tell, nowhere have courts worried aloud that permitting white men to bring compound “reverse discrimination” claims would be preferential treatment, open a Pandora’s Box, or raise questions of plaintiffs’ representativeness on the view that white men are “only” white men so cannot adequately represent all white people and in particular white women.16 Because courts have not subjected white men’s “reverse discrimination” lawsuits to a compound discrimination analysis, those lawsuits have not engendered the same kinds of prohibitory responses as Black women’s lawsuits. Though we develop our claim about white men’s standing in antidiscrimination law elsewhere,17 we note here that the both/and representational capital that white men possess in antidiscrimination law is precisely what Black women have struggled to acquire: the standing to frame their injury as representative of the general (for example, as women) or of the particular (for example, as Black women). The remainder of our argument proceeds as follows.

Part I first revisits Demarginalizing to foreground the emergence of intersectionality as a prism to analyze both the erasure of Black women’s claims of compound discrimination and courts’ refusal to designate them as class representatives in race- and gender-discrimination cases. Underlying these decisions were baselines of whiteness that narrowed the parameters of race and gender discrimination and positioned Black women’s petitions for antidiscrimination protection as demands for preferential treatment. After laying this theoretical foundation, Part I then turns to Justice Powell’s passage on gender. In explaining the flaws in Justice Powell’s reasoning, we highlight the various intersectional obfuscations on which it rests. In particular, we show how his installation of white women as the baseline for articulating the boundaries of intermediate scrutiny is analogous to the Title VII sex-discrimination cases Demarginalizing discussed.

Part II interrogates the tiers-of-scrutiny framework more generally. Our overarching claim is that this framework is a “single-axis” juridical structure, rooted in the either/or approach to antidiscrimination law that Demarginalizing identified. Part II contends that this race or gender or (more recently) sexual-orientation approach sits in tension with a basic intersectionality insight—that because of the intersectional dimensions of power, people live their lives co-constitutively as “both/and,” rather than fragmentarily as “either/or.” Part II further maintains that the “either/or” approach to equal-protection doctrine limits remedial opportunities for Black people not only under strict scrutiny, but under intermediate scrutiny and rational-basis review as well. This remedial disadvantage derives from the fact that, on one conventional interpretation of equal-protection doctrine, the intersection of race with any other identity triggers strict scrutiny. Under this view, were an institution to adopt an affirmative-action plan focused squarely on race and gender, courts should subject it to strict scrutiny, not intermediate scrutiny. Similarly, an affirmative-action plan focused on race and sexual orientation should be subjected to strict scrutiny, not rational-basis review. The distributional effect of this race-plus-anything-triggers-strict-scrutiny vision of equal-protection doctrine is that affirmative-action plans that are expressly intersectional, and therefore likely to benefit African Americans who are women and/or members of the LGBTQ community, are precisely the ones to which courts may apply strict scrutiny. On the flip side, courts would not subject colorblind affirmative-action plans, which target women or the LGBTQ community and are likely to benefit members of those groups who are white, to strict scrutiny. While courts would treat such colorblind initiatives as presumptively race-neutral and likely find them constitutional, remedial efforts to counteract their exclusionary or disparate impact would face a nearly impassable constitutional barrier. Understood in this way, Part II’s intersectional analysis of tiers of scrutiny is a concrete example of how baselines of whiteness can, under the guise of race neutrality, circumscribe remedial possibilities for African Americans.

