The Regulation of Discrimination by Individuals in the Market

I.     Introduction

In 2014, Airbnb came under fire when a Harvard Business School study showed that property owners were less likely to accept those with black-sounding names as renters and that non-black hosts were able to charge “approximately 12% more than black hosts, holding location, rental characteristics, and quality constant.”1 In the wake of public outcry, the company hired former Attorney General Eric Holder to help it devise a new antidiscrimination policy.2

In an October 2016 working paper conducted by the National Bureau of Economic Research, researchers looking at taxi-services Uber and Lyft showed that the cancellation rate for passengers with black-sounding names was more than twice as high as for those with white-sounding names.3 In response to the study, Senator Al Franken wrote to both Uber’s and Lyft’s CEOs to ask why it was necessary to include passenger names and photos, and what Uber and Lyft could do to end discrimination and better enforce their self-imposed antidiscrimination policies.4 While both CEOs defended the use of names and photos as necessary to create a “digital trust profile,” they agreed to experiment with other ways to prevent discrimination. Franken, however, “remain[s] concerned” that the current systems “do not sufficiently guard against discriminatory conduct.”5

In both of these examples, we see a public quick to condemn and demand fixes for discrimination perpetuated by individuals against other individuals, though facilitated and arguably encouraged by the companies synonymous with the on-demand economy. We also see companies looking to mollify public outcry outside of litigation, with varying degrees of success.6 As I’ve written about in the labor context, public pressure on high-profile and reputation-sensitive companies has long been used to garner gains for workers, even when those workers are not the employees of the targeted company.7 And as we’ve also seen recently, the public is ready and willing to boycott companies to protest companies’ policies and their leaders’ political affiliations.8 But eventually the question arises: what does the law say about these objected-to forms of discrimination? It is well and good if Airbnb and Uber voluntarily create and meaningfully enforce antidiscrimination policies, but are the types of discrimination the public objects to—discrimination engaged in by one individual against another within the commercial sphere—illegal, for the individuals discriminating or for the companies facilitating it? And if our current antidiscrimination laws fail to cover these objected-to forms of discrimination, should they? Can they?

There are at least two ways the law might be used to address the types of individual-on-individual discrimination seen in the on-demand economy: such discrimination may be directly prohibited, as discrimination by an employer against an employee is, or such discrimination may be limited indirectly, through the enactment of laws that prohibit companies from facilitating it. Indirect means can shape consumers’ decision architecture along a spectrum, from providing a nudge toward non-discrimination to making the ability to discriminate impossible altogether. The way antidiscrimination laws work in the home-seeking context provides one example of the nudging end of the spectrum. While today laws are not used to directly prohibit individuals from engaging in race discrimination when choosing where to live, laws directed at landlords, realtors, and developers prohibit these third-parties from facilitating home-seeker discrimination. This in turn raises the cost to home-seekers who try to discriminate.9 Along these lines, in an effort to eliminate discrimination against individuals with black-sounding names, we can imagine laws that prohibit companies like Airbnb, Uber, and eBay from showing users’ real names. Individual buyers and sellers may nevertheless find other ways to engage in discrimination—consciously or not—but with the higher costs of doing so, such laws should result in at least somewhat less discrimination.

There are likely pragmatic reasons to favor an indirect strategy of limiting individual-on-individual discrimination within the market. Companies are bigger, have deeper pockets, and some can be sensitive to reputational harms, which may combine to make enforcement easier.10 In addition, targeting individuals directly may result in forms of backlash or resentment that could result in more discrimination and racial animus—negative effects that might be avoided though a more indirect approach. These are all legitimate concerns that, on their own, may lead us to favor indirect action. But there is a qualitatively different kind of concern that scholars and courts have intimated about direct prohibitions on discrimination by individuals—namely, that individuals have a constitutionally protected right to discriminate generally, which covers when they act in the market as consumers or sellers. For those who posit such a right, the direct regulation of individual discrimination seems off the table, or at minimum constitutionally problematic.

But interestingly, at least some people who have this view think that the indirect regulation of individual discrimination avoids these issues. For instance, in their Article Discrimination by Customers, Katharine Bartlett and Mitu Gulati take the position that individuals qua consumers have autonomy and privacy rights that ought to block the enactment of direct prohibitions on their discrimination.11 However, they think additional indirect restrictions on customer discrimination, via additional restrictions on the companies that facilitate it, are acceptable.12 Similarly, Tarunabh Khaitan argues that antidiscrimination duties are not imposed on employees, tenants, and consumers because they have strong claims to negative liberty whereas employers, landlords, and sellers do not.13 But while Khaitan concludes that individuals have negative liberty claims that make direct prohibitions on their discrimination problematic, he does not see antidiscrimination duties on sellers as also implicating those same individuals’ interests. I have begun to worry about this position.

If individuals have a right to discriminate within the market, be they buyers or sellers, there is at least a prima facie case to be made that third-parties (like employers and sellers) will be able to challenge antidiscrimination laws targeted at them by invoking those individuals’ rights. The Ninth Circuit’s ruling in Fair Housing Council v. Roommates.com14 highlights the problem of conjoining a consumer right to discriminate with prohibitions on third-parties who facilitate or encourage that discriminatory conduct.15 There, Roommates.com challenged the constitutionality of a law that would prohibit it from facilitating its users’ discrimination in the selection of housemates on the grounds that such a law would constitute a burden on the constitutional right of its users to engage in that discrimination.16 The Ninth Circuit, invoking the doctrine of constitutional avoidance, appears to have agreed with Roommates.com’s argument.17 Fennell also recognizes this issue within fair housing law. As she explains, some criticize prohibitions on discriminatory collateral search behavior (the behavior of third-parties who might assist home-seekers in engaging in discrimination when finding and securing a property) on the grounds that such collateral behavior should be legal so long as the underlying discriminatory conduct of the home-seeker is.18 In other words, within fair housing law, some argue that if the underlying discrimination of home-seekers is legal, it should be just as legal for third-parties to help them engage in it.19 As a result, Fennell sees arguments that 42 U.S.C. § 1982 prohibits home-seekers from discriminating—regardless of whether § 1982 is enforced—as buttressing the direct liability for collateral search behavior created through the Fair Housing Act.20

The problem of indirect restrictions on exercises of rights is not limited to antidiscrimination law. Consider the speech context. In Brown v. Entertainment Merchants Ass’n,21 the Court struck down on First Amendment grounds a law that prohibited the sale of violent video games to minors absent parental consent.22 The Court found the law, which directly regulated only the activity of sellers, to violate the free speech rights of the buyers (i.e., the children). Scholars have also identified related First Amendment issues when state actors look to suppress individual speech by enlisting private actors like Facebook, Google, and ISPs.23 There are many other cases where the Court has struck down restrictions on actor A on the grounds that those restrictions infringed on the constitutional rights of actor B.24 In short, if individuals’ ability to discriminate on the basis of race and sex while engaged in market activity is constitutionally protected, then the indirect laws favored by Bartlett and Gulati, and the explanation for why firms can be antidiscrimination duty-bearers while individuals cannot provided by Khaitan, look problematic. Such a right also renders vulnerable some of the antidiscrimination laws we have today.

Given the above and the themes of this Symposium, my interest is in whether individuals qua market actors do or ought to have a right to discriminate on the basis of race or gender. As I hope the above shows, those concerned about the future of antidiscrimination law within the on-demand economy and beyond need to get clear on this. To that end, my goal in this Essay is modest. I simply want to lay down groundwork for thinking through whether, normatively or descriptively, there is an individual right to discriminate in the market. To that end, in Part I, I sketch some high-level considerations that those who want to assert or deny that individuals ought to have a right to discriminate in the market need to think through. In Part II, I look at how the law currently treats discrimination by individuals within the market and how that treatment speaks whether there is an individual right to engage in that conduct today.

II.    Theoretical Considerations

Here I focus on two issues surrounding the legitimacy of government regulation of individual discrimination: (1) what makes the targeted forms of discrimination legitimate targets and (2) the significance of the market in that determination.

