Breaking Down Bigotry

37 Const. Comm. 97

WHO’S THE BIGOT? LEARNING FROM CONFLICTS OVER MARRIAGE AND CIVIL RIGHTS LAW. By Linda C. McClain.* New York, NY: Oxford University Press. 2020. Pp. ix + 304. $39.95 (hardcover).

Amy J. Sepinwall[1]

In Who’s the Bigot? Learning from Conflicts Over Marriage and Civil Rights Law, Linda McClain offers a fascinating study of the rhetoric around the term “bigot.” Drawing from an impressively vast array of sources—including parenting guides and marriage manuals, sermons and speeches, legislative histories and court battles—McClain traces the use of the term through some of the most important civil rights battles in this and the last century. She focuses in particular on marriage, putting the controversies surrounding interfaith and interracial marriage in conversation with the controversies around same-sex marriage. The book is lucid, vital, highly readable, and brimming with insights into the minds of the various parties in the culture wars.

McClain seeks to recover, and draw out the tensions between, the way people use “bigot” and its cognate terms in discourse. As she says, her “method is to investigate [the] puzzles [around the term] by tracing how people spoke about bigots and bigotry in a series of past and present controversies over marriage and civil rights” (p. 18). Her results are intriguing and important. They reveal the ways in which we have shifted our moral attitudes, such that norms once considered decent (e.g., segregation in places of public accommodation; beliefs that homosexuality was evil and perverse) are now considered beyond the pale. They also allow us to gain clarity on the rhetorical force of leveling a charge of bigotry: is the imputation likely to promote moral progress or does it instead threaten to end conversations that might otherwise conduce to that progress? McClain’s insights also invite us to think about the relevance of bigotry to today’s efforts to combat inequality: should we focus on prejudice or instead systemic injustice and unconscious bias?

All of this is eminently valuable. The sociologist, rhetorician, and historian will find much to satisfy their curiosity here. At the same time, the focus on rhetoric can have an unsettling effect on the reader keen to know how the relevant terms ought to be used. What is the conceptual relationship between “bigotry,” “animus,” and “discrimination”? What are the necessary and sufficient conditions for conduct to count as an instance of one or more of those terms? And, once we have one or more of these species of conduct on our hands, how should we understand the moral responsibility of the state if it fails to condemn or eradicate the troubling conduct?

These are obviously fraught and difficult questions. Many philosophers and legal scholars have aimed to resolve them.[2] I will not promise anything so grand. But I do aim to make some progress on each. I begin in Part I with an effort to conceptualize the relationship between animus, bigotry, and discrimination. Concluding that bigotry need not involve animus and that discrimination need not involve bigotry, I nonetheless seek to foreground the importance of bigotry in our discrimination jurisprudence. I then turn, in Part II, to the Supreme Court’s decision in Bostock v. Clayton County. In Bostock, the Court held that Title VII’s provisions prohibiting sex-based discrimination in employment applied to discrimination on the basis of sexual orientation and gender identity. While there is much to celebrate in that outcome, I aim to expose the ways Bostock flattens the discrimination landscape, and minimizes the injury that bigotry inflicts. Part III moves from purely private discrimination to conceiving of the role of the state therein. My aim is to gain clarity on the notion of ratification, as concerns about ratifying arise among both private and state actors. Part IV concludes.

  1. What Is Bigotry?

McClain’s book provides a vast array of occasions when “bigotry” has been bandied about.[3] While one might have thought that only those who target a historically oppressed group count as bigoted, McClain intriguingly recovers instances when each side of the culture wars charges the other with bigotry. In this melee, segregationists or opponents of same-sex marriage are bigoted against Blacks or LGBTQ+ individuals, while progressives advocating for rights to marry whomever one pleases are bigoted against those whose religious convictions will not permit them to countenance interracial or same-sex marriage.[4]

At the same time, McClain explores whether all of these uses of bigotry are apt. Two questions concern her specifically. First, where sincere religious belief grounds opposition to interracial or same-sex marriage, does that belief obviate a charge of bigotry? Second, does bigotry require that one act from a bigoted motive or is it sufficient that one treat two individuals differently solely on the basis of an identity-defining characteristic (race, sex, religion, sexual orientation, etc.)? This second question is broader than the first since it allows that the source of the belief leading to differential treatment need not reside exclusively in religion. For example, a foreman who denies a woman a job on a construction site out of concern for the woman’s genteel nature operates with a benign albeit secular belief.[5] Notwithstanding this difference, I shall treat the two questions together, asking simply, does bigotry require animus? And if not, is all differential treatment, or discrimination, bigoted, or does “bigotry” instead pick out a subset of discriminatory conduct? If the latter, what is the hallmark of that conduct?

As a first pass, I shall understand “discrimination” to describe any form of unfavorable treatment as regards access to public accommodations, employment, scarce resources or opportunities, etc., where that unfavorable treatment is grounded in the disfavored person’s membership in a status-based group. To locate the ground of unfavorable treatment in group membership is already to diverge from understandings of discrimination that see it as an individual, rather than a group-based, harm.[6] On these individualist accounts, discrimination arises whenever a person is subject to unfavorable treatment on the basis of an arbitrary and ascriptive characteristic. To my mind, two considerations militate against the individualist understanding. First, it would count unfavorable treatment based on, say, the day of the week on which one was born, to be a form of discrimination. But common usage suggests that it would be strained to describe unfavorable treatment so grounded as “discrimination.” That term seems to connote a subset of unfair treatment—namely, treatment resulting from one’s possession of a protected trait, where the traits warranting protection are those that correspond to the distinguishing features of protected groups.[7] Second, the differential treatment that should concern us should disparage the target of discrimination because of the status-based characteristic.[8] We will see that Bostock poses a troubling challenge to this second consideration in what follows.

The most obvious, and obviously wrong, instances of discrimination involve animus.[9] I shall understand “animus” in the context of discrimination[10] in the following way:

A harbors animus toward B if and only if:

  1. A believes B to be inferior because of B’s membership in a status-based group;[11]
  2. A is disposed to act unfavorably toward B because of this belief; and
  • Where A acts on that disposition, A intends to set back B’s interests.[12]

Given these three conditions, benign discrimination cannot count as animus. Thus, the foreman who refuses to hire women out of concern for their sensibilities does not harbor animus. This is not to deny that his belief about women’s genteel natures isn’t, at some level, a belief in women’s inferiority. He might associate gentility with a softness that, even while virtuous, is not as honorable as whatever it is that makes men strong. The important feature in the example, though, is that the foreman does not intend to set back the woman’s interests in refusing to hire her.[13] To the contrary, he may well believe himself to be protecting, and thereby promoting, her interests. Paternalistic discrimination is just a species of benign discrimination.[14]

Does bigotry require animus? To the extent that McClain weighs in at all, she appears to suggest that it does not.[15] She reports that, at least in the marriage context, the Supreme Court’s jurisprudence has moved from “condemning bad motives—animosity—to condemning practices whose social meaning is to deny equal liberty to groups who are worthy of the status of equal citizenship” (p. 211). I am inclined to agree that bigotry does not require animus.[16] McClain provides vivid examples in early- and mid-twentieth-century statements describing Black people as less evolved than white people.[17] Those who held these beliefs likely satisfied the first two conditions of animus adduced above, but not all of them need have satisfied the third condition. In particular, they would not have satisfied it if their differential treatment was not pursued in order to set back the interests of Black people. It is arguable that some of them sincerely believed that they were promoting segregation as a matter of securing the best interests of both Black and white people.[18]

McClain seems to go further though, suggesting that bigotry can arise not only where condition (III) is absent but also where condition (I) is absent. Here we might distinguish between two cases. The first involves discriminating on the basis of demeaning stereotypes. I take the case of excluding women from construction work to be an example. While a belief in women’s refinement might appear benign and perhaps even flattering, women often find it patronizing,[19] and the limits it imposes on their opportunities oppressive.[20] Ruth Bader Ginsburg argued as much in her brief in Reed v. Reed.[21]

Or again consider an incident Tamar Gendler recounts, where a white woman diner at the Cosmos Club (a private dining club) presented a Black male guest with her coat check stub, assuming that the man must have been an employee.[22] The woman need not have harbored a belief that Black people are of inferior worth; she might simply have been operating with the belief, based on plausible anecdotal evidence, that they did not typically belong to the Club. Still, lurking in the shadows of that empirical hypothesis are the makings of a belief of inferiority: Black people are not sufficiently wealthy or well-connected—both badges of respectability in our culture—to belong. It seems not undue to charge the woman with bigotry in this kind of case, at least of the negligent variety. If she were invited to assiduously probe the ground of her assumption, she would likely uncover its denigrating bases. In this way, her belief is like the belief that Jews are good with money. On its face, the belief associates Jewish people with a positive trait, but the belief is rooted in nefarious assumptions about greed and trickery.[23] I think it not undue to count these cases as instances of bigotry even if the offending party is not in the grip of a belief of the inferiority of the offended party at the moment of the offense. It is enough that the belief motivating the offending conduct rests on other beliefs, and these other beliefs are denigrating.

Importantly, in the cases just described, it is the content of the stereotype that demeans. In a second kind of case, it is instead the stereotyping itself that demeans. To see this, consider that it can be problematic to impute positive characteristics to someone on the basis of their membership in a status-based group. For example, we mistreat an Asian person when we assume they are good at math.[24] This is a form of group-based differential treatment but it is not one plausibly described as being grounded in a belief in the inferiority of the person who is treated differently. And indeed it seems a stretch to call the person who believes that Asians are good at math a “bigot.” Still, there is undoubtedly something morally problematic afoot. We fail to take seriously the separateness of persons—to give people the respect they deserve qua individuals—when we judge them on the basis of the generalizations constituting stereotypes.[25] Moreover, this treating tokens as types may have a special sting when the basis for the generalization is a person’s membership in a status-based group.[26] Thus, it is worse to assume that someone will be relatively better than the average person at basketball because they are Black than because they are tall, even if the statistical evidence were to bear out both generalizations.

To recap, we have then two kinds of cases involving discrimination based on stereotypes, but only one seems to count as bigotry: The bigoted kind of stereotyping involves differential treatment on the basis of (assumed) undesirable features of members of a status-based group (e.g., women are genteel (and hence weak); Jews are good with money (because they are greedy)). The second kind of stereotyping also involves differential treatment on the basis of generalizations, but the generalizations themselves are benign (e.g., Asian people are good at math). I contend that bigotry requires a belief—however suppressed—in the inferiority of the bigot’s target. (This is just condition (I) above.) That is, the first kind of case of stereotyping counts as bigoted, but the second does not.

