Price’s Progress: Sex Stereotyping and Its Potential for Antidiscrimination Law
abstract.American antidiscrimination law has addressed harmful stereotypes since, at least, the Civil Rights Act of 1964. Stereotypes about the different abilities of men and women, or of black and white workers, lay underneath much of the segregation and workplace inequality that Title VII sought to correct. However, Price Waterhouse v. Hopkins has radically expanded our conception of stereotypes as discrimination, and in doing so has introduced revolutionary ideas to the workplace and the courts. Prior to Price Waterhouse, Title VII had been thought to apply only to ascriptive stereotyping—to monolithic misconceptions burdening all members of a disfavored group. Price Waterhouse’s extension of Title VII protection to victims of prescriptive stereotype has constituted a massive, and heretofore unstudied, conceptual leap. This Note examines how Price Waterhouse’s prohibition against stereotyping can transform American workplace law and analyzes one area where it already has—antigay discrimination. By contrasting the requirements for proving antigay discrimination under a Price Waterhouse sex stereotyping theory with the traditional Title VII methods that many states use to protect LGBT workers, I show both how Price Waterhouse can complement proposed LGBT-specific protections such as the Employment Non-Discrimination Act (ENDA), and how its normative vision is a vital addition to existing antidiscrimination law.
author. Ph.D. Candidate, Columbia University; Yale Law School, J.D., 2014. As much as this Note arises from my own longstanding obsessions, it also reflects the support of wonderful friends, mentors, and colleagues. I offer sincere thanks to Bill Eskridge, Vicki Schultz, and Tom Tyler of the Yale Law School, for whose guidance I am deeply grateful. Jessie Weber graciously shared her wisdom, experience, and hospitality with me. Ida Araya-Brumskine, Geoff Mort, Alex Reed, Rachel Shalev, Julie Veroff, Kenji Yoshino, and Noah Zatz offered invaluable feedback at various stages of the editing process. My editor, Devon Porter, has been thoughtful and generous, as have her colleagues Meng Jia Yang and Rachel Bayefsky. Finally, this Note would never have existed without my parents, Robert Herz and Rosalind Fink, for the obvious reasons and also because of their tireless encouragement and support. All errors are, of course, my own.
On January 31, 1968, Isabell Slack was stereotyped.1 Slack, an African-American industrial worker at Havens International, was asked to spend the morning cleaning her department’s workspace.2 Slack’s coworker, a white woman, was excused.3 When Isabell Slack asked why she was being expected to do the work of a cleaning lady, her supervisor explained that “colored people are hired to clean because they clean better.”4 The Ninth Circuit found this to be race discrimination, with perhaps unremarkable brevity;5 the court felt no need to explain how these statements were racially motivated because no other motivation existed. Ms. Slack’s supervisor reduced her to her race—we know nothing about her own aptitude for cleaning, the cleanliness of her workspace, or whether she would be better or worse at such work than her colleague. All we know is that Slack is black. Her supervisor assumed that all African-American women were skilled at domestic work and ascribed that characteristic to Isabell Slack with no further thought. This sort of thinking lies at the core of what Title VII sought to prevent, and has fallen uncontroversially within its ambit since the 1960s.
On May 1, 1989, however, the Supreme Court radically expanded our conception of Title VII stereotype in Price Waterhouse v. Hopkins.6 Ann Hopkins was repeatedly told by her employers to dress, speak, and act in a manner more appropriate to her sex.7 This was stereotype, too, but of a vastly different form. Isabell Slack’s employer paid no attention to her particular characteristics, while Ann Hopkins’s employer obsessed over them. Havens International assumed that Isabell Slack was like other African-Americans, while Price Waterhouse saw that Ann Hopkins was not like other women and held it against her. Havens International, by assigning a characteristic to Isabell Slack without judging her as an individual, engaged in ascriptive stereotyping. Price Waterhouse, by correctly perceiving Ann Hopkins’s individual traits but then judging them against an inappropriately gendered baseline, engaged in prescriptive stereotyping. While Price Waterhouse has been incredibly important in Title VII case law and scholarship,8 this simple difference—between assuming members of a group fit a certain stereotype, on the one hand, and demanding that they do so, on the other—is drastically understudied, and it implicates two major problems in antidiscrimination discourse today.
First, a broader application of Price Waterhouse’s view of discrimination has the potential to resolve, or at least to ameliorate, a serious problem in American antidiscrimination law—the inability of traditional Title VII approaches to address the realities of modern workplace bias. While the Civil Rights Act of 1964 led to huge immediate gains for black9 and female10 workers, those gains have recently stalled. Discrimination has, to quote Zachary Kramer, “become highly individualized;”11 specifically, bias is increasingly expressed as a single factor in complex and multivariate individual evaluation, and is thus increasingly difficult to fit into the specific, historically contingent model of open race and gender hierarchies that animated the Civil Rights Act of 1964. This conceptual mismatch has led to shockingly poor outcomes for employment discrimination plaintiffs and persistent workplace inequality.12 Professor Kramer stands in a long tradition of theorists considering this mismatch and its possible causes. Some view the problem as institutional, arising from structural bias13 or assimilationist work culture.14 Other scholars consider individual decision makers’ roles in creating these structural inequalities, with some highlighting subconscious individual bias as a potential problem for Title VII law15 and others considering how structural discrimination can result from conscious, if covert, biased thinking.16 While these arguments describe slightly different mechanisms of employment discrimination, they all identify the same basic problem: workplace discrimination now largely arises in contexts in which plaintiffs simply cannot prove a violation of Title VII as traditionally understood.
Building on this work, Suzanne Goldberg argues that current evidentiary requirements make complex discrimination exceedingly difficult to prove; by requiring cross-status comparators to demonstrate causation,17 judges implicitly demand that employers treat members of a group differently on the whole in order to generate a cognizable claim, regardless of the impact discrimination might have on the individual.18 Goldberg proposes methods of proof that would permit courts to consider context, and claims that subtle discrimination instantiating group hierarchies can occur without generating the clear in-group/out-group distinctions judges currently require.19 This Note elaborates on how Price Waterhouse doctrines respond to this pressing concern. The context that Goldberg identifies, and that traditional Title VII doctrines cannot reach, is a workplace dominated by gendered and raced prescriptions about how people should behave; Price Waterhouse sex stereotyping doctrine, explicitly designed to fight prescriptive stereotype, is the tool best suited to address this problem.
Furthermore, a nuanced understanding of Price Waterhouse can inform our current debates over how best to protect victims of antigay workplace discrimination. Although there are currently no explicit statutory protections in federal law against antigay discrimination, many courts have held that Price Waterhouse’s ban on “assuming or insisting that [employees] match the stereotype associated with their group”20 forbids employers from discriminating against gender-deviant LGBT employees.21 Furthermore, many states (and soon, possibly the federal government) protect against antigay discrimination using language explicitly modeled on Title VII.22 This division between sex-stereotyping litigation and new, sexual-orientation-specific legislation has put various scholars, lawyers, and activists at cross-purposes. While Lambda Legal and the Human Rights Campaign minimize existing protections against discrimination and urge passage of a federal ENDA,23 and while President Obama has signed an executive order protecting LGBT federal employees,24 the Equal Employment Opportunity Commission (EEOC) has ruled that discrimination against transgender workers categorically violates Price Waterhouse,25 and EEOC Commissioner Chai Feldblum has publicly stated that discrimination against LGB workers does the same.26 Similarly, there is a dispute on the merits of Price Waterhouse and ENDA within the legal academy. While recent work argues that Price Waterhouse protection is a flawed substitute for ENDA,27 and many scholars claim that locating LGBT protections within sex stereotyping jurisprudence is theoretically and practically superior to a separate statute,28 a recent piece proposes abandoning traditional frameworks entirely for a theory of reasonable accommodation of gender performance.29
This Note shows that such arguments fundamentally misunderstand the nature of prescriptive sex stereotyping. The doctrine that has sprung up in Price Waterhouse’s wake does not extend the Civil Rights Act so much as radically reimagine its scope. Sex stereotyping doctrines ask courts to examine just the sorts of subjective and individualized workplace evaluations that Title VII has historically ignored. Courts consider these cases through a two-step process, establishing the importance of a particular workplace norm to the plaintiff’s firing or harassment, and then asking whether that norm reflects biased thinking about sex. The resulting doctrine is not a mere substitute for, or superior version of, protection through traditional30 Title VII claims; the two approaches work very differently, and LGBT workers are best protected by having access to both.
This Note proceeds in five parts. Part I considers how American antidiscrimination law has addressed stereotype both before and after Price Waterhouse, and in particular how courts expect plaintiffs to show differences in treatment of in-group and out-group employees under ascriptive and prescriptive theories of stereotyping. This issue of comparison is a serious hurdle facing Title VII plaintiffs, and one that the Price Waterhouse framework can handle more nimbly and effectively than other Title VII doctrines. Part II considers Price Waterhouse in the arena of LGBT workplace rights, where Price Waterhouse protections exist without any explicit protection based on traditional theories. This Part considers cases where Price Waterhouse doctrine could help LGBT workers whose claims would fail under ENDA—these claims demonstrate not only the differences in coverage between Price Waterhouse and ENDA, but also how the analytical moves made in sex-stereotyping jurisprudence could benefit victims of other types of workplace bias. Part III then considers how Price Waterhouse and ENDA can work together to combat anti-LGBT bias in the workplace. Part V offers a brief conclusion.