A.    Government Legitimacy: Harm or Injustice

Liberalism is committed to there being principled limits to legitimate government authority. At the broadest level, liberals rely on either concepts of harm or of injustice to mark those limits.

Under what is known as the Millian Harm Principle, “the only purpose for which power can be rightfully exercised over any member of a civilized society, against his will, is to prevent harm to others.”25 Once the Harm Principle is triggered through the infliction of a third-party harm, legal regulations aimed at curbing that harmful act are legitimate and the regulated-actor has no liberty-based objection to those regulations as such.26

John Rawls is associated with the alternative way of limiting the state’s legitimate power. On this view, the state can legitimately act to ensure justice, including both corrective and distributive forms.27 Because corrective justice looks a lot like remedies for harms, a simplified version might just say that liberals differ according to whether they think (1) the state can legitimately act to not only deal with harms but also distributive injustices and (2) if the answer to (1) is yes, whether, and if so how, the methods the government can use to respond to harms (i.e. the targets of corrective justice) and injustices (i.e. the targets of distributional justice) differ.28

The above matters for whether the state can prohibit individuals from discriminating against each other, in the market and more generally. If we assume Bartlett, Gulati, and Khaitan are operating within a Millian framework when they say that prohibiting forms of consumer discrimination would violate those consumers’ negative liberty interests, they have committed themselves to the position that consumer discrimination does not constitute a harm and thus government prohibitions on that discrimination would be illegitimate.29 If one wants to resist this conclusion, one needs to (1) show that the relevant forms of discrimination do constitute harms or (2) reject the Millian framework, embrace the Rawlsian alternative, and then show that the relevant forms of discrimination result in distributional injustice, and thus the state can legitimately act to prevent or cure them.

At the same time and so long as we are committed to liberal theory, the account of why the relevant forms of individual discrimination can be legitimately regulated (be it on a Millian or Rawlsian framework) cannot be one that decimates the principled limits of legitimate government authority.30 As I said at the outset, liberalism is committed to there being principled limits to legitimate government authority. Given this, it would likely be problematic to take the position that individual discrimination is harmful (and thus legitimately proscribed) because we harm each other whenever we fail to treat each other “equally.” In a liberal society individuals are permitted to act on their personal preferences and pursue their own projects in ways that will mean privileging some over others. If the state can require us to treat each other equally in all of our dealings, it looks like there is no space left at all for individual decision-making free of state control (i.e. we are no longer within a liberal state).31 But this is all by way of warning—the literature on what makes various types of discrimination wrong is vast and an account can be given as to why the regulation of certain forms of discrimination by individuals, at least within certain contexts, is legitimate (i.e. the discrimination is either harmful or results in distributional injustice).32 The point here is twofold: such an account must be given for each kind of discrimination we want to target and we must be careful to not give an account that does away with our background commitment to a limited state.

B.    The Significance of the Market for Government Legitimacy33

This Symposium is focused on changes to labor and work and so I have focused on whether individuals acting in the market have a right to discriminate. If they do, regulating the sorts of discrimination seen in the on-demand economy will be problematic and an increasing share of American workers may fall outside of the purview of our antidiscrimination laws. But while there are pragmatic reasons I have focused on discrimination within the market, there is a separate question of whether the market/non-market distinction is theoretically significant. One might think the market/non-market distinction tracks something significant with regards to the legitimacy of antidiscrimination law. Then again, one might think the opposite—that there is nothing special about the market and accordingly there is nothing special about discrimination that occurs there. In thinking through whether the state can legitimately prohibit discrimination by individuals within the market (my narrow focus in this Essay) we may need to know whether the fact that the discrimination is occurring within the market matters for purposes of that analysis.

One inroad into the significance of the market/non-market distinction requires a bit of a historical detour. As discussed above, liberals are committed to there being limits to governmental authority. In the United States, we have come up with different frameworks with different limiting principles to mark these limits. During Reconstruction, both those opposed to and in favor of various antidiscrimination laws agreed to a tripartite theory of rights framework.34 There were civil rights, political rights, and social rights.35 Everybody basically agreed that while the government could legitimately act to ensure civil and political forms of equality, it was forbidden—that is, it would be illegitimate to—protect social equality. With this framework in place, the debate surrounding particular antidiscrimination laws centered on whether the laws were addressing civil, political, or social issues.36 Those who objected to certain laws—like those prohibiting race discrimination in public accommodations—argued that such laws were trying to illegitimately “enforce social . . . equality.”37 Within this tripartite framework, the market as such was not a special category. Some market/economic rights were understood to fall squarely within the civil domain, like the capacity to hold property and enter contracts, while others that also dealt with commercial activity, like deciding with whom to associate when doing business, were thought by some to be social and thus beyond the government’s legitimate reach.38 When the Court finally did uphold extensions of antidiscrimination laws to housing and private contracting,39 such cases were understood as “the triumph of civil rights.”40 That is, a triumph of getting the regulated activity to be understood as implicating civil and not social equality, and thus becoming legitimate targets of regulation.

Over time the tripartite theory fell away but the underlying and substantive tensions surrounding the legitimacy of government regulation of private (i.e. non-government) discrimination continued.41 In the language of modern liberal theory we can put the concern like this: how can we have a state with limited powers and a space free from government intervention while also increasingly regulating the choices non-government actors make in deciding with whom to associate and on what basis? With that same concern remaining, alternative frameworks came about for trying to mark the limits of legitimate government authority. More in use today is the public/private distinction, which is as contested and malleable as the tripartite theory before it.42 Another closely-related distinction, and the one I focus on here, draws a circle around economic activity and finds it a legitimate target of regulation. We can call this a market/non-market distinction. While I cannot construct an entire history here, I see the idea that things taking place within the commercial domain can be treated differently for purpose of state regulation in, e.g., the historical treatment of commercial speech,43 fair use analysis in copyright,44 continued prohibitions on prostitution after Lawrence,45 the repudiation of Lochner,46 the repudiation of the constitutional right to freedom of contract/economic substantive due process Lochner contained,47 and the resulting expansion of the administrative state with the New Deal.

We have also seen the market/non-market distinction in antidiscrimination law. In Roberts v. United States Jaycees,48 Justice O’Connor found “only minimal constitutional protection of the freedom of commercial associations” as “[t]he Constitution does not guarantee a right to choose employees, customers, suppliers, or those with whom one engages in simple commercial transactions, without restraint from the State.”49 As Samuel Bagenstos explains, Justice O’Connor’s position in Jaycees has been used to create an expressive-commercial distinction.50 This distinction, in turn, has been used to limit the Court’s ruling in Boy Scouts of America v. Dale,51 where it held that an organization’s First Amendment rights could trump the application of antidiscrimination laws.52

However, like the tripartite theory of rights, the public/private distinction, efforts to cabin the harm principle, and construction of a negligence standard, the market/non-market distinction is an attempt to make sense of the limits of legitimate government authority in our lives and come to some kind of agreement about what we owe each other qua members of a community. These are not, I suspect, metaphysical kinds. As a result, the commercial-expressive and market/non-market distinctions, with government interference with activities categorized in the former thought legitimate in ways it would not be for activities in the latter, is susceptible to critique.53

The Court’s recent treatment of the commercial-expressive and market/non-market distinctions suggests its relationship to the distinctions is increasingly ambivalent. Lawrence is an example of the Court embracing the distinction, where it pointed out that the case involved sex that took place outside the market.54 It is on this basis that lower courts have not found Lawrence to entail a constitutional right to prostitution.55 In contrast, within the First Amendment domain the mar-ket/non-market distinction has increasingly eroded.56 Justice O’Connor’s reliance on a market/non-market distinction in Jaycees and attempts to cabin Dale aside, the Court’s increasing skepticism of the market/non-market distinction may soon find its way into antidiscrimination laws within the market in Masterpiece Cakeshop, Ltd. v. Colorado Civil Rights Commission.57

But putting aside the Court’s actual treatment of the market/non-market framework, one might still think it the best heuristic available for making sense of the limits of government power.58 The point I want to make is this: to think through whether the government can legitimately prohibit individuals from discriminating in the market, one needs to decide whether the “in the market” qualification matters. One might embrace the market/non-market distinction, tenuous as it is. Or one might reject it. The tripartite theory, too, has faded away. But at bottom I suspect we need some principled account of the legitimate limits of government regulation and how antidiscrimination laws square with them.59

III.  A Consumer Right to Discriminate under Current Law

So far I have discussed two larger legitimacy concerns regarding the regulation of individual discrimination in the market. Below I engage the law on the ground. Specifically, I look at two statutes—Section 1981 (briefly) and Title VII (more extensively)—to see how the law currently treats race and sex discrimination by individuals in the market.

