Discrimination is a Comparative Injustice: A Reply to Hellman

Volume 102

102 Va. L. Rev. Online 85
Download PDF
In Two Concepts of Discrimination, Professor Hellman lucidly and systematically explains the difference between comparative and noncomparative conceptions of discrimination.[1] Although other legal scholars and philosophers have addressed the distinction between comparative and noncomparative justice,[2] she profitably applies the distinction to current controversies about the meaning and scope of antidiscrimination norms in statutory and equal protection law. Her approach is largely conceptual and interpretive, identifying the categories and reasoning that courts and legal scholars have employed. However, she also critiques aspects of that reasoning.

Hellman believes that her analysis illuminates a number of issues in contemporary constitutional discrimination jurisprudence. In her view, it explains why the supposed clash between equal protection doctrine and Title VII’s disparate impact approach is illusory, why equal protection doctrine is ambivalent about whether irrational government action is constitutionally problematic, and why equal protection and due process reasoning should only sometimes be combined.

There is much to admire in Hellman’s article. She carefully elucidates important conceptual and doctrinal distinctions, demonstrates a subtle and insightful appreciation of the complexities of equal protection doctrine, and is scrupulously fair in enunciating arguments that she ultimately rejects.

This Essay offers some friendly criticisms of her approach. I share her belief that the distinction between comparative and noncomparative justice is critical for understanding constitutional doctrine.[3] I disagree, however, about how, and even whether, that distinction should apply to antidiscrimination norms. And I do not believe that her analysis fully succeeds in explaining the three contemporary issues that she highlights.

A. The Distinction Between Comparative and Noncomparative Injustice

A comparative conception of injustice asserts that whether X has been unjustly treated depends on how others have been (or would have been) treated. A noncomparative conception asserts that whether X has been unjustly treated does not depend on the treatment of others. Hellman applies this distinction to the moral and legal problem of wrongful discrimination, arguing that courts and scholars have employed both conceptions here as well.

Why does this distinction matter? For several reasons, according to Hellman (with which I largely agree). First, for purposes of understanding the scope and content of the right, a comparative right is defined by reference to how others are treated. Second, as a substantive matter, deeper egalitarian norms explain why this formal or structural feature matters morally or legally. Here, Hellman “proposes” that the relevant substantive value is the duty to treat people “as equals.”[4] This, she suggests, is the substantive value upon which our equal protection doctrine actually relies.[5] However, as we will see, appealing to this particular substantive understanding of the norm underlying equal protection is more controversial than Hellman may realize.

Third, Hellman points out that comparative and noncomparative rights entail different remedies.[6] A comparative injustice can be remedied either by leveling up or by leveling down (or, I would add, by any intermediate remedy that corrects the inequality[7]), while remedying a noncomparative injustice requires leveling up, in the sense of giving the aggrieved person what she is entitled to. Thus, if a judge is sentencing two individuals A and B whose participation in a crime is identical in all relevant respects, but the judge has discretion to give any sentence between five and ten years, the judge commits a comparative injustice if he sentences A to five years and B to ten years. The inequality can be remedied either by increasing A’s sentence to ten years or by decreasing B’s to five (or by imposing any equal, intermediate sentence). But if, in a different, noncomparative case, a judge sentences C to seven years when, as a matter of statutory or constitutional law, C is entitled to a five-year sentence, the only proper remedy is to impose the five-year sentence.

I agree that different remedies flow from comparative and noncomparative rights,[8] but I think that Hellman’s account overstates the difference. If a noncomparative right actually creates a moral or legal entitlement to a particular treatment (as in the example of C’s sentence), then of course it follows that only that treatment satisfies the right. But not all noncomparative rights create specific entitlements. In the sentencing example, suppose the judge has discretion to sentence C to between three and five years, and further suppose that a sentence within that range is consistent with C’s noncomparative right to be punished according to his just deserts. Then, any sentence within that range respects C’s noncomparative rights (both C’s right that the sentencing judge respect the statutory limits and C’s right to be treated according to C’s just deserts). And conversely, a comparative right can create a specific entitlement. For example, if a person violates the Age Discrimination in Employment Act or the Equal Pay Act, the remedy requires leveling up.[9]