I. black women, title vii, and colorblind intersectionality

A. An Intersectional Reading of DeGraffenreid v. General Motors

One of Demarginalizing’scentral observations is that in the 1970s and 1980s, courts denied Black women the opportunity to articulate their claims not only as Black women, but as women generally. When Black women demanded protection against what they claimed was compound discrimination, some courts responded that recognizing such claims would treat Black women preferentially. When Black women asserted their claims as women, some courts responded that Black women were too different to represent all women. One of the cases that demonstrated this marginalization of Black women in antidiscrimination law was DeGraffenreid v. General Motors.18

In DeGraffenreid, Black women brought a discrimination claim under Title VII of the Civil Rights Act of 1964. The women alleged that General Motors had discriminated against them on the basis of both their race and sex. The district court roundly rejected their claim, reasoning that

plaintiffs have failed to cite any decisions which have stated that [B]lack women are a special class to be protected from discrimination. The Court’s own research has failed to disclose such a decision. The plaintiffs are clearly entitled to a remedy if they have been discriminated against. However, they should not be allowed to combine statutory remedies to create a new “super-remedy” which would give them relief beyond what the drafters of the relevant statutes intended. Thus, this lawsuit must be examined to see if it states a cause of action for race discrimination, sex discrimination, or alternatively either, but not a combination of both.19

The court then proceeded to perform that “single-axis”20 (race or sex) analysis to ascertain whether the plaintiffs could establish a separate cause of action for sexism or for racism. With respect to sexism, the court maintained that there was no evidence of sex discrimination because General Motors hired white women during the period in which Black women claimed the company had discriminated against them.21 The court’s conclusion seemed to be premised on the assumption that for an institution to run afoul of Title VII, it would have to treat all women the same. Under this reasoning, the fact that General Motors hired white women negated Black women’s allegations of sexism.

The deployment of General Motors’ history of hiring white women to negate the possibility that it discriminated against Black women was not just an evidentiary move. It was a conceptual one that made white women’s experiences the relevant baseline against which the cognizability of sex-discrimination claims was determined. This framed Black women, but not white women, as gendered and raced subjects, rendering them imperfect representatives of sex-discrimination claims.22 Underscoring this point, the case Moore v. Hughes Helicopters, Inc. rejected a Black woman’s attempt to certify a class of plaintiffs that included white women because the Black female plaintiff “had never claimed that she was discriminated against as a female, but only as a [B]lack female.23What the court was really saying was that the Black female plaintiff had never claimed that she was discriminated against as a white female, but only as a Black female. As Demarginalizing argued,

For white women claiming sex discrimination is simply a statement that but for gender, they would not have been disadvantaged. For them there is no need to specify discrimination as white females because their race does not contribute to the disadvantage for which they seek redress. The view of discrimination that is derived from this grounding takes race privilege as a given.24

In other words, a white female plaintiff would not have to “worry about being ‘only’ a white woman over and against some more generalizable female subjectivity.”25 Her white female identity functioned not as “a particularity of gender but gender itself.”26

If we are right that the courts’ preference for formulating a sex-discrimination claim “as a female” might be rearticulated to mean “as a white female,” but not “as a [B]lack female,” then the courts’ approach was decidedly not race neutral. It is an example of how whiteness can operate as an invisible baseline for gender in ways that privilege white women. As Section I.B discusses, this baselining of whiteness is at play in Justice Powell’s Bakke concurrence as well.

B. An Intersectional Reading of an Overlooked Passage

Our intersectional engagement of Justice Powell’s Bakke concurrence begins with the following passage:

Nor is petitioner’s view as to the applicable standard supported by the fact that gender-based classifications are not subjected to this level of scrutiny. Gender-based distinctions are less likely to create the analytical and practical problems present in preferential programs premised on racial or ethnic criteria. With respect to gender there are only two possible classifications. The incidence of the burdens imposed by preferential classifications is clear. There are no rival groups which can claim that they, too, are entitled to preferential treatment. Classwide questions as to the group suffering previous injury and groups which fairly can be burdened are relatively manageable for reviewing courts. The resolution of these same questions in the context of racial and ethnic preferences presents far more complex and intractable problems than gender-based classifications. More importantly, the perception of racial classifications as inherently odious stems from a lengthy and tragic history that gender-based classifications do not share. In sum, the Court has never viewed such classification as inherently suspect or as comparable to racial or ethnic classifications for the purpose of equal protection analysis.27