A.    Section 1981

The history of § 1981 of the Civil Rights Act of 1866 is long and I will not recount it all here. In broad strokes, one of the main abuses the framers of the Civil Rights Act of 1866 were aware of was not the abuse of large employers but of individual white land owners contracting with free blacks in ways that recreated the conditions of slavery.60 The public/private or private/state action distinctions we find significant today were not of concern.61 Instead, “a principle purpose of the Civil Rights Act of 1866” was to “grant to the Freedmen basic economic rights—to make and enforce contracts, to sue and be sued, and to purchase and lease property. These rights would enable them to act as autonomous, productive workers, who could hope to accumulate some material wealth.”62 While originally enacted to enforce the Thirteenth Amendment, there was uncertainty about whether § 1981 could reach private discrimination after it was reenacted under the Fourteenth Amendment.63 However, the later Runyon v. McCrary64 Court held that it did.65 In Runyon, the Court held that when a private school discriminated against blacks in admissions, it violated § 1981. The Court found that refusing to contract with a black person violated the black person’s right, under § 1981, to have the “same opportunity to enter into contracts as” is “extend[ed] to white offerees.”66 In concluding, the Court thought a school refusing to contract with blacks for educational services constitute a “classic violation of section 1981.”67 The Court came to this conclusion in the face of school and parental claims that they had a constitutional right to discriminate in their associations.68 Section 1981 became, after Runyon, “a powerful weapon against private racial discrimination in a wide variety of relationships which may be defined as contractual in nature.”69

Later, Congress passed the Civil Rights Act of 1991,70 which endorsed the Court’s interpretation of § 1981.71 As “it currently reads, § 1981 prohibits racial discrimination in all forms of contracting, no matter how minor or personal.”72 While the statute has been “unjustly neglected as a source of protection for civil rights,” it remains on the books and suggests that we have historically upheld prohibitions on individual discrimination within the market. Like Fennell’s reading of § 1982, § 1981’s prohibition on race discrimination in contracting buttresses the legitimacy of laws that seek to limit consumer discrimination by prohibiting third-parties from facilitating or selling it.73

B.    Title VII and the BFOQ Exception

Intentional discrimination in employment is illegal—usually.74 Title VII prohibits, among other things, employers from intentionally discriminating against employees on the basis of race and sex.75 But while that prohibition is absolute when it comes to disparate treatment on the basis of race, sex discrimination is permitted when sex qualifies as a bona fide occupational requirement (“BFOQ”) “reasonably necessary to the normal operation of that particular business or enterprise.”76 What counts as a BFOQ and how does it relate to an individual right to discriminate in the market? EEOC Guidelines state that “the refusal to hire an individual because of the preferences of . . . clients or customers” does not warrant the application of the BFOQ exception unless that discrimination is “necessary for the purpose of authenticity or genuineness.”77 This language suggests that consumers do not have a right to engage in race and sex discrimination, or at least that that right is not violated when Title VII prohibits employers from selling that discrimination. However, in a 2002 opinion letter the EEOC recognized that this categorical statement was not entirely accurate; sometimes courts have permitted employers to successfully argue for a BFOQ exception by referring to the discriminatory preferences of consumers that go beyond issues of authenticity or genuineness.78 But not all consumer discriminatory preferences are treated equally. So what does this mean for a recognized consumer right to discriminate?

Given that the BFOQ exception does not include race, consumer racial preferences never legitimate a company’s racially discriminatory policy.79 As it looks and as § 1981 already suggests, there is simply no right to buy or sell race discrimination. Putting § 1981 aside and focusing solely on Title VII, there is no consumer right to engage in race discrimination that blocks laws that prohibit third-parties from selling racial discrimination to them.80

Things are more complicated when it comes to consumer demand for sex or gender-based discrimination. Consumer sex-discrimination demands can be broken down into two categories: those concerning sexual titillation and those concerning privacy.81 Courts treat these demands differently. When employers argue that consumers want to engage in sex-discrimination for privacy-based reasons, courts are much more willing to find sex a legitimate BFOQ than in cases where employers argue consumers want to engage in sex-discrimination for purposes of sexual titillation. Indeed, while in the former case some courts have raised—though not fully explored or committed to there being—consumers’ constitutional privacy rights at stake, the latter are resolved without any reference to consumer rights at all. Let’s now turn to these two categories in more detail.

1.    Sex discrimination for sexual titillation

In sexual titillation cases, “employers argue that they must discriminate on the basis of sex in hiring to provide customers with the type of sexual arousal their businesses promise.”82 Courts’ treatment of these cases fall along a continuum: on one end are cases where employers are selling explicit sexual gratification (e.g., prostitution or lap dances) and particular types of bodies are thought necessary.83 Here scholars assume courts would permit sex-based hiring as a BFOQ.84 In the middle of the spectrum are cases involving the sale of sexual arousal through the exclusive sale of sexual gaze objects (e.g., strippers and centerfolds).85 Here as well, courts and commentators have assumed sex a BFOQ.86 On the other end of the spectrum are “plus-sex” businesses, which sell nonsexual goods bundled with sexual arousal through the provision of sexual gaze objects (e.g., flights plus sexy flight attendants, restaurants plus sexy servers). Here, courts almost invariably reject sex as a legitimate BFOQ, finding sex not “essential” to the business operation, with courts confidently deigning the essence of the business the non-sexy part.87

What do courts’ treatment of employer BFOQ arguments premised on consumer demand for sex-discriminatory sexual titillation tells us about a consumer right to discriminate? First, when considering these BFOQ arguments courts do not view the ability (or inability) of individuals to buy sexual arousal (the sale of which is assumed to require sex discrimination in hiring) to impinge on buyers’ constitutional rights. Take Wilson v. Southwest Airlines Co.,88 where Southwest’s BFOQ argument for hiring only attractive women flight attendants was rejected.89 Southwest presented market data showing that customers wanted to buy air travel bundled with sexual gratification.90 In rejecting the argument, the court never entertained the notion that consumers have a right to either buy sexual gratification specifically or engage in sex discrimination generally while in the market, such that employees’ Title VII antidiscrimination rights and the correlative antidiscrimination duties on employers had to be shaped in light of them.

That courts see no constitutional issue with limiting the ability of consumers to buy sex discrimination for sexual titillation purposes—be it bundled with other goods or not—is in one way not surprising. In the United States, there is no general right to buy sex. Prostitution, the most obvious form of this, is prohibited in most jurisdictions.91 Given this, one might take the position that if there’s no right to buy sex—if the state can ban the sale of literal sex—then the state is also free to ban as much or as little of the sale of in-person sexual gratification or titillation as it deems necessary, and consumers have no rights-based objection to it.92

A few thoughts about this. On the above view, the state’s right to block the sale of in-person sexual gratification must be stronger than any individual’s right to buy sex. The state’s right to block the sale of in-person sexual gratification also must be stronger than any individual’s right to buy sex discrimination. For both of these to be true, it seems necessary to subscribe to the market/non-market distinction and framework I discussed earlier.93 Here’s why: in striking down an anti-sodomy law, the Lawrence Court identified a substantive due process liberty interest in choosing one’s sexual partners.94 In order for bans on prostitution (and perhaps the sale of sexual titillation and gratification more generally) to be constitutional (or at minimum not subject to constitutional challenge), we have to think that the Lawrence liberty interest in choosing a sexual partner (a liberty interest in being free to discriminate on the basis of sex when having sex) is limited to the non-commercial sphere. Some reject the market/non-market framework, think there is nothing special about the market, and thus conclude that bans on prostitution are, post-Lawrence, unconstitutional. So far, this argument has been unsuccessful,95 though given the Court continues to erode the market/non-market distinction, 96 it is conceivable that in the future the framework will be wholly rejected and Lawrence will be read to provide a right to have consensual sex with whomever we want for money. And if that happens, it seems to entail, at minimum, a right to engage in sex discrimination when choosing one’s commercial sex partner (since that is what Lawrence was about). It is an open question whether this newly-extended right also includes a right to engage in race discrimination for the same.97 If it does, and erotic exceptionalism enters the market, it is possible that some will try to extend that right even further, to construct a general right to engage in sex and race discrimination within all commercial transactions.