Although it should not be controversial that some constitutional and legal rights are concerned with comparative injustice and others with noncomparative injustice, what is controversial is Hellman’s central claim in this article—that the wrong of discrimination can be explicated either as comparative or noncomparative. By definition, wrongful discrimination refers to unjustified distinctions between persons. How can this wrong be understood as noncomparative? The very basis of the complaint is the claimed injustice of differential treatment. As Hellman notes, “[T]he independent [noncomparative] conception of discrimination makes the term ‘discrimination’ lose its moral resonance.”[10]

But Hellman is on solid ground in reporting that some scholars and judges do appear to characterize the wrong of discrimination as noncomparative. This is especially true, she notes, of many advocates of an “anticlassification” approach to racial discrimination, an approach that treats differential treatment in favor of minorities (such as affirmative action) with as much suspicion or disfavor as differential treatment that burdens minorities.[11] Those who take this stance sometimes claim that it is wrongful for a decision maker (such as a government, university, or employer) to consider race in any way in making a decision about allocating benefits or burdens. And, at first blush, a right not to have race considered in a decision affecting you looks like a noncomparative right: We need not examine how others are treated in order to determine whether your right was violated.

Nevertheless, I believe that it is implausible to characterize a supposed right to be free of race-based or gender-based decision making as a noncomparative right. At the same time, the reasons why this characterization might seem plausible deserve careful attention, for they reveal a greater complexity in the structure and justification of comparative rights than first appears.

Let us examine more closely the claim that a right to have race ignored is a noncomparative right. How would this claim arise? A white applicant to a university complains that he was rejected while a minority applicant with the same qualifications was admitted. This, of course, is a complaint of unfair comparative disadvantage. Take away that disadvantage and the complaint evaporates. The complaint is not that race was considered simpliciter. It is that race was considered to the complainant’s detriment.[12]

I suspect that Hellman finds a comparative account of the “color-blind” principle to be unpersuasive because she finds the principle itself unpersuasive. However, the question she is addressing is not normative (whether the principle itself is justifiable), but interpretive (how advocates of the principle explain and justify it). Many advocates of the principle do interpret it as a comparative principle, forbidding (or subjecting to serious scrutiny) any classification that disadvantages persons on the basis of race, and it is indeed a coherent exemplar of a comparative right. Other advocates worry that permitting racial preferences will aggravate racial and social divisions,[13] will stigmatize the recipients as inferior,[14] or will require the use of offensive racial criteria reminiscent of Nazi criteria for identifying Jews.[15] These, too, are justifications based on the comparative conception of discrimination, for they focus on the perceived undesirable, inegalitarian effects of permitting race-conscious programs. Although I ultimately share Hellman’s view that the “color-blind” principle is not the best understanding of our constitutional doctrine or of the egalitarian norms that should govern us, these are entirely different questions.

We have seen that the typical complaint against programs of racial preference depends on the complainant suffering a detriment relative to the preferred group. But can we imagine a case in which a racial classification is employed yet members of a particular race are not disadvantaged? Suppose that in 2017, the IRS requires blacks to file taxes on April 1, and whites on April 15. The following year, whites must file on April 1, and blacks on April 15, and so on for future years. This policy does seem troublesome, and perhaps it does reflect a noncomparative right not to have race employed as a criterion in government decision making.[16] However, the difficulty of constructing an example of this sort suggests two things. First, I seriously doubt that the principal concern of those who object to affirmative action programs is the mere use of race. Rather, they are disturbed by the favorable treatment of minority applicants relative to white applicants in securing a competitive position (such as a job or admission to a university). Second, even in this unusual IRS example, the reason that the use of an explicit racial classification is troublesome is, at least in part, a concern about unjustified inequality. The social significance of racial categories in American history cannot be overstated. In light of that tragic and divisive history, employing racial categories for administrative convenience when so many other, less divisive categories could just as easily be used is problematic: It demonstrates indifference to that history and might perpetuate and aggravate racial divisions. In short, I believe that the deeper explanation for our unease at employing racial categories, even those that do not create winners and losers, is an egalitarian, comparative principle.