As we stated in the Introduction, scholars have largely overlooked the race and gender logics on which this passage rests.28 We interrogate those logics here not just because they reveal Justice Powell’s rudimentary understanding of race and gender, but also because Justice Powell’s conceptualization of race and gender reflected and shaped the conceptual infrastructure on which tiers-of-scrutiny analyses are grounded.29 Moreover, the race and gender thinking that descriptively and normatively anchors the passage transcends equal-protection doctrine and likely explains why, across doctrinal regimes, Black women have historically been “impossible subjects”30 in antidiscrimination law. That is to say, judges have pushed Black women’s social disadvantages and vulnerabilities beyond the boundaries of legal remediation.

We begin with Justice Powell’s assertion that “the perception of racial classifications as inherently odious stems from a lengthy and tragic history that gender-based classifications do not share.” This reflects what one of us has called an “intersectional failure” in at least two respects.31 First, the claim ignores the fact that slavery and Jim Crow—both of which presumably are part of the “lengthy and tragic history” to which he refers—were gendered in ways that shaped how Black women experienced themselves as women.32 Implicit in Justice Powell’s account is the view that what happened to Black women in the context of Jim Crow and slavery was not gendered or, alternatively, that the gendered dimensions of slavery were somehow distinguishable. His analysis obscures that slavery and Jim Crow were constituted through regimes of both race and sex. This obfuscation strips Black women of their female identity and banishes their experiences from constitutionally relevant narratives of gender inequality. In sum, Justice Powell failed to see that race and gender are intersectional social forces. It is precisely this intersectionality that explains both the racialized sexual violence Black women endured during slavery (including the sexual exploitation of Black women’s bodies to produce a “peculiar”33 kind of property—Black slaves),34 and the racialized domesticity Black women experienced under Jim Crow (including their economic exploitation as caregivers and domestic workers).35

Another reason Justice Powell’s analysis might be described as an intersectional failure is that it reflects the view that Black women’s experiences do not (and presumably cannot) stand in for or represent the category of women per se. He seems to suggest that gender discrimination is that which disadvantages or limits the opportunities of white women. Underwriting Justice Powell’s reasoning is the idea that because slavery and Jim Crow did not happen to white women, neither can be understood as subordinating regimes that impact the lives of women. Borrowing from a classic Black feminist text, in Justice Powell’s analysis, “all the women are white.”36

Something akin to an “intrafemale separate sphere” ideology is at play here. Unlike the traditional separate-spheres ideology, which policed the boundaries of the public and private in ways that made the home the “natural” place for women,37 the intrafemale separate sphere polices the boundaries of female identity in ways that make white women the “natural” embodiment of women. Under this logic, Black women’s experiences cannot stand in for women’s experiences because Black women cannot stand in for white women. Because white women’s experiences define the boundaries of gender in Justice Powell’s analysis, the gendered dimensions of Black women’s “lengthy and tragic history” are, constitutionally speaking, illegible. His reasoning consigns Black women to a separate sphere in which their race undermines their claim to a female identity.

Of course, at no point does Justice Powell expressly say that white women are “real” women and Black women are not. But his claim that women have not experienced a “lengthy and tragic history” of discrimination analogous to slavery and Jim Crow is based on one of two assumptions (or both): (1) that slavery and Jim Crow did not marginalize Black women and (2) that evidence of women’s marginalization resides outside the “lengthy and tragic history” of racial discrimination. These assumptions, separately and together, effectively make white women the “proper” subjects of gender. In short, Justice Powell’s analysis searches for gender inequality outside of the two significant racial regimes in which Black women experienced it: slavery and Jim Crow. It is as if, in assessing the historical realities of gender, Justice Powell is thinking about Scarlett O’Hara rather than Sojourner Truth.38 He centers white women’s experiences and makes them the baseline against which gender inequality is constitutionally legible or not.