In short, the treatment of sex titillation BFOQ cases suggests, at minimum, that individuals have no right to buy in-person sexual gratification. It does not show that consumers have no right to engage in sex discrimination for things they can buy. If Lawrence results in the future protection of prostitution, then we seem to get a protected right to engage in sex, and possibly race, discrimination while in the market. Once introduced, Lawrence becomes a deregulatory device for those looking to limit and roll back antidiscrimination laws in the market. This is, borrowing from Bagentos, the unrelenting libertarian challenge to antidiscrimination law.98

2.    Sex discrimination for privacy

In contrast to the purchase of sexual titillation, employers are more able to engage in sex-discrimination when they argue that consumers’ privacy interests would be offended or infringed if they did not.99 As Yuracko describes, these cases also fall on a BFOQ privacy continuum, measured by degrees of physical and visual contact.100 The strongest BFOQ claims involve jobs requiring either actual physical contact with or inspection of the consumer’s naked body.101 These claims are the most likely to succeed.102 In the middle of the spectrum are jobs where employees see but do not touch consumers’ bodies in various stages of undress.103 In general, courts find customer interests in not being seen naked by members of the opposite sex sufficiently compelling as to justify sex-discriminatory employment practices.104 The weakest privacy cases do not necessarily involve employees seeing or touching consumers’ naked bodies but are thought to merely involve consumer or co-employee embarrassment or discomfort when performed by members of the opposite sex.105 These claims are only sometimes successful.106

In most privacy-based cases where courts find a legitimate BFOQ for sex discrimination, courts do not explain it by reference to a consumer constitutional right to privacy that would be infringed if the firm’s BFOQ argument were denied. However, that is not always the case. In Fesel v. Masonic Home of Delaware, Inc.,107 a court did tie the privacy-based BFOQ exception for sex discrimination to a consumer privacy right.108 Below, I briefly lay out Fesel and then turn to the complications around the consumer privacy-based right to engage in discrimination that it finds.

a.    Fesel

In Fesel, a retirement home refused to hire a male nurse’s aide because it claimed its women residents would not consent to having their “personal needs attended to by a male” and would in fact leave the facility if there were male nurse’s aides.109 As the court rightly saw, the women refused to be aided by male nurse’s aides “undoubtedly” as a result of sex stereotyping.110 This refusal was a form of “customer preference,” which on its own could not justify a job qualification based on sex.111 However, in upholding the home’s BFOQ defense, the court understood this particular discriminatory preference as not just a discriminatory preference but a privacy interest protected by law.112

As for the specific activities that implicated the relevant privacy interest, the court pointed to the intimacy of the personal care, which included bathing, toilet assistance, geriatric pad changes, and catheter care.113 The court concluded that each of these functions “involves a personal touching as to which each guest is privileged by law to discriminate on any basis.”114 The court located the source of this legal privilege in tort and criminal law, though never pointed to any specific law. Instead, the court asserted generally that “[b]ecause our tort and criminal laws recognize these personal privacy interests, the Home cannot legally force its female guests to accept personal care from males.”115

b.    Complications around a consumer privacy-based right to engage in discrimination

I take the Fesel court’s thinking to be something like this: as both tort and criminal law show, individuals get to choose who touches them and they get to make that decision based on whatever they like. That is, when someone touches you without your consent, it does not matter (for purposes of criminal and tort law) if you refused consent because that person was a man, black, or simply wore shoes you found unpleasant—each person is “privileged by law to discriminate on any basis.”116 And from this the court concluded that since “the Home cannot legally force its female guests to accept personal care from males” without violating those guests’ touching-related rights, the state cannot prohibit the home from hiring only female nurses.117 Thus, a BFOQ exception is required.

While perhaps intuitively appealing on first pass, there are problems with the Fesel court’s analysis. While it is true that the law of battery protects patients from unwanted physical contact, which a patient can demonstrate in the medical domain by showing (1) that she did not consent to either the touching that occurred or (2) that the touching was done by an unwanted medical provider, antidiscrimination laws do not raise this problem; the nurse’s aide—whatever their sex, gender, and race—will only touch the guest once that guest has consented.118 Crucially, “every legal right that patients have to privacy is rendered irrelevant by the fact that patients must consent to medical procedures.”119 Denying the Home’s BFOQ defense only means that the Home is unable to guarantee its guests that the available aide will be a woman. If there are only male nurse’s aides available, the individual guest can—indeed has a constitutionally protected right to—refuse that aid.120 Given this, the privacy interest the court is worried about—the one connected to tort and criminal law—is in fact not in play at all. Regardless of whether the court accepts the BFOQ defense, the Home is never “legally forc[ing] its female guests to accept personal care from males.”121

Second, the court’s statement that the aid administered “involves a personal touching as to which each guest is privileged by law to discriminate on any basis” is false.122 As Robert Post points out, “if the nursing home residents in Fesel had claimed a privacy right not to be touched by nurse’s aides who were African American, their expectations would no doubt properly and ruthlessly be overridden by Title VII.”123

It is easy to see why, at first blush, a particular privacy right may seem imperiled by a particular antidiscrimination law. But, our first-pass intuitions are supposed to be the start and not the end of our inquiry. Upon careful attention to both the contours of the relevant interest and the mechanisms of the relevant law we can see, as the discussion of Fesel shows, that sometimes our intuitions are wrong; perceived conflicts can turn out to be illusory ones.124

c.     Privacy rights and the legitimacy of antidiscrimination laws in the market

That consumer sex-discriminatory preferences couched in the language of privacy are often the most successful suggests that if consumers have a constitutional objection to Title VII’s restrictions on employers’ ability to sell them discrimination, the right derives from either the Fourth Amendment or the hazier constellation of constitutional sources from which other privacy rights derive.125 A few brief thoughts on this.

First, a consumer right undergirded by a vague idea of Fourth Amendment-based privacy rights may help explain why consumers’ sex but not race discrimination is sometimes legitimated. In the Fourth Amendment context, courts still occasionally find the sex but not race of individuals relevant in determining whether a search is unreasonable.126 For instance, in Byrd v. Maricopa County Sheriff ’s Department,127 the Ninth Circuit sitting en banc held that as a matter of law a strip search of a male prisoner by a female cadet was unreasonable in violation of the Fourth Amendment.128 The Ninth Circuit underscored the “longstanding recognition that ‘[t]he desire to shield one’s unclothed figure from [the] view of strangers, and particularly strangers of the opposite sex, is impelled by elementary self-respect and personal dignity.’”129 Given this, the court found cross-sex strip searches especially degrading, humiliating, and in so being, unreasonable absent an emergency.130 Although the Byrd court found the opposite-sex strip search to violate the inmate’s Fourth Amendment rights, it is hard to imagine any court finding that inmates have a cognizable Fourth Amendment claim on the basis of being strip-searched by someone of a different race. In 2017, we no longer recognize as reasonable—at least for purposes of Fourth Amendment analysis—privacy-based arguments for racial segregation.131

The complication here is that while sex-discriminatory preferences are sometimes part of a Fourth Amendment analysis, Fourth Amendment-based privacy interests are not obviously present in cases where firms are prohibited from satisfying the discriminatory preferences of consumers. BFOQ cases concern the privacy interests consumers have in relation to other non-government actors, where the Fourth Amendment does not typically apply. Instead, as the court in Fesel noted, to know our privacy rights vis-à-vis other non-state actors, we refer to tort and criminal laws. And in contrast to Fourth Amendment cases, where courts sometimes recognize a sex-discriminatory privacy interest, tort and criminal laws concerning privacy are not crafted in sex-discriminatory terms. The success of privacy-related tort does not depend on the sex or gender of the actors involved.