To be sure, some advocates of the color-blind or anticlassification view articulate another rationale for their view: the principle that citizens should be treated as individuals, rather than as members of a racial group. Hellman reminds us that this was an important rationale for Justice Lewis Powell’s famous and dispositive opinion in the affirmative action case Regents of the University of California v. Bakke.[17] Hellman characterizes this rationale as noncomparative. I demur. The rationale is comparative, because it rests on the impropriety of treating a person as equal to others in the same (racial) category. At the same time, the underlying claim of comparative right here is not to equal treatment or treatment as an equal. Rather, the claim is to be treated differently from others in the category, in accordance with and in proportion to one’s individual (nonracial) characteristics.[18]

Like Hellman, I believe that the supposed right to individual treatment is not the best explanation of opposition to preferential treatment programs. Opponents surely do not believe that the state is generally disabled from using rational but imperfect proxies for relevant qualities, such as minimum age requirements for driving or voting, or grade point averages and standardized test scores as criteria for university admission.[19] Rather, they judge the use of racial categories to be especially problematic. Nevertheless, if someone truly believes that using overinclusive categories or failing to consider a wide range of individual characteristics is an affront to justice, it is important to see that this belief presupposes a comparative conception of justice.[20]

Following her discussion of the supposedly noncomparative right to be free from race-based classifications, Hellman turns to a different right that, she claims, is grounded in a noncomparative conception of the wrong of discrimination—the right to be free from gender stereotyping.[21] She rightly points out that in many gender discrimination cases, the Supreme Court focuses not on whether a law relies on a generalization that is too loose or inaccurate, but instead on the content of the generalization, and whether the gender stereotype confines individuals to particular gender roles—assuming, for example, that women are more caring than men and thus more suitable nurses, or that men are better suited for the rough and adversarial educational environment in a military academy. Hellman then asserts: “The antistereotyping principle found in sex discrimination cases rests on the view that each person (male or female) has an independent, noncomparative right to define his or her gender identity for him or herself.”[22]

The assertion does not follow from Hellman’s analysis. The Court has not recognized a general right of an individual to define his or her own identity in all respects. Yes, it is plausible to view the Court’s cases as recognizing the more specific right to define one’s gender identity, but notice that such a right is inherently comparative. It is a right not to be limited to “fixed notions concerning the roles and abilities of males and females”[23] such as “the pervasive sex-role stereotype that caring for family members is women’s work.”[24] One cannot make sense of the wrong of confining individuals to fixed gender roles without appreciating that such roles draw distinctions, presumptively treating men differently from women. Once again, a supposedly noncomparative right turns out to be comparative in structure and content.

To be sure, the right to define one’s own gender identity is a right that all citizens enjoy. But a universal right is not necessarily a noncomparative right. If, as in this instance, the rationale for the right is to avoid comparative injustice, then the right should be characterized as comparative.

B. The Doctrinal Implications of Hellman’s Analysis

In the final sections of her article, Hellman reviews three doctrinal issues that, she argues, reveal the payoff of the earlier analysis. I will suggest that her analysis helpfully illuminates the third issue she examines, but sheds less light upon the first and second.

1. Is Disparate Impact at Odds with Equal Protection?

First, Hellman addresses the question whether disparate impact and equal protection are at odds. Some scholars and judges, she observes, have proclaimed that these doctrines are in serious tension: Title VII’s disparate impact doctrine encourages employers to focus on whether an employment practice produces disparities disadvantaging minorities or women, but equal protection doctrine may prohibit employers from changing their practices for the purpose of reducing disparities. Equal protection might have this implication if it is understood as prohibiting, or subjecting to strict scrutiny, any race-based classification. Much more troubling, Hellman notes, is a further implication of this approach: Even facially neutral programs, such as efforts by schools to reduce the racial achievement gap by offering universal pre-kindergarten or efforts to reduce racial disparities in housing or healthcare, might violate equal protection, because the intent to reduce racial disparities is arguably an illegitimate purpose on this color-blind understanding of the demands of equal protection.