McClain seems not to agree. She contends that social meaning alone can ground bigotry—that is, an attitude toward a member of a protected class can be bigoted even if it rests on benign beliefs all the way down. Thus she allows that Justice Kennedy may have been correct in assuming that many who believe that marriage is exclusively a “gender-differentiated union of man and woman” hold that belief “in good faith” and count as “reasonable and sincere.”[27] Still, she suggests, wedding vendors who act on these decent beliefs count as bigoted, she suggests.[28] I am not convinced that this is always true. Instead, I suspect that there is a suppressed element here, which is necessary for making out the claim of bigotry: The conduct must mirror that of others who do act on denigrating beliefs, and the person engaging in that conduct must have been silent about his own relatively benign motives.

To see what I have in mind, consider the Shmites, a fictitious insular minority consisting of light-skinned people who share a religion. No one thinks the Shmites are inferior—no one, that is, except the Shmites themselves. For that reason, no one who is not a Shmite would think a Shmite person unmarriageable. But the Shmites believe that they alone are possessed of an original sin that renders them of inferior moral worth relative to all non-Shmites. The Shmites want to protect others from bearing their taint and so they commit to marrying only among themselves. If a Shmite baker could not, in good conscience, bake a cake for a wedding between a Shmite woman and a Black man, or a Shmite man and a white man, the baker’s refusal would, on its face, mirror the refusals of those who condemn interracial or same-sex marriage—refusals whose social meaning is oftentimes correctly construed as bigoted. But the Shmite baker’s motive ought to, and I believe would, undercut the bigoted social meaning. The example shows, I believe, that we cannot infer social meaning purely from conduct. At the very least, what is required is the differential treatment plus no effort to explain it away on non-denigrating grounds.[29]

In sum, not all cases of discrimination count as bigotry and not all cases of bigotry require animus. Instead, caselaw, commentary, and critical reflection point toward a definition of bigotry according to which bigotry arises where, but only where, one party subjects another to unfavorable treatment because of a belief that the party so subjected is inferior in virtue of that party’s membership in a status-based group. That finding may be straightforward enough. But it has been problematized by the Court’s recent treatment of sexual orientation discrimination.

  1. Bostock and Bigotry

Understanding bigotry as the conjunction of (I) a belief that the target is inferior because of their membership in a status-based group, and (II) a disposition to act unfavorably toward the target as a result of that belief accurately captures, I believe, the nature of the injury sustained by Gerald Bostock when he was fired for being gay. Bostock was an exemplary child welfare advocate in Clayton County, Georgia, for ten years. Under his leadership, the county won national awards. But when word got out that he played on a gay softball team, prominent members of the community made disparaging remarks about him and he was fired for conduct “unbecoming” an employee of the county.[30] In this way, his employer acted on bigotry toward LGBTQ people;[31] the wrong he endured consists in his employer’s so acting.

But that is not the injury the Supreme Court recognized in its landmark decision holding that Title VII’s prohibition on sex-based employment discrimination extends to discrimination on the basis of sexual orientation. [32] By the lights of Justice Gorsuch’s opinion for the Court, Bostock was fired because he was a man. Would Justice Gorsuch’s opinion have the resources to capture the fact that Bostock was a victim of bigotry (rather than simply the differential treatment in which all discrimination consists)? I do not believe that it would. In this Part, I aim to expose the decision’s troubling dimensions, however gratifying its result.[33]

  1. Is Sexual Orientation Discrimination
    “Because of” Sex?

Recall that bigotry requires the discriminating party to act on a belief that the target of discrimination is inferior because of the target’s membership in a status-based group. Since the Court found that the relevant status-based group was defined in terms of sex, it could make out a charge of bigotry only if it had contended that Bostock’s employer believed Bostock to be inferior because Bostock was a man and men are inferior. But there is no evidence that Clayton County believed men to be inferior. Nor does Justice Gorsuch’s opinion ever intimate that it did. So in finding that Bostock was fired because he was a man, the Court would seem to ignore the fact that Bostock was fired on bigoted grounds. And indeed it seems fair to interpret the Court’s opinion as eliding the bigotry Bostock sustained. That elision is the necessary consequence of shoehorning sexual orientation discrimination into Title VII’s sex discrimination prong, as we will see.

Justice Gorsuch defines Title VII’s prohibition on sex-based discrimination in this way: “an employer who intentionally treats a person worse because of sex— such as by firing the person for actions or attributes it would tolerate in an individual of another sex—discriminates against that person in violation of Title VII.”[34] But, he continues, “it is impossible to discriminate against a person for being homosexual . . . without discriminating against that individual based on sex.”[35] This is so because the ground of the adverse employment action—namely, that the employee is attracted, say, to men—counts against him even while being attracted to men would not count against his female colleague.[36] “Put differently, the employer intentionally singles out an employee to fire based in part on the employee’s sex, and the affected employee’s sex is a but-for cause of his discharge.”[37]

For the Court, then, sexual orientation gets linked up to sex because sexual orientation makes unavoidable reference to sex:[38] To know that a person is gay, one must know that they are attracted to members of their own sex. But then it is necessary to know the sex of the person in question. So to fire someone because they are gay is to fire them because of their sex.

The argument is logically flawed: It mistakes a necessary precondition for a cause. One can see the error if one constructs an argument with the same structure as Justice Gorsuch’s: To know that a person is brown-eyed, one must know that they have eyes. So to fire someone because they are brown-eyed is to fire them because they have eyes. Having eyes is a necessary precondition for having brown eyes, just as being a man is a necessary precondition for being a gay man. And, sure, in both cases the necessary precondition is a but-for cause of the firing. But that is because all necessary preconditions are but-for causes—that the fired person was born to these parents, that he was born at all, that he met his now-partner in virtue of whom he ended up in the town where he held the job from which he has now been fired. . . . [39] It does not follow that he was fired because of any of these preconditions.[40]

Nor is this merely a problem of logic. It is also a problem of morality—more specifically, it is a problem that goes to the moral significance of legal conclusions. The law should not merely yield the result that individuals fired because of their sexual orientation have been fired illegally. It should accurately recognize the nature of the wronging so that it can condemn that wronging. To understand what happened to an employee when he was fired for being gay as his having been fired because he is a man is to miss the injury that he sustained.

The foregoing concerns about Justice Gorsuch’s logic echo an objection Justice Alito raises in his Bostock dissent. Justice Alito invites us to imagine an employer who will not hire anyone who is homosexual. That employer does not need to know whether an applicant for the job is a man or a woman; so long as the employer knows that the applicant is homosexual, the applicant will not be hired. But in that case, the applicant’s sex cannot be a factor in the employer’s decision; after all, the employer does not even know their sex.[41]

Some progressive legal scholars have sought to vindicate Justice Gorsuch’s argument as against Justice Alito’s hypothetical. For example, Amanda Shanor writes:

Suppose an employer had a purportedly neutral rule that she would not hire any employee who was in an interfaith marriage. She would hire Catholics, Jews, and Muslims alike—just not anyone who married across religious lines. Such an employer would not have to know the religion of a particular candidate to refuse to hire those in interfaith marriages, just that they were in an interfaith relationship. Yet that would clearly be religious discrimination.[42]

Andrew Koppelman offers a similar hypothetical, focusing on interracial rather than interfaith marriages.[43] As he writes, any case where a person is fired, or not hired, because they married someone of a different race is a case where that “person is discriminated against for being the wrong race . . . ”[44] One does not need to know the particular race of the employee for the wrong to have been perpetrated.

Let us assume what seems eminently plausible—that the employer who opposes interfaith or interracial marriage harbors hostility to at least one of the religions or races at issue. For example, Koppelman’s hypothetical employer almost surely views interracial marriage as a threat to white supremacy.[45] But that assumption reveals an important asymmetry between Shanor’s and Koppelman’s cases and cases of sexual orientation discrimination: In the latter, the employer need not harbor hostility to either sex. As such, the employer’s conduct expresses nothing about the worth of being a man or woman, etc., in its own right. This difference entails that Shanor’s and Koppelman’s examples are not true analogies to cases involving same-sex couples. Their examples do not show what they need to show—namely, that sexual orientation discrimination is sex discrimination (even while they do show that opposition to interfaith or interracial marriage is based in hostility to one or more religious or racial groups, respectively). An employer is being racist when he refuses to hire someone in an interracial marriage—even if the employer does not know the race of the applicant—because the employer opposes interracial marriage on racially supremacist grounds (e.g., on grounds of white supremacy). By contrast, an employer is not being sexist when he refuses to hire someone in a same-sex marriage—whether or not the employer knows the sex of the applicant—because the employer does not oppose same-sex marriage on sexually supremacist grounds (e.g., on grounds of male superiority).

Now, in response, Shanor or Koppelman might contend that sexual orientation discrimination is a species of sexism because the former is predicated upon sex stereotyping. The invidious connection between sexual orientation and sex arises because a person who desires members of their own sex supposedly acts in a way that members of their sex ought not to act. In sanctioning employees for defying this stereotype, the employer acts illicitly on the basis of sex.[46] Two comments in response: First, Justice Gorsuch does not adduce this argument for connecting sexual orientation discrimination and sex discrimination. Second, I’m not convinced that this best captures the wrong of discriminating against people on the basis of sexual orientation. The injury of condemning someone for being a gay man seems not to reside primarily, let alone exclusively, in a judgment that he is unmanly; instead, it seems to reside primarily in a judgment that a central feature of his life somehow renders him less worthy of respect.[47] If anti-discrimination law is meant to recognize the symbolic injuries discrimination victims sustain, and teach citizens what attitudes the state will not tolerate, it should offer accurate and particular constructions of the distinct harms discrimination against different groups inflicts.

In sum, Title VII’s disparate treatment prong ought to be understood as prohibiting bigotry, as that term is defined here—i.e., adverse employment actions taken as a result of negative beliefs about the members of the groups it seeks to protect. That prong punishes racism, not race-based unfavorable treatment; it punishes sexism, not sex-based unfavorable treatment. And it should punish heterosexism too (aka anti-gay hostility or sexual prejudice),[48] which is why progressives,[49] and now the Biden administration,[50] have lobbied to have the term “sexual orientation” added to the list of protected categories under Title VII. Construing unfavorable treatment of LGBQ employees as sex-based discrimination, rather than heterosexism, elides the cruelty of being condemned for one’s sexual orientation, and it demeans the targets of both sexism and heterosexism by lumping the two together.

