Identity as Proxy

Identity as Proxy

As presently constructed, equal protection doctrine is an identity-based jurisprudence, meaning that the level of scrutiny applied to an alleged act of discrimination turns on the identity category at issue. In that sense, equal protection relies on identity as a proxy, standing in to signify the types of discrimination we find most troubling.

Equal protection’s current use of identity as proxy leads to a number of problems, including difficulties in defining identity categories; the tendency to privilege a dominant-identity narrative; failure to distinguish among the experiences of subgroups within larger identity categories; and psychological and emotional harm that can result from being forced to identify in a particular way to lay claim to legal protection. Moreover, because the Court’s identity-as-proxy juris­prudence relies on superficial notions of identity to fulfill a substantive commitment to equality, it is susceptible to co-option by majority groups.

This Essay aims to engage readers in a thought experiment, to envision what equal protection doctrine might look like if it were structured to reflect the values identity is intended to serve without explicitly invoking identity categories as a way to delineate permissible and impermissible forms of discrimination. In doing so, it aims to incorporate directly into equal protection jurisprudence the notion that identities like race and gender are not merely a collection of individual traits, but the product of structural forces that create and maintain subordination. Under the “value-based” approach proposed herein, the primary concern of equal protection is not to eliminate differential treatment, but instead to deconstruct status hierarchies. Therefore, rather than applying heightened scrutiny to government actions based on race or gender, it applies heightened scrutiny to government actions that have the effect of perpetuating or exacerbating a history of discrimination or that frustrate access to the political process.

The clearest impact of such a model would be in the context of affirmative action, where a majority plaintiff could no longer simply claim discrimination on the basis of race. Yet, the potential of a value-based model extends to other contexts as well—for example, challenges to voter identification laws, in which political exclusion would displace discriminatory intent and disparate impact as the relevant measure for analysis; and the treatment of pregnant women, in which discrimi­nation on the basis of pregnancy would no longer have to align with gender to receive heightened scrutiny.

This shift has several advantages: It allows the law to make important distinctions between groups and within groups; it alleviates the need for comparative treatment and solutions that favor taking from all over giving to some; it is less likely to generate identity-based harms; it is fact-driven rather than identity-driven and thus better suited to the judicial function; and it serves an important rhetorical function by changing the nature of rights discourse.


    1. Doctrinal Consequences of the Categorical Model
    2. Individual and Societal Effects of the Categorical Model
    1. Defining the Value-Based Approach
    2. Impact of the Value-Based Approach
    3. Operationalizing the Value-Based Approach
    1. Advantages of a Value-Based Approach
    2. Addressing Critiques of a Value-Based Approach




“[R]ights are only as strong as the so-called natural . . . identities by which they are upheld, and ultimately as strong or weak as the socially constructed legal definitions assigned to those natural identities.” 1 Marcia Alesan Dawkins, Clearly Invisible: Racial Passing and the Color of Cultural Identity 155 (2012).

As presently constructed, equal protection doctrine is an identity-based jurisprudence. 2 See Jessica Knouse, From Identity Politics to Ideology Politics, 2009 Utah L. Rev. 749, 764, 770–85 (describing how “[f]rom a conceptual perspective . . . the doctrine has remained static, always structured entirely around identity groups” and arguing this focus on identity groups renders it “ill-equipped to address subordination arising from any other source”). The degree of protection afforded by the Equal Protection Clause often turns on the identity trait—such as race or gender—implicated by an alleged act of discrimination. 3 The level of scrutiny applied to an alleged instance of discrimination turns on whether a fundamental right or suspect classification is involved. See Plyler v. Doe, 457 U.S. 202, 216–17 (1982) (explaining application of heightened scrutiny turns on whether classification at issue infringes on fundamental right or disadvantages suspect class). A “suspect class” is, in essence, a group of individuals sharing an identity trait—for example, race or gender—found to satisfy certain doctrinally specified criteria. See infra note 34 (listing currently recognized suspect classes). In those instances, the identity trait stands in to represent a set of assumptions about the group it describes: Because that group has been subjected to discrimination in the past or politically marginalized, or because its identifying characteristic is irrelevant to its members’ ability to contribute to or participate in society, the law is particularly sensitive to state action that targets such groups. 4 See Susannah W. Pollvogt, Beyond Suspect Classifications, 16 U. Pa. J. Const. L. 739, 742 (2014) (noting suspect classification analysis asks if group “(1) constitutes a discrete and insular minority; (2) has suffered a history of discrimination; (3) is politically powerless; (4) is defined by an immutable trait; and (5) is defined by a trait that is generally irrelevant to one’s ability to function in society”); Brent E. Simmons, Reconsidering Strict Scrutiny of Affirmative Action, 2 Mich. J. Race & L. 51, 94 (1996) (explaining degree of judicial scrutiny turns on claimant’s “race, gender, or physical abilities”). Section II.A describes in further detail the doctrinal framework for determining whether a class should be deemed suspect.

There is nothing in the Equal Protection Clause itself that suggests the need to use identity as a filter to analyze whether the law is being applied equally. 5 Section 1 of the Fourteenth Amendment provides:

“No state shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any state deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws.”

U.S. Const. amend. XIV, § 1.
Yet most of the equal protection literature does not question this basic premise. 6 Sonu Bedi, Beyond Race, Sex, and Sexual Orientation: Legal Equality Without Identity 32 (2013) (“Although critiques of identity are well established, scholars have largely ignored their relationship to constitutional jurisprudence.”). The problem with a model that subsumes a substantive inquiry regarding equality within the notion of identity is that it risks the distortion of identity and also the possibility that the proxy becomes an end in and of itself. In other words, the purpose identity is meant to serve may become lost in the idea that any classification based on identity is undesirable, regardless of the nature of the “discrimi­nation” at issue. Thus identity may be—and arguably has been—rede­ployed to satisfy different ends than originally intended: 7 Compare infra note 26 (describing historical roots of Equal Protection Clause), with infra note 88 and accompanying text (describing successful use of equal protection by white plaintiffs). transitioning from doctrinal shorthand for marginalization to a superficial marker that is readily associated with inequality regardless of the context.

This Essay suggests that, within equal protection doctrine, identity is important not as an end within itself, but because of what it represents. 8 Reva Siegel has observed that “[t]o this day, equal protection law remains unclear about the nature of the harm it is rectifying and the values it is vindicating.” Reva B. Siegel, Equality Talk: Antisubordination and Anticlassification Values in Constitutional Struggles over Brown, 117 Harv. L. Rev. 1470, 1546 (2004). In contrast, this Essay proposes an equal protection framework that not only would be clearer in its articulation of its underlying values, but also would be organized more directly around the fulfillment of those values. In other words, the law is more skeptical of discrimination based on race or gender not because of some abstract set of beliefs about the category, but because of the role it has played in obstructing access to the political process and in denying state-provided benefits and protections. Discrimination on the basis of race or gender is perceived as more pernicious—yet a description of the category itself, without any further elaboration on its relation to others or to society, does not tell us why. Nor does the prohibition of discrimination on the basis of identity necessarily ensure that the most malicious forms of discrimination will be eradicated.

If one is particularly skeptical of state laws or actions targeting individuals on the basis of their race or gender, it very well may be because those categories have been constructed as subordinate and marginalized in their ability to counteract such subordination. Concerns regarding subordination are structural and focused on the forces that act upon individuals; these forces play a role in the very creation of identity itself. 9 See Richard T. Ford, Race as Culture? Why Not?, 47 UCLA L. Rev. 1803, 1805 (2000) [hereinafter Ford, Race as Culture] (“Racism must not be understood as a set of practices that targets a group because of some preexisting characteristic of its members, but instead as a set of practices that establishes racial hierarchy and assigns individuals to distinctive statuses within that hierarchy.”); see also Richard Thompson Ford, Racial Culture 122 (2005) [hereinafter Ford, Racial Culture] (describing racial identity as “necessarily a product of racial power”); Ian Haney López, White by Law: The Legal Construction of Race 9 (rev. & updated 10th anniversary ed. 2006) [hereinafter López, White by Law] (“[R]ace is not an independent given on which the law acts, but rather a social construction at least in part fashioned by law.”); Khiara M. Bridges, The Dangerous Law of Biological Race, 82 Fordham L. Rev. 21, 38 (2013) (“[T]he idea of race does not produce racism; instead, racism produces the idea of race.”); Jayne Chong-Soon Lee, Navigating the Topology of Race, 46 Stan. L. Rev. 747, 750 (1994) [hereinafter Lee, Topology of Race] (“Current statutory and constitutional doctrine presupposes that the law merely reflects or recognizes race’s independent ‘reality.’”). Thus, in using identity as its organizing principle, law is confined to operating within the very structures that subordinate and is similarly confined to focusing on the product rather than the cause of inequality.

The logic of the current doctrinal framework—which I refer to herein as the “categorical model,” due to its reliance on identity categories—and its formula for identifying suspect classifications assumes that by protecting against actions based on identity, larger structural concerns such as historical subordination and political marginalization 10 See infra notes 144–145 and accompanying text (discussing structural concerns underlying equal protection doctrine). will be addressed. However, the concerns presented by identity are distinct, and sometimes at odds with 11 For example, while it may be in the individual’s interest to articulate her identity in a specific fashion, her desired articulation may conflict with the need—doctrinal or otherwise—to define identity in a broader, socially ascribed way. See Ford, Race as Culture, supra note 9, at 1807–08 (describing how identity politics—and recognition of specific social identities—are sometimes at odds with goals of antisubordination, because former is focused on object of discrimination while latter is focused on structural phenomena that lead to such discrimination). —or a product of 12 Id. at 1808 (“[S]ubjective self-identification is necessarily a product of social discourse—including the oppressive discourses of the social hierarchies of race.”); see also Ford, Racial Culture, supra note 9, at 20 (“[R]acial and analogous social identities are never autonomously adopted or intrinsic; they are always, at least in part, the effect of . . . social regulation.”). —the structural. Moreover, as described in this Essay, there are a number of other negative effects, both within and outside equal protection doctrine, that follow from using identity as a proxy to represent or vindicate other substantive concerns.

One way to reclaim the meaning of equal protection as it relates to equality of status and destabilizing social hierarchies, 13 One might interpret the Equal Protection Clause to provide not mere equality of treatment, but rather an affirmative mandate that existing hierarchies be deconstructed. See J.M. Balkin, The Constitution of Status, 106 Yale L.J. 2313, 2343 (1997) (“[T]he Constitution does more than simply provide fair ground rules for cultural struggle. It also actively intervenes in some status hierarchies and requires that they be dismantled, or at the very least, that the support of law be withdrawn from them. The Constitution . . . demand[s] for equality of social status . . . .”). therefore, may be to supplant identity with the substantive goals it is intended to vindicate. Under this approach, the primary inquiry is not whether all identity groups are treated the same, but whether the alleged discrimination affects the claimants in a way that exacerbates their subordination. In other words, the goal is not to eliminate all differential treatment, but to destabilize status hierarchies and effectively counter forces giving rise to subordination. 14 The model suggested herein is similar in some ways to Martha Fineman’s vulnerability theory, but it also has some significant differences. Like vulnerability theory, it is skeptical of frameworks based on identity and yearns for a more “substantive vision of equality.” Martha Albertson Fineman, The Vulnerable Subject: Anchoring Equality in the Human Condition, 20 Yale J.L. & Feminism 1, 1 (2008). The two also share the belief that the state has a greater and more affirmative role to play in countering subordination and that formal equality has proven inadequate in this regard. See id. at 3 (noting formal equality is “weak in its ability to address and correct the disparities in economic and social wellbeing among various groups in our society”). Yet, vulnerability theory assumes a universalist approach that is not part of the value-based model. See id. at 1 (describing vulnerability as “universal and constant, inherent in the human condition”). The model proposed here does not view all groups or individuals as equally vulnerable. Although it counsels against the use of identity as proxy, it is still concerned with equality and remedying the important differences in treatment that certain groups and individuals receive. Rather than removing the focus from discrimination, it aims to reconfigure the metric by which permissible and impermissible forms of discrimination are defined.

The switch to such a model—which I refer to as herein as the “value-based” approach—need not present a hard break with the current doctrinal framework. This is in part because some of the criteria used to define suspect classifications—past historical discrimination and political powerlessness—reflect these values. While there may be other factors indicative of subordination (and which could therefore play a part in the below inquiry), this Essay focuses on these factors because of their recognition and familiarity under current doctrine, 15 Because the value-based approach is rooted in normative values underlying existing jurisprudence, such a model may gain more traction than the persistent call from progressives to elevate new categories to the status of suspect classification or to alter the application of existing identity categories within the current framework. See Katie R. Eyer, Ideological Drift and the Forgotten History of Intent, 51 Harv. C.R.-C.L. L. Rev. (forthcoming 2015) (manuscript at 85) [hereinafter Eyer, Ideological Drift] (on file with the Columbia Law Review) (arguing normative redeployment of existing constitutional devices may be more effective than direct resistance). and because they present a metric by which courts may assess subordination. In contrast, other factors currently used by the doctrine to single out certain identity categories for heightened scrutiny, such as immutability and trait irrelevance, are less relevant under the value-based approach. This is because those factors are focused on the trait itself and not on the structural forces that create inequality. 16 As Khiara Bridges has explained, the Court’s tendency to focus on group-based traits implies a biological rooting and places fault for certain conditions, whether directly or indirectly, on the individual rather than on society, where fault should lie. See Bridges, supra note 9, at 24 (“Biological race is also dangerous because it argues that racial minorities—namely, black people—remain subordinated because of genetic inferiority, and not because of structural and institutional processes.”). The fact that a trait is irrelevant to the context at hand may suggest underlying prejudice or bias; however, there are many traits that might be irrelevant to the task at hand and yet not signify a deeper pattern of subordination. Like any other law, a law drawing a distinction on the basis of an irrelevant trait—for example, physical attractiveness or one’s number of tattoos—would be subjected to rational basis review. What is at stake in the choice between the categorical model and the value-based model is not whether the law is constitutional or not—if discrimination on the basis of an irrelevant trait is irrational, it will be unconstitutional under any framework—but whether a reviewing court should apply heightened scrutiny.

To illustrate, one might think of contemporary equal protection doctrine as consisting of a two-step analysis: First, did the government discriminate against the plaintiff based on his or her inclusion in X category (where X category serves as a proxy for certain equal protection values); and second, applying the level of scrutiny attached to X category, was the government action justified? Rather than focusing the inquiry on categorical judgments, a value-based approach might supplant the first question with the following: Does the challenged government action deny access to state-provided benefits or protection in ways that perpetuate or exacerbate historical patterns of discrimination against a particular group (as defined by the plaintiff or plaintiffs)? Alternatively, does the challenged government action create or maintain barriers to political access for a particular group? Where the court can answer one of these questions in the affirmative, heightened scrutiny should apply. 17 This Essay does not attempt to resolve specifically what level of scrutiny would apply, although that issue is discussed briefly below. See infra text accompanying notes 172–176. Instead, this Essay focuses primarily on the shift made with regard to the first part of the equal protection inquiry.

So, in the context of laws requiring potential voters to present identification before casting a vote, for example, the primary inquiry would be not whether the law was racially motivated (which is difficult to prove), or even whether the law has a disparate racial impact, 18 Although the factors listed here as part of the value-based approach are similar to disparate impact analysis in that they focus on the effects of discrimination, a disparate impact approach can be distinguished in that it is focused on a quantitative showing of how discrimination affects a particular identity group, rather than a qualitative analysis of how an alleged act of discrimination affects a particular group of plaintiffs. See, e.g., Washington v. Davis, 426 U.S. 229, 237 (1976) (evaluating as disproportionate impact argument fact that four times as many blacks as whites failed written personnel test). but rather whether the law results in political exclusion or obstructs access to the political process. In the case of a pregnant woman who claims she has been wrongly terminated from her government job, the application of heightened scrutiny would turn not on her ability to prove that she was terminated as a result of her gender, but on whether her termination furthered or worsened patterns of discrimination against pregnant women. Admittedly, the greatest impact of such a shift would be in the context of affirmative action, where mere differential treatment on the basis of race or gender would no longer be sufficient to sustain an equal protection claim.

To provide a recent example, a group titled Students for Fair Admissions filed a complaint in federal district court last November claiming that Harvard College’s undergraduate admissions policy discri­minates against Asian Americans. 19 Complaint at 34–37, 43–57, Students for Fair Admissions, Inc. v. President & Fellows of Harvard Coll., No. 1:14-CV-14176, 2014 WL 6241935 (D. Mass. Nov. 17, 2014) (alleging Harvard has intentionally discriminated, and continues to intentionally discriminate, against Asian American applicants by applying “informal ceiling” on Asian American enrollment, despite their high rates of academic achievement). This lawsuit could be seen as merely another iteration of familiar debates regarding affirmative action, pitting those who favor the use of racial classifications against those who advance a colorblind approach. 20 See, e.g., John E. Morrison, Colorblindness, Individuality, and Merit: An Analysis of the Rhetoric Against Affirmative Action, 79 Iowa L. Rev. 313, 314 (1994) (describing how opponents of affirmative action insist it is “not colorblind, because it intentionally invokes racial classifications”); Peter H. Schuck, Affirmative Action: Past, Present, and Future, 20 Yale L. & Pol’y Rev. 1, 6 (2002) (describing how proponents of affirmative action view it as necessary to achieve equal opportunity while opponents view minority preferences and quotas as obstacles to equal opportunity). Given its posture, however, the case raises interesting questions about how we think about other minority-race plaintiffs—not just white plaintiffs—alleging discrimination in the affirmative action context. The current approach used to analyze such a claim would suggest that the fates of Asian Americans and African Americans—and perhaps even whites—in this context are inextricably linked; racial preferences are either permissible, or they are not. In contrast, the model suggested herein offers a substantive way to articulate the difference between the experiences of different applicants of color and therefore justify a policy that may treat them distinctively.

The value-based approach clearly runs counter to anticlassification theory—the dominant approach under current equal protection jurisprudence—which discourages or disfavors the use of identity-based classifications across all contexts. 21 See Kimberly Jenkins Robinson, The Constitutional Future of Race-Neutral Efforts to Achieve Diversity and Avoid Racial Isolation in Elementary and Secondary Schools, 50 B.C. L. Rev. 277, 315 (2009) (“The Court’s current approach to equal protection, which has been labeled an antidiscrimination, anticlassification, or colorblind approach, emphasizes the impropriety of government use of racial classifications.”). In that sense, the value-based approach shares antisubordination theory’s normative view that not all uses of identity-based classifications are undesirable; it is only those uses of identity that serve to subordinate or oppress a particular group that are rendered unconstitutional. 22 See infra notes 219–220 and accompanying text (describing antisubordination theory and its corresponding conception of equal protection as concerned primarily with hierarchy rather than mere use of race). Where the value-based model diverges from antisubordination theory, however, is in the belief that identity is an appropriate or the most effective vehicle by which to reach the normative ends that antisubordination theory aims to achieve. Although the use of identity need not inevitably lead one to an anticlassification approach, it is more easily manipulated toward those ends than a model that builds certain normative assumptions or substantive inquiries into the analysis itself. Moreover, it is more difficult to make arguments advancing colorblindness—the idea that race is irrelevant—in the context of a model that does not turn on identity, or on color.

Perhaps more important—and the greatest distinction between the model suggested herein and other theories that have dominated equal protection debates—are the discursive benefits to be gained by such an approach. By its very essence, the value-based model precludes superficial analysis of discrimination. And, regardless of outcome, the rhetoric used to discuss discrimination matters. It is a very different endeavor to talk about discrimination in terms of broad identity categories, and whether or not we should tolerate distinctions made on the basis of such categories, than it is to engage in a substantive discussion about how the alleged claim of discrimination impacts access for the group at issue and how it may perpetuate earlier denials of access.

By suggesting a new approach to equal protection, this Essay does not purport to offer a silver-bullet solution to the various problems plaguing equal protection today, nor does it always claim to be outcome determinative. Rather, it hopes to engage readers in a thought experiment to explore what equal protection might look like if it were structured to reflect the values identity is intended to serve without explicitly invoking identity categories as a way to delineate permissible and impermissible forms of discrimination.

In Part I, the Essay describes both the doctrinal and extralegal consequences that flow from identity-as-proxy jurisprudence. Part II describes how equal protection analysis would change if it were guided not by categories of identity, but instead by substantive equal protection values underlying the current use of identity. Part III addresses the advantages and possible critiques of an approach to equal protection that no longer uses identity as a proxy for such values.