Hellman argues that this supposed tension, asserted by advocates of the anticlassification approach, between equal protection and disparate impact doctrine (and other policies intended to reduce racial disparities) rests on a conceptual mistake. These advocates are improperly combining a legitimate comparative right (the right not to be subject to a policy intended to harm or disadvantage a racial group) with a controversial noncomparative right (the right not to have race taken into account in how one is treated). As she explains:

[T]he claim that awareness of racial impact violates equal protection[[25]] derives from a flawed conflation of the two conceptions of the wrongful discrimination. It takes the prohibition on racial categories from the noncomparative view and the focus on intention from the comparative view and grafts them together to form a hybrid argument that, though full of sound and fury, signifies nothing.

    . . . .

            . . . [T]hese two facets of our doctrine (the focus on intention and the prohibition on racial classification) derive from different, competing accounts of what makes discrimination wrong. They cannot be fruitfully combined.[26]

The focus on intention, she claims, is a plausible comparative right when it means an intent to harm or disadvantage, but not when it means an intent to classify on the basis of race, because “[t]here is simply nothing obviously wrong with intending to classify on the basis of race.”[27]

I agree with Hellman’s conclusion, that equal protection should not be understood to prohibit race-conscious efforts to reduce racial hierarchies or racial disparities, especially when the programs adopted to further these ends are themselves racially neutral. But I do not believe that the conclusion follows from her premises. In the first place, as I suggested above, the right not to have race taken into account is quite plausibly understood as a comparative rather than a noncomparative right. In claiming that there is nothing obviously wrong with intending to classify on the basis of race, Hellman is asserting her own (quite defensible) normative view, but she thereby departs from the interpretive, rather than normative, stance that she otherwise takes in this article. (Note, too, that she could similarly state that there is nothing obviously wrong with classifying on the basis of race; yet she is willing to credit the anticlassification principle as a coherent, though unpersuasive, understanding of the wrong of discrimination.)

Secondly, even if the anticlassification principle is understood as a noncomparative right, Hellman has offered no argument why comparative and noncomparative right claims should not be combined, either in this context or more generally.[28] In many situations, combining these different types of rights is perfectly acceptable. Recall the earlier example in which comparative equality principles provide that defendant A should receive the same sentence as defendant B because their participation in a criminal endeavor is identical. Now suppose the judge has already sentenced A to three years, but the judge then determines that she should have sentenced A to five years. And suppose that a noncomparative principle provides that, once A has been sentenced, A’s sentence may not be increased. Then, the judge could legitimately combine the comparative and noncomparative principles and sentence B to three years.

2. Why We Are Confused About Rationality Review

Hellman next addresses why courts and commentators are so confused about whether rationality review is toothless or instead has some bite, and, if the latter, about when it has bite. She concludes that the Equal Protection Clause protects individuals from irrational government action under the noncomparative conception of discrimination but not under the comparative conception. Under the latter, she believes, genuine arbitrariness is not a constitutional problem.

Specifically, she reasons that if a law irrationally differentiates—for example, by forbidding opticians from fitting lenses without a prescription but permitting makers of ready-to-wear glasses to fit lenses[29]—then those who are burdened are burdened for no good reason. “If this is a constitutional problem, it is a problem because people have an independent, noncomparative right that state action that burdens their liberty does so for a reason.”[30] However, such irrationality is not, she claims, a comparative wrong.

These claims are surprising and, in the end, unconvincing. To be sure, some laws are irrational and constitutionally problematic for noncomparative reasons. The clearest example would be a law that burdens everyone for no reason. Realistic instances are difficult to conjure up, but suppose a dictator, on a whim, requires everyone (including himself) to stand still for an hour every day. More realistically, suppose that a government official refuses to do his duty unless he is bribed. He does have a reason, but an illegitimate one, for demanding a bribe. And we might reasonably conclude that he wrongs each citizen from whom he demands a bribe, regardless of how he treats other citizens.

But why does Hellman conclude that a similarly arbitrary or illegitimate reason for comparatively unequal treatment is not a distinctive constitutional problem? After all, if the dictator requires only some people to stand still every day, or demands bribes only of some citizens, he has evidently introduced a new problem: the problem of unjust inequality in treatment.