Price Waterhouse v. Hopkins31 has profoundly affected American discrimination law, but its doctrines are still misunderstood and unnecessarily cabined. When the Supreme Court decided that Ann Hopkins had faced discrimination on the basis of sex because she was told to “walk more femininely, talk more femininely, dress more femininely, wear make-up, have her hair styled, and wear jewelry,”32 the Court vastly expanded protections against sex discrimination in the context of a general case law drawing far narrower lines. The resulting doctrine, while still inconsistently applied, has the potential in its broadest form to reach plaintiffs traditionally excluded from Title VII protection. After discussing the state of Title VII sex-discrimination law prior to Price Waterhouse, this Part will show how Price Waterhouse expanded the Court’s existing doctrine on sex-based stereotypes to include prescriptive stereotypes, as opposed to the ascriptive sexism (and racism) already forbidden by Title VII; how lower courts have struggled to reconcile Price Waterhouse with traditional interpretations of Title VII; and how Price Waterhouse’s own text, as well as Congress’s later treatment of the issue in the Civil Rights Act of 1991, support a broad reading of the case.
Price Waterhouse v. Hopkins is often considered the genesis of “sex stereotyping.”33 As Kimberly Yuracko and others have noted, this is false.34 Title VII jurisprudence has focused on stereotypes almost from the moment of the Civil Rights Act’s passage, and understanding how Price Waterhouse expands traditional anti-stereotyping doctrine is critical for understanding its radical implications.
One of the first cases in which the Supreme Court addressed stereotypes in Title VII is Phillips v. Martin Marietta Corp.35 When Ida Phillips applied for a job with the Martin Marietta Corporation in September of 1966,36 her prospective employer appeared friendly to women. The workforce Phillips wished to join was nearly eighty percent female.37 Nevertheless, despite being fully qualified, Phillips was turned away. Martin Marietta assumed that any woman with pre-school-age children would face “domestic complication[s]”38 preventing her from working effectively, and therefore turned away women (and not men) with young children.39 The Fifth Circuit held that Title VII could reach only discrimination that was triggered solely because of sex and that, as long as women who did not have children were treated no worse than men, Phillips could not claim discrimination.40 The Supreme Court reversed and remanded, holding in a brief per curiamdecision that Martin Marietta could not maintain such a policy unless family obligations could be shown to be “demonstrablymore relevant to job performance for a woman than for a man.”41 One can view Martin Marietta as resolving a simple conflict about universality; the Fifth Circuit held that an employer had to discriminate against the entire group, whereas the Supreme Court extended Title VII to cover discrimination against discrete “subgroups.”42
But how discrete, how clearly defined, must these subgroups be? Sex-plus analysis, by definition, requires the characteristics an employer considers (say, sex and parenthood) to be sufficiently general to apply identically to multiple parties. Otherwise, there is no way for men and women to “share” the characteristic and therefore no discrimination.43 In fact, a close reading of the case reveals that this very generality, and Martin Marietta’s deployment of generalizing ascriptive stereotype, lie at the heart of Phillips’s claim. Martin Marietta claimed it was using its ban on mothers of pre-school-age children to exclude employees who would be distracted, prone to absenteeism, or facing frequent emergencies at home.44 By definition, this type of stereotypical assumption can work only if the category triggering the assumption is both broad and somewhat monolithic. Ascriptive stereotype consists of treating a large group of people alike, and erasing individual differences (for example, does this mother have an unemployed husband to take care of the children? Is this childless woman the guardian of her young niece?) in favor of applying the same rule to all members of a specific group.45
This type of stereotyping is impermissible even when it is demonstrated to be broadly accurate. Consider Los Angeles Department of Water and Power v. Manhart,46 in which the City of Los Angeles demanded greater pension fund contributions from female employees on the grounds that women, on the whole, lived longer than men. Although the stereotype here was generally correct,47 the Court held that any employment decision based on gender violated Title VII and suggested that any ascriptive stereotype unacceptably subordinated individual to group identity: “Practices that classify employees in terms of religion, race, or sex tend to preserve traditional assumptions about groups rather than thoughtful scrutiny of individuals.”48 The Manhart Court read Title VII to require consideration of individuals49 as opposed to groups—stereotypes about how one sex will react to childrearing, or how long one sex will live compared to another, violate the statute’s central aim.50
Of course, assumptions are not the only way to stereotype. The Manhart Court described Los Angeles’s pension plan in terms of “sex stereotypes”51 years before Price Waterhouse v. Hopkins, but the phrase meant something very different. Martin Marietta and Manhart both concerned employers who ascribed actual characteristics to their employees based on their membership in a protected class. By contrast, Price Waterhouse forbade prescriptive stereotyping, in which an employer analyzes an employee’s individual characteristics, as Manhart seems to require, but in reference to an unacceptably biased norm.
Price Waterhouse expanded and (depending on its interpretation) radically altered Title VII stereotyping jurisprudence. Compared to the defendants in Manhart and Martin Marietta, Price Waterhouse assessed Ann Hopkins as an individual. While supporters and opponents of her candidacy for partnership disagreed on how to evaluate the traits they perceived, all parties agreed about her actual behavior.52 There was no inaccuracy or factual assumption involved, and Hopkins did not seriously contest her employer’s factual observations.53 Instead, Hopkins alleged that her employer had applied a flawed evaluative system to its correct perception of her behavior. Instead of viewing Hopkins’s aggression as a simple facet of her personality—or a desirable attribute in a Price Waterhouse partner—Price Waterhouse evaluated Hopkins on a biased rubric; was she behaving as a woman “should”?54 This is prescriptive stereotyping, in which assumptions about men and women do not inform the employer’s understanding of the employee’s actual nature(that Phillips will be distracted by her children, or that Manhart will live longer than the average worker) but instead dictate their evaluation of that nature (that Ann Hopkins should “walk more femininely, talk more femininely, dress more femininely”).55 Therefore, a Title VII that protects against prescriptive stereotyping is a subtly different beast than the one that came before.
The basic idea of Price Waterhouse—that “we are beyond the day when an employer could evaluate employees by assuming or insisting that they matched the stereotype associated with their group”56—is uncontroversial, but it has been applied differently by different circuits. Before considering these variant interpretations, however, it is important to understand the conceptual ramifications of a shift from ascriptive to prescriptive stereotyping in discrimination law. This dichotomy is not itself novel. The distinction between ascriptive stereotypes and prescriptive stereotypes has been present in the legal literature for decades.57 However, surprisingly little attention has been paid to the impact of this difference on employment discrimination litigation.58 In contrast to the stereotyping in Martin Marietta and Manhart—stereotyping that made factual assumptions based on membership in a protected class—prescriptive stereotyping takes a problematic evaluative mechanism and applies it to employees based on their individual characteristics.59 Consequently, prescriptive stereotype cases do not fit neatly into the similarly-situated-comparator model used elsewhere in Title VII jurisprudence. As Suzanne Goldberg has noted, courts have increasingly come to view comparative evidence, or evidence showing a difference in treatment between a plaintiff and another employee who is similarly situated except for membership in a protected group, as something approaching a requirement for Title VII litigation.60 Ascriptive stereotypes, which apply to all or nearly all members of the stereotyped group and only to members of the group, can be demonstrated by comparative evidence—Isabell Slack’s employer assumed she was a skilled cleaner and made no such assumption about her white coworker. Prescriptive stereotypes, on the other hand, will be applied only against employees who violate the biased norm in specific, individual ways. Because a prescriptive stereotype does not necessarily cause employers to treat all members of a group equally, and because the prescriptive stereotype will frequently apply in one form or another to employees outside of a protected status group, plaintiffs facing prescriptive sex stereotyping may be unable to prove the discrimination condemned in Price Waterhouse through traditional methods. Courts have generally addressed this problem in two ways, either by recasting prescriptive stereotyping as a bar to advancement for women alone61—thereby analogizing sex stereotyping to traditional forms of sex discrimination—or by adopting radically different comparative models focused on the use of sexist heuristics in evaluating employee conduct or presentation.62
Some courts have interpreted Price Waterhouse quite narrowly, requiring that prescriptive stereotypes be shown to harm members of one sex—usually women—as a class before treating them as actionable under Title VII. The facts of Price Waterhouse could support such a theory: Ann Hopkins was able to show not only that she suffered due to her deviance from stereotyped gender norms, but also that women on the whole were systematically denied promotion to partner.63 Between this disparity and the reluctance of many partners to promote any woman whatsoever,64 Hopkins’s treatment can be thought of as straightforward sex discrimination, or as a barrier to promotion affecting women on the whole. If, on the other hand, we consider Hopkins’s denial of partnership as a punishment for her own, idiosyncratic gender deviance, then we must determine how an evaluation of one woman’s individual traits can constitute discrimination against her as a member of a class.