I. Identity as Proxy

Under the identity-as-proxy paradigm of equal protection, the level of scrutiny applied to alleged claims of discrimination turns on the identity category at issue. 23 As a result, some claimants alleging discrimination have attempted to align themselves with a class that the Court has already deemed worthy of special protection. For a thorough discussion of the feminist movement’s attempts to analogize itself to the civil rights movement and compare the treatment of gender to the treatment of race, see generally Serena Mayeri, Reasoning from Race: Feminism, Law, and the Civil Rights Revolution (2011). Other groups—including sexual-orientation minorities—have sought to make the most of rational basis review, forgoing the seemingly futile task of seeking status as a suspect classification. The Court seems to have adopted a similar strategy: As recently as this past term, the Court provided additional protection to sexual-orientation minorities without explicitly applying a higher level of scrutiny. See Obergefell v. Hodges, 135 S. Ct. 2584, 2607–08 (2015) (holding Fourteenth Amendment requires states to issue marriage licenses to same-sex couples and to recognize same-sex marriages lawfully per­formed in another state, but declining to specify applicable level of scrutiny); cf. id. at *32 (Roberts, C.J., dissenting) (applying rational basis review and finding no equal protection violation “because distinguishing between opposite-sex and same-sex couples is rationally related to the States’ ‘legitimate state interest’ in ‘preserving the traditional institution of marriage’”(quoting Lawrence v. Texas, 539 U.S. 558, 585 (2003) (O’Connor, J., concurring in judgment))). The Court has concluded that some cate­gories, like race, are deserving of the highest level of scrutiny, 24 See Korematsu v. United States, 323 U.S. 214, 216 (1944) (explaining restrictions curtailing rights of a racial group are subject to “most rigid scrutiny”). meaning that the government bears a heavier burden to justify classifications made on the basis of race. In Bakke, Justice Powell suggested that the differential treatment of race is sui generis, 25 See infra note 48 (describing Justice Powell’s effort in Bakke to distance treatment of race from “discrete and insular” minority reasoning in Carolene Products). perhaps attributable to the fact that the Reconstruction Amendments were intended to “provide federal protection for newly ‘freed’ [African] Americans.” 26 Rhonda V. Magee Andrews, The Third Reconstruction: An Alternative to Race Consciousness and Colorblindness in Post-Slavery America, 54 Ala. L. Rev. 483, 490–91 (2003); see also William N. Eskridge, Jr., Some Effects of Identity-Based Social Movements on Constitutional Law in the Twentieth Century, 100 Mich. L. Rev. 2062, 2069 (2002) [hereinafter Eskridge, Some Effects] (explaining Reconstruction Amendments’ purpose was to protect people of color against oppression by state governments). Other identity categories, such as gender, are entitled to intermediate scrutiny, 27 See Craig v. Boren, 429 U.S. 190, 197 (1976) (“[C]lassifications by gender must serve important governmental objectives and must be substantially related to achievement of those objectives.”); cf. Frontiero v. Richardson, 411 U.S. 677, 684–88 (1973) (Brennan, J., plurality opinion) (explaining justification for applying heightened scrutiny to classifications based on sex, including “long and unfortunate history of sex discrimi­nation,” discrimination against women in “political arena,” immutability of sex, and “sex characteristic frequently bears no relation to ability to perform or contribute to society”). and all others are subject to rational basis review. 28 Erwin Chemerinsky, Constitutional Law 717 (4th ed. 2013) (explaining rational basis review applies to equal protection claims where intermediate and strict scrutiny are not warranted). A general premise of the equal protection hierarchy, and a significant part of the rationale for treating these categories differently, is that certain classifications are less likely to be rooted in any supportable basis and thus more likely to signal prejudice or bias. 29 See Plyler v. Doe, 457 U.S. 202, 216 n.14 (1982) (“Some classifications are more likely . . . to reflect deep-seated prejudice . . . . Legislation predicated on such prejudice is easily recognized as incompatible with the constitutional understanding that each person is . . . judged individually and . . . entitled to equal justice under the law. Classifications . . . treated as suspect tend to be irrelevant to any proper legislative goal.”). The greater the likelihood that a classification is rooted in prejudice or stereotype, the stronger justification the Court will demand from the government to use such classifications. 30 For example, one reason for treating gender classifications with an intermediate level of scrutiny is that some perceive “real” differences between men and women, justifying differential treatment under certain circumstances. In contrast, there are no “real” differences that can justify differential treatment among racial groups. See, e.g., Anita K. Blair, The Equal Protection Clause and Single-Sex Public Education: United States v. Virginia and Virginia Military Institute, 6 Seton Hall Const. L.J. 999, 1001 (1996) (arguing strict scrutiny is inappropriate for gender classifications because unlike racial classification context, “there are real differences between men and women”). Another rationale, grounded in Carolene Products’s footnote four, 31 United States v. Carolene Prods. Co., 304 U.S. 144, 152–53 n.4 (1938). For further discussion of Carolene Products’s footnote four and its role in defining suspect classifica­tions, see infra notes 48–49, 146, 150–157 and accompanying text. is that certain groups have historically been rendered politically powerless and thus require “extraordinary protection from the majoritarian political process.” 32 San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 28 (1973); see Carolene Prods., 304 U.S. at 152–53 n.4 (1938) (“[P]rejudice against discrete and insular minorities may be a special condition, which tends seriously to curtail the operation of those political processes ordinarily to be relied upon to protect minorities, and which may call for a correspondingly more searching judicial inquiry.”).

By suggesting a focus on the treatment of “discrete and insular minorities,” the Carolene Products footnote also provided support for an equal protection jurisprudence driven by suspect classifications, 33 See infra notes 150–151 and accompanying text (describing possible reasoning set forth by footnote four for treating certain groups with heightened scrutiny). which are typically organized around a defining identity trait. 34 See Anthony R. Enriquez, Assuming Responsibility for Who You Are: The Right to Choose “Immutable” Identity Characteristics, 88 N.Y.U. L. Rev. 373, 377–80 (2013) (explaining changing understandings of identity, and immutability of certain identity characteristics, influence degree to which courts are willing to view such identities as suspect classifications); William N. Eskridge, Jr., Channeling: Identity-Based Social Movements and Public Law, 150 U. Pa. L. Rev. 419, 507 (2001) (explaining how other identity groups, such as women and gays and lesbians, sought to establish their defining traits as suspect in wake of race’s recognition as suspect classification); Mark Strasser, Equal Protection, Same-Sex Marriage, and Classifying on the Basis of Sex, 38 Pepp. L. Rev. 1021, 1025 (2011) (providing examples of race, religion, nationality, and alienage as “suspect” classes and gender and illegitimacy as “quasi-suspect” classes, given lower level of scrutiny applied to latter categories). Through its cases, the Court later identified factors used to define suspect classes, including a history of past discrimination, political powerlessness, immutability, and relevance of the group’s defining trait to the group’s ability to contribute to or participate in society. 35 For a more detailed discussion of the factors used to define suspect classifications, see infra notes 153–159 and accompanying text. The irony of this framework, as others have pointed out, is the Court’s failure to recognize that by constructing doctrine around identity—and determining when these factors justify the designation of suspect classification—it has played an important role in defining identity itself. 36 In other words, when the Court holds that a group lacks political power, or is defined by a trait that is immutable, it makes a powerful statement about the group’s defining trait that reaches beyond doctrine. See, e.g., Lee, Topology of Race, supra note 9, at 774 (“The Court purports to merely recognize, not construct, definitions of race.”); see also id. at 775–76 (discussing Court’s conflation of biology and race). This is true of other state actors as well. Cf. Cristina M. Rodriguez, Against Individualized Consideration, 83 Ind. L.J. 1405, 1417 (2008) (explaining “school admissions officers do not just define Latino for the sake of admissions; they socially construct the category”). In other contexts, including gender, the Court has been more explicit in giving meaning to categorical definitions. See, e.g., Muller v. Oregon, 208 U.S. 412, 422 (1908) (“The two sexes differ in structure of body, in the functions to be performed by each, in the amount of physical strength, in the capacity for long-continued labor . . . the self-reliance which enables one to assert full rights, and in the capacity to maintain the struggle for subsistence.”).

Equal protection’s reliance on identity is understandable in light of its origins. The institution of slavery, from which the Reconstruction Amendments arose, and more modern legal regimes enforcing or permitting overt discrimination, used the law to explicitly oppress individuals sharing certain identity traits. These regimes gave rise to group-based identities, which, in turn, gave rise to group-based social movements. 37 Eskridge, Some Effects, supra note 26, at 2070 (“[T]he existence of pervasive state exclusion, discrimination, and violence was a necessary factor in the formation of the civil rights, women’s, and gay peoples’ mass social movements. For each movement, the law was critical in creating a social identity as ‘colored,’ female, or ‘homosexual’ . . . .”). The group-based exclusionary nature of the law solidified the social identities of these movements. 38 Id. (explaining law’s critical role in creating social identities such as “colored,” female, or “homosexual”); see also Martha Minow, Not Only for Myself 31 (1997) (“The salience of group-based identities in particular also emerges from social movements organized against group-based oppression or discrimination.”). Given the centrality of identity to both legal and social frameworks, it was logical for those aiming to change the law to focus on altering the perception of certain identities 39 As William Eskridge has explained, the initial purpose of civil rights reform was to “persuade the mainstream that the group-defining trait (color, sex, sexual or gender orientation) was a tolerable variation from the norm and that stigmatized persons ought to have minimal rights.” Eskridge, Some Effects, supra note 26, at 2071. and, later, on rendering identity an illegitimate basis for legal differentiation. 40 See id. (describing early approach to civil rights reform as refuting “inferiority of the group’s defining trait: there is no material difference between blacks and whites, except those created by society and law; women can perform any social role that men can; gay is as good as straight”); Vicki Lens, Supreme Court Narratives on Equality and Gender Discrimination in Employment: 1971–2002, 10 Cardozo Women’s L.J. 501, 520 (2004) (describing formal equality as “emphasiz[ing] men and women’s sameness” and “wary of any gender related special accommodations”).

In today’s world, the benefits of using identity as a framing device for equal protection jurisprudence are less clear. 41 Cf. Angela P. Harris, Equality Trouble: Sameness and Difference in Twentieth-Century Race Law, 88 Calif. L. Rev. 1923, 2004 (2000) (contending neither arguments regarding “sameness” between members of different races nor arguments relying on their inherent “differences” have succeeded in achieving racial justice). Moreover, as Russell Robinson has explained, a paradox may exist for those groups that have achieved suspect classification status. Classifications, Robinson explains, can often serve as a “gate keeping device,” all but foreclosing relief once the Court has decided that a group does meet the criteria to qualify as a suspect class. Russell K. Robinson, Unequal Protection, 67 Stan. L. Rev. (forthcoming 2015) (manuscript at 2) [hereinafter Robinson, Unequal Protection], (on file with the Columbia Law Review); see also id. (manuscript at 18) (“As a practical matter, the requirement of a race classification effectively kills the vast majority of race claims brought by people of color.”). In contrast, a more free-flowing inquiry, applied to groups that have not yet been afforded a spot in the equal protection hierarchy, may actually better position such groups to benefit from the existing framework. For example, because they have not yet been characterized as a suspect class, Robinson argues, LGBT people have been able to benefit from certain aspects of the doctrine—like animus—but have not been saddled with many of its negative consequences. Id. (manuscript at 16–18). While a focus on identity may be perceived by some as “necessary in order for a legally disfavored group to make the transition to formal legal equality,” 42 Katie R. Eyer, Have We Arrived Yet? LGBT Rights and the Limits of Formal Equality, 19 Law & Sexuality 159, 164 (2010) [hereinafter Eyer, Have We Arrived Yet] (arguing model based on identity may be necessary to ensure identity groups are treated equally as formal legal matter and that such a step is required before further progress is made toward equality). There are certainly groups that have not yet attained formal legal equality; yet, as Robinson discusses, it is not clear that the suspect classification framework would serve as a net benefit for such groups. See supra note 41 and accompanying text (describing negative effects that may flow from suspect classification status). it has less utility in a regime in which covert and structural discrimination are just as, if not more, pervasive as more overt forms of discrimination. 43 See Eyer, Have We Arrived Yet, supra note 42, at 160–61 (noting greatest obstacle facing LGBT movement in its fight for formal equality is “shift to a regime in which covert or structural discrimination will take the place of overtly expressed bias”); Ian Haney López, Intentional Blindness, 87 N.Y.U. L. Rev. 1779, 1783 (2012) [hereinafter López, Intentional Blindness] (noting modern racial discrimination does not rely on overt racial classifications “since instances of frank and open mistreatment are now virtually nonexistent”); Lauren Sudeall Lucas, A Dilemma of Doctrinal Design: Rights, Identity, and Work-Family Conflict, 8 FIU L. Rev. 379, 381 (2013), reprinted in Women in the Law (Thomson Reuters 2014) (suggesting rights-driven framework may be better suited to first-generation questions of exclusion and overt discrimination than to current second-generation questions arising from work-family conflict); Susan Sturm, Second Generation Employment Discrimination: A Structural Approach, 101 Colum. L. Rev. 458, 460 (2001) (distinguishing between “first generation” patterns of bias, which involve overt acts of discrimination, and “second generation” claims of discrimination, which “involve social practices and patterns of interaction among groups within the workplace that, over time, exclude nondominant groups”); see also Richard Thompson Ford, Rights Gone Wrong 242–43 (2011) (“Civil rights are remarkably effective against overt prejudice perpetrated by identifiable bigots. But they have proven impotent against today’s most severe social injustices, which involve covert and repressed prejudice or the innocent perpetuation of past prejudice.”). As Susan Sturm has observed, “[s]moking guns—the sign on the door that ‘Irish need not apply’ or the rejection explained by the comment that ‘this is no job for a woman’—are largely things of the past.” 44 Sturm, supra note 43, at 459–60 (arguing cognitive bias, unequal structures of decisionmaking, and discriminatory patterns of interaction persist even though deliberate discrimination has mostly subsided). Where discrimination is not overt, it is often not immediately apparent that identity is at issue; 45 The intent requirement only exacerbates this fact; because state actors are aware that discriminatory intent is required to find an equal protection violation, any savvy actor who wishes to discriminate (whether consciously or not) will be sure to provide an identity-neutral motivation for his or her actions. See David A. Strauss, Discriminatory Intent and the Taming of Brown, 56 U. Chi. L. Rev. 935, 953 (1989) [hereinafter Strauss, Discrimina­tory Intent] (discussing “cases in which the government was using a racial classification but, in contrast to the classic Jim Crow laws of Strauder or Plessy, was trying to conceal the fact that it was doing so”). under such circumstances, using identity as the trigger for heightened scrutiny may not serve as the most effective tool to recognize and provide redress for discrimination. 46 See Suzanne B. Goldberg, On Making Anti-Essentialist and Social Constructionist Arguments in Court, 81 Or. L. Rev. 629, 659–60 (2002) [hereinafter Goldberg, Anti-Essentialist] (“‘[N]atural’ or universal definitions of traits may not enable anti-discrimination protections to reach much of the non-explicit identity-based discrimination that occurs today . . . .”); see also Paul Butler, The Court Should Focus on Justice Rather than Rights, N.Y. Times: Room for Debate (July 19, 2013, 6:41 pm) [hereinafter Butler, Justice Rather than Rights],‌-era-over/‌the-court-should-focus-on-justice-‌rather-than-rights [] (“When, however, the gay civil rights movement evolves beyond basic discrimination issues, the court probably will not be as open to its claims.”).

This Part explores the consequences of equal protection doctrine’s current use of identity. Section I.A discusses how equal protection’s use of identity as proxy facilitates and supports doctrinal interpretations that frustrate antisubordination goals and renders the doctrine unable to account for the varied experiences of identity subgroups. Section I.B reveals the broader sphere of problems caused by the use of identity as proxy, including the negative effects it may have on individual identity and on the relationships between identity groups.

A. Doctrinal Consequences of the Categorical Model

Although not always directly linked to identity, certain doctrinal developments are facilitated by equal protection doctrine’s use of identity as proxy. Under the current model of equal protection, referred to as the categorical model herein, government action that targets certain categorical identities, such as race or gender, is viewed with heightened scrutiny. One underlying theory as to why certain identity groups should receive heightened scrutiny, grounded in Carolene Products’s footnote four and advocated by theorists like John Hart Ely, 47 See infra note 146 (describing Ely’s process-oriented approach to constitutional interpretation). stems from concern about the group’s ability to protect itself using the normal political processes or that the group is otherwise disadvantaged by virtue of its status as a “discrete and insular minority.” 48 See infra notes 150–159 and accompanying text (describing role of Carolene Products in defining suspect classification criteria). It must be noted that in other cases, the Court disavowed this rationale as applied to race. See, e.g., Regents of Univ. of Cal. v. Bakke, 438 U.S. 265, 290 (1978) (emphasizing discrete and insular minority rationale “has never been invoked in our decisions as a prerequisite to subjecting racial or ethnic distinctions to strict scrutiny”). In Bakke, the Court expressly noted that it had not held that:

“[D]iscreteness and insularity constitute necessary preconditions to a holding that a particular classification is invidious. These characteristics may be relevant in deciding whether or not to add new types of classifications to the list of “suspect” categories or whether a particular classification survives close examination. Racial and ethnic classifications, however, are subject to stringent examination without regard to these additional characteristics.”

Id. (citations omitted). This Essay explicitly rejects such a distinction, instead opting to apply the same substantive rationale to all forms of discrimination. The alternative approach set forth in Bakke lacks coherence and fails to articulate why such a distinction is necessary. See López, Intentional Blindness, supra note 43, at 1826 (noting, in Bakke, Justice Powell “had to contradict the Court’s precedents” to conclude that strict scrutiny should apply to all express uses of race).
Thus, the law should give special solicitude to distinctions made on the basis of the group’s defining identity trait. 49 Pollvogt and Yoshino have pointed out that the Court has engaged in post-hoc, reverse engineering to develop its explanations of why certain groups are subject to special solicitude (as opposed to applying a systematic methodology from the outset). See Pollvogt, supra note 4, at 742–43 (2014) (describing mechanisms Court has “devised” but that have been “questionable . . . in terms of prompting critical analysis of the dynamics of invidious discrimination”); Kenji Yoshino, The New Equal Protection, 124 Harv. L. Rev. 747, 755 (2011) (describing how Court “has fashioned a framework of tiered scrutiny” to analyze equal protection claims). Regardless of the origin for such justification, however, it appears to be widely accepted, based on the extent to which it has been applied going forward. See, for example, Attorney General Eric Holder’s characterization of the Supreme Court’s definition of a “suspect classification”:

“The Supreme Court has . . . rendered a number of decisions that set forth the criteria that should inform this and any other judgment as to whether heightened scrutiny applies: (1) whether the group in question has suffered a history of discrimination; (2) whether individuals ‘exhibit obvious, immutable, or distinguishing characteristics that define them as a discrete group’; (3) whether the group is a minority or politically powerless; and (4) whether the characteristics distinguishing the group have little relation to legitimate policy objectives or to an individual’s ‘ability to perform or contribute to society.'”

Press Release, Office of Pub. Affairs, Dep’t of Justice, Letter from the Attorney General to Congress on Litigation Involving the Defense of Marriage Act (Feb. 23, 2011), [] (last updated Sept. 15, 2014).
Applying that judgment to the group as a whole allows the category to serve as a type of doctrinal shorthand: By avoiding discrimination on the basis of the category, the law can (at least in theory) ensure that individuals within the group will not be further disadvantaged and that all will have equal access to state-provided benefits and protections.

One problem with this logic, however, is that there are variations within each category, and subcategories have varying levels of access—in large part due to oppression by other subcategories. For example, within the broad category of “race,” blacks and whites have historically experienced different levels of access and, within the former category, those with lighter skin color or recently immigrated parents have had a different experience from those with darker skin and those descended more directly from slave ancestors. 50 See Taunya Lovell Banks, Colorism: A Darker Shade of Pale, 47 UCLA L. Rev. 1705, 1709–10 (2000) (explaining phenomenon of “colorism,” discrimination against dark-skinned but not light-skinned blacks); infra notes 113–115 and accompanying text (describing varying treatment of these categories in university-admissions context). White women and women of color experience gender discrimination differently and cannot be encompassed within one singular definition of what it means to be a woman. 51 See infra notes 60–62 and accompanying text (challenging notion of “essential” woman’s experience and suggesting defining one singular female experience ignores experiences of women outside that narrative). Thus, allowing the category as a whole to stand in for the values it is intended to protect may therefore provide such protection only sporadically. It also distances equal protection analysis from the norma­tive values that animate it.

One can see, then, how equal protection’s focus on identity facilitates the elision of exclusionary discrimination and discrimination meant to foster inclusion or remedy past exclusion. In cases like Adarand Constructors, Inc. v. Pena 52 515 U.S. 200 (1995). and Regents of the University of California v. Bakke, 53 438 U.S. 265 (1978). the Supreme Court made clear that it would not distinguish between “invidious” and “benign” discrimination. 54 See Adarand, 515 U.S. at 227 (holding “all racial classifications . . . must be analyzed . . . under strict scrutiny” and “benign” racial classifications should not be held to different standards); Bakke, 438 U.S. at 294–95 (“Petitioner urges us to . . . hold that discrimination against members of the white ‘majority’ cannot be suspect if its purpose can be characterized as ‘benign.’ The clock of our liberties, however, cannot be turned back to 1868.”). In other words, the Court would treat any form of identity-based discrimination with the same level of scrutiny, regardless of its purpose or context (i.e., de jure segregation or affirmative action). If the triggering factor for equal protection analysis is discrimination on the basis of identity, it can logically follow—and indeed has, under the Court’s jurisprudence—that the simplest way to remedy such discrimination is to avoid any discrimination on the basis of identity. 55 See, e.g., Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 748 (2007) (“The way to stop discrimination on the basis of race is to stop discriminating on the basis of race.”). Under such an interpretation, the nature of the discrimination at issue is necessarily secondary and can more easily be rendered irrelevant.

Another consequence of the categorical model is that identity categories not yet recognized as suspect may attempt to analogize their experience to those categories, or classes, that have already achieved suspect classification status. 56 See Mayeri, supra note 23, at 233 (“Lawyers for gay and lesbian Americans, the disabled, the aged, and others rely on race and sex equality cases, comparing their own causes to those of civil rights and feminist predecessors.”). As Serena Mayeri discusses in detail in Reasoning from Race: Feminism, Law, and the Civil Rights Revolution, while the use of analogy can contribute to effective coalition building, it can also lead to conflict and to the appropriation of doctrinal concepts from one category to another, including those that may have negative ramifications. 57 See id. at 231–32 (explaining how analogizing sex equality to race equality resulted in transferal of related doctrinal elements, including irrelevance of differences between sexes and equation of benign and invidious classifications). For example, Mayeri describes how the feminist movement attempted to build on analogies to the civil rights movement and the fight for racial equality. 58 Id. As a result, as colorblindness began to dominate notions of racial equality, analogic reasoning dictated that differences between men and women must be rendered irrelevant to achieve equality, and affirmative measures would be precluded in the context of gender as well as race. 59 Id. Thus, analogies to race could be blamed for sexblindness and the dominance of formal equality as well as for the law’s inability to adequately recognize the unique experiences of women of color. 60 Id. at 232 (describing how race-sex analogies encouraged minimization of difference between the sexes and elided experiences of women of color). For example, as Kimberlé Crenshaw poignantly describes, these differences are particularly salient for battered women of color who are more likely to lack resources and other coping devices and face a greater degree of societal domination and discrimination; thus, they are more vulnerable to physical abuse and have a harder time recovering from its effects. Kimberle Crenshaw, Mapping the Margins: Intersectionality, Identity Politics, and Violence Against Women of Color, 43 Stan. L. Rev. 1241, 1245–50 (1991) [hereinafter Crenshaw, Mapping the Margins].

What makes identity appealing as a doctrinal sorting device is its ability to serve as shorthand for a set of values or ideals; 61 See Daniel Weinstock, Is ‘Identity’ a Danger to Democracy?, in Identity, Self-Determination and Secession 15, 20–21 (Igor Primoratz & Aleksandar Pavković eds., 2006) (arguing “identity arguments are just shorthand for arguments that could be more fully spelled out in terms of values and interests” and effects of framing arguments in terms of identity are different from those arising from arguments based on values and interests). to do so effectively, however, it must be defined in a monolithic way. The simplification of identity for purposes of doctrinal utilization often leads to essentialism: “the notion that a unitary, ‘essential’ women’s experience can be isolated and described independently of race, class, sexual orientation, and other realities of experience.” 62 Angela P. Harris, Race and Essentialism in Feminist Legal Theory, 42 Stan. L. Rev. 581, 585 (1990) [hereinafter Harris, Race and Essentialism]; see also Jessica A. Clarke, Against Immutability, 125 Yale L.J. (forthcoming 2015) (manuscript at 33), http:// (on file with the Columbia Law Review) (describing how “new immutability” that has emerged relies on demonstration that identity characteristic is “essential”).
As explained by Angela Harris in her seminal article, Race and Essentialism in Feminist Legal Theory, essentialism is attractive for several reasons. 63 Harris, Race and Essentialism, supra note 62, at 605–07. First, it is intellectually easy because it relies on dominant narratives and therefore requires little work to explore and highlight other experiences. 64 See id. at 589 (“Essentialism is intellectually convenient, and to a certain extent cognitively ingrained.”). Second, by fostering a narrative that is shared by all, essentialism can provide a sense of safety and belonging and reduce any sense of conflict. 65 Id. at 605–06 (discussing how “essentialism represents emotional safety” for women in “struggle against gender oppression”). Last, it can provide the basis for presenting a unified front against a mutual oppressor, facilitating a more powerful political or social movement. 66 Id. at 606–07 (explaining essentialist view that “women are women and we are all oppressed by men” and how in face of internal splintering, “solidarity reappears through the threat of a common enemy”). This aspect of essentialism was present in the conflict between the multiracial movement and the NAACP with regard to the creation of the multiracial status on the United States Census. Lauren Sudeall Lucas, Undoing Race? Reconciling Multiracial Identity with Equal Protection, 102 Calif. L. Rev. 1243, 1263 (2014) [hereinafter Lucas, Undoing Race].