In fact, the Court’s efforts to shoehorn sexual orientation discrimination into the category of sex-based discrimination have two further troubling dimensions, to which I now turn:

  1. Meaning, Not Material Consequences

A different way to illuminate the missing feature in Justice Gorsuch’s account is to notice that it focuses only on the material injury, and not the expressive one. In this respect, Justice Gorsuch’s opinion aligns with Sophia Moreau’s understanding of discrimination as an injury to “our interest in a set of . . . deliberative freedoms: that is, freedoms to have our decisions about how to live insulated from the effects of normatively extraneous features of us, such as our skin color or gender.”[51] Moreau’s basic idea is that traits that do not track whether we would make good tenants or employees should not affect our ability or opportunities to make consequential decisions about where to live or work. Justice Gorsuch identifies the core “message” of Title VII in similar terms: “An individual employee’s sex is ‘not relevant to [their] selection, evaluation, or compensation . . . .’”[52]

The interest Moreau identifies is undoubtedly important, and it is good that the law protects it. But it is not the only interest at play in a case of wrongful discrimination. The target of discrimination also experiences an expressive injury. Other accounts of discrimination make this plain. For example, Deborah Hellman argues that discrimination is wrong in the first instance because it is demeaning.[53] Moreau does not deny that discrimination can wage an expressive injury. But she insists that that injury is but a consequence, and not a constituent, of the wrong of discrimination.[54] That is, for Moreau, the wrong of discrimination is the unjust cost it places on individuals’ deliberative freedoms; it just so happens that that wrong sometimes inflicts two injuries—a loss of opportunities and an expressive slight.

Because the expressive injury is, for Moreau, no necessary part of what makes discrimination wrong, Moreau’s account would seem to condemn conduct aimed at both suppressing and promoting historically oppressed groups—for example, racial segregation along with affirmative action. After all, both practices condition meaningful life opportunities on normatively extraneous facts. So in foregrounding deliberative freedoms, Moreau threatens to arrive at an overly inclusive conception of wrongful discrimination.[55]

At the same time, Moreau’s account also threatens to overlook instances of discrimination that do not involve interfering with deliberative freedoms. Take for example the exclusion of Black people from juries (or women from a military draft). It would be a stretch to say that our deliberative freedoms encompass whether we serve on a jury or get conscripted into war; neither is central to the structure or shape of the self-authoring dimensions of our lives. But these exclusions at least arguably constitute wrongful discrimination.

Moreau may have been prompted to minimize the expressive injury because she is keen to recover an account of discrimination that explains why it occasions tort liability. The tort system is individualist, and the wrong of interfering with a person’s deliberative freedoms is an individual injury. By contrast, the expressive injury is group-based. A person is demeaned because of their membership in a group that has historically been singled out for mistreatment. In this way, the employer who fires someone on the basis of race or sex or sexual orientation injures all members of the relevant group. But that group-level injury is not one that tort law can recognize.

Moreau’s account has theoretical merit insofar as it can justify tort liability. The individualist focus of the Court’s decision in Bostock cannot be so readily justified.[56] At the limit, by divorcing the discriminatory treatment from the sting it inflicts, Justice Gorsuch’s account would seem to make mysterious why we should care about differential treatment on the basis of race, sex, religion, etc., rather than any other form of differential treatment grounded in arbitrary characteristics. That is, by the logic of Justice Gorsuch’s opinion, why think that an employer who will hire Mets fans but not Yankees fans is doing anything better than the employer who will hire a man but not a woman? Justice Gorsuch could of course say that sex, but not baseball team allegiance, is what Title VII encompasses. But that is a punt. Our doctrine, and not merely our statutes, should be able to say something meaningful about why there is something uniquely injurious about mistreatment stemming from one’s membership in a status-based group.

  1. Difficulties in Treating Likes Alike

Justice Gorsuch believes that he has arrived at a “simple” test for identifying illegal discrimination:[57] Illegal discrimination arises whenever one party would not have subjected another to adverse treatment but for the affected party’s possession of a protected trait. To see where this is so, we “change one thing at a time and see if the outcome changes. If it does, we have found a but-for cause.”[58] So for example firing “a woman and a fan of the Yankees is a firing ‘because of sex’ if the employer would have tolerated the same allegiance in a male employee.”[59] By contrast, to fire all fans of the Yankees—men and women alike—is not to fire any of them because of sex. The relevant element then is differential treatment as between two individuals alike in all respects except that one is a man and the other a woman (or one is white and the other Black; or one is straight and the other gay).[60]

Now consider the following cases: Antonio is an anti-Semitic baker. He refuses to bake cakes with the words “Mazel Tov.” Is Antonio engaging in discrimination? Not on Justice Gorsuch’s test so long as Antonio would refuse to bake a cake with the words “Mazel Tov” for any customer, no matter their religion. Or again: Shmeinstein is a hotshot movie producer who fires all employees who rebuff his offers to have sex. Does Shmeinstein subject his employees to discrimination (here, harassment) on the basis of sex? Again, it is not clear that one could answer this question in the affirmative on Gorsuch’s test.

The test’s failures can again be traced to its focus on differential treatment rather than the meaning of that treatment. An account that recognizes only comparative discrimination cannot recognize discrimination where there is no comparison group.[61] But the problem with discrimination is not always—or even primarily—that it treats some people worse than others. It is that it treats some people badly no matter how it treats others. If Antonio refuses to bake “Mazel Tov” cakes for anyone because he hates Jews, that is an injury in itself. And if Shmeinstein makes acceding to his sexual advances a condition of continued employment, that is sexual harassment even if he preys equally on men and women. The law should condemn these instances of discrimination for the injuries they are.

  1. Discrimination and Bigotry

One might worry that there is something retrogressive about insisting that the law attend to bigotry. After all, we should all want less discrimination. If loosening the definition of “sex discrimination” has the effect of protecting LGBTQ individuals, as in Bostock, that should be cause for celebration, not concern. Similarly, denying that bigotry need function as the sine qua non of wrongful discrimination opens up the possibility of recognizing forms of discrimination that may be less pointed but far more pervasive, like those stemming from implicit biases or systemic racism.[62]

With all that said, I think the retreat from recognizing bigotry, which Bostock exemplifies, deserves at most one cheer, and that is so for two reasons. First, it is not as if we have reason to trust that the state has entered a phase of great receptivity to the wrongs of discrimination. To the contrary, it is as blind as ever to disparate impact and the structural and systemic injustices that underpin it.[63] More pointedly, the very same features that the Court adduces in Bostock—that the employers acted “intentionally” and that the verboten element in their adverse employment decisions was to treat employees differently on the basis of a protected characteristic—would seem to spell doom for disparate impact cases and affirmative action cases alike.

Second, there is no reason to think that the law can address systemic injustices only if it ignores the distinct moral meaning of bigotry. Similarly, there is no reason to think that the law can condemn sexual orientation discrimination only if it assimilates it to sex-based discrimination. Each inflicts an injury different from the other, and the law should be fine-grained enough to condemn each for the injury it is.

III. What Is Ratification?

Much of McClain’s book focuses on private discrimination, and understandably so. The Equal Protection clause forbids state discrimination, and federal civil rights laws extend that prohibition (imperfectly, to say the least) to other contexts of daily life, including employment, public accommodations, housing, education, and so on.[64] By contrast, the Court recognizes a private sphere into which the strictures of equality need not enter.[65]

On rare occasions, however, the judiciary has pierced the private sphere with an eye to undermining private bigotry. Two cases are routinely identified as the high-water marks of this kind of judicial intervention,[66] and celebrated for their forceful anti-racist underpinnings. Both cases pick up on a theme that recurs in McClain’s work—that of ratification (pp. 156, 159, 178, 199, 203).

In Shelley v. Kraemer, the Supreme Court refused to enforce a racially restrictive covenant, holding that doing so would recruit the state into a violation of the Equal Protection Clause.[67] In Palmore v. Sidoti, the Court overturned a custody arrangement that was based on community hostility to interracial marriage, because the Court found that deferring to that hostility would vindicate it.[68] Both cases, then, stand for the proposition that the state need not sustain private arrangements that contravene its commitment to equality.[69]

Many progressives herald the anti-ratification rationale for which these cases stand as the antidote to bids for exemptions from anti-discrimination laws. For example, when religious wedding vendors seek to refuse service to same-sex couples, in violation of public accommodations laws, these progressives respond by arguing that any exemption from those laws would have the state ratify the vendors’ discrimination.[70] While I have elsewhere argued that the state should not grant wedding vendors exemptions,[71] I do not believe the anti-ratification rationale is the best ground for that view.

As I see it, there are two problems with anti-ratification. First, concerns about ratification are not the state’s alone; the wedding vendors who want to deny service to same-sex couples argue that they would be ratifying same-sex marriage, in violation of their convictions, were they to provide their goods or services. Progressives can reject the vendors’ ratification concerns only at the expense of defining “ratification” so narrowly that Shelley and Palmore would be of no avail. Second, ratification undersells the nature of the state’s complicity when it does permit private discrimination. I elaborate on each of these in turn.

  1. Distinguishing Private and State Ratification

We can understand wedding vendors’ conscientious objections to public accommodations laws that would require they provide their goods and services to same-sex weddings as bids to avoid ratification.[72] Jack Phillips, the baker in Masterpiece Cakeshop v. Colorado, articulated his concern about providing a wedding cake for a same-sex wedding in just these terms, arguing that Colorado’s anti-discrimination law, by requiring him to furnish a wedding cake, would compel him “to express or celebrate what he cannot in good conscience support.”[73]

In this Section, I aim to show that those who support the rule in Shelley while also rejecting the ratification rationale when it is deployed by a wedding vendor rely upon dubious grounds for distinguishing between the two cases.

More specifically, progressives have rejected wedding vendors’ concerns about ratification on three principle grounds: first, the wedding vendors’ contributions are too minor or remote to count as ratification; second, the wedding vendors’ contributions arise in a sphere where concerns for conscience have no place; and, third, even if the wedding vendors did have compelling and appropriately placed concerns, the state ought not to accede to them because doing so would impermissibly inflict harm on third parties. I address each in turn.