Justice Powell’s gender analysis is an example of what one of us has called “colorblind intersectionality.”39 By colorblind intersectionality we mean “instances in which whiteness helps to produce and is part of a cognizable social category but is invisible or unarticulated as an intersectional subject position.”40 Expressed slightly differently, colorblind intersectionality makes whiteness an unstated baseline for gender, rather than a modifier of it. Thus, anytime a court speaks in terms of women but is really referring to white women, it is engaging in colorblind intersectionality.

Justice Powell’s concurrence reflects colorblind intersectionality. Though he purports to make a claim about gender qua gender, the woman at the center of his analysis, who has not experienced “the lengthy and tragic history” of racial subordination, is clearly white. Because in Justice Powell’s opinion whiteness operates invisibly as the default on which he articulates his views about gender inequality, the racial dimensions of his claims about gender are both incorporated (in that whiteness anchors his race and gender analysis) and unincorporated (in that whiteness is not formally expressed as an intersectional feature of gender).41

None of this means that we endorse Justice Powell’s view that white women have not experienced a “lengthy and tragic history” of subordination. As then-professor Ruth Bader Ginsburg observed, commenting on Bakke, the Supreme Court’s prior gender-discrimination cases contradicted the claim that sex discrimination in the United States was neither “lengthy” nor “tragic.”42 Consider, for example, Frontiero v. Richardson.43The case centered on the constitutionality of a statute that made female spouses of male members of the military dependents for purposes of receiving housing and medical benefits, but required male spouses of female military personnel to prove their dependency.44 Justice Brennan, writing for a plurality of the Court, subjected the statute to strict scrutiny and argued that it was unconstitutional. In doing so, Justice Brennan expressly discussed the historical subordination of women. “There can be no doubt that our Nation has had a long and unfortunate history of sex discrimination,” he wrote.45 “Traditionally, such discrimination was rationalized by an attitude of ‘romantic paternalism’ which, in practical effect, put women, not on a pedestal, but in a cage.”46 Justice Brennan then went on to analogize discrimination against women to discrimination against African Americans,47 observing that

our statute books gradually became laden with gross, stereotyped distinctions between the sexes and, indeed, throughout much of the 19th century the position of women in our society was, in many respects, comparable to that of [B]lacks under the pre-Civil War slave codes. Neither slaves nor women could hold office, serve on juries, or bring suit in their own names, and married women traditionally were denied the legal capacity to hold or convey property or to serve as legal guardians of their own children. And although [B]lacks were guaranteed the right to vote in 1870, women were denied even that right—which is itself “preservative of other basic civil and political rights”—until adoption of the Nineteenth Amendment half a century later.48

We highlight Justice Brennan’s plurality opinion in Frontiero because it is one indication that Justice Powell likely understood that his claim that “gender-based classifications” were not rooted in a “lengthy and tragic history” was contestable.49

Justice Brennan’s plurality is noteworthy in another sense too. It, like Justice Powell’s Bakke concurrence, reflects intersectional erasures. In Justice Brennan’s observation that “although [B]lacks were guaranteed the right to vote in 1870, women were denied even that right,”50 the “Blacks” are clearly men. Black women were not granted the right to vote in 1870. Justice Brennan’s “race” analysis thus clearly excluded Black women’s historical relationship to the right to vote. Moreover, the “pedestal-as-cage” metaphor that, at least in part, underwrites Justice Brennan’s views that strict scrutiny should apply to gender-based classifications does not capture Black women’s historical vulnerability as women. For the most part, “romantic paternalism” was not the ideological foundation on which the subordination of Black women rested. Not quite not women,51 Black women were largely excluded from the “separate sphere” that was the hallmark of romantic paternalism. Thus, while Justice Powell rejected race-sex analogies and Justice Brennan embraced them, both moves effectuated intersectional erasures, articulated women’s experiences through the prism of whiteness, and made white women the representative “women” of equal-protection law.