While these aspects of essentialism make it a tempting option, Harris also articulates its dangers and ultimately advocates against its use. As Harris explains, essentialism silences those “who have traditionally been kept from speaking, or who have been ignored when they spoke.” 67 Harris, Race and Essentialism, supra note 62, at 585. In relying on a dominant narrative, essentialism “relegate[s] to the footnotes” contradictory or more nuanced examples and requires that all experiences be evaluated against one “true” norm. 68 Id. at 591–92; see also id. at 595 (“[B]y defining black women as ‘different,’ white women quietly become the norm, or pure, essential woman.”). Harris explains that essentialism can also give way to unconscious racism; in a racist society, the dominant storytellers construct the story such that they are prominently featured and others are not. 69 Id. at 589.

Essentialism is not entirely avoidable in any model that utilizes identity, whether used as proxy (as in the categorical model) or as a more individualized basis for a claim (as in the value-based model). 70 See Diana Fuss, Essentially Speaking: Feminism, Nature & Difference, at xii (1989) (suggesting that due to many different variations within essentialism, “we can only speak of essentialisms,” and constructionism is “sophisticated form of essentialism”). The primary difference between the categorical model and the value-based model with respect to essentialism is who is doing the essentializing, or where the power to essentialize is based. Under the current model, it is the courts that possess the power to define the relevant identity categories and who will fall within or outside of those categories; this is because identity is an organizing principle of the law that courts design and apply. In contrast, under the value-based model, the onus is on the plaintiffs to claim that they have been discriminated against for a particular reason, which may relate to their identity. 71 See infra note 293 and accompanying text (describing evidentiary burden that may accompany such substantive claim). In doing so, however, they have an opportunity to articulate and define that identity for purposes of the legal analysis and thus retain more power to construct their own narrative. 72 Cf. Rodriguez, supra note 36, at 1406–07 (arguing to “give state officials power to define the content of a racial category . . . undermines the integrity of the individual, the protection of which is the supposed rationale for . . . the Court’s consequent deep skepticism of race-conscious decision making”); id. at 1412 (“[T]he content of terms like Hispanic should be defined primarily by those who might bear that designation, in dialogue with other members of the group and the general culture, and not by the state.”).

The law’s current tendency to focus on one dominant narrative, 73 See Lani Guinier & Gerald Torres, The Miner’s Canary: Enlisting Race, Resisting Power, Transforming Democracy 35 (2003) (explaining “existing legal doctrine and theory . . . depend heavily on stock stories”). to the exclusion of others, may result in under or overprotection. As Suzanne Goldberg has observed, the use of essentialism in litigation may “undermine the ultimate aim of protecting from discrimination the full range of expression associated with a trait.” 74 Goldberg, Anti-Essentialist, supra note 46, at 659. Assume that an individual who does not conform to the dominant narrative wishes to make a claim of discrimination based on the same or a similar trait. If that individual views and experiences the world differently, even though she possesses the trait on which the dominant narrative is based, the relief fashioned on the basis of the dominant narrative may be an imperfect fit for the harm she has experienced as a result of discrimination. 75 For example, the limitation of the relevant class as “gays and lesbians” in the context of sexual orientation (and the accompanying rationales that have been developed to render their exclusion unconstitutional) raises the question of how the law would apply to “a transgender[] woman . . . [claiming] a right to use a women’s restroom, or . . . bisexuals [claiming] a right to be housed in prison units designed to protect gay inmates.” Butler, Justice Rather than Rights, supra note 46. For example, if the dominant narrative emerging from gender discrimination doctrine is that women want to be treated like men, the remedy (formal equality) may not be adequate for those who want their differences to be embraced or the specific nature of their oppression to be recognized. Gender discrimination doctrine based on a male–female binary will not always account for the experience of someone who is transgender; 76 As Judith Butler has observed, “[t]he binary regulation of sexuality suppresses the subversive multiplicity of a sexuality that disrupts heterosexual, reproductive, and medicojuridical hegemonies.” Judith Butler, Gender Trouble: Feminism and the Subversion of Identity 26 (1999); see also Darren Rosenblum, “Trapped” in Sing Sing: Transgendered Prisoners Caught in the Gender Binarism, 6 Mich. J. Gender & L. 499, 513–15 (2000) (describing “identity” problems faced by transgender individuals, including transgender people who have not yet received gender-affirming surgery, who cannot afford it, or who choose not to undergo it, including “permanent victimization by the law’s gender binarism that forces people into either the category ‘male’ or ‘female’”). nor will it always account for the experience of someone who is pregnant (as evidenced by the tension within gender discrimination doctrine), 77 See infra notes 224–231 and accompanying text (describing Court’s approach in Geduldig as declining to equate pregnancy with gender and thus applying heightened scrutiny to classifications based on pregnancy). or for a male plaintiff who does not conform to traditional notions of masculinity. 78 See Ann C. McGinley, Masculinities at Work, 83 Or. L. Rev. 359, 401–04 (2004) (discussing hostile work environment claims brought by male plaintiffs who were harassed for failing to conform to traditional notions of masculinity). On the other hand, racial discrimination doctrine assumes a commonality of experience among racial minorities and, as currently constructed, does not distinguish between those races that have historically enjoyed higher levels of privilege (whites) and those who historically have been subordinated (blacks). 79 There may also be definitional issues within what most perceive to be one racial group. See Banks, supra note 50, at 1711 (2000) (“[T]he government’s definition of the racial category black impedes recognition by courts that black people can be differently racialized.” (footnote omitted)). Angela Harris has argued that we must “make our categories explicitly tentative, relational, and unstable” in order to avoid the harms of essentialism. 80 Harris, Race and Essentialism, supra note 62, at 586; see also Michael Omi & Howard Winant, Racial Formation in the United States: From the 1960s to the 1980s, at 68 (1986) (arguing for conception of race not “as an essence, as something fixed, concrete and objective” but instead as “an unstable and ‘decentered’ complex of social meanings constantly being transformed by political struggle”). Yet, this recognition of identity’s fluidity 81 See, e.g., Minow, supra note 38, at 51 (“Through an ongoing process of nego­tiation, identities, even group identities, constantly change.”). may be incompatible with the nature of doctrine, which requires a more fixed notion of identity in order to form a coherent, administrable body of legal rules. 82 Id. at 59 (“Law plays a substantial role in making identities seem fixed, innate, and clearly bounded.”).

Critical race scholars like Kimberlé Crenshaw have emphasized that those who reside at the intersection of various identities have experiences that may differ not only from majority norms but also from other minority experiences. 83 Crenshaw, Mapping the Margins, supra note 60, at 1244 (“Because of their intersectional identity as both women and of color within discourses that are shaped to respond to one or the other, women of color are marginalized within both.”); see also Meera E. Deo, The Ugly Truth About Legal Academia, 80 Brooklyn L. Rev. 943, 950–51 (2015) (“Because of the multiple ‘opportunities’ for oppression, it becomes clear that those who are marginalized in multiple ways have experiences that differ from not only the norm ( . . . the middle- to upper-class, heterosexual, white male), but even from the norms attributed to particular minority groups.”). For example, she explains:

Sex and race discrimination have come to be defined in terms of the experiences of those who are privileged but for their racial or sexual characteristics. Put differently, the paradigm of sex discrimination tends to be based on the experiences of white women; the model of race discrimination tends to be based on the experiences of the most privileged Blacks. 84 Kimberle Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics, 1989 U. Chi. Legal F. 139, 151 [hereinafter Crenshaw, Demarginalizing the Intersection]. While Crenshaw grounds her discussion primarily in statutory antidiscrimination law, her observations apply just as aptly to equal protection.

In that sense, essentialism may further prevent other unique forms of oppression—for example, that experienced by black women—from being recognized and addressed. 85 Sonu Bedi has observed:

“Consider a law that discriminates against black women. One could very well argue that the law does not discriminate against a racial minority because it does not discriminate against all black individuals. And it also does not discriminate against women because it does not discriminate against all women. Does the Court, then, need to create a new suspect class—those encompassing black women—to meet this problem? And what happens when laws discriminate against lesbians but not gay men? Or black lesbians but not others? As long as the constitutional objection under the Equal Protection Clause is characterized in terms of suspect classes, this problem is seemingly unavoidable.”

Bedi, supra note 6, at 50.

The existing categorical model cannot effectively address intersectionality because a framework that analyzes discrimination as occurring along a single categorical axis is “predicated on a discrete set of experiences that often does not accurately reflect the interaction of [multiple categories, such as] race and gender.” 86 Crenshaw, Demarginalizing the Intersection, supra note 84, at 140; see also Bedi, supra note 6, at 48–49 (explaining “identity categories come to represent the experiences of only certain individuals or subgroups”); Mayeri, supra note 23, at 104, 232–33 (noting universalist strategy of some within feminist movement rendered race invisible in sex equality law). For example, Crenshaw discusses the dilemma posed by a group of black women who wish to challenge their discrimination as black women: Antidiscrimination law cannot conceptualize the claim as one based on gender, because other women (white women) were not discriminated against, or based on race, because other blacks (black men) were not discriminated against. 87 Crenshaw, Demarginalizing the Intersection, supra note 84, at 151–52. In the context of equal protection, a similar predicament exists: Should a black woman wish to claim the denial of equal protection based on her existence as a black woman, it is unclear whether such a claim would or should be treated with intermediate scrutiny (applied to gender) or strict scrutiny (applied to race). The requirement that she prove that similarly situated individuals have been treated differently likewise poses a dilemma: If the uniqueness of her experience is adequately recognized, there may be no similarly situated individuals against whom she can contrast her own treatment to vindicate her claim.

A simplified conception of identity treats all members of a given group the same and, perhaps just as troubling, does not allow for distinctions among different groups. As discussed above, in the eyes of the current Court, discrimination against someone on the basis of her whiteness is equivalent to discrimination against someone on account of his blackness. This quality of equal protection doctrine means that it can be co-opted, either intentionally or unwittingly, by a particular group to change the nature of equal protection or to redefine who is included and who is excluded. For example, many of the race discrimination cases decided by the Supreme Court in recent years have involved successful claims by white plaintiffs that they have been discriminated against on the basis of race. 88 See, e.g., Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411, 2415 (2013) (reversing Fifth Circuit judgment against white plaintiff challenging her rejection from University of Texas at Austin); Gratz v. Bollinger, 539 U.S. 244, 251 (2003) (reversing district court judgment against white plaintiffs challenging rejection from University of Michigan); see also Robinson, Unequal Protection, supra note 41 (manuscript at 17) (“[W]hen strict scrutiny appears in the Court’s race jurisprudence today, it is almost invariably on behalf of whites such as Abigail Fisher, who wield it to dismantle affirmative action policies.” (footnote omitted)). This coming term, the Court will again hear arguments in Abigail Fisher’s case against the University of Texas. See Fisher v. Univ. of Tex. at Austin, 758 F.3d 633 (5th Cir. 2014), cert. granted, No. 14-981, 2015 WL 629286 (June 29, 2015). When it issues its opinion, the Court will likely further restrict the ability of colleges and universities to consider race in admissions—if not eliminate it altogether. Should that be the outcome, the arguments made herein for measuring discrimination by a means other than identity will be all the more germane.

This Essay does not suggest that equal protection’s reliance on identity as proxy is the sole driver of the doctrinal manifestations described above or that one cannot imagine a jurisprudence in which identity functions in a different capacity; yet, it is important to recognize the ways in which the identity-as-proxy model facilitates and supports these distinctions. For the same reasons an identity-based framework may be an effective tool for attacking facial discrimination, it may be less able to draw distinctions among different types of discrimination or to provide redress in situations in which discrimination is not overtly identity-based.

B. Individual and Societal Effects of the Categorical Model

In addition to the effects observed within the doctrinal structure, there are extralegal consequences—both individual and societal—that flow from identity-as-proxy jurisprudence. Equal protection’s appropria­tion of identity as a means for vindicating a particular set of values not only impacts the law’s effectiveness in achieving the normative goals underlying equal protection, but it may also force individuals to compromise their own sense of identity and further marginalize certain sub-identity groups.

The existence and application of categories require definition of the categories and, thus, in this context, require definition of identity. The application of heightened scrutiny to race and gender classifications—and perhaps eventually to those based on sexual orientation—requires making judgments, either explicit or implicit, about who falls within (and outside of) these categories. 89 See Rodriguez, supra note 36, at 1410–11 (describing how, under such regime, state actors decide “who counts” and “how race relates to particular individuals and their potential contributions”); cf. Ford, Race as Culture, supra note 9, at 1811 (arguing approach that views race as culture “must invite—in fact it must require—courts to deter­mine which expressions are authentic and therefore deserving of protection” and “[t]he result will often be to discredit anyone who does not fit the culture style ascribed to her racial group”). Hence, there is a risk that in carving out protection for some, the Court will further marginalize others. For example, in the context of cases like Romer v. Evans and Obergefell v. Hodges, which have been heralded as victories for LGBT advocates, the Court’s focus remained specifically on “gays and lesbians,” excluding from the relevant definition other sexual orientation minorities, like bisexuals. 90 See, e.g., Obergefell v. Hodges, 135 S. Ct. 2584, 2596, 2600, 2602, 2604, 2606 (2015) (characterizing relevant class as “gays and lesbians”); Romer v. Evans, 517 U.S. 620, 625 (1996) (identifying class as “homosexual persons or gays and lesbians”); Robinson, Unequal Protection, supra note 41 (manuscript at 48) (noting plaintiffs in both Romer and Windsor “strategically” framed relevant class as “gays and lesbians”).

Defining identity categories in the context of the law is inherently exclusionary, as the law attempts to circumscribe those who should rightly benefit from certain legal protections and those who should not. The need to define identity categories may also force the simplification of an otherwise complex understanding of identity in order to provide facile descriptions that can be used within doctrinal frameworks. 91 Darren Rosenblum provides a helpful example in highlighting Justice Souter’s description of Dee Farmer, the plaintiff in Farmer v. Brennan, as “a transsexual, one who has ‘[a] rare psychiatric disorder [gender dysphoria] in which a person feels persistently uncomfortable about his or her anatomical sex,’ and who typically seeks medical treatment, including hormonal therapy and surgery, to bring about a permanent sex change.” 511 U.S. 825, 829 (1994) (quoting Am. Med. Ass’n, Encyclopedia of Medicine 1006 (1989)). Rosenblum explains that this description, while facile, ignores the complexity of being transgender. Rosenblum, supra note 76, at 506–08 (explaining, with regard to transgender people, “[n]o category can perfectly define its members”). Law strives for clear rules that can then be applied to adjudicate claims. The very nature of identity, however, defies such clarity, particularly for those whose identities rest at the intersection or overlap of established identity categories, or fall outside the bounds of such categories altogether. 92 See Lucas, Undoing Race, supra note 66, at 1263–67 (describing various ways in which multiracial individuals may choose to identify racially); see also Minow, supra note 38, at 83 (“Law, at its best, cannot resolve tensions between connection and betrayal, individuality and group affiliation.”). To the extent that law and society are co-constitutive forces in constructing the meaning and salience of identity, 93 See López, White by Law, supra note 9, at 10 (“[T]o say race is socially constructed is to conclude that race is at least partially legally produced.”). forcing identity to play such a role in the context of the law may stunt its development outside of that context. Subsequently, the use of identity as a legal device may also hinder the effective use of identity as an inclusionary device in the context of social and political movements. 94 See Guinier & Torres, supra note 73, at 31 (arguing inclusive visions of social justice require “different understanding of the meaning of race and its relationships to power”). Sonu Bedi has emphasized the particularly problematic nature of basing not just law (as promul­gated by legislatures) but also the Supreme Court’s interpretation of the Constitution on identity. 95 See Bedi, supra note 6, at 34 (explaining how Court’s use of identity both protects and stigmatizes women and racial and sexual orientation minorities, in part by forcing them to emphasize their differences). By framing the constitutional analysis in terms of identity, the Court incorporates these issues into the highest level of the law, thereby dispersing and cementing these notions throughout all levels of the law. 96 See Bedi, supra note 6, at 33 (noting many equal rights laws and challenges to discriminatory laws rely on “language of identity”).

The constant evolution of socially constructed categories compli­cates the need to define identity, which necessarily underlies an identity-based jurisprudence. 97 See, e.g., Magee Andrews, supra note 26, at 509 (“The concept of race is perpetually undergoing revision within American society . . . .”). In her book, Not Only for Myself, Martha Minow describes the difficulties that may arise from “contemporary challenges to the basic coherence of group definitions”:

The gaps and conflicts among self-identification, internal group membership practices, and external, oppressive assignments have given rise to poignant and persistent narratives of personal and political pain and struggle. These gaps and conflicts also expose the inconsistent meanings of group membership. The persistent failure of group-based categories to yield consistent applications hints at the defects in their boundaries, their origins, their applications, and their ultimate meaningfulness. 98 Minow, supra note 38, at 40–41 (footnote omitted).

In the case of a multiracial individual, for example, the individual must negotiate her self-categorization against the categories to which other actors assign her. While some multiracials choose to align themselves with a single race (“monoracial”), others choose to identity as biracial or multiracial. 99 Lucas, Undoing Race, supra note 66, at 1264. Yet other multiracials will choose to identify differently depending on the context (e.g., they may identify racially one way at home and another at school), and some view themselves as occupying a space where racial categories do not apply. 100 See id. At the time Carolene Products was written, individuals had little say regarding their racial categorization for government purposes. 101 See id. at 1256 (“Until 1960, the census relied on an enumerator to visually survey and identify each individual’s race.”). Although the current regime now relies on self-identification as the means for racial categorization, it does so by offering an array of predetermined racial categories from which the individual may choose. 102 Id. (noting 2000 Census marked first time in history individuals were given option to select more than one racial category—“mark one or more”). This is necessary, in part, because the government still relies on identity as a mechanism for measuring impact and distributing benefits. 103 Beginning in 2000, individuals were allowed to check all the racial categories that they deemed applicable. See Kevin Brown, Should Black Immigrants Be Favored over Black Hispanics and Black Multiracials in the Admissions Processes of Selective Higher Education Programs?, 54 How. L.J. 255, 279 (2011) (describing changes to 2000 Census for multiracial individuals). Kevin Brown describes this regime as an administrative compromise allowing the government to tabulate individuals according to traditionally defined racial categories. See id. (explaining federal government has used racial and ethnic classifications to combat discrimination and differential treatment experienced by historically marginalized groups). Tension may arise when an individual’s self-identification does not align with the way she is classified externally or when she is forced to identify herself in a manner other than that which she would naturally choose in order to benefit from legal protections. 104 See Lucas, Undoing Race, supra note 66, at 1263, 1282–83 (explaining how identifying as multiracial may hurt multiracials in context of affirmative action and how multiracials benefit from antisubordination theory only to extent they associate with single-race groups with well-established history of subordination); see also Devon W. Carbado & Mitu Gulati, Working Identity, 85 Cornell L. Rev. 1259, 1262 (2000) [hereinafter Carbado & Gulati, Working Identity] (highlighting “identity work” in which outsider groups must engage to distance themselves from negative assumptions about their identities). For example, individuals may feel—in part because the law encourages such behavior—that they have to conform to the dominant social understanding of what it means to be a member of a given identity group in order to secure legal protection as a recognized member of that group. 105 Discrimination in the context of national origin provides one example of the need to conform one’s identity to the larger category to receive protection. Jenny Rivera has demonstrated that the more a Latino individual has acquired English language skills or lacks Spanish language skills, the less likely a judge is to perceive a basis for a discrimination claim based on national origin. Jenny Rivera, An Equal Protection Standard for National Origin Subclassifications: The Context that Matters, 82 Wash. L. Rev. 897, 927–29 (2007). In other words, identity in this context poses a similar dilemma: If the only available basis for a claim of discrimination is national origin, the more “American” one becomes, the harder it may be to claim status as a protected class. Id.; cf. Ford, Racial Culture, supra note 9, at 41 (“The idea that minorities should hew to ‘their’ cultural traditions is as hegemonic as the idea that they should assimilate to a mythical white-bread mainstream.”). Moreover, as I have explained in more detail elsewhere, social science research has demonstrated that being forced to identify in a way that does not align with one’s own self-conception may cause psychological and emotional harm. 106 Lucas, Undoing Race, supra note 66, at 1267. For biracial black-white individuals, for example, social science studies have demonstrated that being forced to adopt a monoracial identity can lead to guilt, lowered self-esteem, and decreased motivation while the ability to assume a biracial identity may lead to a more positive sense of identity and greater self-confidence. 107 Id. at 1267–68.

As suggested by the Windsor example, boundary issues also manifest in the category of gender. More explicit categorization questions arise for transgender individuals. For example, should a transgender woman who was assigned male at birth and who has not received gender-affirming surgery be considered a man or a woman under equal protection’s ostensibly binary approach? Although there has been some positive evolution of the law toward a model that is less focused on requiring self-identification as male or female, 108 Although not binding on the judiciary, the Equal Employment Opportunity Commission (EEOC) issued an opinion in 2012 holding that claims of transgender discrimination are cognizable under Title VII of the Civil Rights Act of 1964. See Macy v. Holder, No. 0120120821, 2012 WL 1435995, at *7 (E.E.O.C. Apr. 20, 2012) (explaining, under Title VII, discrimination against transgender individuals constitutes discrimination on basis of sex). In doing so, the EEOC “articulated various ways to state a valid claim,” including “the sex stereotyping approach, which describes discrimination against transgender individuals as rooted in gender stereotypes, and the per se approach, which posits that such discrimination is inherently sex discrimination because it relates to a change in sex.” Recent Case, EEOC Affirms Protections for Transgender Employees—Macy v. Holder, No. 0120120821, 2012 WL 1435995 (E.E.O.C. Apr. 20, 2012), 126 Harv. L. Rev. 1731, 1733 (2013). The EEOC stopped short, however, of articulating other possible theories of discrimination, such as discrimination based on anatomical sex characteristics, which would have addressed a primary focus of transgender discrimination. Id. at 1734. The comment goes on to explain that although “no single formulation perfectly reflects the experiences of all transgender persons, this approach can coherently describe plaintiffs in all stages of transition.” Id. at 1737 (footnote omitted). existing equal protection cases have often required transgender individuals to identify as a particular sex in order to argue that discrimination was based on deviance from the stereotypes associated with that sex. 109 See, e.g., Glenn v. Brumby, 663 F.3d 1312, 1320–21 (11th Cir. 2011) (finding employer discriminated on basis of gender nonconformity where transgender employee alleged she was fired for deviating from male stereotypes by presenting femininely at work); Smith v. City of Salem, 378 F.3d 566, 572 (6th Cir. 2004) (holding transgender fire department employee sufficiently pleaded gender discrimination claims by alleging employer attempted to compel resignation after her conduct and mannerisms did not conform with sex stereotypes of how men should look and behave). Dean Spade has explained that the characteristics used to determine categorization on the basis of gender and the way in which those “categories are defined and applied create[] vectors of vulnerability and security.” 110 Dean Spade, Normal Life: Administrative Violence, Critical Trans Politics, and the Limits of Law 138 (2011); see also id. at 142–50 (describing difficulties posed for transgender individuals under current identity framework). Those who are left out or excluded from benefits (either explicitly or because their population goes unaccounted for in tracking need among various populations) are left vulnerable—“casualties” of a framework that insists on using monolithic categories of identity. 111 Id. at 142 (“The consequences of misclassification or the inability to be fit into the existing classification system are extremely high, particularly in the kinds of institutions and systems that have emerged and grown to target and control poor people and people of color . . . .”); see also id. at 144–46 (“For the many people who feel that neither ‘M’ nor ‘F’ accurately describes their gender, there is no possibility of obtaining records that reflect their self-identities. . . . [This] causes extensive problems.”).