  1. Causal Insignificance

The recent bids for religious conscientious exemptions have met with resistance partly on the ground that the connection between the provider and the project to which the provider objects is too minimal or attenuated to count as ratification.[74] For example, opponents of these exemptions question why an exmployer should believe that it is implicated in its employee’s contraceptive use simply because the employee accesses contraception through the company health plan. Similarly, they question why a baker should take himself to be complicit in the (supposed) sin of same-sex marriage simply because he provided the cake consumed at the wedding reception.

I believe that these opponents pursue the wrong line of question for assessing complicity claims. The state’s response to concerns about ratification should not turn on questions about the strength of the causal connection between the provider and the customer’s project.[75] For one thing, gauging causal responsibility is a murky task in its own right. Is the doctor who prescribes the medical cocktail to execute a prisoner more or less causally implicated in the death penalty than the person who throws the switch on the electric chair? Second, even where we can confidently judge the strength of a causal connection, there is no reason to think that there is any necessary, or even regular, relationship between causal and moral responsibility. Most relevantly here, the deference we owe to conscientious objections may have little to do with how causally implicated the objector is in the conduct to which he objects. The pacifist has a legitimate claim to have his concerns about being drafted taken seriously whether he would be serving in combat or the canteen, or indeed whether he would be serving near the battlefield at all. Those who focus on the magnitude or proximity of the causal contribution conflate the standards we should apply when we are judging another’s complicity with those the conscientious person ought to apply to herself. It is in general a good thing for someone to hold herself to higher standards than we would apply to her. For example, the person who sees herself as implicated in environmental destruction anytime she fails to recycle a plastic bottle operates with a punctiliousness we would do well to emulate, not disdain. We should not dismiss concerns about ratification on the ground that they are too minor or remote.

  1. The Amoral Marketplace

A different strategy for dismissing the wedding vendors’ ratification worries involves insisting that there is no place for conscience in the commercial marketplace[76] (and perhaps public service agencies too).[77] This claim figured prominently in the New Mexico Supreme Court’s disposition of Elane Photography, LLC v. Willock, in which a photographer opposed to same-sex marriage refused to photograph the commitment ceremony of two women.[78] The court denied the photographer’s bid for an accommodation. Justice Bosson’s concurring opinion in the case captures the insistence on a morality-free marketplace: “The Huguenins are free to think, to say, to believe, as they wish . .  in their personal lives . . . . But [they may not act on those beliefs in] the smaller, more focused world of the marketplace, of commerce, . . . ”[79] McClain quotes this passage approvingly
(p. 190). And other commentators have expressed their support for the idea that sellers in the retail marketplace must leave their consciences at home.[80]

To my mind, the claim that the marketplace cannot accommodate conscience is unneccessarily, and troublingly, sweeping. Surely it is a good thing when businesses decline to source their goods from sweatshops, or use their revenues to fund social welfare initiatives. More to the point, we ought to commend, not condemn, many conscientious refusals of service. We should not have to heed Justice Bosson’s exhortation to have vendors set aside their convictions and “leave space for other Americans who believe something different”[81] where that would condemn them to having to host the KKK banquet or cater the local fraternity’s dwarf bowling extravaganza.[82] We should not settle for a blanket prohibition on refusals of service if we can accommodate conscience consistent with anti-discrimination norms, and I believe we can.[83]

  1. Protecting Third Parties

Finally, some argue that the limit on an exercise of conscience arises where that exercise would impose burdens on third parties.[84] McClain cites this as “well-established jurisprudence” (p. 189). I do not share her view of the caselaw. I have argued at length that doctrinal evidence for a prohibition on third-party harms is murky, and not authoritative at any rate.[85] Moreover, even if the law did enshrine an unequivocal prohibition on exemptions where they would harm third parties, it is not clear that those who oppose exemptions should want to rely on that prohibition. If the reason to deny an exemption is that it stands to harm third parties by depriving them of goods and services, then there may be no reason to deny the exemption where there are other vendors nearby who would be willing to provide those same goods and services.[86] Yet, surely the rights of LGBTQ individuals to be free from discrimination should not be vulnerable to such contingencies. Instead, a polity fully committed to equality would want a categorical prohibition on discriminatory refusals of service, rather than one that kicks in only in the event that the exemption would harm others—e.g., as where the customer who would be turned away would be unable to find a feasible alternative to the discriminating business.

  1. Summary

We have considered three possible features that might thwart bids to avoid ratification in public accommodations: that the conduct the vendor would supposedly ratify is too remote for him to be properly concerned about his complicity; that the marketplace obviates complicity; and that even if wedding vendors’ concerns for complicity were warranted, acceding to them would burden third parties, which the law forbids. I have argued that none of these is compelling in its own right, so none can justify denying wedding vendors exemptions. But more than that, each of these three factors, if accepted, would stand to undermine not just wedding vendors’ concerns about ratification, but those of the state too: Egalitarian opposition to school voucher programs turns, at least in part, on the thought that these programs lend state legitimacy to parents’ discriminatory choices. But if ratification is a verboten rationale in the wedding vendor cases, why should we take it seriously in the school voucher cases?[87] Second, Shelley is as much a case in the marketplace as is Masterpiece Cakeshop. And, finally, state refusals to issue exemptions to religious vendors can be read as disdain for the convictions that are at stake, thereby denigrating all those who share the vendors’ belief. So the state’s decision to uphold the anti-discrimination law can burden third parties in the same way that vendor refusals to provide cakes for same-sex weddings burden the wedding couple.

In short, efforts to defeat the vendors’ ratification concerns fail to convince. And if they did, they would then also impugn the state’s claim to avoid ratification.

  1. Restricting Rights Rather Than Ratification

The better move, I now suggest, is not to seek to distinguish state and private ratification. It is instead to argue that the state that lends its resources to private action contrary to state morality does more than merely ratify; it participates in that action.

To motivate this strategy, notice a troubling implication of the ratification rationale. That rationale allows that the underlying conduct—the conduct the state seeks to avoid ratifying—is perfectly permissible so long as it does not involve the state. Thus, in Palmore, the Court explicitly acknowledged that “private biases may be outside the reach of the law.”[88] And in Shelley, the Court insisted that the Fourteenth Amendment “erects no shield against merely private conduct, however discriminatory or wrongful.”[89] These cases thereby suggest that the ratification rationale stops well short of where it ought. It implicitly condones racism by recognizing a private right to be racist.[90]

If, instead, one starts from the premise that the private party acted impermissibly in discriminating, then the state’s role in upholding the discrimination would not accurately be characterized as “ratification.” To see what I have in mind, consider Palmore v. Sidoti again. If the state heeds racial prejudices and so decides not to place the child of divorced parents with the parent who has since entered into an interracial marriage, the state does not merely ratify those prejudices; it participates in the prejudicial treatment. The prejudiced party’s principle of action becomes the principle of the state, as the principle becomes the ground of the custody determination.[91]

One can find support for this understanding of the state’s role in Burton v. Wilmington Parking Authority, a case involving a private restaurant housed in a government building that refused to serve Blacks.[92] In his concurring opinion, Justice Stewart located the state action not in the fact that the city owned the property that housed the restaurant, but instead in the state law that “permit[ted] the proprietor of a restaurant to refuse to serve ‘persons whose reception or entertainment by him would be offensive to the major part of his customers.’”[93] Based on the text alone, one might have concluded that the statute had been intended for benign purposes—e.g., to permit refusing service to patrons who were rowdy, rude to the waitstaff, improperly attired, and so on. The problem, as Justice Stewart noted, was that the Delaware Supreme Court did not give the statute that construction in its opinion below. The court did not find, nor could it have found, that the plaintiff was ineligible for service because he was rowdy, improperly dressed, etc. Instead, he stood to “offen[d]” solely because he was Black.[94] As such, the constitutional infirmity lay in the Delaware Supreme Court’s construing the statute in question such that it provided a right to discriminate on the basis of race.[95] In this way, Delaware did not merely ratify discrimination in finding that the restaurant, “acting in a ‘purely private capacity,’”[96] had a right to refuse service on the basis of race; it enshrined that right in statute, making it the law of the land.

Where the state adopts or interprets a law legalizing racial discrimination, the state does well more than ratify that discrimination, even if it is the restauranteurs themselves who will decide whether to take up the state’s invitation.[97] Indeed, we can see that the state itself discriminates in enacting this law once we recognize that the mere existence of the law injures, even if no restauranteur ever takes up the legal permission it grants.

But what about cases where the discrimination does not arise out of a law of the state? One might seek to distinguish Burton from cases seeking exemptions from public accommodations by noting that, in Burton, Delaware issued a blanket permission to discriminate, whereas, in the wedding vendor cases, the permission would be restricted to those who have religious objections to same-sex marriage.

To my mind, the distinction is spurious. When Florida heeded the prejudices at issue in Palmore, it did not say that custody in an interracial family would everywhere be disfavored; it restricted its decision to the context where the community in which the child would be raised denounced interracial marriage and so would subject the child to stigmatization. Still, the fact that the Florida court’s decision was attentive to the best of interests of this child, given her particular circumstances, does not make that court’s decision to accede to racism less racist.

Perhaps, though, denominating the vendors’ permission as an “exemption” already allows the state to distance itself from the vendors’ conduct. After all, the term “exemption” implies that the recipient of the exemption is being permitted something that the state in general opposes; the exemption is a deviation from state policy.

I am not convinced by this suggestion. The more accurate way to understand exemptions, I believe, is to see them as adding disjunctive clauses onto the statutory text. So, for example, an exemption from an anti-discrimination law ought to be read as “you may not discriminate on the basis of a protected category OR you may discriminate so long as you can establish that your reasons reside in sincere religious beliefs, etc.” Understanding the structure of statutes-plus-exemptions in this way allows us to see that the second disjunct effectively obliterates the moral force of the first.