If vague Fourth Amendment “reasonableness” notions motivate consumer privacy BFOQ cases, upon reflection it seems an odd fit given the lack of state action. If privacy interests protected by tort and criminal law are the source of a consumer right to discriminate, upon reflection it seems equally as odd.

The Fourth Amendment, though, is not the only constitutional source of privacy rights. Contraception cases like Griswold and abortion cases like Roe rely on substantive due process and “zones of privacy” and “penumbras” stemming from several constitutional sources.132 These cases also show that constitutionally-protected privacy interests can be violated by government regulation of third-parties and/or the relationship between individuals and third-parties.133 Might the consumer privacy right to discriminate averred to in BFOQ cases be the same one(s) protected in Griswold or Roe? Perhaps, but it would require taking the position that sex-discriminating in the market is a personal right that is either “fundamental” or “implicit in the concept of ordered liberty.”134 And while Roe listed a number of decisional areas where the relevant privacy right(s) extended—into “activities relating to marriage; procreation; contraception; family relationships; and child rearing and education”—it would be a leap to argue that the list should also include the entire domain of market activity.135

IV.  Conclusion

The fact that individuals discriminate against other individuals within the market is undeniable.136 So far the law has rarely targeted it directly. For a number of reasons, the visibility of discrimination in the on-demand economy has prompted a rethinking of this choice. But even for those who favor some expansion of antidiscrimination laws, there is a worry that individuals might have a right to discriminate in the market. Worry in mind, these same authors suggest that consumer discrimination should instead be targeted indirectly, through additional regulations on the third-parties, like Uber and Airbnb, that facilitate it. Such indirect action could include prohibiting these companies from showing users’ real names or photos.

This Essay was motivated by a concern with this line of thinking. As I explained, if individuals do have a constitutional right to discriminate while in the market, these indirect methods may be much more problematic than these authors seem to think. Thus, it is my suggestion that we take a step back before going much further. Instead of taking as true the vague intuition that directly prohibiting consumer discrimination would violate those consumers’ rights, we need to examine that intuition.137 We need to understand whether the government can, in fact, legitimately prohibit individuals from discriminating in the market. Part of this inquiry requires a fleshed-out account of whether the regulated-discrimination is harmful or unjust. In trying to craft meaningful limits to the government’s legitimate authority within liberal theory, we also need to decide whether a market/non-market distinction is a helpful one. I outlined my early thinking on these questions in Part II. A related question is whether, today, the law suggests individuals have a right to discriminate in the market. And, as in the case of privacy-based BFOQ cases, whether the arguments courts give to explain those rights are sound. Given both § 1981 and a closer inspection of the consumer sexual titillation and privacy interests raised in BFOQ cases, I find it unlikely that individuals have, today, a right to engage in race and sex discrimination while in the market.