Reliance on a monolithic notion of identity—and the uniform application of one level of scrutiny to every sub-identity group within a larger identity group—also makes it difficult for the doctrine to recognize variant experiences and to justify differential treatment within a given group. 112 Cf. Balkin, supra note 13, at 2371 (“Discrimination against darker-skinned blacks by lighter-skinned blacks should not be constitutionally unprotected simply because there is no bright line that separates them.”). For example, as a number of authors have recognized in the context of educational policy, 113 See, e.g., Kevin Brown & Jeannine Bell, Demise of the Talented Tenth: Affirmative Action and the Increasing Underrepresentation of Ascendant Blacks at Selective Higher Educational Institutions, 69 Ohio St. L.J. 1229, 1233–36 (2008) (arguing “ascendant” blacks—those with two native-born black parents or more direct connection to America’s history of racial discrimination—should be treated differently in context of affirmative action from multiracials with some black heritage or black immigrants); Angela Onwuachi-Willig, The Admission of Legacy Blacks, 60 Vand. L. Rev. 1141, 1143 (2007) (questioning whether affirmative action is effective in providing educational opportunity to diverse group of black students). the experiences of blacks who are descendants of American slaves and those who have a mixed-race background or who are recent immigrants can differ significantly; this in turn can impact the effectiveness of affirmative action programs. Kevin Brown and Jeannine Bell suggest that treating all of these groups as black, without further categorization, undermines the original goals of affirmative action and leads to the underrepresentation “of blacks whose predominate ancestry is traceable to the historical oppression in the United States.” 114 Brown & Bell, supra note 113, at 1231. Similarly, Angela Onwuachi-Willig has observed that “legacy Blacks”—those with four grandparents born in the United States and descendent from American slaves—have been increasingly excluded from elite colleges and universities. Onwuachi-Willig, supra note 113, at 1144–49. She maintains, however, that there are important reasons for the continued inclusion of first and second-generation blacks and mixed-race students. Id. at 1181–85. Therefore, they contend there is good reason, in the context of affirmative action, to “classify[] blacks based on their racial/ethnic ancestry, with a focus on attenuating the effects of racial subordination in the United States.” 115 Brown & Bell, supra note 113, at 1232. Similarly, the fact that Asian Americans are often perceived in the same context to be a group without any need for heightened protection may create unexpected harms for Asian Americans who do not conform to the dominant, model-minority narrative. 116 Onwuachi-Willig, supra note 113, at 1143 (discussing model-minority myth and how “view of Asian-Americans as a monolithic group[] may have a negative impact on affirmative action policies for Asian-American students, especially those who are of Cambodian, Hmong, Laotian, and Vietnamese descent” (footnote omitted)). Yet it is hard to see how the current doctrinal framework would justify differential treatment in either example, given its lack of nuance when it comes to categories like race. 117 See Transcript of Oral Argument at 43–46, Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411 (2013) (No. 11-345), 2012 WL 4812586 (demonstrating certain Justices’ unwillingness to embrace Respondents’ argument that to achieve true diversity, university would need to admit minority students from both privileged and underprivileged socioeconomic backgrounds).

Disregard for differences beyond the more superficial characteristics of identity lead to other negative consequences for the most oppressed. For example, Devon Carbado has explained that the focus on group representation without regard for intragroup differences allows admissions officials to fulfill their commitment to diversity by choosing to admit more palatable “types” of African Americans. 118 See Devon W. Carbado, Intraracial Diversity, 60 UCLA L. Rev. 1130, 1139–41 (2013) [hereinafter Carbado, Intraracial Diversity] (“[I]ntraracial selections in the domain of admissions likely will turn on the racial types that institutions find palatable.”); see also Devon W. Carbado & Mitu Gulati, Acting White: Rethinking Race in “Post-Racial” America 2 (2013) (explaining employers may prefer racial performances or expressions of identity that comport with white middle-class professionalism norms); Carbado & Gulati, Working Identity, supra note 104, at 1261–62, 1267–70 (describing strategies individuals from outsider groups must employ to counter negative stereotypes in workplace, including conforming behavior to insider norms). In doing so, they create or sustain hierarchies within racial groups. 119 Carbado, Intraracial Diversity, supra note 118, at 1138 (explaining how such admissions decisions encourage “whitening” of resumes to increase one’s chances of admission, thus reinforcing racial hierarchies). Similarly, Nancy Leong has written of the commodification risks that come from focusing on “thin” conceptions of diversity—those that reduce the importance of race to mere presence, rather than those that would create substantive change. 120 See Nancy Leong, Racial Capitalism, 126 Harv. L. Rev. 2152, 2157–58, 2169–70 (2013) (critiquing commodification of racial identity for “detract[ing] from more meaningful antidiscrimination goals by prioritizing racial representation at its thinnest and most tokenistic”). This, too, is a byproduct of placing undue importance on identity, which now represents a group of surface-level traits rather than a specific set of experiences.

Basing legal protections on identity may create certain undesirable implications regarding the nature of the group or its relationship to other groups. As an illustrative example, Georgia Warnke has argued that by singling out women in attempting to address pregnancy-related discrimination, “the law suggests that women require special rights and accommodations in order to hold jobs others can hold without them. Hiring them can seem likely to employers to be more expensive than hiring others and, worse, women can seem to be constitutionally unsuited to responsible working lives.” 121 Georgia Warnke, After Identity: Rethinking Race, Sex, and Gender 184 (Ian Shapiro ed., 2007). Therefore, Warnke suggests, a preferable way to accommodate the relationship between pregnancy and employment is to “regard reproduction and child-rearing in ways that are neutral with regard to identities as men or women.” 122 Id.; see also id. at 185 (“Neither pregnancy nor child-rearing . . . needs to be understood in sex and gender terms. Instead, those who are pregnant and those who are raising children are adequately understood as, respectively, pregnant people and parents.”). In other words, reproduction and child-rearing should not be viewed as integral to, or inseparable from, female identity. 123 See Ford, Racial Culture, supra note 9, at 113 (“Conflating pregnancy discrimi­nation with sex discrimination . . . is not only conceptually flawed; as an ideological matter it reinforces sexism.”). Yet, under the current equal protection framework, pregnancy discrimination either equates with gender discrimination (the relevant identity category) or, as the Court held in Geduldig v. Aiello, it does not. 124 See infra notes 224–229 and accompanying text for a discussion of how the Court’s reasoning in Geduldig is a result of identity-as-proxy jurisprudence. Under a value-based framework, doctrine could recognize the impact of policies that impact a particular group—i.e., women who are or may become pregnant—without equating the trait with that identity group (or sub-identity group). This may seem like an unsatisfactory result to many, given the reality that the burdens of pregnancy so clearly fall on women; but, from another perspective, the opposite conclusion may be equally troublesome. Associating pregnancy with sex or gender may imply that motherhood is an essential characteristic of womanhood and thus devalue those women who choose not to have children.

In a related vein, Wendy Brown has argued that “[i]n its emergence as a protest against marginalization or subordination, politicized identity . . . becomes attached to its own exclusion . . . because it is premised on this exclusion for its very existence as identity.” 125 Wendy Brown, States of Injury: Power and Freedom in Late Modernity 73 (1995). In other words, when identity is married to exclusion as a categorical matter, invocation of identity automatically reifies the narrative of oppression. 126 Ford, Racial Culture, supra note 9, at 20 (“[P]rotecting the traits associated with groups that have suffered from pervasive social and political oppression . . . reinforce[s] the regulation that produced the identities in the first place . . . .”); see also Yoshino, supra note 49, at 795 (“[W]hen the courts protect a trait as part of a group’s identity, they strengthen the very stereotypes they mean to disestablish.”). And to attain suspect class status under the Equal Protection Clause, therefore, groups must characterize themselves as victims. 127 Bedi, supra note 6, at 47 (“Victimhood must be placed front and center to gain suspect class status.”). Such a frame may also generate unnecessary and potentially harmful assumptions about the nature of the group—for example, that the lack of political influence is related to the nature of the group itself rather than structural elements in society.

Moreover, the notion of equality present in the law implies some level of hierarchy which, when combined with identity, can help to entrench the current racial order. For example, George Lipsitz has explained that whites have a particular investment in identity politics, given the way in which white identity has been defined. In The Possessive Investment in Whiteness, Lipsitz argues that “white Americans are encour­aged to invest in whiteness, to remain true to an identity that provides them with resources, power, and opportunity” and that whiteness is “an identity created and continued with all-too-real consequences for the distribution of wealth, prestige, and opportunity.” 128 George Lipsitz, The Possessive Investment in Whiteness: How White People Profit from Identity Politics, at vii (2006). Thus, as in the con­text of the law, identity is not imbued with normative value suggesting an innate benefit to those who seek to use identity as an organizing force to combat inequality; rather, it can work just as easily to exacerbate oppression as to fight it. Moreover, if the very content associated with whiteness is dominance, 129 Mary Louise Fellows & Sherene Razack, The Race to Innocence: Confronting Hierarchical Relations Among Women, 1 J. Gender, Race & Just. 335, 343 (1998) (“‘[T]here is no content to whiteness outside of domination: whiteness is the ‘empty and terrifying attempt to build an identity based on what one isn’t and on whom one can hold back.’” (quoting David Roediger, Toward the Abolition of Whiteness: Essays on Race, Politics, and Working Class History 13 (1994))). an identity-based jurisprudence will inherently serve to bolster the existing racial hierarchy.

II. Moving from the Categorical Model to a Value-Based Approach

Building on arguments in Part I that identity-as-proxy jurisprudence has contributed to or facilitated undesirable effects within the doctrine and generates harms for both those included in and excluded from heightened protection, this Part suggests an alternative framework for equal protection doctrine. Because it is rooted primarily on ideas already present in the doctrine, but which have been subsumed by identity’s dominance, it offers potential for redeployment of existing doctrine—and a new way to talk about discrimination using familiar concepts—not just a proposal for radical change.

In addition to the issues raised above, there are several other reasons why a shift in thinking about equal protection is needed. First, as other equal protection scholars have recognized, society has become more pluralistic, 130 See Bruce A. Ackerman, Beyond Carolene Products, 98 Harv. L. Rev. 713, 719–20 (1985) (noting pluralist nature of American politics); Yoshino, supra note 49, at 747 (noting increasing anxiety about pluralism and group-based identity politics in American society). Arguably, society has been pluralistic for some time, but it is only in relatively recent history that the law has been willing to provide recognition of that pluralism. and the demarcation of identity categories has become increasingly complex. For example, in the 1940 census, just two years after Carolene Products was decided, Mexican Americans were considered “white” and black­–white multiracials would have been considered black (or, more accurately, “Negro”) without further elaboration. 131 Sharon Lee, Racial Classifications in the U.S. Census: 1890–1990, 16 J. Ethnic & Racial Stud. 75, 77–79 (1993). Evolving notions of racial and gender identity demand a framework that can accommodate and respond to more nuanced claims of discrimination.

Second, in addition to individuals who straddle or who do not conform to traditional identity categories, there is the question of whe­ther the categories themselves are adequately described. For example, it is unclear whether Latinos and those of Middle Eastern descent are considered members of a racial group for purposes of equal protection; courts have, to date, most often perceived discrimination against those classes as discrimination based on national origin. 132 Compare Rivera, supra note 105, at 901 (noting how courts have used national-origin framework for evaluating cases of discrimination against Latinos), with Saint Francis Coll. v. Al-Khazraji, 481 U.S. 604, 613 (1987) (holding plaintiff of Arabian ancestry could claim racial discrimination under 42 U.S.C. § 1981), and George Yancey, Who Is White?: Latinos, Asians, and the New Black/Non Black Divide 143 (2003) (discussing relative assimilation of Latinos and Asian Americans to alienation of African Americans).

In the statutory context, some courts have failed to equate discrimination on the basis of national origin with language discrimination and thus failed to provide protection to Spanish-speaking bilinguals. Juan F. Perea, Buscando América: Why Integration and Equal Protection Fail to Protect Latinos, 117 Harv. L. Rev. 1420, 1437 (2004) [hereinafter Perea, Buscando América]; see also Jens Manuel Krogstad, Census Bureau Explores New Middle East/North Africa Ethnic Category, Pew Research Ctr. (Mar. 24, 2014),‌fact-tank/‌2014/03/24/‌census-bureau-explores-new-middle-eastnorth-africa-ethnic-category [] (discussing people of Middle Eastern and North African descent who have told Census Bureau they do not want to be categorized as “white” any longer and to add new ethnic category on forms).
While national origin currently enjoys the same level of scrutiny as race, the two categories are nonetheless viewed as distinct and thus perceived differently. 133 See U.S. Census Bureau, 2013 American Community Survey, at 2 (Aug. 14, 2012) [] (showing Census’s separation of Hispanic/not-Hispanic question from question of race/ethnicity). Whether someone who is Latino is a member of not just an ethnic minority but also a racial minority is an ongoing debate within equal protection jurisprudence; 134 See, e.g., Laura M. Goodall, The “Otherized” Latino: Edward Said’s Orientalism Theory and Reforming Suspect Class Analysis, 16 U. Pa. J. Const. L. 835, 837 (2014) (describing how diversity within Latino category makes coherent racial classification unlikely, but framework reflecting social rather than biological foundations of race would better accommodate Latino categorization); Patricia Palacios Paredes, Note, Latinos and the Census: Responding to the Race Question, 74 Geo. Wash. L. Rev. 146, 150 (2005) (describing potential for fluidity in describing Latinos as ethnic or racial group); id. at 147 (“The United States does not view Latinos as a racial group, but rather as an ethnicity, with individuals belonging to any race.”). this example illustrates that it is not clear what is gained by the law’s emphasis on the race–ethnicity distinction. By creating such divisions, the law may unwittingly divorce, or place into a hierarchy, forms of discrimination and oppression that share systemic causes and which might crumble more quickly under a joint attack.

Last, and perhaps most important, the identity framework no longer effectively serves the aims it was designed to achieve, or that others have assumed it would. As discussed above, in recent years, it has been used on behalf of groups at the top of the racial and gender hierarchies as often and as successfully (if not more so) than by those at the bottom. 135 See supra note 88 and accompanying text (noting successful cases brought by white plaintiffs challenging consideration of race in affirmative action policies); see also Timothy K. Giordano, Different Treatment for Non-Minority Plaintiffs Under Title VII: A Call for Modification of the Background Circumstances Test to Ensure that Separate Is Equal, 49 Emory L.J. 993, 993 (2000) (“The Supreme Court’s docket is proof of the prevalence of reverse discrimination claims.”); Osamudia R. James, White Like Me: The Negative Impact of the Diversity Rationale on White Identity Formation, 89 N.Y.U. L. Rev. 425, 429 (2014) (observing “characterizations of affirmative action as ‘reverse discrimination’ have intensified in recent years”); Brian K. Landsberg, Race and the Rehnquist Court, 66 Tul. L. Rev. 1267, 1275 (1992) (noting during first five years of Rehnquist Court, white plaintiffs filed thirty-four percent of race discrimination claims, reflecting “increasing white awareness of the possibility of using the Constitution and the civil rights laws as a litigative tool”). Thus, identity may already have proven itself a deeply flawed tool for combating oppressive racial- or gender-based discrimination. 136 There may be more potential for effectively combating discrimination in attempting to redeploy values underlying the current equal protection framework rather than trying to reconfigure the way in which identity itself is used or changing the notion of identity that is used within the existing framework. See supra note 15 (referencing Professor Eyer’s observation that redeployment of elements within existing constitutional doctrine may prove more successful than direct resistance).

Section II.A begins by defining the structural concerns for which identity currently serves as proxy, and which provide the substantive basis for what I have termed the value-based model. 137 See infra notes 140–159 and accompanying text. To remain as consistent as possible with existing doctrine, it looks to those values already articulated in the case law, particularly the need to address a history of past discrimination and political powerlessness. 138 See infra note 153–155 and accompanying text. In doing so, it focuses only on those values that reflect structural causes of inequality, and not those that relate to individual traits isolated from their social and political role. 139 See infra notes 160–166 and accompanying text. Section II.B explains in greater detail—using specific examples involving race, gender, and language discrimination—how the value-based approach would alter the current analysis of equal protection claims. Section II.C provides some initial thoughts as to how this shift might have potential impact beyond the Supreme Court’s equal protection jurisprudence. It suggests, for example, that litigants in lower courts might take advantage of the value-based model’s roots in existing doctrine to deploy novel arguments, and that legislators and policymakers might use the value-based model to reconceptualize affirmative action policies and antidiscrimination law. Perhaps most important, it submits that the value-based model can provide a new framework for discourse on such issues—one that is less susceptible to co-option and colorblindness rhetoric.

A. Defining the Value-Based Approach

In The Empty Idea of Equality, Peter Westen argues that the notion of equality—to the extent it dictates that likes should be treated alike—is “an empty form having no substantive content of its own.” 140 Peter Westen, The Empty Idea of Equality, 95 Harv. L. Rev. 537, 539, 596 (1982). Westen argues that claims cloaked in the language of equality can ultimately (and more effectively) be reinterpreted as the claim to a specific right or entitlement. 141 Id. at 568. In Westen’s view, this comparative element of equal protection, which can be linked to identity, 142 Id. at 560 (“In order to decide whether a state classification treats differently people who are constitutionally deemed to be ‘alike,’ however, we must first possess a constitutional standard for distinguishing those people who are alike from those who are not.”). In the context of the law, it is identity that helps to facilitate that process. A framework that focuses on treating likes alike is premised on the belief that a group of people can be essentialized or considered similar in every respect but one (which provides the basis for the claim of discrimination). Westen highlights this fallacy, emphasizing that “‘[a]ll individuals are similar in some respects and different in others.’” Id. at 566 n.98 (quoting Terrance Sandalow, Racial Preferences in Higher Education: Political Responsibility and the Judicial Role, 42 U. Chi. L. Rev. 653, 655 (1975)). distracts from the more important inquiry as to whether the underlying substantive right has been violated. 143 In discussing Sweatt v. Painter, 339 U.S. 629 (1950), for example—the case of a black law school applicant, Heman Sweatt, who was denied admission to Texas’s law school—Westen observed:

“To argue that race or skin color should not constitutionally be allowed to matter in Sweatt’s particular case, one must show that excluding blacks from law school on the basis of race causes them a kind of injury not caused in cases in which using race is conceded to be acceptable—an injury, furthermore, from which the Constitution gives them a right to be free.”

Westen, supra note 140, at 566.
As to the Equal Protection Clause, Westen noted that the courts have failed to identify “the precise sort of injury that the substance of the equal protection clause is designed to prohibit.” 144 Westen, supra note 140, at 567. Westen suggests that this failure “may itself result from the fact that the clause is stated in the language of equality.” Id. While it could be viewed as a matter of semantics, this Essay is not as quick to dismiss the importance or relevance of equality; instead, it responds by attempting to provide a substantive meaning for equality that Westen identifies as lacking to date. The value-based approach is responsive to Westen’s argument in that it is an attempt to provide substantive meaning to equal protection—not because equality is necessarily meaningless in the absence of such meaning, but because the notion of equality is otherwise highly susceptible to manipulation and distortion.

Rather than framing equal protection claims as the right of a claimant to be treated the same as a person of any other gender or race, this Essay argues that equal protection should be framed primarily in terms of the structural concerns such categories are intended to vindicate. 145 See Yoshino, supra note 49, at 750 (noting Court’s trend away from explicit reliance on equal protection and towards due process, from pure group-based equality claims to “liberty-based dignity claims”); see also id. at 749 (explaining author’s use of term “dignity” encapsulates both equality and liberty and acknowledges “what academic commentary has long apprehended—that constitutional equality and liberty claims are often intertwined”). Thus, discrimination becomes not about categorical deter­minations of protection (or nonprotection), but instead about how individuals or groups have been excluded and the impact such exclusion has had on their access to the political process 146 In his seminal publication, Democracy and Distrust, John Hart Ely declared the democratic process and the importance of participation in that process the guiding principles of his constitutional theory of judicial review. John Hart Ely, Democracy and Distrust: A Theory of Judicial Review 100–01 (1980) (noting “procedural protections” and scheme designed to ensure “decision process will be open to all” as “principal answers” to constitutional inquiry). Thus, for Ely, the Court’s proper role is process-oriented: to prevent those in power from obstructing the political process in order to preserve the status quo, and to prevent representative government from withholding from a minority group, based on hostility or prejudice, the protection it affords other groups. Id. at 103. Ely concedes that to assess constitutional suspiciousness, one must reference the surrounding conditions to determine whether there are other “systemic bars” to access. Id. at 166; see also id. at 167 (“Throughout this discussion . . . I have been concerned with factors more subtle than the lack of a vote.”). Yet, in Ely’s view, the existence of a live debate on the matter seems to suggest that free interchange is possible and thus no constitutional impediment to access exists. Id. at 166 (“Given such open discussion of the traditional stereotypes [about women], the claim that the numerical majority is being ‘dominated,’ that women are in effect ‘slaves’ who have no realistic choice but to assimilate the stereotypes, is one it has become impossible to maintain except at the most inflated rhetorical level.”). In that regard, Ely may have afforded insufficient weight to extrapolitical factors, such as past discrimination and unconscious or systemic bias, which may not visibly taint the process, but which nonetheless frustrate the realization of access to state-created benefits and protections. For example, a law may be enacted without hostility or prejudice toward a minority group, yet perpetuate discrimination against that group because of laws that have come before it or because of norms that have become deeply embedded within the system, whether we are conscious of their presence or not. Ely seems to suggest, in contrast, that once the barriers to political access have been lifted and a certain period of time has passed, there may be little reason to apply heightened scrutiny to a particular class. See id. at 169 (“A case like that of women, where access was blocked in the past but can’t responsibly be said to be so any longer . . . suggests that a less dramatic remedy may be appropriate . . . . [C]onsequently the new law should be upheld as constitutional.”). and to the provision of state-provided benefits and protections. Group affiliation continues to be relevant, but it is contextual, defined by those alleging discrimination and not by the courts, and relevant only insofar as it serves as the basis for the denial of access—not as the determinant of the level of protection that will be provided.