Courts that have followed the narrow reading of Price Waterhouse, building on a long history of interpreting Title VII as equalizing opportunities for men and women,65 have generally conceived of prescriptive stereotypes as leading to sex discrimination in one of two ways: either as unequally applied to men and women, or as placing women specifically in a “double bind.”66 The first theory makes intuitive sense; an employer who punishes women and not men for gender-nonconformity engages in straightforward “sex-plus” discrimination. Much like Martin Marietta’s punishing only female applicants for having pre-school-age children, employers who punish only their female employees for deviating from gender norms commit sex discrimination as traditionally understood.67 At least as applied to Hopkins, however, this “sex plus deviance” theory seems problematic. In order to prove sex discrimination arising from prescriptive stereotype under such a reading of Title VII, Hopkins would have had to show that a similarly situated man—that is, one who violated prescriptive gender stereotypes in the same way she did—was treated better. Unless Price Waterhouse had a history of promoting docile, non-aggressive men in stereotypically feminine clothes, Hopkins would lose. Her victory casts this interpretation of the statute into serious doubt.68 However, the straightforward sex-plus reading of Price Waterhouse has found unexpected life in some courts as a remedy for transwomen; in Rosa v. Park West, for example, the First Circuit held that a bank may have committed sex discrimination by refusing to serve a femininely attired man if they would have served a masculinely attired woman.69 The more common narrow reading of Price Waterhouse, however, and one that applies more neatly to Ann Hopkins’s own case, reads prescriptive stereotypes as existing in equal force for men and women but as affecting women more severely.
According to this theory—commonly known as the “double bind”—prescriptive stereotypes in employment place women, and only women, in a Catch-22. If high-status jobs are seen as requiring aggression, assertiveness, or other stereotypically masculine traits, then women may be left “out of a job if they behave aggressively and out of a job if they do not.”70 Therefore, while these stereotypes may constrain men and women equally, only men are able to demonstrate their suitability for prestigious forms of employment without being seen as gender-deviant. This theory has some empirical support71 and may in fact lie at the root of many persistent sex-based inequalities in the workforce.72 However, as a reading of the Court’s reasoning in Price Waterhouse,the double-bind theory has one serious flaw; it calls for a different doctrinal framework than the one actually used. If we read conformity with gender stereotype as an evaluative measure applied to men and women equally, and as one that is only pernicious inasmuch as it hurts women more than men, then prescriptive stereotype belongs to the “disparate impact” jurisprudence that originated in Griggs v. Duke Power Company.73 The Court never even gestured at a disparate impact holding in Price Waterhouse; Justice Brennan in fact states the relevant standard in terms of whether or not “the employer actually relied on [Hopkins’s] gender in making its decision.”74 While the problematic effects of the double bind obviously troubled Justice Brennan, the case coheres only if we begin from the theory that stereotyping is itself proscribed by Title VII, regardless of its impact on a protected class.
This latter theory—building largely off of Manhart’s holdingthat individuals may not be evaluated in terms of groups75—has given rise to a broader reading of the Price Waterhouse sex stereotyping doctrine. In circuits that have adopted the broader reading, plaintiffs need only show that they were perceived to deviate from prescriptive stereotypes based on their sex and that they were punished for that deviance. Plaintiffs in these cases still use comparative evidence, but compare themselves not to similarly situated employees of a different sex; instead, plaintiffs compare themselves to employees of either sex who are similarly situated but conform to gender stereotypes to a different extent. A difference in treatment between a plaintiff and a gender-conforming comparator shows that the sex-based stereotype played a role in an employer’s decision. A clear if somewhat summary example of this test can be found in Myers v. Cuyahoga County, which held that a transwoman plaintiff could meet her prima facieburden under McDonnell Douglas by showing “that she was replaced with a gender-conforming person.”76 Susan Myers did not have to show that she was treated differently from someone without her Title VII protected status, as the analysis might proceed if Title VII covered gender identity;77 alternatively, under current Title VII law, such an analysis would require that she be replaced by someone of a different sex. Instead, the Myers court asked for evidence that Myers was replaced by someone with a different gender presentation. Although the category of “gender-deviant individuals” is formally sex-blind, plaintiffs in such cases can then prevail by showing how the categorization reflects problematic ideas about the Title VII category of sex.78 The broader theory of sex stereotyping does not require direct evidence of a difference in treatment based on sex alone; it instead inquires into the criteria employers use in comparing employees regardless of their sex, and attaches liability if those criteria themselves reflect sexist thinking.
Of course, this bifurcated inquiry—what attribute led to the plaintiff’s firing, and does the employer’s dislike of that attribute reflect biased thinking about sex?—raises a host of questions about how to establish the role of bias in a particular workplace prescription. These are not easy questions to answer. In fact, this need to establish bias is one reason why Price Waterhouse cases are more difficult and complex than cases alleging group-based discrimination under traditional Title VII doctrines. But the hurdle can be cleared. Historically, Price Waterhouse plaintiffs are often more successful in harassment cases where the harasser’s use of gendered language clearly indicates sex stereotyping.79 Alternatively, plaintiffs can use expert testimony to show the gendered nature of certain workplace demands80 or even seek judicial notice when the stereotype is commonly understood as based on sexist ideas about men’s and women’s roles.81
In addition, Mary Anne Case has recently argued that Price Waterhouse sex stereotypes can be understood as stereotypes that proscribe behavior for one sex while permitting or requiring it for the other. This requirement essentially echoes the “sex plus deviance” standard described above,82 but instead of similar levels of gender nonconformity, Case believes Price Waterhouse plaintiffs need comparators who behave similarly to the plaintiff but are viewed as gender-conforming because they are of a different sex. Case sees Price Waterhouse as requiring that any behavior acceptable for men (such as aggression, marriage to a female partner, or a short haircut) be acceptable for women, and, vice versa, that any trait accepted for women (such as docility or wearing skirts) be accepted for men as well.83 Notably, under this reading of Price Waterhouse, such a mutuality requirement—frequently termed “trait neutrality”84 —lies at the core of sex stereotyping. Case herself admits that this standard, with its attendant skepticism of gender difference in areas such as dress and grooming, has not been adopted by courts.85 Nevertheless, she is clearly correct that this formal structure (if a stereotype applies differently to men and women, it must be based on sex) can be useful as one of many ways to show sex stereotyping. That said, inasmuch as one views violations of trait neutrality as necessary for a Price Waterhouse claim, and inasmuch as Case presents this test as “in general, a more effective way to achieve legal protection for the broadest possible range of sexual identities, gendered traits, and the individuals manifesting them,”86 there is very real room for disagreement.
Case’s trait neutrality standard asks, as a threshold inquiry, whether the employee’s behavior would have been condoned if he or she were a different (generally the opposite) sex.87 This is a straightforward question, and for those employees who can answer it affirmatively (for example, transgender plaintiffs who are being penalized for presenting according to their true sex),88 it would vastly simplify the litigation process. As a sufficient condition to demonstrate the sex-based nature of a stereotype, trait neutrality is important and useful. But for effeminate men and masculine women, the “queers, sissies, dykes, and tomboys”89 that populate sex-stereotyping law, this is the worst possible question to ask. Trait neutrality protects only behaviors that the employers would accept on a member of the opposite sex; in practical terms, such a rule would permit employers to maintain rigid sexed grooming and conduct standards while simply permitting employees to choose which set of standards to follow.
An example may be useful here, taken from Case’s own discussion of gender-nonconforming individuals. Case refers to a memorable episode where then-Representative Barney Frank, emerging from a conference meeting on ENDA, joked that he had engaged in “kind of a sartorial compromise” by pairing a conservative suit with a lavender tie.90 It’s a funny line. That said, imagine if Rep. Frank wore that tie to his job and were fired as a result. In a context where it is clear that a lavender tie represents a conscious rejection of gender commitments, and where forbidding it represents the imposition of those commitments on an employer’s workforce, trait neutrality would likely offer no help. In order to prevail under a trait neutrality reading of Price Waterhouse, Rep. Frank would have to show that his lavender tie would be acceptable for female employees; however, a woman wearing a boldly colored tie would be, if anything, more gender deviant and therefore more likely to encounter difficulties in the workplace. Because Rep. Frank’s “gender-bending” expressed itself in behavior that is stereotypically associated with neither sex, trait neutrality would allow it to be punished without legal sanction.91
To be clear, trait neutrality is an incredibly valuable tool for transgender plaintiffs and other individuals who can use it to demonstrate sex stereotyping cheaply and easily. Similarly, plaintiffs who can show that gender deviance is accepted for members of one sex and not the other in their particular workforce can succeed under the “sex plus deviance” standard adopted by some courts. However, treating a violation of these or other sex-plus readings of Title VII as a necessary condition for sex-stereotyping claims would be just as likely to narrow Price Waterhouse as it would be to expand it. For those who want Title VII to offer broader protections for gender-deviant workers, this is a serious concern.
However, trait neutrality is not a major feature of the litigation landscape for Price Waterhouse plaintiffs.92 Instead, the two readings of sex stereotyping described above largely dominate: a narrow group-based framework, which forbids stereotyping only when it leads to one sex being directly disadvantaged vis-à-vis the other, and a broader reading that prevents employers from choosing between employees based on biased heuristics. It is not clear in all circuits whether the narrow or broad reading of Price Waterhouse applies, which can lead to confusion over whether or not plaintiffs are entitled to relief. Consider the 1998 Second Circuit case Galdieri-Ambrosini v. National Realty & Development Corp. The plaintiff alleged that she had suffered discrimination relative to her more attractive, stereotypically feminine coworker,93 who was given less work and more lenient treatment. Ambrosini’s attorney argued that this difference was based on an “impermissible sexual stereotype”:94
Dana brought Mr. Simon coffee unsolicited and she also cleaned away his coffee cup. This is something that Miss Ambrosini objected to. This is conforming to a sexual stereotype. The jurors may infer that because Dana . . . looked a certain way, acted a certain way, made Mr. Simon’s life more pleasant in the workplace, even if it was something as simple as bringing him coffee, she conformed to the sexual stereotype and she did not complain about it.95
This is a clear sex-discrimination claim under the broad reading of Price Waterhouse. Ambrosini, a gender-nonconforming woman, was treated worse than another woman in the same position who conformed to sexual stereotypes. Ambrosini alleged that this disparity arose from a difference in the two women’s gender presentation, and that National Realty, in basing employment decisions on sexed evaluative criteria, had violated Title VII. The court rejected Ambrosini’s theory, reading Price Waterhouse to require that an employer treat one sex worse than another in order to face liability.96 There was no issue of fact in Ambrosini—a jury had already found for the plaintiff, who was appealing the trial court’s grant of judgment to the defendant as a matter of law97—and therefore the outcome of the case hinged entirely on which reading of Price Waterhouse the Second Circuit adopted. The difference mattered.