Here is her explanation. In the case of stupid, arbitrary laws that burden some but not all people for no reason (or no good or legitimate reason), the right that is infringed is the noncomparative right not to have your liberty restricted without good reason, not the comparative right not to be treated differently without good reason. Hellman offers the following example. Due to stupidity and with no factual basis, a university gives preference in admissions to students who wear glasses. This policy, she says, restricts the liberty of students without glasses for no good reason and thus rests on a noncomparative right. But the policy does not fail to treat those affected as equals (and thus does not violate a comparative right) because it is simply bad luck that some people with good vision are disadvantaged by this crazy policy, and comparative equality rights are not concerned with mere bad luck.

This explanation is unpersuasive. First, it is not clear that the noncomparative “liberty” interests of university applicants are infringed by irrational policies. Hellman here assumes without argument a broad understanding both of “liberty” and of what counts as an infringement.[31] Second, she also assumes without sufficient argument a narrower understanding of what counts as an infringement of a comparative equality right. Why must any coherent conception of a comparative equality right treat disadvantage due to stupid or arbitrary criteria as “bad luck” for which the decision maker is not responsible? It is hardly a matter of “luck” that students with good vision are disadvantaged by the foolish policy: This effect of the policy is entirely foreseeable, and the decision maker can readily avoid the effect simply by repealing the policy. The “mere bad luck” view is even more implausible in scenarios where the disadvantage is more significant than loss of a chance to be admitted to a university: suppose a judge sentences B to a longer sentence than similarly situated A simply because she dislikes B’s hairdo.

I suspect that the real explanation for Hellman’s willingness to recognize a noncomparative right not to be burdened for no reason and unwillingness to recognize an analogous comparative right not to be disadvantaged for no reason is that, as a normative matter, she is disposed to endorse the first right but not the second. She apparently views the Dworkinian right to be treated as an equal as a right that addresses only relatively serious forms of injustice, not trivial inequalities. In her recent book, she espouses the view that this egalitarian norm is best understood as prohibiting classifications that express denigration of a subgroup of citizens,[32] and foolish and ineffective laws are not demeaning. Hellman’s is a very plausible view of when discrimination is especially wrong. But, for purposes of the interpretive task that she sets for herself in the current article, it is unpersuasive to claim that other forms of discrimination, including irrational differential treatment for insufficient reason, are not genuine violations of comparative equality rights.

The final example Hellman offers in her irrational treatment section is the so-called “class of one” problem. Courts have struggled with the question whether, if the government intentionally treats a single citizen B differently from, and less favorably than, the way the government treated another, similarly situated citizen A in the past, this violates equal protection—for example, the government requires an easement from B but not from A as a condition of granting a permit. Hellman characterizes this as a question of the noncomparative right of B not to be treated irrationally. But the issue in these cases is comparative: whether the difference in treatment is irrational. There is typically no question in these cases that the government could rationally (and constitutionally) have decided in the first instance to impose the same burden on A that it later imposed on B. It is thus difficult to see why she believes this category of cases exemplifies a noncomparative rather than comparative right.

To be sure, rational basis tests are controversial. It is a difficult and important question whether, when legislation does not infringe fundamental rights and does not employ suspect or quasi-suspect traits, constitutional constraints should be completely, extremely, or only modestly deferential.[33] But the question is not resolved by whether we characterize the constraint as comparative or noncomparative. A comparative right can be robust or deferential,[34] and the same is true of a noncomparative right.

3. When Should Equal Protection and Due Process Analysis Be Combined?

The final payoff that Hellman identifies from her analysis is how to make sense of cases, including the recent cases of Lawrence v. Texas[35] and Obergefell v. Hodges[36], that combine equal protection and due process analyses. Here, I find Hellman’s approach illuminating. As she explains, a comparative view of the wrong of discrimination treats equal protection as requiring a comparative analysis, while due process requires a noncomparative analysis. In Skinner v. Oklahoma,[37] for example, the comparative view treats procreative capacity as a fundamental interest that requires strict scrutiny if it is differentially burdened, but not simply because it is burdened. By contrast, on a noncomparative view of discrimination, both equal protection and due process require noncomparative analysis. On this view, the problem in Skinner is the noncomparative burden, apart from how others are treated. Similarly, the undifferentiated emphasis on “dignity” in some recent Supreme Court cases is problematic on the comparative view of discrimination, because the term might encompass either a comparative right (to be treated with respect in a nondemeaning way) or a noncomparative right (to have control of one’s own destiny including one’s sexual autonomy).[38] But this lack of differentiation is not problematic on the noncomparative view, because both the Equal Protection and Due Process Clauses focus on whether the claimant has an independent entitlement to be treated in a particular way. On the noncomparative view, Hellman explains, the fact that some are burdened and others are not “may make the underlying rights deprivation salient but it isn’t what makes it wrong.”[39] In the recent Obergefell opinion, for example, although the court refers to both equal protection and due process analysis, its references to equal protection are largely in service of identifying the important liberty interest at stake when gay citizens are denied the right to marry.