This seems to have been the view of the Court in Newman v. Piggie Park, where the Court found “patently frivolous” the notion that one should gain an exemption from anti-discrimination laws because one believes, as a religious matter, that integration “‘contravenes God’s will.’”[98] The Court rejected Piggie Park’s bid for an exemption because it recognized that anti-discrimination laws cannot co-exist with exemptions from them. Any exemption signals that equality is not the state’s highest priority; instead, respect for religion enjoys greater priority
still. Yet anti-discrimination laws exist precisely in order to convey that no interest is more important than equality in
the spheres to which these laws apply. Their purpose is to “eradicate[e] barriers to the equal treatment of all citizens . . . Were [the state] to carve out a patchwork of exceptions for ostensibly justified discrimination, that purpose would be fatally undermined.”[99]

One might think that an insistence on categorical adherence to anti-discrimination laws would foreclose every exemption from them. In fact, I believe matters are more complicated. There is one species of exemption I believe the state should countenance—namely, those where compelled service would recruit sellers into contributing to the promotion of hate that it is the purpose of anti-discrimination laws to combat. To take a paradigmatic example, consider groups like Christian Identity, a religious KKK group that subscribes to “a unique anti[-S]emitic and racist theology.”[100] If a restauranteur refuses to cater, or have her party room used for, a Christian Identity worship service and banquet, she will be violating public accommodations laws. (It is no defense for her to say that her opposition targets the supremacist aspects of the group, not its religious aspects—i.e., that she would not cater an event held by a secular KKK group either—since supremacy is a matter of religious belief and command for Christian Identity. Her claim would then be akin to saying, “I would bake birthday cakes for my LGBTQ customers but not cakes celebrating their weddings.” Where the objectionable conduct is a central feature of the protected identity, a vendor cannot refuse to support the conduct without also discriminating on the basis of that identity.[101]) Should she be permitted to seek an exemption from these laws? I believe she should. I defend that position at length elsewhere.[102] Here, I will merely gesture at two thoughts that support it.

First, note how different the disjunction in this kind of exemption case would look relative to the disjunction articulated above: “You may not discriminate on the basis of a protected category OR you may discriminate on the basis of a protected category [religion, in the Christian Identity KKK case] where providing service would in fact involve your participating in discrimination on the basis of a different protected category [race].” Seeing the disjunction in these terms makes plain that cases like Christian Identity land the vendor in a practical contradiction: the first disjunct mandates that she not discriminate; but accepting the Christian Identity’s business would have her discriminate. Of course, one might say that the two forms of discrimination are not equivalent: the first disjunct prohibits direct discrimination—here, turning the Christian Identity patron away. By contrast, compelling service would not involve the vendor in direct discrimination; it would instead simply have her contribute goods and services to an event at which other people (i.e., Christian Identity adherents) discriminate. But once the state must compel the vendor to provide service—once it, say, enjoins her from refusing—the state comes to participate in the Christian Identity’s activities just as surely as if it were to provide the goods and services itself. That is just what the discussion of participation, described above, is meant to establish.

At the same time—and this is the second thought—it is important that it is through an exemption regime that the state avoids participation. The alternative would be for the state to outright prohibit the hate-promoting activities of groups like Christian Identity. But that alternative would straightforwardly involve the state in a Free Exercise violation, as well as viewpoint discrimination. Under an exemption regime, by contrast, the state takes no stance when vendors willingly serve Christian Identity. (Again, it is a subjective matter whether a vendor sees herself as participating in the event to which she contributes.) It is only when vendors are unwilling to serve, because they believe service would implicate them in discrimination, that the state comes to be involved. In such cases, the state owes it to its subjects—those whom it commands not to discriminate—to support them in their resolve not to discriminate.

I mean to draw out a continuity between anti-discrimination laws and a regime that would allow vendors to refuse service to customers who would use the vendors’ goods or services in projects that promote hate directed at protected classes. Both the laws and the refusals aim to thwart bigotry. But one need not agree that an anti-discrimination regime should, or even may, grant any exemptions at all. The more important point is that the state may not grant exemptions in cases where the putative customer is a member of a protected class, and the project for which he seeks the vendor’s wares is not one that would promote bigotry. The prohibition on exemptions in these cases is necessary to prevent the state from participating in the vendors’ discrimination.

  1. Conclusion

In this era of national reckoning with injustices afflicting virtually all identity-based groups, it is undoubtedly important that we attend to the subtle, systematic, and structural elements of our institutional arrangements and social practices that contribute to inequality. It is a sobering fact that so much inequality is not a matter of bigotry—sobering, because bigots are outliers, whereas the blind spots and unwitting biases explaining so much inequality today may well be pervasive (p. 216). While we should unquestionably commit ourselves to exposing and extirpating the sources of this pervasive inequality, we should also ensure that we do not end up minimizing the evils of bigotry in the process. Black Lives Matter rightly targets systemic racism, but Derek Chauvin is not a well-meaning but simply misguided police officer. Double binds and the motherhood wage gap may well account for much of the gender-based disparity in workplace remuneration and success,[103] but there are still the Harvey Weinsteins to contend with, too. Bigotry may be the tip of the iceberg, but it is no less acute or compelling for that.

At the same time that we resist minimizing bigotry, we should expand our understanding of state action. The state may participate in well more bigotry—as well as much more non-bigoted discrimination—than much of our caselaw or commentary has recognized. Tracking the mechanisms of state contributions to these discriminatory practices may yield fruitful insights into how to eradicate them.

        *     Robert Kent Professor, Boston University School of Law.

        [1].    J.D., Ph.D. (Philosophy). Associate Professor, Legal Studies and Business Ethics, The Wharton School, University of Pennsylvania. I thank the Wharton Dean’s Research Fund for generous support, and Nico Cornell for fruitful exchanges.

        [2].    What follows is a woefully incomplete sample of relevant works: Michele M. Moody Adams, A Commentary on Color Conscious: The Political Morality of Race, 109 Ethics 408 (1999); Elizabeth Anderson, What is the Point of Equality?, 109 Ethics 287 (1999); Owen M. Fiss, Groups and the Equal Protection Clause, 5 Phil. & Pub. Aff. 107, 123–24 (1976); Philosophical Foundations of Discrimination Law (Deborah Hellman & Sophia Moreau eds., 2013); Tommie Shelby, Race and Ethnicity, Race and Social Justice: Rawlsian Considerations, 72 Fordham L. Rev. 1697 (2004); Christopher L. Eisgruber & Lawrence G. Sager, Religious Freedom and the Constitution (2007); Iris Marion Young, Justice and the Politics of Difference (1990).

        [3].    In one of the book’s most fascinating discussions, McClain explores the question of why “bigotry” is so seldomly applied to animus-based discrimination on the basis of sex (pp. 227–229). She concludes that the term “misogyny” has displaced “bigotry,” and she approves of this usage, since “misogyny” more specifically identifies the injury in question.

        [4].    Andrew Koppelman offers a similar diagnosis in his book, Gay Rights v. Religious Liberty?: The Unnecessary Conflict (2020).

        [5].    We might imagine that our foreman is under the sway of something like the belief expressed in Justice Bradley’s concurring opinion in Bradwell v. Illinois: “The natural and proper timidity and delicacy which belongs to the female sex evidently unfits it for many of the occupations of civil life.” 83 U.S. 130, 141 (1872) (Bradley, J., concurring).

        [6].    For discussions describing these two orientations, see, e.g., Fiss, supra note 2, at 126; Richard A. Primus, Equal Protection and Disparate Impact: Round Three, 117 Harv. L. Rev. 493, 553 (2003); Reva B. Siegel, Discrimination in the Eyes of the Law: How “Color-Blindness” Discourse Disrupts and Rationalizes Social Stratification, 88 Cal. L. Rev. 77, 92–93, 98–99, 103–05 (2000). For discussion of the ascendancy of individualism in Supreme Court doctrine, see, e.g., Heather K. Gerken, Understanding the Right to an Undiluted Vote, 114 Harv. L. Rev. 1663, 1718–19 (2001).

        [7].    Cf. City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 470–71 (1985) (Marshall, J., concurring) (“Discrimination, in the Fourteenth Amendment sense, connotes a substantive constitutional judgment that two individuals or groups are entitled to be treated equally with respect to something.  . . .  [In particular, the Amendment] prohibit[s]  . . .  castes created by law along racial or ethnic lines . . . .”) (internal citations omitted; italics added); Connecticut v. Teal, 457 U.S. 440, 459 (1982) (“There can be no  . . .  disparate impact in the absence of disparate impact on a group”).

        [8].    Compare Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 229–30 (1995) (condemning all uses of race-based criteria) with id. at 245 (Stevens, J., dissenting) (“The consistency that the Court espouses would disregard the difference between a ‘No Trespassing’ sign and a welcome mat.”). The view of discrimination that I am championing comes down firmly on Justice Stevens’s side of the debate.

One might have thought that a requirement that discrimination involve disparagement would rule out cases of disparate impact. But disparate impact is often meaningful as evidence of intentional and covert efforts to exclude disfavored groups. See, e.g., Jennifer C. Braceras, Killing the Messenger: The Misuse of Disparate Impact Theory to Challenge High-Stakes Educational Tests, 55 Vand. L. Rev. 1111, 1143–44 (2002) (discussing this rationale). Further, practices with a disparate impact can be disparaging even if wholly unintentional, for they may subtly convey that the members of the group whom they disadvantage do not deserve to occupy the positions or to obtain the benefits from which these members have been excluded. See, e.g., Primus, supra note 6, at 567–85. Cf. Owen M. Fiss, A Theory of Fair Employment Laws, 38 U. Chi. L. Rev. 235 (1971) (arguing that facially neutral practices that disproportionately disadvantage racial minorities operate as the “functional equivalents,” and also the moral equivalents, of race-based criteria).

        [9].    Even in one of the Supreme Court’s most racist decisions, Korematsu v. United States, 323 U.S. 214 (1944), where it affirmed the constitutionality of internment camps for Japanese Americans during World War II, the Court condemned animus-based discrimination: “[p]ressing public necessity may sometimes justify the existence of [racial discrimination]; racial antagonism never can.” Id. at 216.

      [10].    Of course, animus need not be restricted to the context of discrimination. The Montagues and Capulets harbored mutual animosity. See William Shakespeare, The Tragedy of Romeo and Juliet (1597). I leave aside the question of whether animosity unrelated to the hated party’s membership in a status-based group need involve a belief in the inferiority of that party.

      [11].    I follow the Supreme Court in describing the relevant kinds of discrimination as “status-based.” Thus, for example, the Court understands Title VII’s discrimination provision making it an “unlawful employment practice” for an employer “to discriminate against any individual . . . because of such individual’s race, color, religion, sex, or national origin,” § 2000e-2(a), as a prohibition on “status-based” discrimination, Univ. of Texas Sw. Med. Ctr. v. Nassar, 570 U.S. 338, 348 (2013).