However, when we consider the actual harm done to Ambrosini it becomes clear how, without the protection of a broad Price Waterhouse doctrine, victims of obvious discrimination can be left without a legal remedy. In essence, Ambrosini suffered the same harm as did Slack; both were punished for their inability to conform to their supervisors’ raced and sexed realities. But Slack found relief where Ambrosini did not, largely due to two tangential factors: Slack had a white coworker who was not subjected to the same demands, whereas Ambrosini could not find a well-suited male comparator,98 and Slack’s supervisor directly mentioned race in his remarks.99 Whatever harm we think Isabell Slack suffered, it did not arise from the presence of comparators, and the gendered nature of Ambrosini’s treatment should be clear without an employer being foolish enough to say it out loud. A narrow reading of Price Waterhouse, by imposing evidentiary requirements that are unrelated to the actual harm suffered by victims of discrimination, puts real obstacles in front of victims of prescriptive stereotype and makes it harder for them to find relief.
Of course, the difference in locating the harm of prescriptive stereotyping is not only practical—it also presents competing normative visions of the purpose of antidiscrimination law. We can think of traditional ascriptive stereotypes as wrong on three distinct grounds. First, one could say that such stereotypes generate factually inaccurate perceptions about individuals, and thus lead to decisions that are incorrect or insufficiently respectful of individual attributes.100 Second, one could object to these stereotypes on anticlassification or formal-equality grounds: ascriptive stereotypes that assign attributes based on membership in certain groups will make group identity impermissibly salient in the workplace, while also leading to impermissible disparities in employment outcomes between in-group and out-group members.101 Third, one could make a broader substantive-equality or antisubordination claim: regardless of the identity of those to whom a stereotype is applied, that stereotype becomes problematic when it has the effect or purpose of suppressing disfavored groups.102 However, if we are to believe that the problem with stereotype extends to unfair prescription, then we must reject the first of these theories; prescriptive stereotype consists of normative judgments that, ipso facto, can never be factually inaccurate. Similarly, such stereotypes begin with a (theoretically) accurate view of an individual’s traits and behaviors; this would seem to satisfy the autonomy interests implicated by individuality claims. So we are left with formal and substantive equality as two different bases for attacking prescriptive stereotype.
To return to Ambrosini, the court’s holding can be seen as limiting Price Waterhouse to violations of formal equality; because Ambrosini did not show that her treatment differed from that of a man, her employer was sex-blind for purposes of the statute. The plaintiff argued in contrast that the stereotype deployed against her violated Price Waterhouse due to its inherently oppressive content, regardless of its sex-blind application. If we subscribe to the substantive-equality reading of sex stereotyping, then the Title VII violation is clear. A prescription that women should be servile and men dominant,103 even if applied to all employees in a specific workplace, would obviously reinforce a degrading and subordinating workplace culture. The question becomes: after Price Waterhouse, does Title VII forbid employers from using evaluative criteria that reflect biased thinking, or merely from applying their evaluative criteria in biased ways?
It is still unsettled which version of sex stereotyping doctrine reads Price Waterhouse correctly. The opinion itself—or, more properly, opinions—offer some support for both interpretations, but a close reading of Price Waterhouse and of the Civil Rights Act of 1991 (which overrode parts of the decision) support a broad reading. While Price Waterhouse was a plurality opinion, Justices Brennan and O’Connor, representing five total votes, both expressed less concern with disparities in treatment between men and women than with the influence of impermissible considerations in the hiring process. While the bottom-line disparities between men and women were more than sufficient to support an inference of discrimination,104 Justices Brennan and O’Connor both went out of their way to frame the stereotype that affected Ann Hopkins not as a factor that caused an impermissible difference in Price Waterhouse’s treatment of men and women, but as a form of impermissible treatment in and of itself.105 Furthermore, while language in Justice O’Connor’s concurrence expresses discomfort with using Title VII to police evaluations in the total absence of a difference in outcomes,106 this part of her opinion was superseded by the Civil Rights Act of 1991.
The Civil Rights Act of 1991 was explicitly adopted in order to override problematic aspects of several then-recent Supreme Court employment decisions,107 and its treatment of Price Waterhouse aroused far less discussion (and thus less legislative history) than other aspects of the bill.108 That said, what evidence there is suggests that the Civil Rights Act’s “motivating factor” provision was intended to overrule any requirement of disparity in outcome. The Civil Rights Act of 1991 established an “unlawful employment practice . . . when the complaining party demonstrates that race, color, religion, sex, or national origin was a motivating factor for any employment practice.”109 Once a plaintiff has proven a discriminatory motive, liability is assigned—an employer who demonstrates that she would have made the same decision in the absence of the impermissible factor cannot be forced to reinstate the employee or pay damages, but the plaintiff remains entitled to declaratory relief and attorneys’ fees.110 By comparison, Justices Brennan and O’Connor both found that, if the employer could show that the same decision would have been made absent discrimination, then he would be absolved of all liability.111 In the 1991 Civil Rights Act, Congress explicitly stated that impermissible motivations, without impermissible differences in outcome, suffice for Title VII liability. It is difficult to square this reading with a requirement that plaintiffs show disparate outcomes between cross-status comparators in order to merit even fees and declaratory relief. Since the cross-status comparator is most relevant to demonstrating whether or not the employer would have made the same decision absent discrimination, this requirement collapses the two prongs of the statutory test and renders the same-decision language surplusage. While comparative evidence is always useful in proving the salience of one particular factor to a decision, the actual wrong lies in the stereotype itself, and not in its operation to generate workplace disparities.112
This plain-meaning analysis is further supported by the legislative history of the Act. While earlier language had required a plaintiff to show that the discriminatory motive was a “contributing” factor, this language was specifically changed to remove any implication of an impact requirement.113 Unlike “contributing,” which queries the causal chain leading to an adverse decision, “motivating” suggests a specific concern with employers’ reasoning; Justice O’Connor rejected this language precisely due to her discomfort with reading Title VII to require that type of analysis.114 By amending the original draft to include Brennan’s “motivating factor” language, Congress explicitly authorized courts to consider motivation.115 While both Justices Brennan and O’Connor contemplated using Title VII to query employers’ evaluative methods, Congress took this process one step further by permitting liability on the basis of such methods alone, without considering the differences in treatment required by some courts. As amended in 1991, Title VII clearly allows for the “biased criteria” analysis that underlies the broad reading of Price Waterhouse prescriptive stereotype protection.
While this broader interpretation of Price Waterhouse is still not universal, its reach is expanding.116 The Supreme Court itself has not clearly addressed the issue. It offered ambiguous support for the narrow reading of Price Waterhouse in its 1998 decision Oncale v. Sundowner Offshore Services,117 but the decision has only generated further confusion in its progeny, and has not been clarified at the Supreme Court level. Oncale, a case holding that male-on-male sexual harassment was actionable under Title VII, would at first seem to be an ideal test case for disaggregating the two readings of Price Waterhouse.118 Joseph Oncale faced adverse employment action119 based on perceived deviance from gendered norms, but was harassed by members of his own sex.120 Oncale was victimized by prescriptive sex stereotypes in a way that clearly did not disadvantage men as a class.121 The Court’s decision in Oncale, however, punted the issue. Justice Scalia, writing for a unanimous Court, merely reaffirmed that Title VII protects against harassment “because . . . of sex.”122 The Court then provided a nonexhaustive list of methods by which a plaintiff could show causation, all falling under the narrow reading of Price Waterhouse,123 before rejecting Sundowner’s motion to dismiss. The Court reversed even though Oncale had alleged no facts that would suffice to state a claim under a narrow Price Waterhouse theory—Oncale never alleged that he would have received better treatment if he had been a woman124—which strongly suggests that the Court felt Oncale met his burden without having to introduce evidence showing discrimination against men. The Court never clarified further; Oncale settled the case days before further proceedings were to begin.125
Further confusing the matter, five days after its decision in Oncale the Court vacated a Seventh Circuit decision that read Price Waterhouse broadly“for further consideration in light of Oncale v. Sundowner Offshore Services.”126 While this decision might suggest discomfort with the broad reading of Price Waterhouse, later decisions in the Seventh Circuit continue to read Price Waterhouse broadly,127 and in fact the original Belleville decision is still cited approvingly elsewhere.128
The ambiguity surrounding Price Waterhouse has created a sex-stereotyping doctrine that still varies from circuit to circuit, but the decision itself, as well as later legislative and judicial behavior, indicates that Price Waterhouse prohibits stereotypes that work indirectly to subordinate protected groups. Under this reading, Price Waterhouse is a substantial innovation in antidiscrimination law; Price Waterhouse breaks with the comparative, group-disparity model that governs the rest of Title VII jurisprudence and offers a more context-sensitive and nuanced way to think about workplace bias. The next Part will consider Price Waterhouse in the area where it has been perhaps most generative: in cases protecting gender-deviant LGBT employees. In the absence of ENDA, gay and transgender victims of discrimination have turned to Price Waterhouse for relief; now, as states increasingly enact more traditional protections against sexual orientation and gender identity discrimination, LGBT plaintiffs are running on parallel tracks. A close analysis of these cases makes clear not only how sex stereotyping can complement ENDA, but also how the ideas underlying sex stereotyping can complement traditional Title VII protections for all classes of plaintiffs.