Although I largely agree with this analysis, conceiving of a noncomparative right as creating an “independent entitlement” can be misleading for reasons already noted. A noncomparative right does not invariably create an unconditional entitlement. Rather, the entitlement is sometimes conditional, especially when the decision maker has discretion whether to provide the benefit (or alleviate the burden) in question. For example, statutes burdening the fundamental interest in a meaningful appeal of a state criminal conviction can be understood as problematic for noncomparative reasons without assuming that there is a constitutional right to such an appeal. The wrong is still noncomparative if the injustice of conditioning an appeal on the ability to pay a fee does not depend on how others are treated. On this view, such a fee requirement is akin to a requirement that forbids all defendants, rich or poor, from furnishing a trial transcript to the appellate court.[40] Thus, the noncomparative view espoused by Justice Harlan in equal protection-fundamental interest cases could apply even in cases where the interest at issue is constitutionally optional.[41] Similarly, state governments are free to decide to get out of the business of recognizing marriages entirely, leaving “marriage” as a religious rather than secular, civil law category. In that sense, the right to marry is constitutionally optional. But once the state recognizes marriage and permits significant social, legal, and economic benefits to flow to married persons, it has a noncomparative duty not to deny that status to persons unless it has a very weighty reason to do so.

Finally, the Obergefell opinion, although focused on the denial of the freedom to marry, also emphasizes egalitarian concerns, including the concern that excluding gays from marriage demeans and stigmatizes them. And it is telling that the court’s actual holding is framed in comparative terms: “[T]he State laws challenged by Petitioners in these cases are now held invalid to the extent they exclude same-sex couples from civil marriage on the same terms and conditions as opposite-sex couples.”[42] Insofar as states are free to impose a significant range of conditions and restrictions upon marriage, and insofar as states differ considerably from one another in the nature and scope of these restrictions, we must conclude that the noncomparative right to marry, although rhetorically important in Obergefell, is not a complete explanation of the scope of the Court’s ruling.


Despite my disagreements with Professor Hellman about how one should draw the distinction between comparative and noncomparative injustice and about the implications of that distinction, I share her view that the distinction is crucial in identifying the scope and content of, and the justifications for, specific legal rights not to suffer discrimination. Any satisfactory account of the wrong of discrimination must attend to the important questions she raises in this perceptive and original article.


[1]Deborah Hellman, Two Concepts of Discrimination, 102 Va. L. Rev. 895, 897 (2016).

[2]See, e.g., Joel Feinberg, Noncomparative Justice, 83 Phil. Rev. 297, 298 (1974); Kent Greenawalt, “Prescriptive Equality”: Two Steps Forward, 110 Harv. L. Rev. 1265, 1266 (1997); Raleigh Hannah Levine & Russell Pannier, Comparative and Noncomparative Justice: Some Guidelines for Constitutional Adjudication, 14 Wm. & Mary Bill Rts. J. 141, 188 n.152 (2005); Derek Parfit, Equality or Priority?, in The Ideal of Equality 81, 81–125 (Matthew Clayton & Andrew Williams eds., 2000); Peter Westen, The Empty Idea of Equality, 95 Harv. L. Rev. 537, 551–52 (1982).

[3]See Kenneth W. Simons, Equality as a Comparative Right, 65 B.U. L. Rev. 387, 450, 461 (1985) [hereinafter Simons, Comparative Right]; Kenneth W. Simons, The Logic of Egalitarian Norms, 80 B.U. L. Rev. 693, 709–10, 763 (2000) [hereinafter Simons, Logic].