As suggested above, one need not look far for the animating values that could supplant an identity-as-proxy approach. To decide which forms of discrimination are constitutionally permissible, one might look to the Court’s own assessment of why certain identities are deemed deserving of heightened scrutiny and the reasons for that distinction—in other words, the criteria used in defining a classification as “suspect.” 147 As discussed above, there are cases in which the Court has disavowed this reasoning as applied to race, instead treating racial classifications as sui generis. See supra note 48 (describing Justice Powell’s attempt in Bakke to distance race from Carolene Products’s “discrete and insular minority” approach); see also Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 272 (1979) (“Certain classifications, however, in themselves supply a reason to infer antipathy. Race is the paradigm.”). Yet others clearly see a link between the treatment of race and other suspect classes. See, e.g., López, Intentional Blindness, supra note 43, at 1830 (“Reasoning from race, the animating insight of [suspect classification] analysis was that certain social groupings, such as those along gender lines, so closely correlated to illegitimate forms of hierarchy that every state deployment of such classifications warranted a close look.”). Although the origin of these criteria is problematic, the history of their application conflicted, 148 The Court developed its “suspect classification” framework in the 1970s, during a time in which, as a practical matter, the Court had become increasingly hostile to the enforcement of civil rights. See López, Intentional Blindness, supra note 43, at 1832 (describing Court’s “hostility toward race-based remedies” during this time). Thus, at the same time it created such a framework, it “animated suspect class analysis with a rigid colorblindness principle, indifferent to group position and nearly always fatal to remedial measures.” Id. at 1831. and the expansion of their application seem­ingly limited, 149 See infra note 243 and accompanying text (describing Court’s hesitance to recognize additional classes as suspect). I use these criteria because they have become part of the doctrine and therefore offer a bridge from the categorical model to the value-based approach. Under the Court’s logic, certain identity categories are entitled to heightened protection because of their ability to, on a generalized level, satisfy these criteria. In a sense, then, they serve as a proxy for these values, yet because they do so in an imprecise fashion, they leave doctrinal wreckage in their wake.

For many constitutional law scholars, the beginning point for discussing equal protection doctrine is Carolene Products’s footnote four. In United States v. Carolene Products, the Court wrote in what is now one of the most famous footnotes of all time:

It is unnecessary to consider now whether legislation which restricts those political processes which can ordinarily be expected to bring about repeal of undesirable legislation, is to be subjected to more exacting judicial scrutiny under the general prohibitions of the Fourteenth Amendment than are most other types of legislation . . . . Nor need we enquire whether similar considerations enter into the review of statutes directed at particular religious, or national, or racial minorities; whether prejudice against discrete and insular minorities may be a special condition, which tends seriously to curtail the operation of those political processes ordinarily to be relied upon to protect minorities, and which may call for a correspondingly more searching judicial inquiry. 150 304 U.S. 144, 152–53 n.4 (1938) (citations omitted).

The doctrine’s reliance on identity is arguably traceable to Carolene Products. In footnote four, Justice Stone set off from other cases those statutes involving prejudice against “discrete and insular minorities.” 151 See Ackerman, supra note 130, at 741–42 (explaining how this model was natural response to social conditions in which it originated; in the minds of the Justices were examples such as German Jews and black Americans who, at different points in history, had been stripped of their civil rights). As Ackerman writes, “it was—and remains—obvious that the political choice to disenfranchise these groups was made vastly easier by virtue of their discreteness and insularity.” Id. at 742. Even decades ago, however, Ackerman recognized that this framework—which relies on a showing that the group is discrete and insular—would not provide adequate protection in an increasingly pluralist society, where a group’s actual influence on the political process is disproportionate to its numbers. See id. at 742 (citing groups unlikely “to achieve influence remotely proportionate to their numbers,” for example, “groups that are discrete and diffuse (like women), or anonymous and somewhat insular (like homosexuals), or both diffuse and ano­nymous (like the victims of poverty)”). Because Carolene Products’s “approach to minority rights is profoundly shaped by the old politics of exclusion,” Ackerman argues, it “yields systemically misleading cues within the new participatory paradigm.” Id. at 717. Ackerman focuses on the fact that minorities are no longer drastically underrepresented in number and yet they continue to face systemic discrimination; thus, relying on a framework that correlates low numbers or anonymity with marginalization (and the converse with adequate political access) is no longer effective. Yet, as Carolene Products’s own language makes clear, the status of the individual as part of a “discrete and insular minorit[y]”—as a marker of identity—is relevant only insofar as it serves to restrict his or her ability to utilize the political process to secure necessary legal protections. 152 Or, as Jack Balkin has suggested: “Discreteness and insularity are metaphors of division that describe, albeit from a limited perspective, certain features of particularly egregious status hierarchies.” Balkin, supra note 13, at 2371.

In the years following Carolene Products, the Court identified a number of criteria used to define suspect classifications: 153 As Marcy Strauss summarizes:

“[A]lthough described in different ways, the basic factors for determining suspect class status were in place by the early 1980s: (1) prejudice against a discrete and insular minority; (2) history of discrimination against the group; (3) the ability of the group to seek political redress (i.e., political powerlessness); (4) the immutability of the group’s defining trait; and (5) the relevancy of that trait.”

Marcy Strauss, Reevaluating Suspect Classifications, 35 Seattle U. L. Rev. 135, 146 (2011) [hereinafter Strauss, Suspect Classifications].
a history of past discrimination, 154 See Frontiero v. Richardson, 411 U.S. 677, 684 (1973) (“There can be no doubt that our Nation has had a long and unfortunate history of sex discrimination.”). political powerlessness, 155 See Carolene Prods., 304 U.S. at 153 n.4 (noting “prejudice against discrete and insular minorities . . . tends seriously to curtail the operation of those political processes ordinarily to be relied upon to protect minorities, and which may call for a correspondingly more searching inquiry”). the irrelevancy of a trait to an individual’s ability to contribute to or participate in society, 156 See Frontiero, 411 U.S. at 686 (“[W]hat differentiates sex from such nonsuspect statuses as intelligence or physical disability . . . is that the sex characteristic frequently bears no relation to ability to perform or contribute to society.”). and immutability. 157 See City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 442 (1985) (“They are thus different, immutably so . . . and the States’ interest in dealing with and providing for them is plainly a legitimate one.”). There remains some debate about whether immutability is truly one of the defining characteristics of suspect classifications status. See Ely, supra note 146, at 154–55 (arguing although trait is immutable, such as eyesight, it is not necessarily suspicious); see also id. at 150 (“[C]lassifications based on physical disability and intelligence are typically accepted as legitimate, even by judges and commentators who assert that immutability is relevant. The explanation . . . is that those characteristics . . . are often relevant to legitimate purposes. At that point there’s not much left of the immutability theory, is there?” (footnote omitted)). Setting immutability and relevance aside, 158 The Court has demonstrated that immutability is neither necessary nor sufficient to achieve suspect classification status. See Nyquist v. Mauclet, 432 U.S. 1, 9 n.11 (1977) (treating alienage as suspect classification even though alien status may be changed through naturalization); see also Cleburne, 473 U.S. at 445–46 (refusing to treat intel­lectually disabled as suspect class while acknowledging as group, they possess “immutable disabilities”).

As Jessica Clarke explains, a new doctrinal approach to immutability has emerged, which protects not those traits that are incidents of birth or unchangeable, but traits that are fundamental to individual identity. Clarke, supra note 62 (manuscript at 15, 21). This approach to immutability is arguably more problematic to those concerned about essentialism, given the correlating requirement that to receive heightened protection, individuals must argue that the identity characteristic at issue is “essential.” Id. (manuscript at 31). Clarke further argues that immutability fails to meaningfully distinguish between protected and unprotected traits and to adequately explain the reasons why discrimination against certain groups is wrong. Id. (manuscript at 34–35).

Moreover, the question of immutability becomes less relevant once the framework for equal protection does not rely on the privileging of a few select categories. The value-based model relies on the plaintiffs to define the relevant group for purposes of the analysis, which may include not only characteristics of the group itself but also a des­criptive account of the context in which plaintiffs allege that the discrimination occurred.
these criteria represent the substantive reasons for finding certain types of discrimi­nation particularly pernicious. Yet, rather than attempting to root out discrimination that directly implicates such criteria, and following in the lead of the “discrete and insular minority” model, the Court chose to carve out certain categories to represent these criteria and to apply heightened scrutiny to discrimination made along such categorical lines. 159 Susannah Pollvogt has similarly argued that courts should “assess[] the attributes of the laws, not the groups against which they discriminate.” Pollvogt, supra note 4, at 796. Yet the model she suggests would eliminate all group-oriented aspects of suspect classification, transforming the inquiry into an individualized “trait-relevancy analysis.” Id. at 801.

The argument made herein does not focus on immutability or relevance as key values underlying equal protection because both of those factors relate to traits rather than to structural causes of inequality. 160 Ford, Racial Culture, supra note 9, at 31 (arguing focus on difference diverts attention from structural inequality and racism is not “result of objective and intrinsic difference among natural racial groups” but “social institution based on a formal status hierarchy and a set of ideologies that justify that status hierarchy”). Immutability is not independently relevant, but it provides a proxy by which we can gauge personal responsibility for the denial of access. 161 “[T]he imposition of special disabilities upon the members of a particular sex because of their sex would seem to violate ‘the basic concept of our system that legal burdens should bear some relationship to individual responsibility.’” Frontiero, 411 U.S. at 686 (quoting Weber v. Aetna Cas. & Sur. Co., 406 U.S. 164, 175 (1972)). This Essay is concerned not with whether an individual should be held personally accountable for an aspect of her identity, or whether that trait justifies differential treatment, but instead with the structural forces that result in her being treated as different and the effects such treatment may have. 162 Although when viewed as a set of traits, race is one source of differentiation or distinction, and there are many other traits for which the same could be said—for example, hair color and left-handedness. It is the distinction in treatment based on a trait, and society’s response to that trait or traits, that provides reason for distinguishing between the different types of traits. See Lee, Topology of Race, supra note 9, at 772–73 (“Rather than determining whether a definition is oppressive based solely on its content, we can instead examine its effects.”). As Jack Balkin has explained:

The question to ask is not whether a trait is immutable, but whether there has been a history of using the trait to create a system of social meanings, or define a social hierarchy, that helps dominate and oppress people. Any conclusions about the importance of immutability already presuppose a view about background social structure. 163 Balkin, supra note 13, at 2366.

Similarly, the determination whether a trait is relevant to one’s contribution to society cannot be divorced from the way in which society itself has constructed the trait. 164 Take, for example, society’s changing views on sexual orientation. Whereas at one point society viewed homosexuality as relevant to one’s ability to engage in several aspects of societal life, it is increasingly viewed as irrelevant or benign. See Watkins v. U.S. Army, 875 F.2d 699, 725 (9th Cir. 1989) (Norris, J., concurring) (explaining sexual orientation has “no relevance to a person’s ‘ability to perform or contribute to society’”); William N. Eskridge, Jr., Sexual and Gender Variation in American Public Law: From Malignant to Benign to Productive, 57 UCLA L. Rev. 1333, 1349–52 (2010) (describing how treatment of sexual orientation in American law evolved from state-sanctioned discrimination to tolerable or benign trait). Moreover, because identity is a social construction that has been used to create and maintain subordination, 165 See López, White by Law, supra note 9, at 11 (explaining race is primarily function of meaning attributed to certain traits); Samuel A. Marcosson, Multiplicities of Subordination: The Challenge of Real Inter-Group Conflicts of Interest, 71 UMKC L. Rev. 459, 483 (2002) (arguing identity is socially constructed and shaped by those in dominant rather than subordinate position); id. (“[W]hites have been in greater control in the construction of race than African-Americans, and men have constructed gender so as to create and maintain subordination of women.”); supra note 9 and accompanying text (describing subordination’s role in creation of identity); cf. Goldberg, Anti-Essentialist, supra note 46, at 631–33 (debating utility of such arguments in litigation strategy, but acknowledging identity is socially constructed). a system organized around identity will inevitably incorporate its problematic roots. Therefore, the description of a trait in isolation—without regard to the role it has played in a larger social and political context—is not a helpful metric by which to assess when heightened scrutiny is truly warranted. 166 Moreover, as pointed out above, the universe of traits that could be deemed irrelevant to any given context is vast. The value-based approach does not place all of these traits on the same plane; rather, it prioritizes those that have played a more important role from a social and structural perspective. See Balkin, supra note 13, at 2323 (“The issue is social stratification based on traits, not the nature of the traits themselves.”). As Balkin explains:

“My central concern is with those status hierarchies where status identity is a central feature of one’s social existence, and affects many different spheres of one’s life. There may be a status hierarchy between skiers and snowboarders. Being a skier rather than a snowboarder, however, is not a central feature of one’s social identity. It is not something that affects many overlapping aspects of one’s everyday interactions with others, or that has ripple effects in various parts of one’s life, including wealth, social connections, political power, employment prospects, the ability to have intimate relationships and form families, and so on. By contrast, being a black person as opposed to a white person, or being female as opposed to being male, is a central feature of one’s identity, at least in contemporary America. It does affect a large percentage of one’s personal interactions with others, and it has many mutually supporting and overlapping effects.”

Id. at 2360. It is not so much identity that is pivotal, but the way in which certain aspects of identity mesh with external forces at work in society.

While other scholars like Bruce Ackerman 167 See supra note 151 (explaining how Carolene Products’s focus on “discrete and insular minorities” is ill-suited to increasingly pluralistic society, in which marginalization is not necessarily equated with low numbers or anonymity). and Susannah Pollvogt 168 Pollvogt has observed that the problem with suspect classification analysis is that it “asks the wrong question, and scrutinizes the wrong actor.” Pollvogt, supra note 4, at 798. She suggests that the problem with suspect classifications is that they focus on groups and, in doing so, are forced to make assumptions about groups that become “frozen” in the doctrine, resulting in a jurisprudence that is not sufficiently malleable or responsive. Id. at 798, 802. Another problem she identifies is that suspect classification factors like immutability and relevance “internalize rather than externalize subjective judgments of the worthiness of a group.” Id. at 798. have recognized the shortcomings of the Carolene Products approach, none have pinpointed the use of identity as proxy as the source of the problem. Due to the inextricable relationship between equal protection and identity, other scholars like Kenji Yoshino, honing in on the difficulties posed by an increasingly pluralistic society, have suggested a move away from equality altogether (and in Yoshino’s case, towards a liberty-based dignity analysis). 169 See, e.g, Yoshino, supra note 49, at 792–97 (noting liberty-based dignity analysis is more inclusive and “less likely to essentialize identity”). Ultimately, this Essay’s primary objection is not to the use of equality as a vehicle, or to the criteria used by the Court to define suspect classifications (although, certainly, other criteria might be used as well)—but only to the use of identity to embody such criteria. Hence the proposal that equal protection be grounded directly in such values, unconstrained by the strictures of identity.

Under a value-based framework, an individual need not align herself with an established group identity to receive a heightened level of scrutiny; nor need she carve out a new suspect classification. Furthermore, the touchstone for equality is not whether she has been granted or denied access to the same degree as a member of a different identity group who is otherwise similarly situated. Rather, she must show that the nature of the discrimination she has endured reflects one of the non-trait-based criteria triggering heightened scrutiny: a history of past discrimination or political powerlessness. Discrimination that implicates those factors should require strong justification in order to survive constitutional scrutiny.

Ultimately, a value-based framework adopts a structural view of equality. It is less concerned with how one individual is treated in relation to another, or about the nature of the trait on which the discrimination is based. That shift has implications not only for assessing harm, but also for the remedies that may be imposed. A focus on identity will lead to remedies that ensure that the state treats members of different identity groups in the same way. By contrast, this model’s focus on how social systems oppress—and on ensuring that the state does not contribute to discrimination that obstructs certain types of access—may be more likely to invite discussions of structural change. In other words, if the problem is framed not as treating an individual unfairly because of her gender, or the color of her skin, but as one instance of a larger phenomenon that has oppressed everyone in her position, a superficial response may be more likely to seem inadequate. 170 This Essay acknowledges the realist critique that framing alone may be unlikely to cause such a shift. That said, there is still value in constructing doctrine such that it forces engagement with these questions, rather than silently embedding them in the framework of identity. Moreover, a value-based approach avoids the current equal protection loophole that allows the state to deny everyone a benefit rather than provide it on equal terms. 171 For an illustration of such a loophole, see Palmer v. Thompson, 403 U.S. 217, 219 (1971) (holding municipality did not violate Constitution by closing formerly segregated community swimming pools rather than keeping them open as desegregated pools).

That said, a value-based framework does not necessarily diminish the impact that identity—and identity-based prejudice, such as racism or sexism—has on individuals. To the contrary, it is intended to give substantive meaning to distinctions premised on identity and to prevent identity from being doctrinally distorted so that it is no longer capable of providing legal protection to those who need it most. The fact that the value-based model rejects the use of identity as a vehicle for vindicating equal protection values should not be taken as a denial of its importance in other contexts. Identities such as race and gender have always played—and will continue to play—a critical role in creating collective identity and thus in mobilization and political organizing. But the fact that race and gender are such an important part of identity—both personally and politically—does not necessarily mean that they are always the best vehicles for providing legal redress.

Difficulties will inevitably arise in attempting to ascertain the factors that give rise to heightened scrutiny under such a model. For example, it would not be a simple task to determine the requisite level of historical discrimination or political powerlessness that would give rise to heightened scrutiny. 172 See Strauss, Suspect Classifications, supra note 153, at 151 (describing lack of clarity regarding history of past discrimination factor).

Moreover, questions will certainly arise as to the scope and context in which obstruction or adequate levels of access exist. For example, would such an approach render unsuccessful a claim where discrimination may exist more generally, but cannot be proven in the immediately relevant context? Cf. City of Richmond v. J.A. Croson Co., 488 U.S. 469, 511 (1989) (finding municipality failed to provide sufficient proof of identifiable past discrimination in city’s construction industry to justify race-based quotas for municipal construction contracts). Or might a claim be defeated by showing that, in a specific context, a minority racial group wielded an unusual amount of political power? See Ricci v. DeStefano, 557 U.S. 557, 598–604 (2009) (Alito, J., concurring) (suggesting city’s refusal to certify test results that would have resulted in disparate impact on minority firefighters resulted from exercise of political clout by powerful minority constituency). These are difficult questions this piece does not attempt to answer in their entirety; nor does it purport to be setting forth an approach that will reach a different or more favorable result (for plaintiffs) in every case. It is unlikely that the Court will be willing, under any framework, to act on a highly generalized showing of societal discrimination. But the model suggested here would at the very least provide an outlet for plaintiffs to demonstrate the impact of such generalized discrimination. And in the rare case where a minority constituency does wield political clout, the framework suggested herein would provide the majority with fewer tools to thwart the outcome of such a political process, under the assumption that: (i) the outcome is fairly anomalous in its impact; and (ii) the outcome does not, for example, provide grounds for the majority’s own claim of historical discrimination or is demonstrative of a larger obstacle to political access.
That is in part why a spectrum of protection—as Justice Marshall and Justice Stevens have described—would likely be a better fit for such a model than a scheme where rigid levels of scrutiny apply to rigidly defined categories. During his tenure on the Court, Justice Marshall repeatedly expressed his disagreement with the Court’s rigid approach to equal protection analysis, rejecting the application of what was at the time only two predetermined levels of scrutiny. 173 See, e.g., San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 98 (1973) (Marshall, J., dissenting) (“I must once more voice my disagreement with the Court’s rigidified approach to equal protection analysis.”). Instead, he advocated a “spectrum of standards” that would base the degree of care applied on the constitutional and societal importance of the interest affected and the character (or invidiousness) of the classification drawn. 174 Id. at 99; see also Craig v. Boren, 429 U.S. 190, 211–12 (1976) (Stevens, J., concurring) (“There is only one Equal Protection Clause. It requires every State to govern impartially. It does not direct the courts to apply one standard of review in some cases and a different standard in other cases.”). One could similarly envision here that the level of scrutiny applied would correlate to the severity with which access has been denied. 175 Applying tiers of scrutiny in a more rigid manner results in some of the same shortcomings as applying rigid categories of identity; it may force a need for consistency where parallels cannot be drawn and it may be over and underinclusive with respect to the policies it sweeps into the wake of heightened scrutiny. See Suzanne B. Goldberg, Equality Without Tiers, 77 S. Cal. L. Rev. 481, 487–91 (2004) (expressing frustration with rational basis review’s vacillation between deferential and meaningful review and with strict scrutiny’s “categorical use of rigorous review . . . regardless of context”). For example, if the history or pattern of discrimination alleged is particularly recent, significant in degree, or sweeping in its scope, 176 See, e.g., Juan F. Perea, Doctrines of Delusion: How the History of the G.I. Bill and Other Inconvenient Truths Undermine the Supreme Court’s Affirmative Action Jurisprudence, 75 U. Pitt. L. Rev. 583, 584–88 (2014) (arguing for race-based affirmative action to remedy government-sponsored race discrimination, relying on decades of discrimination against blacks encouraged and subsidized by federal government policy in education and housing). the level of scrutiny would likely be higher than if the history of past discrimination was smaller in scale and left less at stake.

This mode of analysis—eschewing a more formalized suspect classification inquiry in favor of a more functional approach—is not com­pletely absent from equal protection doctrine in its current form. Yet the cases employing such an approach are often seen as anomalous, rather than foundational. One such case is Hernandez v. Texas, 177 347 U.S. 475, 478 (1954). which involved a claim that state officials in Fort Bend County, Texas, had engaged in discrimination against Mexican Americans, systematically excluding them from jury service. 178 Ian Haney López & Michael A. Olivas, Jim Crow, Mexican Americans, and the Anti-Subordination Constitution: The Story of Hernandez v. Texas, Race Law Stories 279 (Rachel F. Moran & Devon Wayne Carbado eds., 2008). In analyzing the plaintiff’s claim, the Court acknowl­edged the need for a more flexible, contextualized approach, explaining:

Throughout our history differences in race and color have defined easily identifiable groups which have at times required the aid of courts in securing equal treatment under the laws. But community prejudices are not static, and from time to time other differences from the community norm may define other groups which need the same protection. 179 Hernandez, 347 U.S. at 478.