II. antigay discrimination, or sex stereotyping? comparing price waterhouse and enda in selected employment contexts
Employment protections for LGBT plaintiffs under current law are, to say the least, a piece of work. Courts considering claims by LGBT plaintiffs must balance prohibitions on discrimination based on ideas about how men and women should behave with a total lack of explicit federal protection against antigay discrimination. This riddle bedevils both courts, which struggle to differentiate between gender deviance and sexual orientation, and advocates, who use Price Waterhouse protections to the best of their ability while also fighting for LGBT protections under a different, more traditional model. These advocates have met with some success—many states and municipalities offer formal protection against sexual orientation and gender identity discrimination, creating an environment in which anti-LGBT discrimination can be fought using two vastly different theories. These fact patterns are ideal test cases for exploring just how Price Waterhouse differs from traditional understandings of Title VII. More importantly, these cases show how Price Waterhouse improves on those understandings and offers solutions to truly vexing problems in American antidiscrimination law. After briefly considering the state of current and proposed law addressing sexual orientation, gender identity, and gender deviance discrimination, this Part uses the LGBT example to show how Price Waterhouse can address “corner cases” in which clear discrimination cannot be fit into traditional Title VII standards of proof.
LGBT plaintiffs currently have two separate theories for redressing antigay discrimination, but both are limited and highly contingent. Many courts have interpreted Price Waterhouse to hold that discrimination against LGBT employees is based on their perceived violation of gendered prescriptions and thus constitutes sex stereotyping, but only if the discrimination arises from the plaintiff’s perceived gender deviance and not from anti-LGBT bias per se. Otherwise, a separate, older doctrine bars relief. The first case to address Title VII’s applicability to sexual minorities, DeSantis v. Pacific Telephone & Telegraph Co.,129 used legislative history to foreclose “bootstrapping” sexual orientation protection into Title VII.130 DeSantis has never been addressed by the Supreme Court, but it (and decisions like it in other circuits) bar relief for plaintiffs alleging sexual orientation discrimination. However, plaintiffs who allege that they were seen as violating gender stereotypes may seek relief under a Price Waterhouse theory, even if the harasser perceived his victim as homosexual,131 and even if large parts of the harassment consisted of antigay speech.132
The DeSantis plaintiffs alleged that discrimination against homosexuals constituted “sex-plus” discrimination against men under Martin Marietta133 and also had a disparate impact on men due to higher incidence of homosexuality among males as opposed to females.134 The DeSantis court firstrebuffed the plaintiffs’ claims of “sex-plus” discrimination by asserting that homosexual men and women were being treated equally: “[W]e note that whether dealing with men or women the employer is using the same criterion: it will not hire or promote a person who prefers sexual partners of the same sex.”135 This theory—that ideas about how members of one sex should behave are acceptable as long as they are accompanied by “mirror-image”136 rules for the other—no longer holds in the Ninth Circuit.137 But even in circuits that require asymmetrical disparate treatment of men and women, courts have found that antigay or antitrans harassment can violate Title VII.138
The DeSantis court’s other argument against relief—its appeal to legislative history139—is no more dispositive. DeSantis’s explicitly redistributive reading of Title VII as exclusively focused on “plac[ing] women on an equal footing with men”140 is difficult to square with Oncale,141 to say nothing of Ricci v. DeStefano.142 Going forward, the fear of “bootstrap[ping]” that courts frequently invoke in denying Price Waterhouse protection to gender-deviant homosexual plaintiffs143 is likely misplaced. Nevertheless, DeSantis and its progeny are still widely cited for the proposition that, if Congress had intended Title VII to forbid sexual orientation discrimination, it would have said so. If this seems confusing, it is: courts that adopt the broader reading of Price Waterhouse have struggled to reconcile prohibiting discrimination based on gender deviance with permitting discrimination against individuals whose sexual orientation leads them to be perceived as gender deviant.
Two cases illustrating both sides of this coin are Nichols v. Azteca Restaurant Enterprises144 and Dawson v. Bumble & Bumble.145 Both cases involved gay employees facing adverse employment action. However, because the Nichols plaintiffs were able to frame their claim in formally sexual-orientation-neutral terms, they succeeded where Dawson failed. The Nichols court, focusing on instances in which a gay male server was referred to as “‘she’ and ‘her.’ . . . and a ‘fucking female whore,’”146 found that the harassment, while consisting partly of antigay slurs, “reflected a belief that [the plaintiff] did not act as a man should act”147 and thus gave rise to a colorable claim of Price Waterhouse sex stereotyping. By contrast, because Dawn Dawson “conflated” her claims of sex stereotyping with claims of explicit antigay discrimination,148 and because “[t]he law is well-settled in this circuit and in all others to have reached the question that . . . . Title VII does not prohibit harassment or discrimination because of sexual orientation,”149 Dawson’s claims, according to the court, constituted impermissible bootstrapping of sexual orientation into existing protections.
As Zachary Kramer, Brian Soucek, and others have noted, this case law is both doctrinally and descriptively incoherent; Kramer decries courts’ subsuming all prescriptive sex stereotyping of gay plaintiffs into “sexual orientation simpliciter claims in disguise,”150 whereas Soucek notes that courts attempting to separate antigay bias from ideas about gender deviance do so “solely by fiat.”151 This problematic case law puts LGBT plaintiffs in a delicate position and can doom unsophisticated plaintiffs who do not carefully observe the artificial distinction between antigay bias and gender norms that courts require. However, courts’ insistence on this distinction arises in large part from case law holding that, by refusing to amend Title VII to explicitly protect LGBT workers, Congress has made clear its lack of intent to do so.152
LGBT rights organizations are now attempting to write explicit legal protections for LGBT workers into state and federal law. Recently, gay-rights groups have been able to pass variants of ENDA in more than twenty states and the District of Columbia,153 and they hope to pass a federal version in the next few years. A federal ENDA would have far-reaching implications; in addition to providing relief in straightforward sexual-orientation-discrimination claims,154 such a law would vitiate the doctrine forbidding “bootstrap[ping]”155 sexual orientation claims into Title VII and would allow for a far cleaner application of sex stereotyping to cases involving gay and lesbian plaintiffs. However, the application of ENDA, and the gaps in current law that it seeks to address, are themselves subjects of debate. The EEOC recently ruled that discrimination against transgender workers already violates Price Waterhouse156 and is bringing similar claims on behalf of victims of anti-LGB discrimination.157 Meanwhile, President Obama has signed an executive order instituting protections modeled on ENDA for employees of the federal government and its contractors.158 Consequently, disentangling the potential protections offered by ENDA from those found in Price Waterhouse is a serious challenge, and one that this Part hopes to address.
It is always difficult to interpret hypothetical legislation. However, in order to determine exactly how ENDA would differ from the protections Price Waterhouse provides, we must first sort out what exactly ENDA would do. The most recently proposed legislation states that “[i]t shall be an unlawful employment practice for an employer . . . to discharge any individual . . . because of such individual’s actual or perceived sexual orientation or gender identity,”159 mimicking the language of Title VII itself. However, it is currently unsettled whether ENDA would contain Title VII’s motivating-factor language. The version of the bill introduced in the House of Representatives does not contain an equivalent to the Civil Rights Act’s “motivating factor” test, and in the absence of such language, the Supreme Court has consistently interpreted “because of” as requiring strict but-for causation.160 By contrast, the Senate version of the bill includes Title VII’s motivating-factor and same-decision tests.161 Until legislation is passed, it is impossible to know which version will be written into law. Nevertheless, while the impact of this split could be substantial for plaintiffs like Ann Hopkins, whose employers used both proper and discriminatory evaluative tools,162 courts still generally read this language to require group disparities in treatment in every context but prescriptive sex stereotyping. The motivating-factor provisions of Title VII recognize discrimination without a difference in outcome at the individual level. However, outside of cases that challenge gender conformity demands and are thus understood as falling under Price Waterhouse, this form of discrimination is still defined as a preference for one group of workers over another, and it is still expected to be shown through comparisons between members of different groups.163 Therefore, ENDA would provide a group of potential plaintiffs who currently only have Price Waterhouse claims with the sorts of narrow provisions found elsewhere in Title VII. This makes sexual orientation and gender identity discrimination an excellent test case for understanding how Price Waterhouse builds on traditional understandings of discrimination. Sex stereotyping and traditional protections address fundamentally different types of bias, and they can work together to benefit plaintiffs who have been woefully underserved thus far.