We note parenthetically that Justice Powell’s rejection of race-gender analogies foreclosed not only reasoning from gender to race; it also foreclosed reasoning from race to gender. This limited the application of strict scrutiny from equal-protection race cases to invidious uses of gender (such as the statutory regime at issue in Frontiero) and limited the application of intermediate scrutiny from equal-protection gender cases to benign uses of race (such as the affirmative-action policy implicated in Bakke).52 In this respect, Justice Powell circumscribed the scope of equal-protection remediation for both race- and sex-discrimination claims.53

This is a good place to bring Black women back into focus. We do so to consider how, if at all, Justice Powell’s analysis would have changed if Black women had the representational currency to function as women. Before answering that question, it is helpful to reprise two important doctrinal foundations on which his concurrence rests. The first implicates race ostensibly unmodified by gender—namely, that any use of race is constitutionally suspect.54 Accordingly, any time the government relies on race, even for benign or remedial purposes, courts should apply strict scrutiny.55 The second doctrinal feature is more implicit. It trades on a point of law that Bakke inherits and implicates gender ostensibly unmodified by race—namely, that gender-based classifications are quasi-suspect. Accordingly, anytime the government relies on gender, even for invidious or pernicious purposes, courts should apply the less onerous intermediate scrutiny standard of review.56

To borrow from Anna Julia Cooper,57 where and when do Black women enter the tiers-of-scrutiny landscape? Are policies that focus on Black women a suspect classification (because Black women are Black) or a quasi-suspect classification (because Black women are women)? Had Black women’s experiences figured as women’s experiences in Justice Powell’s analysis, he would have been forced to confront the fact that their identity implicates both strict and intermediate scrutiny. Justice Powell did not have to engage this doctrinal tension because his colorblind intersectionality made Black women illegible as women. This illegibility is part of a broader representational position, whereby Black women’s race either overdetermines their gender (Black women are Black people, not women) or overly particularizes it (Black women are Black women, not women). Both problems limit the ability of Black women’s experiences to stand in for those of women.58

Part II builds on these intersectional insights and expressly performs the race and gender analyses that Justice Powell’s concurrence avoids. Here, we explore how equal-protection jurisprudence should process affirmative-action programs that intersectionally target race and gender, and discuss the implications of our analysis for the tiers-of-scrutiny framework more generally.

II. black women and tiers of scrutiny

This Part situates intersectionality in the context of a doctrinal discussion about tiers of scrutiny. We engage a tension between the fact that tiers of scrutiny are structured around disaggregated articulations of identity—race or gender or sexual orientation, with each triggering a different level of scrutiny—and the fact that people’s existential realities are structured by intersecting regimes of power—for example, racism and sexism and homophobia.59 In addition to raising difficult normative and doctrinal questions, this tension reveals the subtle but significant ways in which whiteness operates as an invisible baseline across tiers of scrutiny.

Our starting point is a hypothetical. Assume that a university history department establishes a gender-studies curriculum and seeks to hire faculty in this area. The department is especially interested in historians whose work reflects an intersectional sensibility. The department is also hoping to use this chance to diversify the representation of women on the school-wide faculty. Currently, the faculty is thirty percent women, almost all of whom are white. The department is hoping that structuring a search around people whose scholarship draws on intersectionality will increase the representation of women of color in the pool. The department chair thinks this is a good idea but also encourages her colleagues to intersectionalize their affirmative-action efforts by taking applicants’ race and gender into account.

After some debate, the department adopts the chair’s proposal and their search concludes with the hire of a well-qualified Black woman who many on the faculty described as a “star” in her field. An unsuccessful applicant organizes a lawsuit against the university. His claim is that the department’s intersectional affirmative-action policy violates the Equal Protection Clause of the Fourteenth Amendment. The court adjudicating the case has to decide the standard of review. Which tier of scrutiny should it apply? Strict? Intermediate? Something else?