      [12].    I take this definition to comport with the Supreme Court’s “initial articulation of animus as ‘a bare . . . desire to harm a politically unpopular group.’” Susannah W. Pollvogt, Unconstitutional Animus, 81 Fordham L. Rev. 887, 888 (2012) (quoting U.S. Dep’t of Agric. v. Moreno, 413 U.S. 528, 534 (1973)). Other theorists have also highlighted intent to harm as central to animus. See, e.g., Joseph S. Jackson, Persons of Equal Worth: Romer v. Evans and the Politics of Equal Protection, 45 UCLA L. Rev. 453, 496 (1997) (understanding animus to arise in laws where “the illicit purpose [is] to disadvantage the group burdened by the law”).

      [13].    In contrast, consider the Colorado constitutional amendment challenged in Romer v. Evans, 517 U.S. 620 (1996). That amendment would have prohibited any political subdivision from passing laws protecting people from discrimination on the basis of sexual orientation. Finding not even a rational basis for this prohibition, the Court found that the amendment seemed “inexplicable by anything but animus toward the class it affects.” Id. at 632.

      [14].    As applied to women, the attitudes of those like the foreman are sometimes described as “benevolent sexism.” See, e.g., Peter Glick et al., Beyond Prejudice as Simply Antipathy: Hostile and Benevolent Sexism Across Cultures, 79 J. Personality & Soc. Psychol. 763, 773 (2000).

      [15].    By contrast, all animus directed toward status-based groups is bigotry. McClain acknowledges this connection (p. 148). See also Civil Rights—Public Accommodations: Hearing on S. 1732 Before the S. Comm. on Commerce, 88th Cong. 22 (1963) (statement of Robert F. Kennedy, Att’y Gen. of the United States) (understanding refusals to serve racial minorities in public accommodations as “insult[s perpetrated]  . . .  for no reason than the arbitrary and immoral logic of bigotry.”); cf. Linda McClain, The Civil Rights Act of 1964 and “Legislating Morality”: On Conscience, Prejudice, and Whether “Stateways” Can Change “Folkways,” 95 B.U. L. Rev. 891, 901–02 (2015) (citing advocates of the Civil Rights Act for the proposition that humiliation was the aim, and not merely a byproduct, of race-based discrimination).

      [16].    Cf. Deborah Hellman, Discrimination and Social Meaning, in The Routledge Handbook of the Ethics of Discrimination 97 (Kasper Lippert-Rasmussen ed., 2017) (“A requirement that the actor be motivated by animus in order for his action to constitute wrongful discrimination also would permit race or sex discrimination in those cases where race or sex is a good proxy for a relevant characteristic.”).

      [17].    See, e.g., p. 84 (describing eugenic view that supposedly justified racism).

      [18].    McClain also provides evidence that opponents of anti-miscegenation laws—for example, the married couple in Loving v. Virginia—undoubtedly decried “racial prejudice” (which would seem to be synonymous with “racial bigotry”), which they characterized as a belief in Blacks’ inferiority combined with efforts to subordinate Blacks, all the while leaving open whether these efforts aimed to injure Blacks; instead they might have sought only to “exalt” whites (p. 185).

      [19].    See, e.g., Peter Glick & Susan T. Fiske, The Ambivalent Sexism Inventory: Differentiating Hostile and Benevolent Sexism, 70 J. Personality & Soc. Psychol. 491, 510 (1996).

      [20].    See, e.g., Marilyn Frye, The Politics of Reality: Essays in Feminist Theory 6 (1983); Drucilla Cornell, Freedom’s Conscience Women, Gays, and the Constitution 63 (1998) (“psychosexual dynamics [ ], in the name of chivalry, oppressed both ‘white’ and ‘Black’ women.”); cf. Courtney Fraser, From “Ladies First” to “Asking for It”: Benevolent Sexism in the Maintenance of Rape Culture, 103 Cal. L. Rev. 141, 144 (2015) (arguing that benevolent sexism, or chivalry, underpins sexual violence).

      [21].    See Carol Pressman, The House That Ruth Built: Justice Ruth Bader Ginsburg, Gender and Justice, 14 N.Y.L. Sch. J. Hum. Rts. 311, 320–21 (1997).

      [22].    Tamar Szabo Gendler, On the Epistemic Costs of Implicit Bias, 156 Phil. Stud. 33 (2011).

      [23].    In one survey, respondents’ five most selected traits for Jews were “Shrewd,” “Mercenary,” “Industrious,” “Grasping,” and “Intelligent”—“multi-directional (both positive and negative) traits that one could indeed trace to the propagation of stereotypes surrounding Jewish control of Wall Street.” Justin D. Levinson et. al., Judging Implicit Bias: A National Empirical Study of Judicial Stereotypes, 69 Fla. L. Rev. 63, 94 (2017); cf. Katherine M. Franke, The Central Mistake of Sex Discrimination Law: The Disaggregation of Sex from Gender, 144 U. Pa. L. Rev. 1, 38 n.148 (1995).

      [24].    Emmalon Davis, Typecasts, Tokens, and Spokespersons: A Case for Credibility Excess as Testimonial Injustice, 31 Hypatia 485 (2016). Doctrine supplies a different example. In City of Los Angeles v. Manhart, 435 U. S. 702 (1978), an employer required women to make larger pension fund contributions than men. The employer sought to justify its disparate treatment on the ground that women tend to live longer than men, and thus are likely to receive more from the pension fund over time. By everyone’s admission, the employer was not guilty of harboring animosity toward women or relying on “a fictional difference between men and women.” Id. at 707. Instead, the employer deployed “a generalization that the parties accept as unquestionably true.” Id. Further, the difference is likely a good thing for women for whom the generalization is true; presumably, a longer life is better than a shorter one. Still, it is disconcerting and maybe even offensive to be treated generically, which is how the pension fund treated all women. See Sarah Moss, Moral Encroachment, 118 Proc. Aristotelian Soc’y 177 (2017).

      [25].    See Moss, supra note 24.

      [26].    Cf. Metro Broad. v. FCC, 497 U.S. 547, 604 (1990) (O’Connor, J., dissenting) (decrying policies that “embody stereotypes that treat individuals as the product of their race, evaluating their thoughts and efforts—their very worth as citizens—according to a criterion barred to the Government by history and the Constitution”); Roberts v. U.S. Jaycees, 468 U.S. 609, 628 (1984) (“the Jaycees relies [sic] solely on unsupported generalizations about the relative interests and perspectives of men and women . . . . Although such generalizations may or may not have a statistical basis in fact with respect to particular positions adopted by the Jaycees, we have repeatedly condemned legal decisionmaking that relies uncritically on such assumptions (citation omitted)).

      [27].    P. 155 (quoting Obergefell v. Hodges, 576 U.S. 644 (2015)).

      [28].    “I have argued that equating bigotry only with hateful motives or actions misses the historical prominence of religious bigotry as a form of bigotry” (p. 212). See also id. (noting the “problem” with denying that a “sincere religious belief cannot possibly be bigoted”).

      [29].    This is of course, just the rationale used to defend the BFOQ doctrine—i.e., employment decisions based on protected characteristics for which there is a bona fide occupational qualification. See 42 U.S.C. § 2000e-2(e)(1).

      [30].    This presentation of the facts follows that in the Supreme Court opinion almost verbatim. See 140 S. Ct. 1731, 1737–38 (2020).

      [31].    Bostock was consolidated with two other cases alleging wrongful termination. One of these also involved sexual orientation discrimination. The other involved gender identity discrimination. 140 S. Ct. at 1738. The question that this section addresses is whether discrimination on the basis of sexual orientation is meaningfully captured under Title VII’s prohibition on discrimination because of sex. I do not consider whether Title VII’s “because of sex” language ought to cover gender identity discrimination, although I am much more sympathetic to the thought that it does.

      [32].    Title VII provides, in pertinent part, that “It shall be an unlawful employment practice for an employer [ ] to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s  . . . sex . . . .” 42 U.S.C. § 2000e-2(a)(1).

      [33].    While this Part offers a critical assessment of the Court’s reasoning, I want to be clear about where I stand on the moral merits of Bostock’s complaint: Firing someone on the basis of their sexual orientation is wholly unacceptable and the law should categorically forbid it. The problem with Bostock, as I see it, is not that it extends protections to an undeserving group; it is instead that it misconstrues—indeed, it demeans—the nature of the injury that members of that group sustain when they are victims of employment discrimination. I aim to make that clear in what follows.

      [34].    Bostock, 140 S. Ct. at 1740.

      [35].    Id. at 1741.

      [36].    Id.

      [37].    Id.

      [38].    “Sex plays a necessary and undisguisable role in the decision, exactly what Title VII forbids.” Id. at 1737.

[39].  Justice Gorsuch might object because the employer need not make “undisguisable” reference to any of these conditions in firing Bostock. But nor need an employer make undisguisable reference—or indeed any reference—to sex in firing someone because of their sexual orientation. Justice Alito offers a hypothetical where a manager asks his clerk about a job applicant’s sexual orientation but not about the applicant’s sex, and then declines to hire the applicant when the clerk tells the manager the applicant is “homosexual.” 140 S. Ct. at 1758–59 (Alito, J., dissenting). Justice Alito’s hypothetical clearly involves discrimination on the basis of sexual orientation all the while bypassing any reference to sex. (I elaborate on the hypothetical infra, text accompanying note 41.) At the same time, there are undoubtedly other elements or details that the employer cannot bypass in their deliberations and yet—like sex—the employer does not fire the employee because of these elements. For example, an employer cannot fire someone unless the employer believes that the person to be fired is currently in their employ. Does that mean that the employer fires the person because of the fact that the person is currently in their employ? Obviously not. In short, whether or not some element must factor in the adverse employment decision says nothing about whether it was meaningfully because of that element that the employer took the decision.

      [40].    Cf. Jesse Hughes et al., A Semantics for Means-End Relations, 158 Synthese 207, 224 (2007) (distinguishing between a necessary precondition for an end and a means to that end). Building on an example of theirs: a Bachelor’s degree is a necessary precondition for law school admission; it is not the cause of one’s attending law school (which, as they might put it, would be “better employment,” id. at 224).