[4]Hellman, supra note 1, at 901–03. This formulation, the right to be treated “as an equal,” derives from Professor Ronald Dworkin and his notion of the right to be treated with equal concern and respect. See Ronald Dworkin, Taking Rights Seriously 227 (1977) (“If I have two children, and one is dying from a disease that is making the other uncomfortable, I do not show equal concern if I flip a coin to decide which should have the remaining dose of a drug. This example shows that the right to treatment as an equal is fundamental, and the right to equal treatment, derivative. In some circumstances the right to treatment as an equal will entail a right to equal treatment, but not, by any means, in all circumstances.”).

[5]Hellman, supra note 1, at 902 n.18.

[6]Id. at 917.

[7]See Simons, Logic, supra note 3, at 711.

[8]One important complication here is that the appropriately flexible remedy depends on the nature of the comparative right. For example, because it is comparatively unjust to exclude black citizens from a public swimming pool, it might seem that the injustice can be remedied either by admitting both blacks and whites or instead by closing the pool. See Palmer v. Thompson, 403 U.S. 217, 227 (1971). However, it is also plausible to characterize the action of denying a benefit to both blacks and whites for an impermissible racist reason as a comparative injustice, as Hellman recognizes. See Hellman, supra note 1, at 917 & n.61; Simons, Comparative Right, supra note 3, at 431–33.

[9]The Age Discrimination in Employment Act, 29 U.S.C. § 623 (2012), provides: “It shall be unlawful for an employer” to discriminate against an employee because of the employee’s age or “to reduce the wage rate of any employee in order to comply with this chapter.” The Equal Pay Act, which addresses certain forms of sex discrimination in employment, similarly forbids leveling down as a remedy. Equal Pay Act of 1963, 29 U.S.C. § 206(d)(1) (2012) (“[A]n employer who is paying a wage rate differential in violation of this subsection shall not, in order to comply with the [nondiscrimination] provisions of this subsection, reduce the wage rate of any employee.”).

[10]Hellman, supra note 1, at 909.

[11]Id. at 914–17.

[12]Hellman states:

The anticlassification approach isn’t focused on the comparison between two cases—on the fact that X, a white applicant, is rejected while Y, a comparable black applicant, is accepted. Rather the focus of the anticlassification approach is on the single case and the fact that race was a factor that affected its outcome.

Id. at 917. But, in order for race to “affect the outcome,” it must be the case that some individuals were advantaged or disadvantaged by race relative to others. Moreover, it is difficult to believe that most critics of affirmative action and other preferential treatment programs are unconcerned with the fact that nonpreferred applicants lose a benefit that preferred applicants obtain.

[13]Justice Powell raises this concern in his opinion in Regents of the University of California v. Bakke, 438 U.S. 265, 298–99 (1978), discussed below: “Disparate constitutional tolerance of [benign as opposed to invidious racial] classifications well may serve to exacerbate racial and ethnic antagonisms rather than alleviate them.” Similarly, in the Supreme Court’s most recent affirmative action decision, the dissent characterizes the use of racial criteria as “noxious” and “pernicious.” Fisher v. Univ. of Tex. at Austin, No. 14-981, slip op. at 32, 51 (U.S. June 23, 2016) (Alito, J., dissenting).

[14]See Grutter v. Bollinger, 539 U.S. 306, 373 (2003) (Thomas, J., dissenting).

[15]See Fullilove v. Klutznick, 448 U.S. 448, 534 n.5 (1980) (Stevens, J., dissenting) (“[T]he very attempt to define with precision a beneficiary’s qualifying racial characteristics is repugnant to our constitutional ideals. . . . If the National Government is to make a serious effort to define racial classes by criteria that can be administered objectively, it must study precedents such as the First Regulation to the Reichs Citizenship Law of November 14, 1935 . . . .”). Justice Stevens then describes the Nuremberg laws, which defined “Jew” according to such criteria as parentage, marriage, and belonging to a Jewish religious community. Id. at 534–35 n.5.