On one view—not ours—the answer is clearly strict scrutiny. Notwithstanding the applicability of intermediate scrutiny to gender-targeted programs,60 strict scrutiny should apply because, for the purposes of constitutional analysis, Black women here are Black, not women. To appreciate the wider reasoning of this perspective, imagine that, instead of structuring its affirmative-action program around race and gender, the department focused on race and geography. The department’s reliance on geography would not negate the application of strict scrutiny, the argument might run, because the race-as-one-factor-among-many rule that forms a core part of Justice Powell’s concurrence in Bakke contemplates that universities will incorporate factors other than race into their selection decisions.61 Under this argument, if an affirmative-action plan that combines race and geography triggers strict scrutiny, a race-and-gender plan should as well. Indeed, race plus anything at all should trigger strict scrutiny. Any other approach would create an end-run around the race per se rule on which strict scrutiny is based.

On another view, which we find more appealing, intermediate scrutiny should apply. Black women lag behind white women with respect to many dimensions of social wellbeing, including income,62 health,63 educational attainment,64 professional status,65 and wealth.66 As a class, Black women are not as well-positioned as white women to take advantage of the employment opportunities that colorblind gender-based affirmative-action policies afford. In this respect, colorblind gender-based affirmative-action policies are not racially neutral; they are more likely to benefit white women. Race-conscious gender-based affirmative action is a way to offset that advantage and level the gender-equality remedial landscape.

There is no principled reason why courts should treat affirmative-action policies that are more likely to benefit Black women more stringently than those that are more likely to benefit white women. Subjecting the latter to intermediate scrutiny and the former to strict scrutiny is to apply different equal-protection standards for Black and white women. Such disparate doctrinal treatment could have broader institutional effects. Because race-conscious affirmative action would be more vulnerable to constitutional challenge than its colorblind counterparts, employers would be more inclined to implement the latter. Additionally, applying strict scrutiny to race-conscious, gender-based affirmative-action policies sends a social message that policies designed to address the marginalization of Black women are more “suspect” than those that are more beneficial to white women. The foregoing reasons are why we argue that courts should apply intermediate (not strict) scrutiny to race-conscious, gender-based affirmative action.67

We recognize that judges presented with a race-conscious, gender-based affirmative-action policy would not be limited to intermediate- or strict-scrutiny review. Setting aside rational-basis review, another doctrinal pathway exists. A court could conclude that, because race triggers strict scrutiny and gender triggers intermediate scrutiny, the doctrinal framework should “compound” those two regimes into a standard of review we might call “strict scrutiny with more bite” or “strict scrutiny-plus.” For several reasons, such an approach would be unsatisfying.

First, this additive approach reproduces the very problematic analysis of the “single-axis” thinking in law that intersectional analysis seeks to contest. In the same way that Black women’s experiences are not simply the sum total of white women’s and Black men’s, remedial interventions that target Black women should not simply be the sum total of the tier of scrutiny for race (ostensibly sans gender) plus the tier of scrutiny for gender (ostensibly sans race). The argument in Demarginalizing that rejected the framing of Black women’s vulnerability to discrimination as an added gloss on what happens to Black men or white women can be redeployed here to challenge an additive approach to the doctrinal rules on remedial efforts that target Black women.

Second, under the additive approach, while the intersection of femaleness and whiteness would trigger intermediate scrutiny, the intersection of femaleness and Blackness would trigger “strict scrutiny-plus.” This outcome would authorize a hierarchical racialization of gender, through which Black women become “super suspect” as remedial subjects while white women remain only “quasi suspect.” This “super suspect” status would further distance Black women from white women in their capacity to access antidiscrimination countermeasures like affirmative action. It would once again legitimize the view that the dimensions of patriarchy that justify societal resources are those faced by women whose narratives of gender disadvantage courts and policymakers understand and articulate without the express reference to race. Put another way, a “super suspect” status formalizes the troubling dicta in the Title VII cases we mentioned earlier. In these cases, courts framed Black women’s appeals for protection from compound discrimination as preferential treatment. More troubling still, formalizing that dicta would itself constitute intersectional discrimination in the application of legal doctrine. Under the additive approach, it is not the intersection of race and gender per se that would trigger “strict scrutiny-plus,” but the intersection of Blackness and female identity.