Perhaps the relationship between sex and sexual orientation is not that the former is a precondition for the latter but instead that the former is a proxy, or provides evidence, for the latter. Even then, however, the Court’s opinion would fare no better. Here, the analogous argument would look something like this: In Nazi Germany, Jews must wear a yellow star. An anti-Semitic shopkeeper refuses service to anyone wearing the star. So the shopkeeper refuses service because of the star. The argument’s inference is false: one should understand the refusal as being “because of the star” only if one adopts a tortured understanding of “because.” It is true that the shopkeeper uses the star to identify the customers to whom he will refuse service. But it is not the star that excites his animosity, it is the religion of the wearer.

      [41].    140 S. Ct. at 1758–59 (Alito, J., dissenting).

      [42].    Amanda Shanor, Sex Discrimination Behind the Veil Is Still Sex Discrimination, Take Care (Oct. 11, 2019),

      [43].    Andrew Koppelman, Justice Alito’s Desperation Move, Balkinization
(Oct. 9, 2019), Koppelman expands upon his argument in Andrew Koppelman, Bostock, LGBT Discrimination, and the Subtractive Moves, 105 Minn. L. Rev. 1, 17–19 (2020).

      [44].    Koppelman, Justice Alito’s Desperation Move, supra note 43.

      [45].    McClain compellingly details the connection between the opposition to racial miscegenation and white supremacy (ch. 7). Anti-miscegenation laws were eventually declared unconstitutional because they were “obviously an endorsement of the doctrine of White Supremacy.” Loving v. Virginia, 388 U.S. 1, 7 (1967).

      [46].    Shanor in fact offers a powerful version of this argument. See supra note 42.

      [47].    See, e.g., United States v. Windsor, 570 U.S. 744, 772 (2013) (the federal Defense of Marriage Act “tells [same-sex] couples, and all the world, that their otherwise valid marriages are unworthy of federal recognition. This places same-sex couples in an unstable position of being in a second-tier marriage. The differentiation demeans the couple . . . .”).

      [48].    See, e.g., Gregory M. Herek, Beyond “Homophobia”: Thinking About Sexual Stigma and Prejudice in the Twenty-First Century, 1 Sexuality Res. & Soc. Pol’y 6 (2004) (defining the relevant terms).

      [49].    See, e.g., Caroline Medina et al., Improving the Lives and Rights of LGBTQ People in America: A Road Map for the Biden Administration, Center for Amer. Progress (Jan. 12, 2021, 9:42 AM),

      [50].    See Equality Act of 2021, H.R. 5, 117th Cong. (2021) (proposing an amendment to the Civil Rights Act of 1964 that would add sexual orientation and gender identity to the protected categories).

      [51].    Sophia Moreau, What Is Discrimination?, 38 Phil. & Pub. Aff. 143, 147 (2010).

      [52].    Bostock, 140 S. Ct. at 1741 (quoting Price Waterhouse v. Hopkins, 490 U. S. 228, 239 (1989) (plurality opinion)).

      [53].    Deborah Hellman, When Is Discrimination Wrong? 7–33 (2008). See also Elizabeth S. Anderson & Richard H. Pildes, Expressive Theories of Law: A General Restatement, 148 U. Pa. L. Rev. 1503 (2000).

      [54].    “On my account, then, although discriminatory actions can certainly express demeaning messages, they only do this because they are restrictions on deliberative freedoms to which a person is entitled. Their demeaning message results from the way in which they wrong individuals. It does not constitute that wrong; rather, it is a side effect of the wrong.” Moreau, supra note 51, at 177–78.

      [55].    Cf. Jason Mazzone, Bostock: Were the Liberal Justices Namudnoed?, Balkinization (July 6, 2020), (predicting that Bostock will lead to prohibitions on affirmative action).

      [56].    Justice Gorsuch’s arguments against the group-based nature of discrimination turn on his understanding the group-based injury to be something like an injury in the aggregate. See Bostock, 140 S. Ct. at 1741. Thus he worries that, on a group-based understanding, a woman who is the victim of sexual harassment will have no complaint if her employer treats most women well. Id. In point of fact, however, there is no necessary conceptual connection between construing discrimination as group-based and caring only about how the group fares in the aggregate, or “overall.” Id.

      [57].    Id. at 1739.

      [58].    Id.

      [59].    Id. at 1742.

      [60].    See id. at 1741 (offering an example where two people are “materially identical” in all respects except the one triggering the discrimination).

      [61].    See, e.g., Suzanne B. Goldberg, Discrimination by Comparison, 120 Yale L.J. 728 (2011).

      [62].    McClain suggests as much (pp. 215–216). Cf. Aziz Huq, Bostock v. BLM, Boston Rev., July 15, 2020, (contrasting two views of equality—one (exemplified by Bostock) is transaction-focused and individual, and the other (exemplified by the BLM movement) is structural and group-based).

      [63].    See, e.g., Federal Housing and Urban Development Department, Final Rule, HUD’s Implementation of the Fair Housing Act’s Disparate Impact Standard, 85 Fed. Reg. 60288 (Sep. 24, 2020) (to be codified at 24 C.F.R. pt. 100); Kleber v. CareFusion Corp., 140 S. Ct. 306 (2019) (denying cert petition for an appeal from a 7th Circuit decision holding that outside employees may not sue on a disparate impact claim under the ADEA). In this vein, it may be worth noting that even Palmore v. Sidoti, a case once heralded for its anti-racist reach (see infra text accompanying note 68), is now used to support an anti-progressive race agenda that forbids taking account of race for any purpose. See, e.g., Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 277 (1986) (prohibiting state from pursuing affirmative action in its hiring of public school teachers); Grutter v. Bollinger, 539 U.S. 306, 342 (2003) (quoting the same passage in expressing concerns about the constitutionality of affirmative action for law school admissions). This has also been the fate of Shelley v. Kraemer, another high-water mark of anti-racist jurisprudence (see infra note 67 and text accompanying note 69). Shelley has been used not to prohibit the state from promoting discrimination but instead to to thwart affirmative action, see, e.g., Reva B. Siegel, Foreword: Equality Divided, 127 Harv. L. Rev. 1, 30 (2013), or to underscore that private acts are private, and so no business of the state, see, e.g., Catharine A. MacKinnon, Disputing Male Sovereignty: On United States v. Morrison, 114 Harv. L. Rev. 135, 171 (2000).

      [64].    See, e.g., Civil Rights Act of 1964, Pub.L. 88–352, 78 Stat. 241.

      [65].    See, e.g., Virginia v. Rives, 100 U.S. 313, 318 (1879); Shelley v. Kraemer, 334 U.S. 1, 13 (1948); Lugar v. Edmondson Oil Co., 457 U.S. 922, 936 (1982) (“Careful adherence to the ‘state action’ requirement preserves an area of individual freedom by limiting the reach of federal law and federal judicial power”); Burton v. Wilmington Parking Auth., 365 U.S. 715, 722 (1961) (“private conduct abridging individual rights does no violence to the Equal Protection Clause unless to some significant extent the State in any of its manifestations has been found to have become involved in it.”).

      [66].    See, e.g., Nestor M. Davidson, Judicial Takings and State Action: Rereading Shelley After Stop the Beach Renourishment, 6 Duke J. Const. L. & Pub. Pol’y 75, 77 (2011) (“a high-water mark for the realist conception of property came in the famous 1948 decision Shelley v. Kraemer”).

      [67].    “The Constitution confers upon no individual the right to demand action by the State which results in the denial of equal protection of the laws to other individuals.” Shelley, 334 U.S. at 22.

      [68].    “The Constitution cannot control [racial] prejudices but neither can it tolerate them. Private biases may be outside the reach of the law, but the law cannot, directly or indirectly, give them effect.” Palmore v. Sidoti, 466 U.S. 429, 433 (1984). The Court issued a similar declaration in City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 435 (1985) (finding unconstitutional the refusal to grant a zoning permit for a residence that would house “mentally retarded” individuals, stating: “[T]he City may not avoid the strictures of th[e Equal Protection] Clause by deferring to the wishes or objections of some fraction of the body politic.”).

      [69].    See Pollvogt, supra note 12, at 906–08 (arguing that the Court’s commitment to abolishing animus led it to overturn the custody arrangement in Palmore, even though that arrangement would have passed strict scrutiny).

      [70].    See, e.g., Ronald J. Krotoszynski, Jr., Agora, Dignity, and Discrimination: On the Constitutional Shortcomings of “Conscience” Laws That Promote Inequality in the Public Marketplace, 20 Lewis & Clark L. Rev. 1221, 1227 (2017); cf. Jonathan Kahn, The 911 Covenant: Policing Black Bodies in White Spaces and the Limits of Implicit Bias As A Tool of Racial Justice, 15 Stan. J. Civ. Rts. & Civ. Liberties 1, 40 (2019) (extending the idea, from Shelley and Palmore, that the state may not support private discrimination to the context of private businesses seeking to eject Black customers on dubious claims of loitering); Russell K. Robinson & David M. Frost, ”Playing It Safe” with Empirical Evidence: Selective Use of Social Science in Supreme Court Cases About Racial Justice and Marriage Equality, 112 Nw. U. L. Rev. 1565, 1582 (2018) (suggesting that proponents of same-sex marriage ought to have relied more heavily on the Palmore logic).

Elsewhere, McClain herself loosely connects Palmore and the wedding vendor cases. Linda C. McClain, Prejudice, Constitutional Moral Progress, and Being “On the Right Side of History”: Reflections on Loving v. Virginia at Fifty, 86 Fordham L. Rev. 2701, 2714 (2018). It is also worth noting that state ratification has been invoked as a ground for refusing exemptions from the ACA’s contraceptive mandate. See, e.g., Brief of Amicus Curiae Catholics for Choice, Feb. 16, 2016, at *22 (“Petitioners may believe that women who choose contraception are sinners; they are not entitled to government ratification and enforcement of their belief.”).

      [71].    Amy J. Sepinwall, Conscience in Commerce: Conceptualizing Discrimination in Public Accommodations, 53 Conn. L. Rev. 1 (2021).

      [72].    See, e.g., Adam Liptak, Cake Is His ‘Art.’ So Can He Deny One to a Gay Couple?, N.Y. Times (Sept. 16, 2017), (“The government, Mr. Phillips contends, should not be allowed to compel him to endorse a message at odds with his beliefs.”).