• 1Benjamin Edelman and Michael Luca, Digital Discrimination: The Case of Airbnb.com, (Harv. Bus. Sch., Working Paper No. 14-054, 2014), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2377353, https://perma.cc/XZX9-3CMV.
• 2See Brian Chesky, An Update on the Airbnb Anti-Discrimination Review, Airbnb Blog (July 20, 2016), http://blog.airbnb.com/an-update-on-the-airbnb-anti-discrimination-review/, https://perma.cc/M8K9-UWRU.
• 3See Yanbo Ge et al., Racial and Gender Discrimination in Transportation Network Companies, (NBER Working Paper, No. 22776, 2016); see also Mark Scott, Study Finds Some Uber and Lyft Drivers Racially Discriminate, N.Y. Times, (Oct. 31, 2016), https://www.nytimes.com/2016/11/01/technology/uber-lyft-racial-discrimination.html?_r=0, https://perma.cc/X3SM-EDTB.
• 4See Al Franken, Letter to Travis Kalanick and Logan Green (Nov. 2, 2016), https://www.franken.senate.gov/files/letter/161102_UberLyft.pdf, https://perma.cc/5FML-9MXH; see also Megan Rose Dickey, Sen. Al Franken Asks Uber and Lyft about Alleged Racial Discrimination Against Passengers, TechCrunch (Nov. 3, 2016), https://techcrunch.com/2016/11/03/sen-al-franken-asks-uber-and-lyft-about-alleged-racial-discrimination-against-passengers, https://perma.cc/88QK-3SP9.
• 5See Megan Rose Dickey, In Light of Discrimination Concerns, Uber and Lyft Defend their Policies to Show Rider Names and Photos, TechCrunch (Dec. 29, 2016), https://techcrunch.com/2016/12/29/uber-lyft-respond-al-franken-about-discrimination, https://perma.cc/44HL-ZD6J.
• 6That is not to say there have not been some lawsuits. In June 2016, Gregory Selden, the African American Airbnb user who first created the #airbnbwhileblack hashtag, sued Airbnb for race discrimination on behalf of himself and other African-American Airbnb users. He argued that Airbnb was liable under federal civil rights law as a public accommodation. However, the district court never reached the substantive claim, instead granting Airbnb’s motion to compel arbitration consistent with the Terms of Service. See Memorandum Opinion, Selden et al. v. Airbnb, Inc., No. 16-cv-00933, 2016 WL 6476934 (D.C. Cir. Nov. 1, 2016).
• 7See Heather M. Whitney, Rethinking the Ban on Company—Labor Organization Cooperation, 37 Cardozo L. Rev. 1455 (2016).
• 8See Alana Semuels, Why #DeleteUber and Other Boycotts Matter, The Atlantic (Feb. 22, 2017), https://www.theatlantic.com/business/archive/2017/02/why-deleteuber-and-other-boycotts-matter/517416/, https://perma.cc/2ZBT-UHKX; Cam Wolf, Pro-Trump Shoppers are Boycotting Nordstrom, Netflix, and More, Racked (Feb. 10, 2017), https://www.racked.com/2017/2/10/14577910/pro-trump-boycotting-nordstrom-netflix-starbucks-tj-maxx, https://perma.cc/9ZRZ-3QSN; Paul Farhi, Breitbart News seems to be cleaning house after readers and advertisers drift away, Washington Post (June 7, 2017), https://www.washingtonpost.com/lifestyle/style/is-breitbart-news-veering-away-from-the-farthest-far-right/2017/06/06/35f91160-4ad1-11e7-a186-60c031eab644_story.html?utm_term=.f55603158d39, https://perma.cc/T76S-U5RF; Lauren Sherman, Amidst backlash, Ivanka Trump clothing is secretly relabeled as Adrienne Vittadini, Business of Fashion (April 24, 2017), https://www.businessoffashion.com/articles/news-analysis/ivanka-trump-clothing-relabeled-adrienne-vittadini, https://perma.cc/9CAG-ME32.
• 9See generally Lee Anne Fennell, Searching for Fair Housing, 97 B.U. L. Rev. 349 (2017).
• 10See Katharine T. Bartlett & Mitu Gulati, Discrimination by Customers, 102 Iowa L. Rev. 223, 228–31 (2016) (discussing these efficacy reasons).
• 11Id. at 226.
• 12Id. at 249.
• 13Tarunabh Khaitan, A Theory of Discrimination Law 212 (2015).
• 14666 F.3d 1216 (9th Cir. 2012).
• 15Id.
• 16Brief for Defendant-Appellant at 51, Fair Housing Council of San Fernando Valley v. Roommates.com, LLC, No. 09-55272, 2010 WL 2751575 (9th Cir. May 14, 2010) (“The injunction impermissibly inhibits the rights of home-seekers to state and act on preferences, however, by forbidding Roommates from making available to them formatted questions or matching using even voluntary responses. Denying such users of roommates.com the choice of focusing their search based on the characteristics of those with a home to share is not only unwarranted, it is unconstitutional.”) (emphasis in original).
• 17Roommates.com, 666 F.3d at 1222–23.
• 18See Fennell, supra note 9, at 395.
• 19See id.
• 20See id. (“[T]he fact that § 1982 reaches biased home-seeking can also be used to buttress direct liability for collateral search behavior under the FHA itself.”).
• 21564 U.S. 786 (2011).
• 22Id. at 786–87.
• 23See generally Seth K. Kreimer, Censorship by Proxy: The First Amendment, Internet Intermediaries, and the Problem of the Weakest Link, 155 U. Pa. L. Rev. 11 (2006).
• 24See, e.g., Eisenstadt v. Baird, 405 U.S. 438, 445–46 (1972) (holding professor could challenge statute forbidding the distribution of contraceptives to unmarried persons); Craig v. Boren, 429 U.S. 190, 194–95 (1976) (permitting beer vendor to challenge law prohibiting the sale certain alcohol to males under 21 but only to women under 18 by invoking the Equal Protection rights of the buyers); Planned Parenthood of Se. Pa. v. Casey, 505 U.S. 833 (1992) (discussing indirect burdens on abortion rights); Planned Parenthood of Wis., Inc. v. Schimel, 806 F.3d 908 (7th Cir. 2015) (holding Wisconsin law requiring abortion providers to obtain admitting privileges at nearby hospitals an unconstitutional unburden on women’s right to an abortion).
• 25John Stuart Mill, On Liberty, in On Liberty, Utilitarianism and Other Essays 12–13 (2015) (1859).
• 26In other words, once there is a third-party harm, the end—preventing that harm—is legitimate. Under a Millian picture, your liberty is not violated when the state acts to prevent you from inflicting a third-party harm: your liberty interests end where third-party harms begin. One might object to the means the state uses to prevent the third-party harm but there is no liberty-based objection to the end—preventing third-party harm—itself. For a more detailed account of this, see Heather M. Whitney, The Autonomy Defense of Private Discrimination (2017) (unpublished manuscript), https://ssrn.com/abstract=2922241, https://perma.cc/G444-FB55.
• 27John Gardner discusses this in more detail. See John Gardner, Discrimination as Injustice, 16 Oxford J. Legal Stud. 353, 365 (1996).
• 28See John Gardner, Liberals and Unlawful Discrimination, 9 Oxford J. Legal Stud. 1, 2–3 (1989).
• 29For a more careful parsing of these arguments see Whitney, supra note 26.
• 30A libertarian worry lurks here. See David E. Bernstein, Context Matters: A Better Libertarian Approach to Antidiscrimination Law, Cato Unbound (June 16, 2010), http://www.cato-unbound.org/2010/06/16/david-e-bernstein/context-matters-better-libertarian-approach-antidiscrimination-law, https://perma.cc/RR7U-5BEM (“The proliferation of antidiscrimination laws explains why libertarians are loath to concede the principle that the government may ban private sector discrimination. There is no natural limit to the scope of antidiscrimination laws, because the concept of antidiscrimination is almost infinitely malleable.”). Of course, the concept of harm is also susceptible to this criticism and yet libertarians (and liberal theory more generally) are committed to it as a legitimate doctrine.
• 31See Gardner, supra note 28, at 3.
• 32For some of this literature, see, e.g., Tarunabh Khaitan, A Theory of Discrimination Law 117–39 (2015); Kasper Lippert-Rasmussen, Born Free and Equal? A philosophical inquiry into the nature of discrimination 103–192 (2014) (Part II is on the wrongness of discrimination); Philosophical Foundations of Discrimination Law (Deborah Hellman & Sophia Moreau eds., 2013); Larry Alexander, What Makes Wrongful Discrimination Wrong? Biases, Preferences, Stereotypes, and Proxies, 141 U. Pa. L. Rev. 218 (1992). For an example of work that rejects a broader equality of opportunity principle while supporting prohibitions on forms of discrimination, see Matt Cavanagh, Against Equality of Opportunity (2002).
• 33My discussion here is brief. For more, see Heather M. Whitney, Markets, Rights, and Discrimination by Customers, 102 Iowa L. Rev. Online 346 (2017).
• 34For a longer discussion, see Samuel R. Bagenstos, The Unrelenting Libertarian Challenge to Public Accommodations Law, 66 Stan. L. Rev. 1205, 1210 (2014) (“Most public actors at the time seem to have taken for granted that there was a distinction between three classes of rights.”).
• 35See id.
• 36See id. at 1210.
• 37Plessy v. Ferguson, 163 U.S. 537 (1896), overruled by Brown v. Bd. of Educ., 347 U.S. 483 (1954).
• 38See Bagenstos, supra note 34, at 1210. See also Reva Siegel, Why Equal Protection No Longer Protects: The Evolving Forms of Status-Enforcing State Action, 49 Stan. L. Rev. 1111, 1120 (1997).
• 39See Jones v. Alfred H. Mayer Co., 392 U.S. 409, 440–41 (1968) (implicitly finding the Fair Housing Act constitutional); Runyon v. McCrary, 427 U.S. 160, 179 (1976).
• 40Bagenstos, supra note 34, at 1218.
• 41See id. at 1209.
• 42See id. at 1212; see also Duncan Kennedy, The Stages of the Decline of the Public/Private Distinction, 130 U. Pa. L. Rev. 1349 (1982).