The Court also noted that the question “[w]hether such a group exists within a community is a question of fact.” 180 Id. In Hernandez, the Court determined that persons of Mexican descent being excluded from jury service constituted a “separate class” warranting equal protection. 181 Id. at 479–80. Having done so, the Court next asked whether the class had been sub­jected to differential treatment. 182 Id. at 480–82. López and Olivas explain that, as applied, the Hernandez Court “asked, first, whether the group seeking protection suffered from subordination generally and, second, whether the challenged practice amounted to a specific aspect of such oppression.” López & Olivas, supra note 178, at 291. Rather than analyze the case as one involving a racial classification or discussing the issue in racial terms, 183 López and Olivas explain that the parties on both sides of the Hernandez case classified those of Mexican ancestry as white. López & Olivas, supra note 178, at 291. Classifying Mexican Americans as white was also a means used by the Texas courts to defeat their equal protection claims. Id. at 298. the Court focused on the specific treatment of Mexican Americans in the community—for example, the fact that their involvement in business and community groups had been marginalized and that segregation measures either were or, until recently, had been in place. 184 Hernandez, 347 U.S. at 479; see also White v. Regester, 412 U.S. 755, 767–69 (1973) (describing how “historic and present condition” of Mexican American commu­nity, its cultural and economic marginalization, and legislature’s unresponsiveness to its interests, justified conclusion that Mexican Americans were “effectively removed” and “invidiously excluded . . . from effective participation in political life” (internal quotation marks omitted) (citation omitted)); López & Olivas, supra note 178, at 289–90 (describing Court’s analysis of Mexican American subordination within community). The Court’s reasoning in Hernandez, as Ian Haney López and Michael Olivas have argued, was grounded in group mistreatment and subordination: “The Hernandez test rests on opposition to group hierarchy: It focuses on status and subordination, without being distracted by the irrelevant questions of the exact nature of the group identity or the presence of discriminat­ory intent.” 185 López & Olivas, supra note 178, at 291–92. Thus, it is very much in line with the value-based approach.

Similarly, in Plyler v. Doe, the Court held that the undocumented status of Mexican children could not justify their exclusion from public schools. 186 457 U.S. 202, 226 (1982). Although the Court acknowledged that the children’s immigration status was not wholly irrelevant, it emphasized to a greater degree that the deprivation of a proper education would take an “inestimable toll . . . on the social, economic, intellectual, and psycholo­gical well-being of the individual” and stand in the way of individual achievement. 187 Id. at 222. In both cases, the Court focused primarily on the contextual impact of the claimed discrimination, rather than the nature of the identity at issue.

There will of course be difficulties in measuring the variables that are part of this analysis, just as there are under the current model. 188 For a more thorough discussion of how to best measure or assess political powerlessness, see generally Bertrall L. Ross II & Su Li, Measuring Political Power: Suspect Class Determinations and the Poor, 104 Calif. L. Rev. (forthcoming 2016) (manuscript at 5) (on file with the Columbia Law Review) (arguing court should rely on more holistic measure of political power), and Nicholas O. Stephanopoulos, Political Powerlessness, 90 N.Y.U. L. Rev. (forthcoming 2015) (manuscript at 3) (on file with the Columbia Law Review) (arguing group is relatively politically powerless if policy preferences are less likely to be enacted than those of similar group). But, whereas attempting to define an identity group results in negative externalities (in terms of identity and exclusion), attempting to define these variables has the potential positive effect of highlighting or exposing more subtle aspects of discrimination. Rather than assuming that general antidiscrimination objectives will be fulfilled by treating the members of certain pre-defined groups with heightened scrutiny, this analysis can be applied on a case-by-case basis to ensure both that the relevant criteria apply as accurately as possible and that the groups defined by the court are not over- or under-inclusive.

Eliminating the use of identity as proxy should not be equated, however, with the irrelevance of group-based discrimination. The fact that individuals are often discriminated against because of their affiliation with or relationship to a specific group remains highly relevant. In that regard, one important distinction from the current model centers on who defines the group at issue. Under the current framework, individuals are incentivized to associate themselves with groups that have been defined—and deemed worthy of heightened protection—by the courts. 189 See Magee Andrews, supra note 26, at 516 (“[T]oday, to allege a claim of race discrimination under the Equal Protection Clause, one must identify oneself as a member of a protected class . . . and satisfy the court that a particular practice or classification based on that class amounts to unlawful discrimination.”). In contrast, under the value-based model, the claimant or claimants define the contours of the group themselves and the court merely assesses the claim by applying the substantive criteria outlined above.

B. Impact of the Value-Based Approach

How might this model alter the mode of analysis for equal protection claims? One illustrative example emerges in reconceptualizing the 2007 case Parents Involved in Community Schools v. Seattle School District No. 1. 190 551 U.S. 701 (2007). At issue in Parents Involved were the school assignment plans formulated by school districts in Louisville, Kentucky and Seattle, Washington. Although the two districts had varying histories of racial discrimination—Louisville had been subject to a desegregation order but then found to have achieved unitary status 191 Id. at 715–16. and Seattle had never been subject to de jure racial discrimination 192 Id. at 712; see also id. at 806–07 (Breyer, J., dissenting) (“In Seattle, the plaintiffs alleged that school segregation reflected not only generalized societal discrimination and residential housing patterns, but also school board policies and actions that had helped to create, maintain, and aggravate racial segregation.”). —both districts were characterized by a substantial amount of de facto discrimination. 193 Id. at 806 (Breyer, J., dissenting) (“In both Seattle and Louisville, the local school districts began with schools that were highly segregated in fact.”). In response, and for the purpose of “eradicating earlier school segregation, bringing about integration, or preventing retrogression,” 194 Id. (Breyer, J., dissenting). the two school districts had developed plans that would take account of race in assigning students to schools.

Chief Justice Roberts authored the Court’s opinion declaring both plans unconstitutional under the Equal Protection Clause. 195 Id. at 748 (plurality opinion). The opinion first focuses on the fact that race was a determinative factor in assigning students to schools. 196 Id. at 723. Following in the path of past precedent, the Court found it unnecessary to consider whether the use of race by the plans was “benign” or “invidious.” 197 Id. at 741–42. Triggered by the use of race alone, the Court applied strict scrutiny and rejected the notion that “racial balancing” (even if labeled as “racial diversity”) could constitute a compelling interest to justify the use of race. 198 Id. at 730–33. The plurality opinion also criticized the means used by the school districts in implementing the plans, explaining that the school districts’ failure to consider race-neutral alternatives and their limited use of racial categorizations failed to satisfy strict scrutiny’s narrow tailoring requirement. 199 Id. at 733–35. In this context, the narrow tailoring requirement lends itself to a potential Catch-22. Should the school district’s use of race have too great an impact, that may suggest the use of racial classification is impermissible; however, if the use of race has too minimal an impact on school enrollment, that may also “cast[] doubt on the necessity of using racial classifications.” Id. at 734. Similarly, making distinctions among races (or choosing to be particularly concerned about particular racial distinctions—for example, in light of the unique history of a certain group in the context at hand) may raise concerns under the current approach. Id. at 723 (“Even when it comes to race, the plans here employ only a limited notion of diversity, viewing race exclusively in white/nonwhite terms in Seattle and black/‘other’ terms in Jefferson County.”).

Epitomizing a colorblind approach to equal protection juris­prudence, Chief Justice Roberts’s opinion famously declared, “[t]he way to stop discrimination on the basis of race is to stop discriminating on the basis of race.” 200 Id. at 748. This quote is also a manifestation of the race-as-trait view. If race had been viewed as the product of structural forces, the solution could not be as simple. Justice Sonia Sotomayor perhaps expressed this point most eloquently in her Schuette dissent: “The way to stop discrimination on the basis of race is to speak openly and candidly on the subject of race, and to apply the Constitution with eyes open to the unfortunate effects of centuries of racial discrimination.” 201 Schuette v. Coal. to Defend Affirmative Action, 134 S. Ct. 1623, 1676 (2014) (Sotomayor, J., dissenting). If the problem is characterized simply as discrimination on the basis of identity, then it can be remedied by ceasing to discriminate on the basis of identity. If, in contrast, as Justice Sotomayor suggests, the problem runs much deeper, a solution that aims at a more superficial level will fail to address the underlying causes of discrimination.

Applying a value-based inquiry, the analysis of the equal protection question posed by Parents Involved would be quite different. Rather than focusing on the use of race (the relevant identity category) as the triggering factor, the Court would look to the plaintiff or plaintiffs to demonstrate how the discrimination experienced by them (or by the class that they defined) implicated the values underlying equal protection. For example, the plaintiffs would have to show that the student assignment plans stemmed from or perpetuated a history of past discrimination against the plaintiff class or affected the class’s ability to access the political process as a means to provide for legal protection. In doing so, the plaintiffs would have to first articulate the nature of the discrimination and its effect—here, such a description would have to revolve around the treatment of those who identify as white. Interestingly, Chief Justice Roberts’s opinion makes no mention of the plaintiffs’ race, never specifically identifying them as white. 202 Cf. Wendy Parker, Recognizing Discrimination: Lessons from White Plaintiffs, 65 Fla. L. Rev. 1871, 1873 (2013) (discussing Parents Involved as one example of cases involving white plaintiffs); cf. also Justin Driver, Recognizing Race, 112 Colum. L. Rev. 404, 409 (2012) (exploring how and why judges should identify individuals racially in context of their opinions). This might be taken to suggest that the specifics of the plaintiffs’ race are irrelevant; all that matters is that identity is at stake. In that sense, Chief Justice Roberts’s opinion epitomizes not only colorblindness, but also the problem with the categorical approach.

Under the model proposed herein, a white plaintiff could no longer make a claim based on “race” alone. Given the history of Louisville and Seattle, as outlined in detail in Justice Breyer’s dissent, it is unlikely that such a plaintiff would be able to argue that the plans’ operation somehow exacerbated or perpetuated a history of discrimination against white students in the relevant jurisdiction. 203 Cf. Deborah Hellman, When Is Discrimination Wrong? 80 (2008) (discussing Ronald Dworkin’s critique of Allan Bakke’s claim and suggesting that, to extent Bakke’s claim was exclusion on basis of racial prejudice or contempt, “[t]here is no history of whites qua whites being excluded that gives this interpretation traction”). It is also unlikely she would be able to argue that her assignment obstructed or diluted her ability to effectively utilize the political process. 204 Although not dispositive, it would likely be relevant here that both Seattle and Louisville are at least seventy percent white. U.S. Census Bureau, American Community Survey Demographic and Housing Estimates: 2005–2007 (on file with the Columbia Law Review) (estimating population of Seattle, Washington, to be 73.6% white and population of Louisville/Jefferson County, Kentucky, to be 77.4% white from 2005 to 2007).

The focus of this approach takes on particular significance when applied to the contributions made by Justice Thomas’s and Justice Breyer’s opinions in Parents Involved. In his concurrence, Justice Thomas wrote:

Disfavoring a colorblind interpretation of the Constitution, the dissent would give school boards a free hand to make decisions on the basis of race—an approach reminiscent of that advo­cated by the segregationists in Brown v. Board of Education. This approach is just as wrong today as it was a half century ago. 205 Parents Involved, 551 U.S. at 748 (Thomas, J., concurring) (citation omitted).

As the dissenters emphasized, Justice Thomas’s approach ignores the vastly different context in which the two cases were decided. In contrast, Justice Breyer’s seventy-seven-page-long dissent traced the history both of the school districts themselves and of the doctrine, emphasizing the differences in nature between de jure segregation and the actions taken by Seattle and Louisville’s school boards. In conclusion, Justice Breyer wrote:

The lesson of history is not that efforts to continue racial segregation are constitutionally indistinguishable from efforts to achieve racial integration. Indeed, it is a cruel distortion of history to compare Topeka, Kansas, in the 1950’s to Louisville and Seattle in the modern day—to equate the plight of Linda Brown (who was ordered to attend a Jim Crow school) to the circumstances of Joshua McDonald [plaintiff in the instant case] (whose request to transfer to a school closer to home was initially declined). This is not to deny that there is a cost in applying “a state-mandated racial label.” But that cost does not approach, in degree or in kind, the terrible harms of slavery, the resulting caste system, and 80 years of legal racial segregation. 206 Id. at 867 (Breyer, J., dissenting) (citations omitted).

By shifting the focus of the inquiry to the substantive criteria suggested above, and away from a more superficial inquiry about race, the Court would be forced to grapple with the evidence Justice Breyer sets forth; Justice Thomas’s comparison of the Louisville and Seattle school boards to the actions of segregationists then becomes far less tenable. This is not to say that the result would necessarily be different, particularly given the current composition of the Court, but the tenor and focus of the Court’s primary opinion would surely reflect a different level of engagement.

Another area in which the shift from a categorical model to a value-based model might have a significant impact is in the voting rights context. As Richard Hasen has pointed out, in the context of analyzing the legality of voting regulations, race and party often coincide and yet laws seen as discriminating on the basis of political party are likely to stand whereas those based on race are likely to fall. 207 Richard L. Hasen, Race or Party?: How Courts Should Think About Republican Efforts to Make It Harder to Vote in North Carolina and Elsewhere, 127 Harv. L. Rev. Forum 58, 61 (2014) (“[I]f courts call this a law about party politics and view it through the lens of partisan competition, then the law is more likely to stand . . . . If courts call this a law about race . . . . then the law is likely to fall . . . .”). Hasen suggests in response that rather than focus on race—for example, asking whether laws requiring potential voters to present identification are racially motivated—the Court should apply a standard akin to strict scrutiny (“strict scrutiny light”) to voting laws that discriminate against a party’s voters or otherwise burden voters. 208 Id. at 71–72 (arguing legislature passing burdensome laws should be required to “produce real and substantial evidence that it has a good reason for burdening voters and that its means are closely connected to achieving those ends”). Hasen acknowledges that some may perceive his argument as not giving race “a sufficiently explicit role in policing elections”; his response is that the Court is more likely to protect the voting process from partisan manipulation than it is to endorse laws policing racial discrimination in voting. Id. at 73. Hasen’s proposal is similar to the value-based model in that it relies on a substantive (or factual) rather than an identity-based triggering inquiry. In other words, the primary task is not to determine whether identity category X is involved or serves as the motivation for the law, 209 Hasen suggests one problem with this approach is that it requires courts to “make decisions about what is in legislators’ hearts” and that “[a] search for racist intent is not the most productive way to think of these issues.” Id. at 71. but instead whether the law has some detrimental effect on access to the political process and political power.

The debate surrounding majority–minority districts is another voting rights issue that raises important questions about the doctrine’s reliance on identity. The reliance on majority–minority voting districts as a means of protecting minority voting rights relies to a large extent on the assumption that minority voters are monolithic in their voting preferences. 210 See Janai S. Nelson, White Challengers, Black Majorities: Reconciling Competition in Majority-Minority Districts with the Promise of the Voting Rights Act, 95 Geo. L.J. 1287, 1297–98 (2007) (describing how reliance on “political cohesion” as necessary element to establish violation of Voting Rights Act penalizes blacks for defying monolithic perception of their political preferences). The Court recognized this tension in Shaw v. Reno, finding that a redistricting plan that aggregated minority voters in majority–minority districts “reinforce[d] the perception that members of the same racial group—regardless of their age, education, economic status, or the community in which they live—think alike, share the same political interests, and will prefer the same candidates at the polls.” 211 509 U.S. 630, 647 (1993). Jayne Chong-Soon Lee notes how, in her opinion, Justice O’Connor “limits race to a biological definition, and evokes the opposition between the biological and the social to undermine the validity of race-consciousness, and thus of race-conscious remedies.” Lee, Topology of Race, supra note 9, at 776. Lee goes on to explain that “[t]he [Shaw] majority’s argument that all racial classi­fications cause harm depends on the conflation of biology and race and the use of only one definition of race, and invites us to view every acknowledgement of race as racism.” Id. Given the redistricting legislation’s clear reliance on racial stereotypes, its remedial aims were not enough to avoid strict scrutiny under the Equal Protection Clause. 212 Shaw, 509 U.S. at 642–44, 653. The Court’s holding in Shaw was consistent with a broader understanding of identity, yet in pushing back against a particular narrative, the decision constructed a barrier for those attempting to protect minority voting rights. Had the Court been less focused on preserving a particular conception of racial identity, and more on the impact of the law on political exclusion, its analysis surely would have been different. 213 See, e.g., id. at 666 (White, J., dissenting) (“[I]t strains credulity to suggest that North Carolina’s purpose in creating a second majority-minority district . . . was to discriminate against members of the majority group by ‘impair[ing] or burden[ing their] opportunity . . . to participate in the political process.’” (quoting United Jewish Orgs. of Williamsburgh, Inc. v. Carey, 430 U.S. 144, 179 (1977))); id. at 681–82 (Souter, J., dissenting) (“In districting . . . the mere placement of an individual in one district instead of another denies no one a right or benefit provided to others.”).

The last context in which this shift might have an impact is with regard to those characteristics that align with existing suspect classifications but fall short of precise correlation. One such example is language discrimination. In Hernandez v. New York, a prosecutor had struck Spanish-speaking bilingual jurors, claiming they would not be able to accept the interpreter’s version of testimony given by Spanish-speaking witnesses. 214 500 U.S. 352, 356–57 (1991). The Court ultimately held that the prosecutor’s reason for striking the jurors was race-neutral, in part because both categories at issue—those who might have difficulty accepting the translator’s version of Spanish-language testimony and those who would not—would include Latinos. 215 Id. at 361. This reasoning is analogous to the relationship between pregnancy and gender as analyzed in Geduldig v. Aiello. See infra notes 224–231 and accompanying text (discussing Geduldig as emblematic of Court’s need to associate group with established suspect classification before affording heightened scrutiny). Because, in the Court’s view, language did not operate here as a “surrogate for race,” there was no equal protection violation. 216 Hernandez, 500 U.S. at 371. Under a value-based model, there would be no similar need to align a trait such as language with race or any other established suspect classification. 217 Similar to the argument made here, Richard Ford has argued against a trait-based approach to rights. See Ford, Racial Culture, supra note 9, at 125 (“[G]roup difference is not intrinsic to members of social groups but rather contingent of the social practices surrounding group identification . . . . [A]nti-discrimination law need not ‘protect group traits’ in order to prohibit discrimination on the basis of group statuses.”). Ford explains that protecting groups on the basis of a trait necessarily requires a decision “about the merits or demerits of the trait or practice” in addition to a decision “about the merits or demerits of encouraging the association of the trait or practice with a[n] ascribed social identity or status.” Id. Instead, the focus of the equal protection inquiry would be on the nature of the discrimination at issue. Thus, a plaintiff might argue that the group at issue has experienced a history of discrimination based specifically on language, 218 See Perea, Buscando América, supra note 132, at 1426–46 (describing long history of language-based subordination of Spanish-speaking individuals in United States). or that the alleged discrimination has a negative impact on their ability to engage in civic participation (e.g., jury service).

One might ask how this approach is any different from antisubordination theory, which would not invalidate just any use of race, but only those which serve to oppress or subordinate members of a particular racial group. 219 See, e.g., Ruth Colker, Anti-Subordination Above All: Sex, Race, and Equal Protection, 61 N.Y.U. L. Rev. 1003, 1007–08 (1986) (“From an anti-subordination perspective, both facially differentiating and facially neutral policies are invidious only if they perpetuate racial or sexual hierarchy.”); Owen M. Fiss, Groups and the Equal Protection Clause, 5 Phil. & Pub. Aff. 107, 108 (1976) (proposing “group-disadvantaging principle” because it “represent[s] the ideal of equality,” “takes a fuller account of social reality,” and “focuses the issues that must be decided in equal protection cases”); Cass R. Sunstein, The Anticaste Principle, 92 Mich. L. Rev. 2410, 2411–12 (1994) (forbidding “social and legal practices from translating highly visible and morally irrelevant differences into systemic social disadvantage, unless there is a very good reason”). In substance, the value-based model is in fact very similar to the antisubordination approach; it may even be thought of as a way to operationalize antisubordination theory. As often understood and applied, however, antisubordination may be subject to some of the same critiques as described in Part I above. Antisubordination does not necessarily eschew the categories of race and gender as commonly defined; rather, it embraces their role as proxy for a set of specific social and institutional experiences and approves of action on the basis of those categories under certain circumstances 220 See Colker, supra note 219, at 1009–10 (“Anti-subordination proponents there­fore advocate the use of race- or sex-specific policies, such as affirmative action, when those policies redress the subordination of racial minorities or women.”). (in contrast to anticlassifi­cation’s wholesale disapproval of action on the basis of such categories). While the approach advocated herein shares the ultimate goal of dismantling existing hierarchies and combating subordination, it diverges from antisubordination—and other approaches—in using identity categories as the doctrinal means for achieving that end. In other words, rather than assuming that acting on the basis of race or gender will address the subordination problem, it would tackle the subordination question directly.

Ruth Colker uses the following as a demonstrative example of anti­subordination’s impact: “[A] policy excluding persons who have primary child care responsibilities from consideration for employment, although phrased in sex-neutral terms, would have a disparate impact on women. It would also perpetuate a history of sexual hierarchy by penalizing women for their societally imposed child care responsibilities.” 221 Id. at 1008 n.14. Thus, she suggests, the policy would be viewed as invidious under an anti­subordination approach. 222 See id. at 1007–08 (arguing policies are invidious under an antisubordination perspective when they perpetuate racial or sexual hirerarchy). The analysis and result of the instant case—depending on the court’s willingness to apply current doctrine as Colker suggests—could be the same under either the categorical or a value-based approach. The point is that application of the identity framework adds little benefit; for purposes of the instant policy, the plaintiffs could define themselves as women who bear the primary responsibility for child care, thereby making a clearer case for contextual discrimination and avoiding the implication (or further entrenching the assumption) that bearing the primary responsibility for child care is an essential trait of womanhood. 223 Another point of clarification: The approach offered here is—or can be—just as much a group-based approach as antisubordination; it need not be interpreted in the individualistic mold of anticlassification or antidifferentiation theory. Cf. id. at 1008 (“In contrast to the anti-differentiation approach, the anti-subordination perspective is a group-based perspective . . . .”). The primary distinction is that under the nonidentity narrative, the claimants define the relevant group for purposes of the claim. This need not under­mine the notion that certain groups are doctrinally relevant because of their historically or legally subordinate status; a plaintiff or group of plaintiffs wishing to draw on that history may still do so under the substantive criteria outlined herein.