Instead of considering ENDA as an improvement on, or substitute for, stereotyping protections,164 this section considers how one can fill gaps left behind by the other. A traditional discrimination framework can offer superior protection in many circumstances.165 However, ENDA itself would be unable to address certain kinds of antigay workplace bias, just as narrow readings of Title VII fail to address such bias in other contexts. Here, I consider two types of plaintiffs who, although suffering discrimination that is immediately cognizable as antigay, would be unable to find relief under ENDA. I then discuss how Price Waterhouse might offer these plaintiffs a better chance at relief. These types of plaintiffs are offered as examples. While they represent only one part of the Price Waterhouse litigation landscape, considering how these plaintiffs would be treated under ENDA and Price Waterhouse shows how the theories of discrimination accepted in sex stereotyping jurisprudencecan address stubborn, broader problems in antidiscrimination law.
One useful example of the way in which straightforward sexual-orientation discrimination claims fail in cases where sex stereotyping would succeed is that of the “gayer” plaintiff—in other words, of an LGBT person who is treated worse than another employee of the same sexual orientation who behaves in such a way as to deflect attention from her status.166 Probably the most obvious example is that of a particularly flamboyant gay male employee, or a lesbian with a short haircut, but other cases raise a broadly similar problem. Consider the case of Shahar v. Bowers, in which a lesbian employee of the state of Georgia had her job offer revoked after she married another woman.167 In response to Shahar’s allegation that the adverse action was motivated by mere animus against homosexuals and thus violated the Equal Protection Clause under Romer v. Evans,168 Judge Edmondson held that “[c]onsidering . . . public reaction to a future Staff Attorney’s conduct in taking part in a same-sex ‘wedding’ and subsequent ‘marriage’ is not the same kind of decision as an across-the-board denial of legal protection to a group because of their condition, that is, sexual orientation or preference.”169
In other words, Bowers’s actions were not discriminatory because they were not “across-the-board”; since these actions only disadvantaged individual homosexuals, not homosexuals as a class, they could not constitute antigay discrimination.170 In other words, because of intragroup differentiation,171 or differences in treatment across out-group members, Shahar could not demonstrate legally cognizable discrimination. It should be clear that this problem would vanish if Shahar’s potential employer used ascriptive rather than prescriptive stereotyping to motivate his decisionmaking. If Shahar’s job offer were rescinded based on an ascriptive stereotype that all gay employees were lazy or deceitful, then there would be clear differences between the treatment of gay and straight employees generally that could support a claim. It is this difference that would doom a theoretical ENDA suit—an employer who is offended by an employee’s same-sex marriage, and fires her after the wedding, does not commit “sexual orientation” discrimination as traditionally understood. After all, the employee did not suddenly become a lesbian,172 and the employer can point to an openly gay worker (the plaintiff herself, prior to the wedding) who did not face the same action. Generally speaking, such a comparator kills an employment claim: if similarly situated employees who share the employee’s protected status are treated better, then courts assume that the difference must be something other than status.173 Courts make this assumption even though the adverse employment action reveals what would commonly be perceived as discriminatory animus towards the group generally, as well as a preference for members of that group who work to minimize their perceived affiliation therewith. Theoretically, in order for employees to claim sexual orientation discrimination in these circumstances, they would have to claim that whatever behavior led to the adverse action (flamboyance, a marriage, or perhaps “gay” behavior outside of the office)174 actually constituted part of their sexual orientation.
That said, courts have generally frowned on attempts to read conduct that is neither universal to a group nor limited to its members as functionally equivalent to a protected Title VII status—one case that sheds light on how such claims might fare under ENDA, and on the broader impact of Title VII’s discomfort with intragroup differentiation, is Rogers v. American Airlines, Inc.175 Renee Rogers challenged “a grooming policy of the defendant American Airlines that prohibits employees in certain employment categories from wearing an all-braided hairstyle.”176 Notably, Rogers did not claim that the no-braids policy had a disparate impact on African-Americans due to biological difference or even due to disparate hairstyle preferences among members of different races.177 Instead, she claimed that her hairstyle was itself part of her race.178 This is the same type of claim that flamboyant gay plaintiffs alleging discrimination under traditional doctrines would have to raise—that discrimination based on certain kinds of conduct is discrimination based on a protected status, even if this conduct is not universal to members of the protected status category—and it failed in Rogers for the same reason. Judge Sofaer critiqued Rogers’s case on several grounds, but two are relevant here: Rogers’s hairstyle was not “immutable,”179 and white women could also engage in the same practice.180 Both of these critiques apply to flamboyant LGBT plaintiffs; the behavior giving rise to discrimination is neither immutable (one can avoid getting married, speak with a deeper voice, or discuss different topics at work), nor coterminous with membership in the protected class.181 Furthermore, both of these claims again reflect the difficulty of applying Title VII to prescriptive stereotype. Here, because the prescription against traditionally African-American hairstyles did not affect all black women, and did affect some white women, the judge’s opinion did not take full account of the biased thinking it might reflect. Rogers is still good law, and a good normative case can be made that it should be: as Richard Ford notes in his seminal book on race and culture, a black woman who does not understand her identity as requiring braids would find that Rogers’s theory “not only hinders her . . . but it also adds insult to injury by proclaiming that cornrows are her cultural essence.”182 Nevertheless, from a litigation perspective, Rogers and the principle it stands for—that employers can punish behaviors “socioculturally associated with”183 a protected status without discriminating—hamstring employees who face prescriptive workplace norms that express a clear preference for practices commonly associated with one group over those associated with another, but that nevertheless remain formally status-blind.184 Anecdotally, the Rogers principle is a frequent hurdle for plaintiffs in states that offer sexual orientation protection based on the traditional model. Jessie Weber, a civil rights attorney working in Maryland, describes how defendants accused of antigay harassment frequently show a preference for gender-conforming LGB employees, and use this difference in treatment to show that their discrimination was not based on “sexual orientation” in a legally cognizable fashion, even if the harassment itself consisted of antigay speech.185 In these cases, while the discrimination clearly arises from antigay bias, the employer’s preference for gender-conforming, “less gay” coworkers makes the case unintelligible without a Price Waterhouse framework.
But is that such a bad thing? As long as it is possible for the victim of discrimination to conform to her employer’s expectations—as long as some gay people can function in the workplace, or black employees can succeed at the cost of a hairstyle—should we worry about these expectations?186 It is clear that the current regime, which protects stereotypically “gay” conduct without protecting LGBT status, raises serious problems. Soucek is undoubtedly correct that sex stereotyping, by protecting only employees who combine LGBT status with other forms of gender-deviant flamboyance, can encourage a problematic form of “reverse covering.”187 However, his proposed remedy—conceiving of antigay discrimination solely in the narrower terms found elsewhere in Title VII—leads to a different and equally problematic result. By penalizing gay employees whose gender expression subjects them to greater antigay animus than gender-conforming coworkers, traditional understandings of antidiscrimination law simply ratify the sorts of prescriptions that were condoned by the Rogers court and are still widely accepted in other contexts. By contrast, Price Waterhouse allows plaintiffs to attack these prescriptions directly. In doing so, Price Waterhouse doctrine responds effectively to two very different normative critiques of workplace culture.
On one hand, demands that employees act in a way that is stereotypically associated with their status (such as those made of Ann Hopkins or Isabell Slack) impose difference on workers who simply want to engage in the same behavior as everybody else, and do so in a way that obviously falls disproportionately on members of the stereotyped group.188 On the other, demands that employees not engage in behavior that they associate with their own culture (such as those made of Renee Rogers) may be a simple reflection of egalitarian professionalism demands,189 but may also reflect an employer’s distaste for what is perceived as minority expression,190 and employees are then forced to conform their behavior to that distaste. These critiques may seem diametrically opposed, but they both arise from the same basic problem—employers are evaluating employees not according to any job-relevant criterion, but according to discriminatory ideas about how men and women should behave, or how “black” hairstyles look in an office setting. The concern here is not in how these prescriptions apply to members of different groups, but in the ideas motivating the prescriptions themselves. Regardless of whether one is worried about stereotypes creating difference, on the one hand, or erasing it, on the other, Price Waterhouse allows plaintiffs to question the reasoning behind policies asking them to conform their behavior to sexist or racist ideas.
These biased conformity demands themselves run deeply counter to the antisubordination norms animating American antidiscrimination law. In their seminal article about performing minority identity in the workforce, Devon Carbado and Mitu Gulati discuss some of the normative problems posed by a workplace that conditions acceptance of minority employees on their adherence to majority norms.191 Carbado and Gulati note the disparate psychic harm of this adherence; minority employees may find assimilating to the norm, even if theoretically possible, to be draining or alienating in a way that nonminority employees do not.192 This can either be understood as an objectively heavier burden—actions which would seem innocuous when performed by a white woman might seem more troubling when performed by a black male employee193—or, alternatively, as a neutral norm that becomes heavier on minority employees due to different norms governing off-work behavior.194 Returning to the case of LGBT discrimination, although many straight employees who engage in flamboyant behavior are simply treated as gay,195 a culture of adherence to traditional heterosexual norms may be more troubling to employees who are used to different forms of interaction elsewhere in their lives, or whose deviance is more salient due to their LGBT status. Furthermore, this environment provides cover for more straightforward discrimination, and employees who attempt to conform to the prescriptions governing their workplace may find themselves powerless to attack that discrimination in court.196 Punishing only employers who discriminate against an entire class, and not just against individuals who exhibit behaviors associated with that class, allows employers to continue evaluating workers according to bigoted criteria.197 Given the Civil Rights Act of 1991’s punishment of bigoted decisionmaking processes, even if these processes do not lead to discriminatory actions as traditionally understood,198 it is difficult to see how employers who punish LGBT plaintiffs for failing to adhere to a norm that is itself based on discriminatory ideas should escape liability.199
Price Waterhouse can provide this sort of protection. Unlike traditional discrimination claims, which presume a plaintiff’s membership in a large, undifferentiated group and a defendant’s discrimination against that group,200 claims of prescriptive sex stereotyping assume that each employee is being evaluated individually,and then assert that the criteria used in that evaluation reflect impermissible bias. The difference is subtle, but important for resolving the intragroup differentiation puzzle. Instead of having to show that all LGBT workers were treated poorly in comparison to heterosexuals, the plaintiff can, under a Price Waterhouse theory, isolate the specific conduct that she believes led to her firing and demonstrate its importance by comparing herself to other workers (homosexual or otherwise) who acted differently. If the plaintiff can successfully demonstrate that the defendant was motivated by his dislike of these behaviors, the question then becomes whether that dislike was motivated by discriminatory ideas about how different sexes should behave. This is a hard question for plaintiffs to answer, but it is a fair question for courts to ask, and it gets at the heart of what makes sex stereotyping so pernicious. Price Waterhouse claims, by focusing on specific behaviors rather than group identification, allow courts to reach subtler and more individuated forms of sex, and sexual orientation, discrimination.