• 43At one point, commercial speech received no First Amendment protection. See Valentine v. Chrestensen, 316 U.S. 52 (1946), overruled in Va. State Pharmacy Bd. v. Va. Citizens Consumer Council, 425 U.S. 748 (1976).
• 44Sony Corp. of Am. v. Universal City Studios, 464 U.S. 417, 449 (1984) (stating that “a commercial or profit-making purpose” was “presumptively” unfair). The Court walked this back somewhat ten years later. See Harper & Row Publ’rs, Inc. v. Nation Enters., 471 U.S. 539, 562 (1985) (stating that when copying is commercial it “tends to weigh against a finding of fair use”) (emphasis added). Almost ten years after, the Court made clear that commercial uses are not determinative in a fair use analysis, though still can weigh against a fair use finding if the new work not sufficiently transformative. See Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 579 (1994) (“[T]he more transformative the new work, the less will be the significance of other factors, like commercialism, that may weigh against a finding of fair use.”). Even today, though, commerciality is used to argue strongly against a fair use determination. See Opening Brief and Addendum of Plaintiff-Appellant at 27–28, Oracle v. Google, No. 17-1118 (Fed Cir. Feb. 10, 2017).
• 45Lawrence v. Texas, 539 U.S. 558, 578 (2003) (pointing out that the case “does not involve public conduct or prostitution”).
• 46Lochner v. New York, 198 U.S. 45 (1905).
• 47See West Coast Hotel Co. v. Parrish, 300 U.S. 379 (1937) (upholding constitutionality of state minimum wage law).
• 48468 U.S. 609 (1984).
• 49Id. at 633–34 (O’Connor, J. concurring in part and concurring in the judgment).
• 50Bagenstos, supra note 34, at 1230. I see this as a close cousin to the market/non-market framework.
• 51530 U.S. 640 (2000).
• 52Id. at 640. The distinction works as a limit because the organization in question was not commercial.
• 53Bagentos goes through some of these distinctions. See Bagentos, supra note 34, at 1231–40. I discuss some other problems in Whitney, supra note 26.
• 54See Lawrence v. Texas, 539 U.S. 558, 578 (2003) (pointing out that the case “does not involve public conduct or prostitution”).
• 55See, e.g., State v. Freitag, 130 P.3d 544, 546 (Ariz. Ct. App. 2006) (“We thus join other state courts that have specifically rejected any constitutionally protected fundamental liberty or privacy interest in soliciting or engaging in prostitution.”) (case citations omitted).
• 56See Matal v. Tam, 137 S.Ct. 1744, 1767 (2017) ((“‘Commercial speech is no exception,’ to the principle that the First Amendment ‘requires heightened scrutiny whenever the government creates a regulation of speech because of disagreement with the message it conveys.’”) (citing Sorrell v. IMS Health Inc., 564 U.S. 552 (2011)) (Kennedy, J., with whom Ginsburg, J., Sotomayor, J., and Kagan, J. joined, concurring in part and concurring in the judgment)). I also discuss this erosion elsewhere. See Whitney, supra note 33, at 28.
• 57Masterpiece Cakeshop, Ltd. v. Colo. Civil Rights Comm’n, docket no. 16-111, cert. granted June 26, 2017. The question presented is whether Colorado’s public accommodation law, which requires the petitioner to provide a wedding cake for a same-sex couple, unconstitutionally violates his free exercise or free speech rights under the First Amendment.
• 58Some philosophers continue to defend the market/non-market distinction as a meaningful one. See, e.g., Debra Satz, Why Some Things Should Not Be For Sale (2010); Margaret Jane Radin, Contested Commodities 132–36 (1996); Elizabeth Anderson, Value in Ethics and Economics 150–58 (1993). Others within the libertarian tradition have disagreed and found the market/non-market distinction mistaken. See Jason Brennan & Peter Martin Jaworski, Markets Without Symbolic Limits, 125 Ethics 1053 (2015).
• 59As an example of another approach, some have thought antidiscrimination duties can be legitimately targeted at entities that distribute primary social goods or are agents of distributive justice. See, e.g., Gardner, supra note 27, at 363 (discussing employers as agents of distributive justice). While at first blush this seems to exclude the imposition of antidiscrimination duties on individuals, some have argued that individuals do distribute primary social goods—for instance, they argue that romantic relationships are primary social goods. Thus, the regulation of online dating sites is legitimate. See Sonu Bedi, Sexual Racism: Intimacy as a Matter of Justice, 77 J. Pol. 998 (2015).
• 60See Barry Sullivan, Historical Reconstruction, Reconstruction History, and the Proper Scope of Section 1981, 98 Yale L.J. 541, 554 (1989) (“[F]ormer masters generally did not refuse to contract with their former slaves, but recognized the use of labor contracts as a means of reintroducing practical slavery, insisting on terms so onerous as to leave the Freedmen worse off than they had been under slavery.”). See also Danielle Tarantolo, From Employment to Contract: Section 1981 and Antidiscrimination Law for the Independent Contractor Workforce, 116 Yale L.J. 170, 186 (2006).
• 61See Sullivan, supra note 60, at 545 (“The Thirty-ninth Congress was little concerned with any theoretical distinction between public and private action in early 1866, but was committed to giving effect to the Thirteenth Amendment by ‘secur[ing] to all persons within the United States practical freedom.’”) (citations omitted).
• 62Tarantolo, supra note 60, at 550.
• 63See Sullivan, supra note 60.
• 64427 U.S. 160 (1976).
• 65Id. at 176 (“[T]he Constitution . . . places no value on discrimination, and . . . [i]nvidious private discrimination . . . has never been accorded affirmative constitutional protection.”). See also Jones v. Alfred H. Mayber Co., 392 U.S. 409 (1968) (interpreting § 1982 as applying to individual discrimination).
• 66Runyon v. McCrary, 427 U.S. 160, 170–71 (1976).
• 67Id. at 172.
• 68Id. at 175–76.
• 69Note, The Expanding Scope of Section 1981: Assault on Private Discrimination and a Cloud on Affirmative Action, 90 Harv. L. Rev. 412, 422 (1976).
• 70Civil Rights Act of 1991, Pub. L. 102-166, 105 Stat. 1071 (1991).
• 71See George Ruterglen, The Improbable History of Section 1981: Clio still bemused and confused, 2003 Sup. Ct. Rev. 303, 306 (2003).
• 72Id.
• 73See Fennell, supra note 9, at 395.
• 74See generally Kimberly A. Yuracko, Private Nurses and Playboy Bunnies: Explaining Permissible Sex Discrimination, 92 Cal. L. Rev. 147 (2004) (paraphrasing Kimberly Yuracko).
• 7542 U.S.C.§ 2000e-2 (2000). Title VII prohibits both intentional (disparate treatment) as well as some unintentional (disparate impact) discrimination. See 42 U.S.C. § 2000e-2(k)(1)(A)(i). See also Griggs v. Duke Power, 401 U.S. 424, 436 (1971) (allowing an employer to defend against a disparate impact claim by showing the policy or practice is “demonstrably a reasonable measure of job performance”); Ricci v. DeStefano, 557 U.S. 557, 577 (2009). Below I focus on whether consumers have a right to intentionally discriminate.
• 7642 U.S.C. § 2000e-2(e)(1) (2000). The fact that discrimination is ever permissible is in direct conflict with the plurality’s broad language in Price Waterhouse v. Hopkins, 490 U.S. 228 (1989). There, invoking a sex-and-gender-blindness conception of Title VII, the Court interpreted the Act “to mean that gender must be irrelevant to employment decisions.” Id. at 240 (Brennan, J., plurality) (emphasis added).
• 7729 C.F.R. 1604.2(a)(1)(iii) (emphasis added).
• 78EEOC Office of Legal Counsel Opinion Letter Re: Title VII: BFOQ (March 5, 2002), https://www.eeoc.gov/eeoc/foia/letters/2002/titlevii_bfoq.html, https://perma.cc/3PV4-YJ38.
• 79In using “preferences” I do not mean to subscribe to a Gary Becker-type conception of race discrimination as a matter of “taste.” See Gary Becker, The Economics of Discrimination 14–15 (2d ed. 1971). Race discrimination is not a matter of mere individualized taste. For a more thorough explanation of why, see Richard H. McAdams, Cooperation and Conflict: The Economics of Group Status Production and Race Discrimination, 108 Harv. L. Rev. 1003 (1995).
• 80Of course, there are more subtle ways to sell racial homogeneity. See Lior Jacob Strahilevitz, Exclusionary Amenities in Residential Communities, 92 Va. L. Rev. 437 (2006).
• 81See Yuracko, supra note 74.
• 82Id. at 155–56.
• 83Id. at 156.
• 84Id.
• 85Id. at 157.
• 86Id. at 158. See also, e.g., Wilson v. Southwest Airlines Co., 517 F. Supp. 292, 301 (N.D. Tex. 1981) (“[I]n jobs where sex or vicarious sexual recreation is the primary service provided, e.g. a social escort or topless dancer, the job automatically calls for one sex exclusively.”).
• 87Yuracko, supra note 74, at 158. Yuracko criticizes courts’ attempt to determine the essence of these businesses. See id. Pan American and Southwest are often thought a paragon of this. See Diaz v. Pan Am. World Airways, Inc., 442 F.2d 385 (5th Cir. 1971). There, the Fifth Circuit rejected Pan Am’s BFOQ argument that it hired only female flight attendants because, in part, customers preferred female stewardesses for their skill in “non-mechanical aspects” of the job, such as “providing reassurance to anxious passengers, giving courteous personalized service and, in general, making flights as pleasurable as possible within the limitations imposed by aircraft operations.” Id. at 387. While the court acknowledged that woman flight attendants increased the pleasantness of the environment through their “obvious cosmetic effect,” this was non-essential to the business, which was “provid[ing] safe transportation from one place to another.” Id. at 388. See also Southwest Airlines, 517 F. Supp. at 302 (holding that, even though Southwest explicitly advertised as the ‘love airline’ and on that basis hired only female flight attendants, sex was not a legitimate BFOQ because the company’s primary function was to “transport passengers safely and quickly”).