The last point serves as an effective segue to another case that has proved problematic under the current framework: Geduldig v. Aiello. 224 417 U.S. 484 (1974). Although effectively overruled by congressional enactment of the Pregnancy Discrimination Act of 1978, 42 U.S.C. § 2000e(k) (2006), Geduldig has never been overruled by the Supreme Court. At issue in Geduldig was California’s disability insurance program, which paid benefits to those temporarily disabled from work, but excluded pregnancy-related disabilities. 225 See Geduldig, 417 U.S. at 488–89 (stating issue in case is exclusion of disabilities relating to pregnancy under § 2626 of Unemployment Insurance Code). The Court held that the statute did not exclude anyone from benefits on the basis of gender, but “merely remove[d] one physical condition—pregnancy—from the list of compensable disabilities.” 226 Id. at 496–97 n.20. While the opinion conceded that only women could become pregnant, it declined to equate classifications based on pregnancy with sex-based classifications. 227 See id. (“While it is true that only women can become pregnant, it does not follow that every legislative classification concerning pregnancy is a sex-based classification . . . .”). The quandary raised by Geduldig is a manifestation of identity-based jurisprudence: Because heightened protection can only be afforded to those who fall within a specific identity category, the primary inquiry is whether the trait at issue can be wholly associated with a protected identity category. As the Court emphasized:

The lack of identity between the excluded disability and gender as such under this insurance program becomes clear upon the most cursory analysis. The program divides potential recipients into two groups—pregnant women and nonpregnant persons. While the first group is exclusively female, the second includes members of both sexes. 228 Id. A more recent manifestation of the same reasoning can be observed in Coleman v. Court of Appeals of Md., 132 S. Ct. 1327 (2012), in which the Court held that the self-care provision of the Family and Medical Leave Act of 1993 (FMLA), which is particularly relevant for women needing leave from employment for pregnancy or maternity-related reasons, was not directed at an identified pattern of gender-based discrimination and was not congruent and proportional to any pattern of sex-based discrimination by the States; thus, it was an unconstitutional exercise of Congress’s Section 5 power under the Fourteenth Amendment. See id. at 1330–31; cf. id. at 1344–45 (Ginsburg, J., dissenting) (suggesting Court “revisit” its conclusion that “discrimination on the basis of pregnancy is not discrimination on the basis of sex”).

Thus, the requirement that protection be based on identity means that a trait on which discrimination is alleged must either equate to identity or it is unlikely to provide a basis for unlawful discrimination. Given this binary analysis, the Geduldig Court was able to conclude as a logical matter that not all women are pregnant or will become pregnant; thus discrimination on the basis of pregnancy does not equate to gender discrimination and does not warrant heightened scrutiny. 229 See Geduldig, 417 U.S. at 496–97, 496 n.20 (concluding given “lack of identity between the excluded disability and gender,” state’s decision not to include such disability in its insurance program did not “discriminate against any definable group or class” and thus was not valid equal protection claim). To equate pregnancy and gender has the potential to essentialize all women. 230 See Vicki Schultz, Taking Sex Discrimination Seriously, 91 Denv. U. L. Rev. 995 (forthcoming 2015) (manuscript at 1080) (on file with the Columbia Law Review) (“In portraying pregnancy as a quintessentially female experi­ence, the uniqueness view presented an artificial sense of unity among women around their reproductive experiences, an exaggeration that later became the subject of extensive critique by other feminists, including feminists of color.”). Yet to refuse such an association denies the fact that those affected most directly by the exclusion will inevitably be women. Under the value-based model, such a dilemma would be avoided, as the plaintiff class could contextualize its own identity and avoid the need to conform the classification at issue to a broader identity-based framework. 231 For example, women who feel they have been discriminated against in the workplace because they have children could bring a claim without having to account for the fact that other women in the same workplace may not have children, therefore avoiding the problem that the alleged discriminatory conduct does not correlate perfectly with gender. If the former group could show that women who are also mothers have experienced a unique history of discrimination in the context at hand, heightened scrutiny would apply to the alleged act of discrimination.

Perhaps the greatest fear raised by the shift from a categorical model to a value-based approach is the notion that identity provides cover for those who on their own could not establish a substantive claim to relief, but have the power to do so as part of a larger group. In other words, the current framework has done most of the work for those who can simply find a way to wedge themselves into a covered class. While a justifiable concern, it is not clear that the current framework actually offers much protection for those most in need of such protection. While such individuals may still be clambering for cover under the umbrella of equal protection, the Court is not interested in expanding the umbrella’s shade. And in fact, it may be that those previously given cover are increasingly being crowded out as the umbrella is repurposed to provide protection only to a privileged few. 232 See Cheryl I. Harris & Kimberly West-Faulcon, Reading Ricci: Whitening Discrimination, Racing Test Fairness, 58 UCLA L. Rev. 73, 109–11 (2010) (arguing Ricci Court shifted doctrinal rules to favor white plaintiffs alleging discrimination); Girardeau A. Spann, Disintegration, 46 U. Louisville L. Rev. 565, 592–93 (2008) (arguing Parents Involved overruled Brown and sacrificed interests of racial minorities for benefit of disgruntled white plaintiffs); see also Butler, Justice Rather than Rights, supra note 46 (arguing current Court “seems friendlier to gays than to people of color”).

C. Operationalizing the Value-Based Approach

Given that the Court is unlikely to radically alter its framework for equal protection analysis, this Essay offers a few brief thoughts on several other vehicles through which a value-based approach might be operationalized or provide a valuable contribution nonetheless.

First, in litigating such issues in the lower courts, plaintiffs could adopt such an approach, drawing on the substantive underpinnings of cases like Carolene Products and Frontiero v. Richardson. Rather than arguing for recognition as a suspect class, however, plaintiffs might request that the court simply apply the substantive criteria (without using an identity filter) in the specific context of the claim at issue. In doing so, a plaintiff would not be attempting to contradict Supreme Court precedent—by asking, for example, that the court apply strict scrutiny when it is clear that intermediate scrutiny applies to the category—but instead, the plaintiff would be asking the court to rely on the reasoning underlying existing precedent as applied to the situation at hand. In other words, a plaintiff alleging discrimination as a single, pregnant mother would allege not that she had been discriminated against on the basis of gender but rather that she, and the class she represents, have experienced a history of discrimination in the context at hand or are not in a position to effectively achieve change through the political process. Thus, she would argue, the court should require a greater level of justification for government action taken to disadvantage the class at issue. While many courts would likely impose the dominant identity-based framework regardless, a more receptive court could entertain plaintiffs’ arguments without running directly afoul of Supreme Court precedent.

Alternatively, outside of the litigation context, policymakers might attempt to take advantage of the fact that the doctrinal model suggested here follows a different inquiry than that required by the existing model. For example, rather than using proxies or correlates for race, such as socioeconomic status, officials attempting to design an affirmative action plan might use the substantive criteria set forth here. This would both shield the plan against constitutional attack (because it does not use “race” as a triggering factor) and flesh out meaningful differences that cannot be addressed even under a race-based preference plan. 233 See Onwuachi-Willig, supra note 113, at 1160–61 (discussing how traditional affirmative action policies do not differentiate between first-generation black immigrant students, second-generation black American students, and mixed-race students with one black parent among larger group of “Black” applicants); see also Brown & Bell, supra note 113, at 1229–30 (discussing how traditional affirmative action policies do not differentiate between “Black/White Biracials,” “Black Immigrants,” and “Ascendants”). While the constitutionality of race-neutral affirmative action programs that serve as a mere façade for race-based preferences may be questionable, 234 See Brian T. Fitzpatrick, Is the Future of Affirmative Action Race Neutral?, in A Nation of Widening Opportunities: The Civil Rights Act at Fifty (Samuel Bagenstos & Ellen Katz eds., forthcoming) (manuscript at 17–18 & 17 n.61) (on file with the Columbia Law Review) (arguing Court’s colorblind approach and focus on intent suggest “race-neutral affirmative action” (i.e., programs designed to increase minority representation without directly invoking race) must also be subject to strict scrutiny). a policy that aims to satisfy the substantive criteria to applicants as they present themselves may have a better chance of survival. For example, a policy might rely more heavily on personal statements and holistic determinations of an applicant’s relative privilege or level of access—perhaps by requesting more information regarding the applicant’s family history—rather than racial markers standing in as a proxy for an essentialized experience.

On the legislative front, within the confines of its Section 5 enforcement power, 235 U.S. Const. amend. XIV, § 5 (“The Congress shall have power to enforce, by appropriate legislation, the provisions of this article.”); see also City of Boerne v. Flores, 521 U.S. 507, 517–19 (1997) (explaining Congress may act to prevent or remedy constitutional violations, but not to determine what constitutes a constitutional violation). Given Boerne’s requirement that Congress act in accordance with the Court’s constitutional interpretations, it would be helpful that the value-based framework ultimately relies on the Court’s own underlying rationales for applying heightened scrutiny to certain suspect classifications. See id. at 519 (“[Congress] has been given the power ‘to enforce,’ not the power to determine what constitutes a constitutional violation”); see also infra notes 255–256 and accompanying text (describing how value-based model redeploys existing elements of existing framework rather than starting anew). Congress could legislate to provide protection to groups not adequately captured by the identity categories often utilized in antidiscrimination law. Much as Congress responded to the Court’s decision in Geduldig by statutorily amending the definition of sex discrimination to include discrimination on the basis of pregnancy, childbirth, or related medical conditions, 236 Insofar as it would attempt to protect categories not perfectly aligned with existing suspect classes, this response would be similar to the legislative “fix” provided by the Pregnancy Discrimination Act (PDA), issued after the Court’s decision in Geduldig. See 42 U.S.C. § 2000e(k) (2006); see also supra note 224. it could statutorily prevent discrimination that thwarts political access or exacerbates existing patterns of historical discrimination.

Last, the value-based model also has potential to influence important debates occurring with regard to state legislation. For example, in the affirmative action context, though many lament its inevi­table demise in the Supreme Court, 237 See, e.g., Kermit Roosevelt III, The Ironies of Affirmative Action, 17 U. Pa. J. Const. L. 729, 729–30 (2015) (describing how under Court’s recent approach to affirmative action, “almost all race-based affirmative action programs are likely unconstitutional”); see also López, Intentional Blindness, supra note 43, at 1782 (“Given the emergence of Justice Anthony Kennedy as the swing vote in racial cases, there is also good reason to fear that the Court will soon end affirmative action in higher education.”). several states, including California and Michigan, have already prohibited affirmative action in multiple contexts. 238 Cal. Const., art. I, § 31(a) (originally Proposition 209) (“The State shall not discriminate against, or grant preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity, or national origin in the operation of public employment, public education, or public contracting.”); Mich. Const. art. I, § 26 (Proposal 2) (prohibiting affirmative action in public education, employment, and contracting). To the extent that the battle over affirmative action will be fought in the states through popular initiatives, referenda, and legi­slation, the value-based approach provides supporters of affirmative action with new rhetoric and a different way to frame those debates—one that is less susceptible to popular appeals toward colorblindness.

III. Advantages and Critiques

This Part endeavors to explore some of the possible advantages and critiques of the model described in Part II.

Section III.A describes the advantages of a value-based model, including: (1) the capacity to recognize important distinctions both within and among different groups within a broadly constructed identity category; (2) flexibility in its case-based and more contextual approach; (3) the ability to address intersectionality concerns; (4) eradication of the comparative and intent elements of existing equal protection doct­rine; (5) improved utilization of the judicial function in asking courts to analyze empirical data rather than define or police identity; and (6) its important discursive function in changing the rhetoric used to talk about discrimination.

Section III.B attempts to address some of the critiques that might be levied against such an approach, including the arguments that it deemphasizes the importance of identity and persistent overt discrimi­nation and may be susceptible to other forms of co-option.

A. Advantages of a Value-Based Approach

One of the key advantages of a framework that eschews identity as a guiding principle is its ability to make distinctions between groups and within groups. It can account both for differences in historical treatment—for example, between blacks and whites—and for the different ways in which groups may be treated given the context. For example, while the current model may struggle to distinguish within the larger group of Asian Americans, a value-based approach could distinguish between the experiences of a Chinese American woman in Los Angeles and a Hmong woman in central California. 239 Cf. Onwuachi-Willig, supra note 113, at 1143 (“Scholars have examined how the model minority myth, in particular the view of Asian-Americans as a monolithic group, may have a negative impact on affirmative action policies for Asian-American students, especially those who are of Cambodian, Hmong, Laotian, and Vietnamese descent.”). To provide another example, some scholars have argued that in the context of affirmative action, all blacks are not created equal 240 Id. at 1157 (describing, in context of admissions, not all blacks are created equal and studies have shown “educational, economic, and cultural differences between legacy Blacks and non-legacy Blacks”). and that “ascendant” blacks—those with two native-born black parents or a more direct connection to America’s history of racial discrimination—should be treated differently from multiracials with some black heritage or black immigrants. 241 See Brown & Bell, supra note 113, at 1231 (asserting consideration of blacks as one racial group for admission purposes creates underrepresentation of blacks whose ethnic and racial heritage can be traced to historical oppression of blacks in the United States). Kevin Brown and Jeannine Bell contend that treating all of the groups as “[b]lack,” without any further categorization, undermines the original goals of affirmative action and leads to underrepresentation of blacks with a traceable connection to historical oppression in the United States. 242 Id. Angela Onwuachi-Willig has observed similarly that “legacy blacks”—those with four grandparents born in the United States and descended from American slaves—are underrepresented as compared to their first and second generation counterparts at elite colleges and universities. Onwuachi-Willig, supra note 113, at 1160. She maintains, however, that there are important reasons for the continued inclusion of first and second-generation blacks and mixed-race students. Id. at 1180. Regardless of one’s opinion on the merits, a framework not based on a monolithic version of identity would encourage such discussions, guided not by identity but instead by a substantive set of priorities.

The value-based model also offers more flexibility and does not require courts to draw sweeping or generalized conclusions. The Court has been extremely hesitant to expand the number or scope of suspect classifications, 243 Yoshino, supra note 49, at 755 (noting in past several decades, Court has limited its equal protection jurisprudence in “at least three ways—it has limited the number of formally protected classifications, it has curtailed its solicitude for classes within already protected classifications, and it has restricted Congress’s power to enact antidiscrimination legislation”); id. at 757 (“At least with respect to federal equal protection jurisprudence, this canon has closed.”). in part because of the broad impact that making such a finding would necessarily have under the current regime. 244 Under the current model, once a class is deemed suspect, discrimination against that class is treated with heightened scrutiny regardless of the deprivation at issue. Another concern the Court has expressed is that beyond those categories already established as suspect, it would be difficult to draw principled distinctions. City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 445 (1985) (“[I]f the large and amorphous class of the mentally retarded were deemed quasi-suspect for the reasons given by the Court of Appeals, it would be difficult to find a principled way to distinguish a variety of other groups . . . .”). Under a more contextualized model, conclusions made with regard to a plaintiff class would not necessarily dictate the result in every future case, possibly giving courts more freedom to recognize discrimination without fearing the consequences that may follow. 245 Of course, even with such a contextual approach, some efficiencies will inevitably arise in the process. This, however, may actually prove useful: For example, data produced in one case to demonstrate a history of discrimination might be relied upon by other cases as well. In contrast to the current regime, whose concept of identity is fixed in a period of time that does not reflect current realities, a value-based model offers more flexibility and can adapt to evolving notions of identity and the increasing legal recognition of social and cultural pluralism. Perhaps more important, it can adapt to the way in which society’s response to such changes and discrimination itself may evolve.

There are some experiences that everyone raced as black or classified as female will likely share, but others that will vary widely among members of the group. Because the value-based model allows the group to define itself on its own terms, there is less risk of over or underinclusion. It also avoids the identity harms that can result from forcing individuals into categories that do not align with their own self-conception. 246 See Lucas, Undoing Race, supra note 66, at 1267–68 (describing psychological and emotional harm generated by refusing multiracial individuals ability to identify themselves as multiracial). For example, there are some individuals, such as those of Middle Eastern descent, who are legally required under the current framework to identify as white but who “argue that their experiences with race discrimination make their experiences more similar to racial minorities.” 247 Camille Gear Rich, Elective Race: Recognizing Race Discrimination in the Era
of Racial Self-Identification, 102 Geo. L.J. 1501, 1543 (2014) [hereinafter Rich, Elective Race].
The same difficulty confronts Latinos who “resist being categorized as white.” 248 Id. Juan Perea has argued for heightened protection against language discrimination, given the long history of linguistic subordination of Spanish speakers in the United States. Perea, Buscando América, supra note 132, at 1425 (“Indeed, for Latinos and other language minority groups, language discrimination is race discrimination, and courts should treat it as such.”). Under a value-based model, one would not be forced to compare or equate language and race discrimination, but could still apply heightened scrutiny to language discrimination, based on the unique contours of its history. See id. at 1426–38 (demonstrating how “Spanish speakers, historically and in the present, have been treated as inferior and discriminated against by English-speaking America”). Those who do not “feel that the current configuration of racial categories adequately describes their personal (private) views about race . . . are forced to describe themselves imperfectly, and they do so in ways that may cause problems if they later raise discrimination claims.” 249 Rich, Elective Race, supra note 247, at 1542. Similarly, individuals whose gender iden­tity and biological gender diverge need not align themselves with one side of a binary gender framework. 250 See Terry S. Kogan, Transsexuals in Public Restrooms: Law, Cultural Geography and Etsitty v. Utah Transit Authority, 18 Temp. Pol. & C.R.L. Rev. 673, 686 (2009) (“Social, architectural, and legal norms combine to provide the structural and spatial framework through which society . . . enforce[s] its cultural commitment to the binary vision of sex/
gender, engendering a system that has little flexibility to encompass people who fail to fit neatly in one of the two categories.”).
Given its emphasis on experience over label and function over form, 251 Camille Gear Rich, Affirmative Action in the Era of Elective Race: Racial Commodification and the Promise of the New Functionalism, 102 Geo. L.J. 179, 185–86 (2013) (describing functional approach toward racial categorization, which aims to incorporate individuals’ unique and diverse experiences with racialization). the framework offered here provides recourse for those in a racially or gender-liminal position. 252 Camille Gear Rich uses the term “racial liminals” to refer to those who conscien­tiously object to American definitions of racial identity. See Rich, Elective Race, supra note 247, at 1542.

The more contextualized and nuanced nature of the value-based approach would also make it more effective in addressing concerns regarding intersectionality. Whereas the categorical model is either exclusionary of intersectionality concerns or forces individuals to conform their claims to the single-axis framework, 253 See supra notes 86–88 and accompanying text (describing dilemma posed by current doctrine for individual wishing to make claim based on her specific treatment as black woman). the value-based approach empowers claimants to describe the identity of the individual or group and describe the contours of the discrimination at issue. In doing so, it provides a forum for claimants to explore the intersection of various forms of discrimination or oppression and to show how, in some cases, that intersection may lead to even more pernicious discrimination, with correlating results. Such discrimination would otherwise be rendered invisible under the current framework. 254 See, e.g., Colker, supra note 219, at 1029–32 (discussing how existing model could not effectively address economic and social inequities that have plagued black single mothers).

Unlike more radical approaches, which would seek to wholly supplant existing doctrine, 255 See, e.g., Charles R. Lawrence III, The Id, the Ego, and Equal Protection: Reckoning with Unconscious Racism, 39 Stan. L. Rev. 317, 324 (1987) (suggesting Court examine cultural meaning of laws to determine presence of collective, unconscious racism, rather than look for discriminatory motive). this model draws on underlying elements of the current framework. Thus, it has the potential for redeployment that has been exploited in other contexts, such as the reappropriation of equal protection’s intent requirement to serve politically conservative ends. 256 Eyer, Ideological Drift, supra note 15 (manuscript at 76) (describing history and development of equal protection’s intent requirement, demonstrating trajectory from device for racial justice advocates to tool for opponents of racial justice). Such redeployment has the potential to give new life to the discriminatory effects strand of equal protection jurisprudence. It also has the potential to avoid doctrinal pitfalls that have developed under the existing framework. For example, a value-based model alleviates the need for comparative treatment 257 Magee Andrews, supra note 26, at 507 (highlighting equal protection’s overreliance on comparative notion of equality). —i.e., demonstrating that like have been treated alike—because the focus is not on how one identity group is being treated in relation to all others. In addition to more pragmatic difficulties it poses, 258 Comparative treatment, or the “different treatment” approach, “requires black female plaintiffs [challenging policy allowing termination if employees exceed their allotted sick leave and which has affected only black women] to show that the defendant is treating them differently than it is treating similarly situated white or male workers.” Colker, supra note 219, at 1029. As Colker explains, however, the “black women . . . could not use this approach to challenge the policy because there are no similarly situated white or male workers who are being treated differently.” Id. the similarly situated requirement makes possible the perverse result that, to provide redress in the face of a possible equal protection violation, the state may withdraw all benefits rather than provide them on an equal basis. 259 See, e.g., Palmer v. Thompson, 403 U.S. 217, 219 (1971) (noting how, faced with possible equal protection problem, city closed all swimming pools rather than integrate them). In contrast, under a value-based approach, the question is not how an individual has been treated in relationship to others, but whether the way in which she has been treated implicates the values the doctrine has prioritized.

For progressives, another frustration with equal protection doctrine is the requirement that a plaintiff must prove discriminatory intent in order to make out an equal protection violation. 260 See López, Intentional Blindness, supra note 43, at 1783 ([T]he requirement that malice be proved is so exacting that . . . it has never been met—not even once.”); Reva B. Siegel, Foreword: Equality Divided, 127 Harv. L. Rev. 1, 2–3 (2013) (“When minorities challenge laws of general application and argue that government has segregated or profiled on the basis of race, plaintiffs must show that government acted for a discriminatory purpose, a standard that doctrine has made extraordinarily difficult to satisfy.”); Strauss, Discriminatory Intent, supra note 45, at 957 (“A court applying the discriminatory intent standard should ask: suppose the adverse effects of the challenged government decision fell on whites instead of blacks . . . [w]ould the decision have been different? If the answer is yes, then the decision was made with discriminatory intent.”); see also López & Olivas, supra note 178, at 305 (asserting “Court’s pinched conception of race lends support to an intent test [and] allows the Court to equate race-conscious remedies with racism”). To the extent race has no substantive meaning from a legal perspective, it is because the Court has deprived it of any such meaning; the Court has willed colorblindness by saying that effects alone are usually insufficient in assessing claims of discrimination. 261 See Washington v. Davis, 426 U.S. 229, 242 (1976) (holding discriminatory effects, without showing of discriminatory intent, are insufficient to prove equal protection violation); see also Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 279 (1979) (holding to demonstrate discriminatory purpose, plaintiff must show decisionmaker “selected or reaffirmed a particular course of action at least in part ‘because of,’ not merely ‘in spite of,’ its adverse effects upon an identifiable group”). Under a value-based model, the Court no longer assigns meaning to identity, and the facially racial/facially neutral distinction is no longer relevant because categories, like race and gender, no longer serve as the triggering factor for heightened scrutiny. The purpose of the intent requirement is to prove that a statute that appears neutral on its face is, in actuality, acting as a racial classification. 262 Daniel R. Ortiz, The Myth of Intent in Equal Protection, 41 Stan. L. Rev. 1105, 1118 (1989) (describing how, when faced with facially neutral law, intent can “uncover covert classifications”). Ultimately, it matters that race is at issue because of the values that such discrimination represents. If instead, the inquiry focuses directly on whether state action implicates certain values, it does not necessarily matter whether such implication was intended or not. In other words, under the current framework, courts use intent to ferret out a distinction that is imbued with specific meaning.