      [73].    Brief for Petitioners at 17, Masterpiece Cakeshop, Ltd. v. Colo. Civil Rights Comm’n, 138 S. Ct. 1719 (2018) (No. 16-111),–ts.pdf. Endorsement also grounded Kim Davis’s refusal to issue marriage licenses to same-sex couples. Davis was the county clerk who “opposed same-sex marriage on religious grounds” and so worried that “her name’s appearance
[on the marriage license] was the equivalent of her personal endorsement.” Petition
for Writ of Certiorari, at 99a, Davis v. Ermold, 936 F.3d 439 (6th Cir. 2019) (No. 19-926), pdf.

      [74].    For instances of this argument in the contraceptive mandate cases, see generally Amy J. Sepinwall, Conscience and Complicity: Assessing Pleas for Religious Exemptions in Hobby Lobby’s Wake, 82 U. Chi. L. Rev. 1897, 1938–44 (2015); in the wedding vendor context, see, e.g., Michael C. Dorf, The Troublingly Widening Gyre of Complicity Claims, Verdict (Nov. 1, 2017), Sometimes the concerns are voiced as objections not to generic complicity claims but instead to the more specific form complicity takes when expressive interests are present. See, e.g., John Corvino,”Bake Me a Cake”: Three Paths for Balancing Liberty and Equality, What’s Wrong? (Oct. 15, 2015), For an example of this line of argument in the public services context, see the First Amendment Scholars’ amicus brief in Fulton v. Philadelphia, when they suggest that it is “outlandish” for Catholic Social Services to worry that Philadelphia would be compelling it to speak and act according to Philadelphia’s beliefs were it to abide by Philadelphia’s non-discrimination law. Brief of First Amendment Scholars As Amici Curiae in Support of Respondents at 26, Fulton v. City of Philadelphia, 141 S. Ct. 1868 (2021) (No. 19-123).

      [75].    The arguments in this paragraph draw upon Sepinwall, Conscience and Complicity, supra note 74.

      [76].    See, e.g., Steven J. Heyman, A Struggle for Recognition: The Controversy over Religious Liberty, Civil Rights, and Same-Sex Marriage, 14 First Amend. L. Rev. 1, 88 (2015) (“[A]n enterprise that offers to serve the public becomes part of the social realm of commerce. Such an enterprise properly can be regarded as a place of public accommodation with a duty to serve everyone.”); Joseph William Singer, No Right to Exclude: Public Accommodations and Private Property, 90 Nw. U. L. Rev. 1283, 1410 (1996).

      [77].    See Brief for City Respondents at 2, Fulton v. City of Philadelphia, 141 S. Ct. 1868 (2001) (No. 19-123).

      [78].    Elane Photography, LLC v. Willock, 309 P.3d 53 (N.M. 2013).

      [79].    Elane Photography, 309 P.3d at 79–80 (Bosson, J., concurring).

      [80].    See, e.g., sources cited supra note 76.

      [81].    Elane Photography, 309 P.3d at 80 (Bosson, J., concurring).

      [82].    One might think current public accommodations law already permits these refusals, as the party turned away is not a member of a protected class. But there is reason to think otherwise. For one thing, some white supremacist groups are religious—their supremacy is taken to be religiously mandated. See infra note 100 and accompanying text. Further, even where the party who would be refused service is not a member of a protected class, that fact is of no avail to the theorist under consideration here. After all, her claim, like Justice Bosson’s, is that conscience has no place at all in the retail sphere; it is not that conscience has no place in the retail sphere when its exercise would negatively affect members of protected classes.

      [83].    Sepinwall, Conscience in Commerce, supra note 71.

      [84].    In her dissent in Hobby Lobby, Justice Ginsburg railed against the majority for failing to attend to third-party harms. Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 740 (2014) (Ginsburg, J., dissenting). Some commentators ground the claim that the state may not accommodate a religious adherent at the expense of third parties in the Establishment Clause. See, e.g., Frederick Mark Gedicks & Rebecca G. Van Tassell, RFRA Exemptions from the Contraception Mandate: An Unconstitutional Accommodation of Religion, 49 Harv. C.R.-C.L. L. Rev. 343, 356–61 (2014); Micah Schwartzman & Nelson Tebbe, Obamacare and Religion and Arguing Off the Wall, Slate (Nov. 26, 2013, 2:32 PM),

      [85].    Sepinwall, Conscience and Complicity, supra note 74, at 1961–66.

      [86].    The prospect that, e.g., a gay couple could find others to serve them figures prominently in libertarian defenses of a robust accommodation regime. See, e.g., Richard A. Epstein, Public Accommodations Under the Civil Rights Act of 1964: Why Freedom of Association Counts as a Human Right, 66 Stan. L. Rev. 1241, 1290 (2014) (arguing that business owners should be permitted to discriminate so long as would-be customers can be served elsewhere).

      [87].    The argument could instead be run the other way: Opponents of exemptions in the wedding vendor cases deny that a couple’s decision to marry redounds to the wedding vendor. But this is just the rationale for denying that school voucher programs involve the state in discrimination. See, e.g., Zelman v. Simmons-Harris, 536 U.S. 639, 652 (2002) (“[The school voucher program] permits government aid to reach religious institutions only by way of the deliberate choices of numerous individual recipients.”).

      [88].    Palmore, 466 U.S. at 433.

      [89].    Shelley, 334 U.S. at 13.

      [90].    This line of argument is strongly influenced by Nico Cornell’s assessment of Shelley v. Kraemer:

While one might say that racially restrictive covenants cannot be enforced by the government, this seems to dodge the important issue. The more illuminating explanation for the unenforceability of racially restrictive covenants—which Shelley fails to offer—is that the racist who demands enforcement actually has no legitimate complaint that she was wronged. The Shelley strategy implies that the landowner does commit a wrong by selling to a black family—just a wrong with no remedy. I think that, morally speaking, this cannot be correct. We should say that the racist has no moral complaint, not that she has a complaint that the government is unable to recognize because it is bound by other commitments.

Nicolas Cornell, A Complainant-Oriented Approach to Unconscionability and Contract Law, 164 U. Pa. L. Rev. 1131, 1155 (2016) (footnote omitted).

On Cornell’s account, Kraemer could not have been wronged when Shelley bought the house because the covenant prohibited conduct that no one was morally permitted to prohibit. In that way, it functioned in just the way efforts to hold seats function at events where there is open seating. Cornell focuses, then, on how we should understand what is morally at stake as between the landowners. For that reason, he does not consider how we should judge the state that treats Kraemer as if he has an actionable complaint. In what follows, I extend Cornell’s analysis: once we appreciate that Kraemer sustains no wronging when Shelley buys the house, would the state that nonetheless finds in favor of Kraemer merely ratify Kraemer’s racism, or would it engage in racism itself? I aim to establish that the latter more faithfully captures the nature of the state’s contribution.

      [91].    But cf. Leslie Kendrick & Micah Schwartzman, The Etiquette of Animus, 132 Harv. L. Rev. 133, 144–45 (2018) (“A private homeowner may eject entrants on the basis of offense, even animus. In enforcing trespass law, the state does not thereby acquire the motives of the homeowner. Similarly, commercial public accommodations can refuse service on bases not prohibited by law. In enforcing their ability to do so, the state does not acquire their motives.”) I do not disagree. But it makes a significant difference whether the exclusion is permissible or not. Where the private or commercial party is not permitted to exclude, I believe the state that supports the exclusion does participate.

I do not seek to determine whether participation requires adopting the motives of the party with which one participates. I will say for now that I am doubtful that it does. (The doctor who participates in a pregnancy termination might be indifferent as to whether the pregnancy is carried to term, and she certainly need not have taken on board the motives for the woman’s decision to terminate.)

      [92].    365 U.S. 715 (1961). While, in what follows, I focus on the form state participation took as articulated in Justice Stewart’s concurring opinion, I note that the majority opinion understands the state’s action to have taken the form of participation, too. The majority focuses on the interdependence between the state and the restaurant—the fact that the state and restaurant were involved in a “joint venture,” Reitman v. Mulkey, 387 U.S. 369, 393 (1967)—and so it comes to the notion of participation directly. See Burton, 365 U.S. at 724.

      [93].    Burton, 365 U.S. at 726 (Stewart, J., concurring) (quoting 24 Del. Code §1501).

      [94].    Id. at 726.

      [95].    Id. at 726–27.

      [96].    Id. at 716–17 (quoting from the decision of the Delaware Supreme Court, 157 A.2d 902).

      [97].    See also Mulkey v. Reitman, 64 Cal. 2d 529, 543 (1966), aff’d, 387 U.S. 369 (1967) (overturning a California constitutional amendment that recognized property owners’ rights to refuse to sell or rent on any basis whatsoever, on the ground that the “legislative action ‘which authorized private discrimination’ [ ] made the State ‘at least a partner in the instant act of discrimination.’” (italics added)); cf. Bell v. Maryland 378 U.S. 226, 329 (1964) (Black, J., dissenting) (contending that racially restrictive covenants “constituted a restraint on alienation of property, sometimes in perpetuity, which, if valid, was in reality the equivalent of and had the effect of state and municipal zoning laws, accomplishing the same kind of racial discrimination as if the State had passed a statute instead of leaving this objective to be accomplished by a system of private contracts, enforced by the State.”).

      [98].      Newman v. Piggie Park Enter., Inc., 390 U.S. 400, 403 n.5 (1968).

      [99].    State v. Arlene’s Flowers, 441 P.3d 1203, 1235 (Wash. 2019).

    [100].    Christian Identity, S. Poverty L. Ctr., (last visited Aug. 11, 2020).

    [101].      See, e.g., Christian Legal Soc’y v. Martinez, 561 U.S. 661, 695 (2010); Lawrence v. Texas, 539 U.S. 558, 575 (2003) (“When homosexual conduct is made criminal by the law of the State, that declaration in and of itself is an invitation to subject homosexual persons to discrimination . . . .”). Cf. Bray v. Alexandria Women’s Health Clinic, 506 U.S. 263, 270 (1993) (“A tax on wearing yarmulkes is a tax on Jews.”).

    [102].      See Sepinwall, Conscience in Commerce, supra note 71.

    [103].      On double binds as they affect women, the locus classicus is Frye, supra note 20. On the motherhood wage gap, see, e.g., Courtney Connley, U.S. Moms Working Full-Time Are Paid $0.75 for Every Dollar Paid to Fathers, Leading to Devastating Economic Losses, Make It (May 5, 2021, 3:43 PM),