• 88517 F. Supp. 292 (N.D. Tex. 1981).
• 89Id. at 304.
• 90Id. at 295.
• 91Some might disagree with this claim. See, e.g., Mary Wollstonecraft, A Vindication of the Rights of Woman with Strictures on Political and Moral Subjects 338 (1796) (describing marriage as “legal prostitution”).
• 92I say “in-person” because the First Amendment is understood to protect the sale and purchase of non-obscene pornography (speech).
• 93See section I.B. See also Whitney, supra note 33, at 26–37.
• 94Lawrence v. Texas, 539 U.S. 558 (2003).
• 95See, e.g., People v. Williams, 811 N.E.2d 1197 (Ill. App. 3 Dist. 2004); State v. Freitag, 130 P.3d 544, 546 (Ariz. Ct. App. 2006) (“We thus join other state courts that have specifically rejected any constitutionally protected fundamental liberty or privacy interest in soliciting or engaging in prostitution.”) (citations omitted). Again, likely in part because of the Court’s dicta in Lawrence. See Lawrence, 539 U.S. at 578 (“The present case . . . does not involve public conduct or prostitution. The petitioners are entitled to respect for their private lives.”).
• 96See section I.B. See also Whitney, supra note 33, at 26–37.
• 97For more discussion of the complications of race and sex discrimination in the realm of sex work, see Adrienne D. Davis, Regulating Sex Work: Erotic Assimilationism, Erotic Exceptionalism, and the Challenge of Intimate Labor, 103 Cal. L. Rev. 1195, 1262–73 (2015).
• 98See Bagentos, supra note 34.
• 99See Yuracko, supra note 74, at 156 (“Courts have generally been quite permissive toward sex discrimination committed on behalf of customer privacy.”).
• 100Id. We might also add to the continuum the presence of children, since courts may be more likely to find same-sex hiring essential in those spaces.
• 101See id. Cf. Norwood v. Dale Maint. Sys., 590 F. Supp. 1410, 1416 (N.D. Ill. 1984) (“[I]n certain situations the privacy rights of individuals justify sex-based hiring by an employer” and these situations generally involve occupations that require workers to “work with or around individuals whose bodies are exposed in varying degrees.”).
• 102See Yuracko, supra note 74, at 156.
• 103See id. at 157.
• 104Id. While not consumers, cases involving opposite-sex prison guards in sex-segregated prisons have been especially thorny, regardless of whether the guards see prisoners naked. Compare Ambat v. City and County of San Francisco, 757 F.3d 1017, 1030 (9th Cir. 2014) (finding the record to not demonstrate that the presence of male deputies violated female inmate privacy given policies that prevented male deputies from taking on duties, like strip searches, that might violate those interests), Henry v. Milwaukee County, 539 F.3d 573, 581–86 (7th Cir. 2008) (holding male sex not a BFOQ in juvenile men’s unit because job responsibilities could be rearranged to protect inmate privacy), Griffin v. Michigan Department of Corrections, 654 F. Supp. 690 (E.D. Mich. 1982) (holding sex not a BFOQ for officers in an all-male prison regardless of whether female officers saw them naked because male inmates did not have a protected privacy right to not be viewed naked by female officers), and Gunther v. Iowa State Men’s Reformatory, 612 F.2d 1079, 1085–87 (8th Cir. 1980) (same in prison), with Everson v. Mich. Dep’t of Corr., 391 F.3d 737, 747–61 (6th Cir. 2004) (holding female sex a BFOQ for certain guard positions in women’s prison), Torres v. Wis. Dep’t of Health & Social Service, 859 F.2d 1523, 1528–32 (7th Cir. 1988) (en banc) (holding female sex may be a BFOQ based on female inmates’ need), and Dothard v. Rawlinson, 433 U.S. 321, 334–37 (1977) (holding sex a BFOQ for guards in men’s prison).
• 105See Yuracko, supra note 74, 157.
• 106See id. Compare Norwood v. Dale Maint. Syst., Inc., 590 F. Supp. 1410, 1418 (N.D. Ill. 1984) (holding that sex was a BFOQ for staffing attendants who cleaned single-sex restrooms in a large office building because office workers would feel “embarrassment” and “increased stress” if they expected to see members of the opposite sex in the washroom) with EEOC v. Hi 40 Corp., 953 F. Supp. 301, 302–305 (W.D. Mo. 1996) (holding that sex was not a BFOQ in the hiring of weight-loss counselors even though 95% of the company’s clients were women, most of whom felt uncomfortable with males taking their body-fat measurements and discussing the various emotional and physiological issues associated with weight loss with them; and finding the privacy interest only “minimal”).
• 107447 F. Supp. 1346 (D. Del. 1978), aff ’d 591 F.2d 1334 (3d Cir. 1979).
• 108Id. at 1352. See also Backus v. Baptist Medical Center, 510 F. Supp. 1191, 1193–94 (E.D. Ark. 1981) (identifying patients’ interest in female-only delivery nurses on the basis of a constitutional right to privacy).
• 109Fesel, 447 F. Supp. at 1352.
• 110Id.
• 111Id.
• 112See id.
• 113Id.
• 114Id. at 1353 (emphasis added).
• 115Id.
• 116Id.
• 117Id.
• 118For more, see Kimani Paul-Emile, Patients’ Racial Preferences and the Medical Culture of Accommodation, 60 UCLA L. Rev. 462 (2012).
• 119Amy Kapczynski, Same-Sex Privacy and the Limits of Antidiscrimination Law, 112 Yale L.J. 1257, 1268 (2003).
• 120See Cruzan by Cruzan v. Director, Mo. Dept. of Health, 497 U.S. 261, 278 (1990) (“The principle that a competent person has a constitutionally protected liberty interest in refusing unwanted medical treatment may be inferred from our prior decisions.”).
• 121Fesel, 447 F. Supp. at 1353. One might push back here and say there is de facto coercion, given the medical context. I’m sympathetic to this more robust conception of coercion but leave for another time the many difficulties it raises. For instance, if that is right, then must private companies always have available doctors and other staff who can satisfy the prejudices of patients or else be guilty of coercing those patients to touching they do not want? I doubt it.
• 122Id. at 1353 (emphasis added).
• 123Robert C. Post et. al., Prejudicial Appearances: The Logic of American Antidiscrimination Law 1–53 (2001)
• 124For more in this vein, see Amy Kapczynski, Same-Sex Privacy and the Limits of Antidiscrimination Law, 112 Yale L.J. 1257, 1273–74 (2003) (“Same-sex privacy doctrine cannot be defended with recourse to the law of privacy. In many cases, the third parties in question have no relevant privacy rights. Even where they do have such rights, it is not self-evident that there is or ought to be a link between those privacy rights and sex.”); Deborah Calloway, Equal Employment and Third Party Privacy Interests: An Analytical Framework for Reconciling Competing Rights, 54 Fordham L. Rev. 327 (1985).
• 125While the Constitution “does not explicitly mention any right of privacy” the Court has recognized constitutionally protected privacy interests stemming from a variety of constitutional sources. Katz v. United States, 389 U.S. 347 (1967), most canonically ties Fourth Amendment protection to individuals’ reasonable expectations of privacy. See id. at 361 (Harlan, J. concurring) (“My understanding of the rule . . . is that there is a twofold requirement, first that a person have exhibited an actual (subjective) expectation of privacy and, second, that the expectation be one that society is prepared to recognize as ‘reasonable.’”). Outside of the Fourth Amendment, the Roe Court identified the “roots” of a right of personal privacy, “or a guarantee of certain areas or zones of privacy” as stemming from the First, Fourth, Fifth, Ninth, and Fourteenth Amendments, in addition to locating it in “the penumbras of the Bill of Rights.” Roe v. Wade, 410 U.S. 113, 152 (1973).
• 126Though, as many scholars critique the legitimization of consumer sex-discrimination via the BFOQ exception, so do others critique the legitimization of sex-discrimination via the Fourth Amendment finding of reasonableness. See, e.g., I. Bennett Capers, Unsexing the Fourth Amendment, 48 U.C. Davis L. Rev. 855 (2015).
• 127629 F.3d 1135 (9th Cir. 2011).
• 128Id. at 1136.
• 129Id. at 1141 (citing York v. Story, 324 F.2d 450, 455 (9th Cir. 1963)).
• 130Id. at 1143–47 (emphasis added).
• 131Though it is worth pointing out that consumers have historically raised privacy objections to racial integration. See Race and Retail: Consumption Across the Color Line (2015).
• 132See Griswold v. Connecticut, 381 U.S. 479, 484 (1965); Roe v. Wade, 410 U.S. 113, 152 (1973).
• 133See generally Griswold, 381 U.S. 479.
• 134Roe, 410 U.S. at 152–53 (1973).
• 135Id.
• 136The legal literature has acknowledged consumer discrimination in tipping since at least 2005. See Ian Ayres, Fredrick E. Vars, and Nasser Zakariya, To Insure Prejudice: Racial Disparities in Taxicab Tipping, 114 Yale L.J. 1613 (2005).
• 137I develop this analysis further in other work. See Whitney, supra note 33.