If the intermediary category of identity is eliminated, a more complicated and more involved inquiry must ensue to determine whether the substantive criteria have in fact been met in the context at hand. But the use of a doctrinal shorthand comes with a cost—namely, that the shorthand can be dismissed more easily through the application of the intent–effects test. Rather than showing that the substantive criteria have not been met—which could itself be part of an important political and social dialogue—a defendant can merely demonstrate that the plaintiff has not shown that race or gender was the clear cause of the differential treatment. It is true that the question of evidentiary standards is distinct from the predicate for a claim; in other words, intent could be required under any approach. Yet, the very nature of the value-based approach implies that intent is not relevant; ultimately, the model’s primary concern is about the effects of the discrimination on the group at issue.

The value-based approach shifts attention from the individual act to structural or systemic phenomena that can cause an otherwise unobjectionable act to have constitutional import. Dean Spade has argued that, in its current form, antidiscrimination law individualizes racism and is “about bad individuals who intentionally make discriminatory choices and must be punished. In this (mis)understanding, structural or systemic racism is rendered invisible.” 263 Spade, supra note 110, at 84; see also Alan David Freeman, Legitimizing Racial Discrimination Through Antidiscrimination Law: A Critical Review of Supreme Court Doctrine, 62 Minn. L. Rev. 1049, 1054 (1978) (describing how antidiscrimination law “views racial discrimination not as a social phenomenon, but merely as the misguided conduct of particular actors”). Similarly, unconscious bias or racism goes unnoticed or unaddressed because it is difficult to establish the causal chain between actor and illegal act. 264 See Lawrence, supra note 255, at 323–25 (explaining intent requirement provides means for assigning fault and demonstrates decision was based on race, but ignores true nature of how human mind works). In that respect, however, unconscious racism might be subjected to the same set of questions that resurface throughout this piece: Is unconscious racism troubling because it quietly and not explicitly invokes race (in which case the primary concern is ferreting out race as the underlying basis for action), or is it that unconscious biases aid in perpetuating certain dynamics that serve to maintain hierarchy and obstruct access?

Because it forces dialogue to remain at a fairly superficial level, the current framework short-circuits the potential for full political debate under a guise of colorblindness. While the questions demanding a response under the value-based model would not be easy, the substantive discussion that would result is sorely needed. In his seminal work, The Id, the Ego, and Equal Protection: Reckoning with Unconscious Racism, Charles Lawrence wrote:

Blacks and other historically stigmatized and excluded groups have no small stake in the promotion of an explicitly normative debate. While their version of shared values or fundamental principles—the victim’s perspective—may not hold sway at the moment, the courts can become a legitimate forum for the persuasive articulation of that version. And once the debate is made explicit, the hegemonic function of the law is diminished. This is not to say that the courts should become the exclusive or even the primary forum for normative debate, but rather that, by making the debate over fundamental principles explicitly political, one expands the arena for that debate. 265 Id. at 386.

Ultimately, the substantive nature of the framework suggested here would facilitate more transparent and open dialogue about when heightened scrutiny should apply. 266 Robinson, Unequal Protection, supra note 41 (manuscript at 15) (calling for more transparency); cf. Laurence H. Tribe, American Constitutional Law 1444–45 (2d ed. 1988) (criticizing Cleburne’s more covert approach to equal protection analysis and calling for more explicit debate about whether traits warrant application of heightened scrutiny). Even if more substantive analyses by the courts do not lead to greater judicial protection, they may serve as a vehicle to unearth factual realities that can facilitate action in the political and legislative arenas. Moreover, the language of the value-based approach offers a needed reprieve from the binary nature of the current approach and a new tactic for advocates battling against colorblindness. Once the dialogue has shifted from “color matters/color doesn’t matter” to a discussion of substantive treatment, the rhetoric of colorblindness is no longer applicable and anticlassificationists can, in essence, no longer be anticlassificationists (because the framework no longer revolves around classification). Instead, they must debate plaintiffs on the substantive terms of their claims.

On an institutional level, courts may be better suited to answer the questions posed under a value-based model than they are to do what is expected of them under the current model. Under either model, courts will play a gate-keeping function; their role will always be to identify legal violations and to award the appropriate relief. Under a categorical model, the Court also assumes the awkward position of policing identity categories. 267 See supra note 89 and accompanying text (describing need under current doctrine to define identity category at issue and, thus, who is included and excluded from that category). As Justice Kennedy has recently suggested, for courts to undertake the venture of defining race-based categories would not only lack “clear legal standards or accepted sources to guide judicial decision but also it would result in, or at least impose a high risk of, inquiries and categories dependent upon demeaning stereotypes, classifications of questionable constitutionality on their own terms.” 268 Schuette v. Coal. to Defend Affirmative Action, 134 S. Ct. 1623, 1635 (2014). Moreover, for the Court to make such a judgment in the context of one case, which will then govern all other instances of discrimination based on the same category, rather than assessing the role of identity on a case-by-case basis, increases the risk that the experience described by the plaintiff in that case will become the dominant narrative for all plaintiffs sharing the same identity trait. There may be a further danger in that, if and when the Court makes such determinations, they remain fixed in the doctrine and inevitably affect myriad other cases—at every level of the judicial system—as well as legislation that is based on Congress’s enforcement power under the Fourteenth Amendment. 269 See supra note 235 (describing Congress’s Section 5 power and Boerne limitations); see also supra notes 95–96 and accompanying text (outlining danger arising from fixing certain notions of identity in doctrine).

In contrast, under a value-based model, courts are tasked with determining whether certain substantive criteria have been met. While any judicial determination will retain some level of subjectivity, courts may be better designed to analyze empirical evidence to determine whether a particular phenomenon has been demonstrated—for example, a history of discrimination, or the denial of access to the political process—than they are to define race or gender and determine who should qualify as a member of either category. 270 See sections I.A–B (presenting numerous difficulties resulting from judiciary’s attempts to define identity); section II.B (illustrating how courts could instead rely on less subjective evidence). Moreover, because identity categories are creatures of social construction, they are imbued with meaning that the courts cannot expunge. 271 See supra notes 162–166 and accompanying text (discussing social construction of identity). Thus, by utilizing identity as the foundation for equal protection doctrine, courts seeking to effect change necessarily cede control to society and lose the ability to enforce normative commitments that may underlie the distinction of certain identity categories as deserving of special protection. In contrast, although the nomenclature suggests otherwise, the questions inherent to a value-based approach can ultimately be thought of as factual questions—e.g., is there a demonstrated history of discrimination against this group? 272 Allison Orr Larsen has written about the questionable fact-finding practices utilized by the Supreme Court, which may make one hesitant to base equal protection analysis on similar findings of “fact.” See Allison Orr Larsen, Confronting Supreme Court Fact Finding, 98 Va. L. Rev. 1255, 1290–305 (2012) (arguing when judges engage in legislative fact finding they increase risk of introducing bias and mistakes and undermine fairness of adjudicative process). If not relying on facts, however, the Court will be left to rely on subjective judgment, and many would surely be just as uncomfortable with that scenario. Larsen acknowledges that, desirable or not, constitutional law is increasingly reli­ant on facts and suggests in response that judges either limit their use of facts or engage in wider fact finding to limit bias. See id. at 1290–91, 1305–12 (detailing proliferation of independent judicial research due to accessibility of digital material and suggesting alter­native approaches to judicial fact finding).

Taking the power to define categories and assign individuals to those categories away from the state protects the equal protection doctrine from manipulation. For example, as the Hernandez story demonstrates, 273 See supra note 183 and accompanying text (explaining how classifying Mexican Americans as white deprived them of potential basis for equal protection claims). under the current framework, a court’s (or the government’s) decision to classify individuals in a certain way—for example, classifying Mexican Americans as white—may deprive them of the means to prove an equal protection violation. Similarly, the Court’s decision to dissociate gender and pregnancy left the plaintiff in Geduldig in the same disadvantaged position. 274 See supra notes 224–231 and accompanying text (analyzing Geduldig). Under a value-based approach, a court’s role is not to define categories or decide who meets the criteria for any given category; rather a court’s role is to make a factual determination—much as the Hernandez Court did—as to the treatment of a specific group within the applicable context.

Last, the shift to a value-based model of equal protection would serve a broader educative function for both the judiciary and the public. The considerations that led to the designation of race and gender as suspect classifications are presently assumptions buried deep in the doctrine. Thus most people have little occasion or reason to ponder why we actually care more about discrimination on the basis of race or gender than about other forms of discrimination. By forcing litigants to frame their arguments to emphasize the nature of discrimination and allowing them to contextualize such discrimination—rather than merely rely on superficial distinctions—this model forces courts (and perhaps the public) to grapple with the substantive impact of discrimination. 275 As with any framework that eschews bright-line rules (or perhaps any legal framework), the quality of the arguments made will likely depend on the availability of evidence and the skill of the lawyers involved. While I recognize such realities, I would also submit that all plaintiffs would face a similar burden and that any framework will always be subject to possible manipulation. And by allowing for various narratives rather than requiring conformance to one dominant narrative, this model facilitates a more textured and nuanced exploration of discrimination.

B. Addressing Critiques of a Value-Based Approach

Perhaps the broadest critique levied at such an approach would question its eschewal of identity as a framing device, given the inherent value in highlighting the importance of categories such as race and gender. As Rhonda V. Magee Andrews has noted, “many liberal and critical race theorists argue . . . that, to ensure redress of substantive conditions of oppression, the last thing we need is less of a focus on race. Instead, we need to focus more attention on explicitly racial concerns.” 276 Magee Andrews, supra note 26, at 511 (footnote omitted). Critical race theorists, like Kimberlé Crenshaw, have underscored that “the most valuable political asset of the Black community has been its ability to assert a collective identity and to name its collective political reality.” 277 Kimberlé Williams Crenshaw, Race, Reform, and Retrenchment: Transformation and Legitimation in Antidiscrimination Law, 101 Harv. L. Rev. 1331, 1336 (1988).

In response to critiques in that vein, this Essay does not suggest that identity itself is irrelevant; nor does it by any means suggest that we have transitioned to a postracial era. 278 See Sumi Cho, Post-Racialism, 94 Iowa L. Rev. 1589, 1595 (2009) (describing post-racialism as ideology under which “race does not matter, and should not be taken into account or even noticed”); see also Mario L. Barnes et al., A Post-Race Equal Protection?, 98 Geo. L.J. 967, 977–92 (2010) (documenting myriad ways in which society is not postracial). Yet as an isolated concept, without any necessary elaboration as to why it is important, identity may fail to serve its intended purpose or, worse, be manipulated to unintended ends. As Lani Guinier and Gerald Torres have written, “[i]n the view of the neoconservatives, race is merely skin color and is thus meaningless and ignorable.” 279 Guinier & Torres, supra note 73, at 13. In other words, one’s identity as black or as a woman has only as much meaning in the legal context as the law is willing to ascribe to it; supplanting identity with the substantive ends it is intended to serve may therefore make it more difficult to dismiss its relevance.

Identity is incredibly important, both to an individual’s sense of self and to social movements, but in those contexts, it plays a very different role than the role it plays in the context of the law. 280 See Lucas, Undoing Race, supra note 66, at 1291–92 (arguing against conflation of identity and doctrine, given different purposes served by each). In the hands of individuals, identity may serve to define one’s place in the world and in relation to other individuals. For social and political organizers, identity serves as a critical bond and a basis to rise up collectively against oppressive forces. In contrast, law is a system designed to negotiate relationships between individuals and groups. From the perspective of social and political organizers, it is often advantageous to paint identity with a broad brush, allowing it to be as inclusive as possible. 281 See id. at 1263 (discussing NAACP’s opposition to creation of multiracial status because of concern said status would undermine antidiscrimination law, weaken ability for NAACP to police racial identity, and jeopardize race-based protections). Yet, in thinking about the law’s role in providing remedies to address the harm suffered, a broad definition of identity may be unproductive for individuals who experience harm differently from those who more closely conform to the dominant identity narrative.

To the extent that law plays a critical role in “reproduc[ing] the structures and practices of racial domination,” 282 See Gotanda et al., Critical Race Theory: The Key Writings that Formed the Movement, at xxv (Kimberlé Crenshaw et al. eds., 1995) (describing “[r]acial power” as “sum total of the pervasive ways in which law shapes and is shaped by ‘race relations’ across the social plane”). allowing the law to use identity as a tool risks its own distortion. 283 As Magee Andrews notes:

“[T]he present framework inevitably embodies a conception of race that both maintains its nineteenth century meaning as an objective fact beyond the power of the court to eradicate (thereby capturing the most pernicious aspects of the meaning of race) and also strips race of its sociohistorical implications, permitting, for example, whites, who continue to control America’s major institutions and the vast majority of the country’s wealth, to claim equal if not greater vulnerability to race-based oppression in present day America.”

Magee Andrews, supra note 26, at 514.
The value-based model leaves the power to define identity and to articulate its harms and consequences in the hands of the claimants. 284 The value-based model does not deny the importance of group affiliation, but it leaves defining the group at issue and making the group-based claim, based on the criteria described herein, to the plaintiff or plaintiffs (rather than allowing courts to define the group and, in the process, distorting the category itself and/or excluding some from protection). To the extent that law shapes social movements and society and culture more broadly, 285 See Magee Andrews, supra note 26, at 490 n.19 (“[R]ules and society are mutu­ally constitutive and reinforcing; that is, law is ‘both agent and object’ of society’s normative impulses.” (citing Owen M. Fiss, The Death of the Law?, 72 Cornell L. Rev. 1, 15 (1986))). a value-based approach has potential to diminish the polarization caused by identity politics and instead to focus society’s attention on the effects of discrimination and disparate levels of access experienced by marginalized groups and individuals.

Of course, overt discrimination does still occur, and categorization on the basis of identity still does a great deal of work in oppressing individuals and groups—for example, when a young black man is stopped by a police officer simply because he is black. The value-based model does not intend to obscure the importance of identity in that sense, or the fact that in many cases, it is the category itself that triggers subordination. Because the greatest contribution of the value-based model is arguably its shift in rhetoric, it is least needed in instances where the basis for discrimination (or subordination) can easily be articulated or is irrational. Its role in those instances may be to emphasize that one cannot fairly see “race” or even “identity” as a universally applicable concept. In other words, even when one’s primary complaint is discrimi­nation on the basis of identity, as in the example of the police stop above, that does not necessarily suggest that all discrimination on the basis of identity is undesirable. Thus, the value-based model can help elucidate and articulate why certain types of discrimination on the basis of identity are more troubling than others and have a more devastating impact on the community subject to those types of discrimination. It also avoids the conclusion that discriminating more broadly, among additional identity groups, would cure the problem posed by racial profiling. Viewing such interactions from the perspective of a value-based model may also be helpful in that it makes such incidents seem less like one-time occurrences or products of singular bad actors (or bad departments), but instead part of a larger story in which members of a group continually endure oppression in a number of different and overlapping ways.

Another possible concern is that a value-based approach risks over-fragmentation of identity groups and the possibility of infighting among subgroups that currently share the same identity marker (e.g., all of those encompassed within the term “black”). One might counter, acknowledg­ing that identity and law are co-constituitive, 286 See López, White by Law, supra note 9, at 9–11 (explaining “legal construction of race”). that it is preferable to force solidarity by grouping these subgroups together under one label. If one views the context in which equal protection operates as a zero-sum game, it is of course possible that subgroups would attempt to battle it out among themselves on the terms of the substantive criteria raised by the value-based model. Yet, if the zero-sum assumption is true (which it may not necessarily be), then under the current model, not everyone within the larger group can benefit either. It may be advantageous to utilize a model that will prioritize those who have the greatest barriers to access or who have been most severely subordinated rather than those in the current model who are most likely to emerge victorious—those who rise to the top of the hierarchy within the subgroup or who can most closely align themselves with the dominant narrative (e.g., lighter-skinned blacks and women adhering to male norms).

There may also be a concern that members of the majority could co-opt this model as well. But given its substantive nature and what it requires of plaintiffs in articulating and arguing their claims (e.g., a plaintiff cannot just claim discrimination on the basis of whiteness, but must then explain how that discrimination meets the substantive criteria under the model), the value-based model is less susceptible to the colorblindness brand of co-option that has emerged under present doctrine. As to what it might mean for a majority plaintiff to succeed on the argument that she has in fact been subject to a pattern of historical discrimination, for example, the response is two-fold. First, this model does allow for a white plaintiff, or a plaintiff of any race or gender, to make a contextualized argument that may be limited to a particular set of facts. 287 Note 172, supra, addresses the question of scope (i.e., would a plaintiff in such a position be able to rely on contextualized discrimination without being forced to address broader dynamics of white privilege) to some extent. However this, like other questions of implementation, or how the model might apply in every imaginable case, is beyond the compass of this piece. Second, the current model allows for the same type of manipu­lation—possibly distorting the nature of history and its underlying relevance—but does so under the guise of applying identity categories in a universal manner.

In his defense of a liberty-based dignity approach, Kenji Yoshino highlights a potential criticism that would also apply here: By requiring a case-by-case approach to protection and eliminating the ability to attain heightened protection across multiple contexts in one fell swoop, this model is weaker in countering subordination. 288 See Yoshino, supra note 49, at 799–800 (addressing argument that current equal protection model offers advantage of applying across-the-board—i.e., to any instance of discrimination against minority group at issue, in any context); see also William N. Eskridge, Jr., Destabilizing Due Process and Evolutive Equal Protection, 47 UCLA L. Rev. 1183, 1216 (2000) (“[T]he Equal Protection Clause alone offers a minority group a potential constitutional jackpot at the wholesale level, that is, in challenges to an array of interconnected discriminations in state benefits as well as burdens.”). The two responses Yoshino provides have been touched upon above and would apply here as well. First, although equal protection does offer the potential for such sweeping protection, pragmatic realities—and the “closure of the heightened scrutiny canon of classifications”—have rendered that “jackpot” highly unlikely. 289 Yoshino, supra note 49, at 800. Even for those who have arguably already hit the proverbial jackpot, the perverse consequence of having done so may mean that affirmative action measures utilizing the same trait will be subject to heightened scrutiny as well. Second, given the existing model’s focus on comparative equality, remedies may level up or level down; it is possible, therefore, that discrimination may be eliminated by denying rather than providing entitlements to all. 290 Id.

Another pragmatic and inevitable critique may be that courts unwilling to advance justice under the current framework will be no more willing to do so under a different set of rules. In the last several decades, the Court has pulled back from declaring additional classes “suspect,” 291 See supra note 243 (describing Court’s unwillingness to recognize any additional suspect classifications). due in part to a fear that too many other groups could then claim the same level of protection. 292 In City of Cleburne v. Cleburne Living Center, the Court held:

“[I]f the large and amorphous class of the mentally retarded were deemed quasi-suspect for the reasons given by the Court of Appeals, it would be difficult to find a principled way to distinguish a variety of other groups who have perhaps immutable disabilities setting them off from others, who cannot themselves mandate the desired legislative responses, and who can claim some degree of prejudice from at least part of the public at large. One need mention in this respect only the aging, the disabled, the mentally ill, and the infirm. We are reluctant to set out on that course, and we decline to do so.”

473 U.S. 432, 445–46 (1984).
Any approach will demand the imposition of limits and, under the current framework, the Court’s response has been to curb the proliferation of suspect classifications. Yet, the definition of identity groups for purposes of equal protection will often be under- or overinclusive, risking the possibility that truly problematic instances of discrimination will fall on the unprotected side of the line. Thus, the Court’s decision to set limits based on identity may not align perfectly with prohibiting the most pernicious forms of discrimination. Similar line drawing would have to occur under a value-based approach, but because the analysis under such an approach is value-based rather than proxy-based, the basis for the line would be less arbitrary and more likely the result of substantive, reasoned judgments.

There are a number of other possible issues that could arise when thinking about implementation of the value-based approach—for example, the fact that arguing such substantive claims, which arguably impose a higher evidentiary burden, may require a large investment of time and resources, something poor and underresourced litigants do not have. Unfortunately, that specific concern is universal and not unique to this context—there are many other contexts where arguing a claim will require adequate resources to do so. With regard to those individuals, the value-based model at least offers a potential benefit in that it has greater capacity to recognize wealth (or the lack thereof) as a constitutional concern. There are certainly contexts in which poverty could meet the substantive criteria set forth in the value-based approach, whereas it has been foreclosed from suspect classification under the current model. 293 See San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 2 (1973) (holding those who live in comparatively poor school district do not comprise suspect class). The primary project of this Essay, however, is not to address every question regarding implementation, of which there would undoubtedly be many. Instead, it is to convince readers of the initial premise that an identity-as-proxy jurisprudence is less preferable to one grounded in more structural concerns.

As stated above, the model suggested herein is not intended as a panacea; nor is it expected to consistently result in better outcomes from the perspective of minority interests. To a large extent, factors beyond the control of any legal framework will dictate the outcome of any given case. But a jurisprudence divorced from identity will at the very least avoid some of the harms described in Part I and strive for a substantive notion of equality less susceptible to manipulation or distortion.


Identity has played an integral part in the formation of equal protection doctrine and rightfully so. But as identity becomes more complex and the nature of discrimination evolves, so must the legal models used to assess and redress discrimination. Identity is fluid and yet a jurisprudence based on identity fixes certain definitions of identity in a manner that is far less so. Increasing legal recognition of racial and gender identities outside of the black–white and male–female binaries has placed greater emphasis on who is in and who is out, and whether the alleged perpetrator intended to discriminate against an individual specifically because of his or her identity, than on the more important question of why we are concerned about protecting against identity-based discrimination. Discrimination against certain identity groups is more pernicious not because the distinction is based on identity, but because of the role that distinction has played on the social and political landscape. It is the latter that should be the focus of equal protection doctrine to ensure that it effectively provides equal opportunity to all.

Although the doctrinal impetus for singling out race and gender as suspect classifications focused on the treatment of African Americans and women, current jurisprudence has reconceptualized identity-based jurisprudence as universally applicable. Thus, discrimination against any form of identity within a given identity category is viewed as equal. Yet, many do not perceive all forms of discrimination to be equal because law does not operate in a vacuum, but instead against a social and political backdrop that involves historical subordination and oppression of only some identity groups.

This Essay suggests that one way to ensure equal protection remains capable of dismantling existing hierarchies may be to reclaim—and possibly revisit—the justifications used to utilize identity as a means for implementing equal protection. While identity categories like race and gender are fluid and socially constructed, the social, political, and economic meaning of race and gender are far more rigid and persistent; thus, the latter is a more appropriate basis than the former for doctrinal analysis. By adopting an approach that is based not on identity categories but on the reasons why discrimination against such categories is particularly troubling, courts will be less likely to generate negative externalities for those who do not conform to dominant identity narratives, and will be more capable of recognizing the specific harms faced by victims of discrimination. Moreover, a return to first principles may prevent co-option of the doctrine by those less in need of protection. While identity’s influence is inescapable, and its importance in the sociopolitical arena undeniable, its ability to protect marginalized individuals within the context of the law has diminished. Reclaiming the normative values that lead doctrine to privilege certain identities has the potential to recalibrate the meaning of equal protection and provide for a doctrinal structure that is more flexible and more responsive to future claims under the Equal Protection Clause.