Furthermore, the problems Price Waterhouse addresses are hardly gay-specific. Inasmuch as any Title VII identity forms a mix of immutable status (I am Jewish because my parents are Jewish, my appearance is Jewish, and my genetic background is Ashkenazi) and voluntary performance (I am Jewish because I choose not to work on the Sabbath, or to fast on Yom Kippur, or even to use Yiddish expressions), Price Waterhouse’s consideration of dislike becomes critical. A workplace where an employee can be fired for engaging in any sort of discretionary cultural performance is a workplace without real diversity, and Price Waterhouse offers a conceptual framework through which courts can address these coercive norms both in and out of the sexual-orientation context.
Price Waterhouse doctrine also improves on traditional discrimination theories in dealing with non-group-identified plaintiffs who nevertheless face harassment based on group bias. In the sex-stereotyping context, this issue is usually raised by heterosexual plaintiffs who face antigay harassment due to their perceived gender deviance. Consider the plaintiffs in Doe by Doe v. City of Belleville,201 whose case appears tailor-made for a straightforward claim of antigay harassment. Although the plaintiffs identified as straight, one was repeatedly called “fag” and “queer,” and the defendants intimated that the plaintiffs were having a homosexual relationship and belonged in “San Francisco with the rest of the queers.”202 These statements look like plain accusations of homosexuality—if ENDA protects plaintiffs on the basis of “actual or perceived” sexual orientation,203 the case should lie. But here, it is not at all clear that the plaintiffs were perceived as homosexual. In one telling incident, a coworker joked that one of the plaintiffs had spread poison ivy to the other through anal sex; other coworkers responded “that if that were the case, then Dawe [the man who had made the original remark] must have contracted a rash as well, since he was always taking H. ‘out to the woods.’”204 Taken literally, Dawe is also being perceived as gay. Nowhere in the case does it suggest that Dawe was gay or faced homophobic harassment; on the contrary, these imputations of homosexual orientation and conduct are clearly, simply, metaphors.205
The plaintiff’s burden is thus made insurmountable. According to the traditional framework of a Title VII claim, in order to make a prima faciecase of employment discrimination, an employee must demonstrate, among other things, that he or she is a member of a protected class.206 Since these plaintiffs would be unable to prove actual homosexuality or bisexuality, they would need to show that they were perceived as homosexual or bisexual; if their harassers were accusing each other of engaging in homosexual sex without actually perceiving themselves to be gay, then it would be difficult for a plaintiff to prove that similar taunts reflected perceived homosexuality when aimed at them. Even if a defendant actually did view his victim as gay, it is unclear how a plaintiff could show that; jokes about a plaintiff’s homosexuality could not support an inference of perceived homosexuality if similar jokes were made about individuals who were definitely not seen as gay, and threats of homosexual sexual assault would, if anything, support a traditional sexual harassment claim.207 Without Price Waterhouse, these employees would fall into a doctrinal black hole; even under ENDA, men who were effeminate enough to be accused of being gay, but who were nonetheless never believed to be gay, would receive no protection (absent actual or threatened sexual violence).208
In these circumstances, ENDA would offer no redress to employees who have suffered behavior that would strike the average observer as not only offensive209 but discriminatory. However, the reason ENDA would fail is that the discrimination at issue is not antigay in any traditional sense. These employees are not perceived as homosexual men—Kevin McWilliams’s coworkers told him that “[t]he only woman you could get is one who is deaf, dumb, and blind”210—but as failed men; the sexual violence and taunting described in Belleville and McWilliams follow from that initial perception. In each of these cases, the plaintiffs seem to have been evaluated on their adherence to masculine prescriptive norms: one of the plaintiffs in Belleville wore an earring, the other was ridiculed for physical weakness, and Kevin McWilliams suffered from cognitive and behavioral disabilities.211 Again, because the discrimination is focused on individuated perception and evaluation, as opposed to stereotypes about how men as a class actually behave, traditional interpretations of Title VII cannot grasp the biased heuristics that led to these plaintiffs’ harassment. The bias becomes apparent, however, upon even a cursory search. This interplay between gender deviance and presumed homosexuality is perhaps more obvious in cases involving masculine women; women who violate sex stereotypes are often accused of homosexuality, and courts understand these accusations to arise from sex rather than sexual orientation discrimination.212 Gender-deviant heterosexual men, however, face aproblem—if their harassment does not harm LGBT workers or male workers as a class, then how can it be legally cognized?
Courts in these cases can, and should, rely on sex-stereotyping doctrine. Assuming that the imputation of homosexuality was triggered by a perceived deviance from prescriptive sex stereotypes, plaintiffs can allege Price Waterhouse discrimination regardless of whether they were actually perceived as homosexual or merely accused of being so. While gay slurs could still be used to show that the harassment created “an objectively hostile or abusive work environment,”213a sex-stereotyping claim deemphasizes the plaintiff’s perceived membership in a victimized group and focuses on the harasser’s bigoted evaluation.
Once again, case law at the state level shows the importance of sex stereotyping for plaintiffs who face the language of antigay harassment without identifying or being perceived as gay. The case that originated sex-stereotyping doctrine in New Jersey, Zalewski v. Overlook Hospital,214 presented a male-on-male sexual harassment claim with no evidence of perceived homosexuality;215 the court adopted a Price Waterhouse theory of sex stereotyping precisely because the remedy for sexual orientation discrimination was not available.216 DePiano v. Atlantic County217 makes the distinction between sex stereotyping and antigay harassment even more explicit, as it comes after New Jersey amended its Law Against Discrimination (LAD) to include sexual orientation. Atlantic County argued that the plaintiff, a cross-dressing heterosexual male,218 could not prevail under LAD without evidence that he was actually perceived to be homosexual or bisexual.219 The court, however, denied the defendant’s motion for summary judgment and accepted DePiano’s claim that “under the LAD’s prohibition of discrimination on the basis of sex a plaintiff need not be perceived as homosexual. Rather . . . a plaintiff may bring an LAD claim for discrimination or harassment based on gender role stereotyping.”220 In cases where insisting on a sexual orientation discrimination theory would present an unacceptable evidentiary burden, courts accept sex-stereotyping claims instead. Were an ENDA to pass, it is easy to see how this application of the Price Waterhouse doctrine could—and should—survive.
But what does this teach us about Price Waterhouse’s applicability beyond sexual orientation? In some ways, not much—sexual orientation is arguably unique in how it serves as a metonym for broader gender transgressions. The problems that sex stereotyping addresses in New Jersey do not come up in other contexts. For example, an employee who does not participate in racist workplace norms will not suddenly be considered African-American. However, this point could also be framed as an argument for Price Waterhouse’s greater relevance outside of the sexual orientation context. Workplaces that police gender norms punish outlaws using the language of antigay discrimination, leaving themselves at least somewhat vulnerable under traditional frameworks. What about, say, a white employee who is punished for supporting black civil rights?221 That employee has no hope of a standard claim under McDonnell Douglas, but Noah Zatz’s argument that employees who refuse to endorse a workplace’s discriminatory action face “a form of stereotyping especially repugnant to Title VII values” seems particularly on point here.222 Cross-status comparators get this employee nowhere, since this employer is unlikely to treat black employees any better and will probably be much friendlier to racist white workers. Nonetheless, if the employee can demonstrate that her support for civil rights caused the adverse employment action (quite likely through comparison to other white employees with different racial attitudes),223 Price Waterhouse would permit relief, as long as the prescriptive stereotype that employees should not support civil rights reflected racist thought.224 The question of whether a given action reflects discriminatory thought seems like the better question to ask, both analytically (since such a theory better reflects what is actually occurring) and normatively. Discursively imposing homosexuality on gender-deviant employees sends disturbing and subordinating messages: homosexuality is effeminacy,225 effeminacy is failure,226 and effeminacy is weakness.227 The fact that these ideas are expressed through harassment of straight employees does not make them less harmful, and sex stereotyping allows us to pierce through the irrelevant distraction of perceived group membership to see the degrading force of the norms at work.
Sex stereotypingis a nimble doctrine—its ability to look past groups to find contextual bias makes it a potentially valuable tool to fight discriminatory dynamics in the modern workplace. The following Part considers in more detail how sex stereotyping can interact with federal or state ENDAs to protect workers more effectively.