Finally, “strict scrutiny-plus” likely would be “strict in theory and fatal in fact.”68 As such, it would place Black women in a doctrinal category effectively beyond remediation, creating an equal-protection landscape in which less vulnerable women would receive the most remedial attention.

Our observations about the compromised position of Black women under the tiers-of-scrutiny framework take us back to one of the problems intersectionality was initially mobilized to address.69 Some courts interpreted Title VII’s prohibition against sex discrimination roughly to include only discrimination claims asserted by women that did not also expressly reference race.70 To be cognizable, gender-based discrimination claims had to be colorblind. But as we have tried to demonstrate, this approach is not racially neutral. It relies on the fact that white women do not have to specify their race to name their gender vulnerability and marginalization. Against the backdrop of that colorblind intersectionality, courts framed Black women’s explicit reference to race when asserting gender-based discrimination claims as a demand for preferential treatment.71

Our worry is that a version of this perception of preferential treatment could manifest itself in affirmative-action jurisprudence. That is, courts could conclude that affirmative-action plans that expressly take race and gender into account operate as racial preferences for Black women or other women of color, triggering strict scrutiny. Such a view obscures how colorblind gender-based affirmative-action policies are themselves intersectional projects that can function as unmarked racial preferences for white women.72

In some ways, our effort in the preceding discussion to apply the tiers-of-scrutiny framework to level the equal-protection playing field for Black women does not sufficiently challenge the fact that the tiers-of-scrutiny regime is fundamentally at odds with how power functions. A critical feature of the tiers-of-scrutiny architecture is that courts should subject particular social categories to particular levels of review: strict scrutiny for race; intermediate scrutiny for gender; and rational basis for sexual orientation. The theory of intersectionality stands in opposition to that idea.

We are not saying, to be clear, that there are no moments in which Black women think of themselves as women or as Black people without further particularity.73 There are indeed such moments, and they are necessary both for community-building and for political organizing and resistance.74 Our worry is that the tiers-of-scrutiny approach legitimizes the existential predicament about which Audre Lorde so powerfully wrote:

As a Black lesbian feminist comfortable with the many different ingredients of my identity, and a woman committed to racial and social freedom from oppression, I find I am constantly being encouraged to pluck out some one aspect of myself and present this as the meaningful whole, eclipsing or denying the other parts of self.75

The tiers-of-scrutiny regime invites precisely the plucking away at identity that Lorde describes—and not just at the individual level,76 but also at the level of institutional governance. For example, a college or university that wants a court to subject its gender-targeted affirmative-action policy to intermediate scrutiny rather than strict scrutiny has every incentive to “pluck” race out of that policy and present it as formally colorblind. This might explain why we are unaware of a single affirmative-action policy, from any institutional setting, that expressly takes Black women’s race and gender into account. In this regard, Black women are invisible as subjects of remedial concern not only on the pages of Justice Powell’s opinion but also in the institutional policy-making spaces where affirmative-action initiatives are developed. In those contexts as well, colorblindness works to “pluck” Black women’s experiences and antidiscrimination needs out of gender-based affirmative-action plans.

Our modest proposal is that courts apply intermediate scrutiny to race-conscious gender-based affirmative-action policies. In advancing this approach, we recognize that we are participating in the very regime of power we have been contesting. We view our approach as a provisional strategy in the event that an intersectional affirmative-action case comes before the courts. This “in the meantime” strategy is precisely the approach many of us have found ourselves adopting when defending the diversity rationale for affirmative action, even though we believe that courts should not subject affirmative action to strict scrutiny, and that under strict scrutiny, there are stronger justifications for it (such as combating societal discrimination). In other words, we articulate our intermediate scrutiny proposal under conditions of doctrinal constraint.