[16]This example bears some similarity to Professor Paul Brest’s famous example of a school principal instructing black students to sit on the left side of the auditorium and white students on the right side for purportedly aesthetic reasons. In Hellman’s view, such an order demeans black students. Deborah Hellman, When is Discrimination Wrong? 25–27 (2008).

[17]438 U.S. at 299.

[18]For discussion of the comparative right to proportional treatment, see Simons, Comparative Right, supra note 3, at 437–46.

[19]See Hellman, supra note 16, at 116, 136–37.

[20]The short-lived “irrebuttable presumption” due process doctrine reflected this concern about overbroad classifications that fail to treat a person as an individual. See Kenneth W. Simons, Overinclusion and Underinclusion: A New Model, 36 UCLA L. Rev. 447, 514–18 (1989).

    The idea that one should be treated according to all of one’s relevant individual characteristics is, of course, highly impractical. How could a decision maker feasibly measure all those characteristics? This idea also fails to solve the problem of imperfect proxies. Suppose a university admissions office permits applicants to provide any information that they believe is relevant to admission, and the office carefully reviews all of that information. The office would still need to determine criteria of relevance and significance in comparing this applicant to others. But whatever criteria are used can then be criticized as imperfectly furthering or constituting the university’s goals or mission. For further discussion, see Hellman, supra note 16, ch. 5; Simons, Comparative Right, supra note 3, at 439–40.

[21]Hellman, supra note 1, at 918–21.

[22]Id. at 920.

[23]Id. at 919 (quoting Miss. Univ. for Women v. Hogan, 458 U.S. 718, 725 (1982)).

[24]Id. (quoting Nev. Dep’t of Human Res. v. Hibbs, 538 U.S. 721, 731 (2003)).

[25]Here, Hellman frames her conclusion in a misleading and exaggerated way. She objects to “the claim that awareness of racial impact violates equal protection,” id. at 924, and to the argument “that equal protection prohibits the awareness of racial impact that disparate impact requires.” Id. at 900. But no one claims that awareness alone is unconstitutional. Rather, the claim is that actors violate equal protection if they possess that awareness and then engage in intentional efforts to reduce or ameliorate racial disparities. This claim, although still ultimately unpersuasive, is much more plausible.

[26]Id. at 924–28.

[27]Id. at 929.

[28]Indeed, later in the article, Hellman states: “[T]he fundamental interests-equal protection line of cases offers an example of how comparative and noncomparative claims can be fruitfully combined—in a manner that is clear about the distinct values that are in play.” Id. at 944.

[29]Williamson v. Lee Optical, 348 U.S. 483, 488–89 (1955).

[30]Hellman, supra note 1, at 933.

[31]Earlier, Hellman asserts that a “promising candidate” for the source of a noncomparative right not to suffer discrimination is the “right to (some degree of) freedom or autonomy.” Id. at 913–14. I agree that such a right is very often a plausible source, but property rights are another source, as in the “class of one” cases discussed infra.

[32]Hellman, supra note 16, ch. 2; see Hellman, supra note 1, at 905–06.

[33]For a recent argument that the rational basis test should be strengthened to require proof of an actual rather than conceivable permissible purpose and of a meaningful relationship between means and ends, see Erwin Chemerinsky, The Rational Basis Test Is Constitutional (and Desirable), Geo. J.L. & Pub. Pol’y (forthcoming 2016) (manuscript at 10–11), http://ssrn.com/abstrac‌t=2782109 [https://perma.cc/XB27-M7FE].

[34]See Simons, Logic, supra note 3, at 759–61.

[35]539 U.S. 558, 564 (2003).

[36]135 S. Ct. 2584, 2602–03 (2015).

[37]316 U.S. 535, 536 (1942).

[38] See, e.g., United States v. Windsor, 133 S. Ct. 2675, 2692 (finding an injury to a person’s dignity to be a violation of the Fifth Amendment).

[39]Hellman, supra note 1, at 947.

[40]See Simons, Comparative Right, supra note 3, at 470–71.

[41]Hellman, supra note 1, at 948.

[42]135 S. Ct. at 2605 (emphasis added).

Click on a link below to access the full text of this article. These are third-party content providers and may require a separate subscription for access.