To date, the debate over how to best protect LGBT workers has been markedly zero-sum; scholars and advocates urging the passage of ENDA have treated it as either superseding a flawed substitute doctrine,228or offering protections in a field where none currently exist.229 This need not be so. A federal ENDA can work with sex-stereotyping protections to fight workplace bias more effectively than either doctrine could alone, and the normative considerations underlying both theories of discrimination can function in harmony.
To begin with the obvious point, many cases of discrimination really are so straightforward that ENDA uncontroversially applies. For example, the recent New York decision Salemi v. Gloria’s Tribeca, Inc.230 held that an employer who repeatedly referred to homosexuality as a “sin” and said that homosexuals “were going to hell”231 discriminated on the basis of sexual orientation under New York City’s Human Rights Law. Here, the employer’s statements refer to gay people as a single undifferentiated class, and the animus they reflect applies equally to all members of that class. There is no need to interrogate the prescriptions that underlie the statements, because they so clearly harm gay people based solely upon their sexual orientation and thus constitute sexual orientation discrimination under even restrictive readings of antidiscrimination law. This employer clearly harassed gay employees “because of” sexual orientation, no matter how one interprets the phrase.
While the interpretation of Price Waterhouse urged by the EEOC and adopted in Terveer could still cognize this behavior as harmful stereotyping,232 it would do so using a framework that perfectly mimics traditional Title VII litigation—in order to establish what particular trait led to the adverse action, the plaintiff would still need to show the employer’s animus towards homosexuals. However, folding sexual orientation entirely into sex-stereotyping jurisprudence would interpose an extra step, requiring each plaintiff to argue, once again, that antigay bias reflects harmful ideas about how men and women should behave. Furthermore, treating all gay plaintiffs as gender-deviant imposes a real, if subtle, dignitary harm. Plaintiffs who perceive themselves as gender-conforming should be able to claim protections without using a framework many feel reifies offensive stereotypes.233 Finally, and not inconsequentially, proving sex stereotyping is not cheap. Price Waterhouse cases often rely on expert testimony to show the gendered nature of the employer’s behavior;234 by comparison, if an employer makes blatantly antigay remarks or fires a worker immediately upon their disclosure of LGBT status, then ENDA allows a far more practical route to relief. Just as Price Waterhouse provides a more effective tool for combating subtle, context-specific, and individual discrimination, traditional protections have real power against open, class-based animus.
One question, however, remains: if ENDA’s “because of” provisions and Price Waterhouse each attack different kinds of bias, then how can they interact? Paradoxically, one of ENDA’s most immediate effects would be radically strengthening sex-stereotyping protections—as discussed above, one of the biggest hurdles facing Price Waterhouse plaintiffs is Congress’s perceived intent not to protect against sexual orientation discrimination.235 ENDA would remove this barrier and allow gender-deviant LGBT plaintiffs to bring sex stereotyping and ENDA claims concurrently. In practice, these claims are often most powerful together; defendants frequently claim that adverse actions were based on sexual orientation in response to a sex stereotyping claim,236 or defend against a state law sexual-orientation claim by citing the employee’s gender-deviant behavior or manner of dress as disruptive or upsetting to others.237 As long as only one set of protections is in place, employers will claim that behavior that would violate the other constituted a legitimate, nondiscriminatory reason for the adverse action.238 Both theories work in concert, however, to protect employees who suffer from antigay discrimination as it frequently appears in the modern workplace: as an inchoate mix of antigay bias and heterosexist ideas about how men and women should behave. ENDA works within a legal framework designed to equalize outcomes across easily identifiable groups, whereas sex stereotyping seeks to ensure that the workplace is free of subtler, more contextual bias. In practice, these approaches often go hand in hand. Allowing plaintiffs to present evidence of both theories of discrimination gives courts better information, allows plaintiffs the dignity of having their entire story heard in court, and prevents biased employers from hiding behind mutually exclusive defenses.
Another potential avenue for stereotyping suits is through ENDA itself, using either the motivating-factor provisions set out in the Senate version of ENDA239 or the new gender-identity protections common to both.240 I hope that I have shown how an LGBT plaintiff could bring a hypothetical “sexual-orientation-stereotyping” claim if that survives, but the gender-identity provisions of ENDA are novel and deserve a closer look. While ENDA’s prohibition on firing or harassment “because of . . . gender identity”241 is generally considered a remedy for a discrete group of plaintiffs who identify as transgender,242 the statutory definition of “gender identity” is sufficiently broad to encompass something superficially very much like Price Waterhouse protection. Specifically, ENDA defines gender identity as “the gender-related identity, appearance, or mannerisms or other gender-related characteristics of an individual, with or without regard to the individual’s designated sex at birth.”243 ENDA gender-identity protections could pose many of the same problems that Price Waterhouse was designed to address—plaintiffs who face discrimination on the basis of gender-deviant mannerisms will likely be unable to show their membership in a broad enough category to generate the sort of comparative evidence that courts typically require. Instead, they would have to show the salience of a particular mannerism in their own firing and then prove, at a second stage, that the mannerism in question formed a part of their gender identity.
Such a process seems, at first, like a perfect analogy to a Price Waterhouse stereotyping claim—it would certainly seem to offer relief to plaintiffs like Brian Prowel, whose behaviors were avowedly effeminate,244 or even Gregory DePiano, whose deviance from the norm of his workplace had a clear gendered component.245 But what about plaintiff H. in Doe by Doe v. City of Belleville, who may have been expressing any number of cultural commitments by wearing an earring;246 or Mark McWilliams, who was harassed because of a developmental disability;247 or Joseph Oncale, whose gender deviance similarly existed only in the minds of his tormentors?248
This sex-stereotyping theory looks, instead, very much like the alternate proposal made recently by Zachary Kramer, who proposes a reasonable-accommodation model built on Title VII’s current protections for religious practices. Under Kramer’s theory, a plaintiff would be expected to show that a given mannerism or appearance reflected his or her sincere gender commitments and that a defendant was unwilling to reasonably accommodate those commitments.249 The difference between Kramer’s proposed solution and the one that already exists in Price Waterhouse is the fundamental difference between a Title VII that protects gender-deviant plaintiffs and a Title VII that prohibits gender-based evaluation. Claims proceeding under Kramer’s accommodation model must begin by showing that the plaintiff’s idiosyncrasy is both “sex based” and “sincerely held.”250 Under this approach, plaintiffs like Doe, McWilliams, and Oncale would be unable to show that they viewed the traits that led to their harassment as “sex based,” and thus to make out even a prima facie claim.
Of course, this could be the right answer—such an approach certainly ensures that claims will only be raised by individuals whose sincere gender commitments stand sufficiently far outside the norm to require accommodation. If we read Title VII as Justice O’Connor did in Price Waterhouse, as concerned with the protection of out-group members and not with the tools employers use more generally, then Kramer has the right solution; Congress cares about discrimination only when it directly impacts a member of a protected group. But Congress overrode Justice O’Connor. The Civil Rights Act of 1991 reveals a broader commitment—to stamp out bias and bigotry in the workplace no matter whom it affects. Stereotypes can be deeply subordinating towards disfavored groups even when applied to group nonmembers—as long as an employee is harassed based on outdated and offensive ideas about sex, gender, race, or any other Title VII category, there is no need to bother with the secondary inquiry of the employee’s own commitments.251 LGBT plaintiffs may succeed under Kramer’s theory—they may be able to demonstrate that their disfavored traits arise from a disfavored gender identity. But, as Jessie Weber put it to me, “the beauty of the sex stereotyping claim is that it applies to everyone.”252
It is impossible to overstate the impact of Price Waterhouse on the American workplace, but stating it accurately isn’t much easier. Statistics on unreported discrimination suits are notoriously unreliable, to say nothing of the claims settled before an EEOC inquiry, or the changes that workplaces make in order to prevent violations from occurring. Nevertheless, the conceptual changes wrought by Price Waterhouse and the Civil Rights Act of 1991 are tremendous in scope, and they offer a vision of antidiscrimination law that courts still only partially recognize. Because federal antidiscrimination law is formally blind to sexual orientation and gender identity, courts have been unable to use traditional tools of group protection to justify protecting LGBT workers. Instead, Price Waterhouse demands that judges interrogate the cause or purposeof employers’ differentiation between their employees.253 It questions whether the heuristics that employers use to evaluate their employees themselves reflect impermissible bias. This doctrine has bloomed from arid soil; both Price Waterhouse and its progeny dealing with anti-LGBT discrimination arose from courts seeing clear discrimination without a clear statutory solution, and the analytic moves underlying prescriptive sex stereotype show remarkable sensitivity to the nature of the modern workplace.
In particular, what Price Waterhouse can acknowledge, and what traditional group-based remedies do not specifically address, is the extent to which oppressive stereotypes are mala in se, regardless of the disparate group outcomes they might or might not create. These stereotypes hurt everybody. Irrational and degrading norms about how men and women should behave, how different races should interact, and what cultural practices are fit for professional life limit everyone’s identity, not just members of subordinated groups.254 Price Waterhouse allows plaintiffs to focus on these harmful norms, but it is currently available only in the limited sphere of gender deviance. Sex stereotyping is a valuable part of our Title VII jurisprudence and should survive the passage of traditional sexual orientation and gender identity protections. More than that, however, the ideas that Price Waterhouse espouses should be better understood and embraced far, far more broadly. In an often hostile world, Price Waterhouse has given gender-deviant plaintiffs a strange and precious gift. Now it is time to share.