At the same time, it is not our view that eliminating tiers of scrutiny altogether in favor of a contextual approach that reviews governmental action for subordination would necessarily solve the intersectional problems we have raised. Though we are sympathetic to arguments that seek to abolish tiers of scrutiny, our view is that this would not itself disrupt the hierarchical ordering that makes some claims to equality more suspect than others. The problems this Essay foregrounds in the formalistic interpretation of tiers of scrutiny also appears in Degraffenreid’s formalistic interpretation of the word “or” in Title VII’s protected grounds of discrimination. The more fundamental problem is that both doctrinal areas trade on either/or conceptualizations of power and identity that create and obscure baselines against which African Americans in general and Black women in particular are perceived to be suspect as people seeking justice.

Our challenge, then, does not merely confront legal standards and constitutional rules, but interrogates particular ways of conceptualizing race and gender both within and beyond legal discourse. Advocates and stakeholders within discursive communities who are not bound by the constraints of such constitutional rules routinely reproduce precisely the “single-axis” frameworks that privilege and foreground group members whose narratives of injustice fit the either/or parameters of equality claims. If our own thinking and practice cannot transcend hierarchical tiers, and if our social-justice interventions fail to address baselines of privilege and power, then a mere shift in the articulation of the legal rules will not move the justice needle in the “both/and” direction that the theory of intersectionality contemplates.


Our purpose in this Essay was to revisit Demarginalizing and examine a passage in Justice Powell’s Bakke concurrence to which scholars have paid scant attention.77 By articulating racial and gender inequality as separate and apart from each other, Justice Powell rendered Black women, juridically speaking, not quite women and doubly suspect. They are suspect under strict scrutiny (because remedial efforts directed at Black people trigger strict scrutiny) and suspect as beneficiaries of intermediate scrutiny (because they are not, like the women at the center of Justice Powell’s analysis, white).

Interrogating how Black women figure into the Bakke passage raises broader concerns about tiers-of-scrutiny and equal-protection doctrine. One concern is whether, under existing doctrine, remedial projects that intersectionally target race and gender should be subject to intermediate scrutiny rather than strict scrutiny, as we argue they should. Another is whether equal-protection law is already an intersectional project, albeit an unmarked one, that privileges the intersectional identities of white antidiscrimination claimants. Still a third concern is that the core analytical structure of equal-protection doctrine (separate tiers of scrutiny for separate dimensions of identity) sits in tension with the concept of intersectionality. While we have engaged each of these issues and discussed the implications of our analysis for equal-protection law, the ideas we have expressed are decidedly provisional and intended to generate further debate. Such a debate is warranted, we think, because three decades after the publication of Demarginalizing, equal-protection scholarship and teaching has yet to take an intersectional turn. We do not mean by this that scholars should take our view of the intersectional issues this Essay describes. Rather, our modest invitation is to the scholars, judges, and lawyers who routinely navigate an equal-protection landscapestructured around compartmentalized tiers of power and identity: ask the intersectionality question.

Devon W. Carbado is the Honorable Harry Pregerson Professor of Law, UCLA School of Law, and the William H. Neukom Fellows Research Chair in Diversity and Law, the American Bar Foundation (2018-19). Kimberlé W. Crenshaw is the Promise Institute Chair in Human Rights and Distinguished Professor of Law, UCLA School of Law, and the Isidor and Seville Sulzbacher Professor of Law, Columbia Law School. For conversations about and comments on this Essay, we thank Cheryl Harris, Luke Harris, Robert Goldstein, Melissa Murray, Richard Re, Russell Robinson, the participants at the African American Policy Forum’s Social Justice Writers Retreat and UC Davis’s 2019 symposium on Bakke. UCLA School of Law, Columbia Law School, the American Bar Foundation, and the Bunche Center for African American Studies provided funding for this Essay. Matthew Gardner and Michael Kramer provided invaluable research assistance. All errors are our own.