Exceptional Proff 600

ARTICLE: THE NEW SEX DISCRIMINATION

January, 2014

Reporter 63 Duke L.J. 891 * Length: 28939 words Author: Zachary A. Kramer+

Copyright © 2014 Zachary A. Kramer.

+ Professor of Law, Arizona State University, Sandra Day O'Connor College of Law. For helpful comments and conversations along the way, thanks to Paul Berman, Luke Boso, Karen Bradshaw Schulz, Erin Buzuvis, Adam Chodorow, Laura Clymer, Rhiannon Cornelius, David DePianto, Aaron Fellmeth, David Gartner, Liz Glazer, Zach Gubler, Andy Hessick, Carissa Hessick, Art Hinshaw, Lily Kahng, Marcy Karin, Tim Koch, Liz Kukura, Rhett Larson, Patrick Luff, Roger Perlstadt, Mary Sigler, Judy Stinson, Doug Sylvester, and Jim Weinstein. The Article benefited immensely from comments during presentations at the Williams Institute's Work-in-Progress Series at UCLA School of Law and ASU's junior faculty colloquium. Western New England University School of Law graciously invited me to present an early version of this Article as the Clason Lecture in the spring of 2012, and I am grateful to their faculty for their hospitality and for their tremendously useful feedback on my project. I must also thank Chris Bryant, Whitney Bosworth, and the rest of the Duke Law Journal's editorial staff for their diligent work on this Article. Finally, I am especially thankful to Craig B. Haircuth for his remarkable research assistance. Any and all mistakes are, of course, my own.

Highlight

Abstract

Sex discrimination law has not kept pace with the lived experience of discrimination. In the early years of Title VII of the Civil Rights Act of 1964, courts settled on an idea of what sex discrimination looks like - formal practices that exclude employees based on their group membership. The problem is that sex discrimination has become highly individualized. Modern sex discrimination does not target all men or all women, nor does it target subgroups of men or women. The victims of modern sex discrimination are particular men and women who face discrimination because they do not or cannot conform to the norms of the workplace. These employees have been shut out of a sex discrimination regime that still expects employees to anchor their claims to a narrative of group subordination.

I argue that the lived experience of discrimination should determine employment discrimination doctrine and not the other way around. Accordingly, I propose a new regime for sex discrimination law. The model for the new sex discrimination regime is religious discrimination law. Unlike other areas of employment discrimination law, religious discrimination law offers a dynamic conception of identity and a greater array of different theories of discrimination. I argue that sex discrimination law can and should work this way, too.

On a broader level, the paper makes a strong normative claim about the substance of Title VII's sex equality project. I argue that sex discrimination law needs to recalibrate its vision of equality. Difference is universal. No two women (or men) are the same, and this is a good thing. Thus the central task of sex discrimination law should be to better recognize - and, in turn, protect - the distinctive ways in which employees express their maleness and femaleness. It is these differences, after all, that shape the way employees experience modern sex discrimination.

Text

 [*893] 

Introduction

In 2000, Darlene Jespersen lost a job that she loved.  1  For over twenty years, Jespersen worked as a bartender at Harrah's Casino in Reno, Nevada.  2  During her career at Harrah's, she accumulated a strong work record, earning the praise of coworkers and customers alike.  3  Things went downhill for Jespersen, however, when Harrah's adopted a new appearance code for its beverage service personnel.  4  The new policy - which Harrah's called the "Personal Best" program - required female employees to, among other things, wear makeup during their shifts.  5  The makeup requirement posed a problem for Jespersen, as she never wore makeup on or off the job.  6  Jespersen explained that wearing makeup made her feel "ill," "degraded," "exposed," and "violated."  7  She felt that makeup robbed her of her "credibility as an individual and as a person," so much so that she could not do her job well if forced to wear makeup during her shifts.  8

Despite her many years of service and her exemplary record, Harrah's would not budge on the makeup requirement, and Jespersen soon found herself out of a job.  9  She responded by bringing a sex discrimination claim under Title VII of the Civil Rights Act of 1964 (Title VII),  10  alleging that Harrah's "Personal Best" program - specifically the makeup requirement - amounted to unlawful sex discrimination.  11  Jespersen's theory of discrimination was that the makeup requirement compelled female employees to conform to a   [*894]   stereotypical standard of femininity,  12  a variation on the gender-stereotyping theory developed by the Supreme Court in Price Waterhouse v. Hopkins.  13  The thrust of the gender-stereotyping theory is, in the words of the Court, that an employer cannot "evaluate employees by assuming or insisting that they matched the stereotype associated with their group."  14

In the end, Jespersen could not sustain her stereotyping claim against Harrah's. The Ninth Circuit concluded that her injuries were too subjective to support her gender-stereotyping claim. According to the court, "The record contains nothing to suggest the grooming standards would objectively inhibit a woman's ability to do the job."  15  The court made clear that, as a general matter, an employee can challenge a grooming code on gender-stereotyping grounds  16  but that the "subjective reaction of a single employee" is not sufficient to support such a claim.  17

There is a coming crisis in sex discrimination law, and employees like Darlene Jespersen are at the center of it.  18  Sex discrimination law has not kept pace with the lived experience of sex discrimination. When Title VII became law, most instances of sex discrimination involved overt discrimination that differentiated between men and women, almost always to the detriment of female employees.  19  And it was not uncommon for employers to justify a discriminatory practice by appealing to perceived or stereotypical differences between men and women. Consider a prominent example. The Water and Power Department for the City of Los Angeles administered a retirement benefits system for its employees, which it funded, in part, through   [*895]   employee contributions.  20  The Department required female employees to contribute greater monthly payments to its pension fund than male employees, which meant that female employees took home less pay than their male coworkers.  21  The Department's reason for making women contribute more than men was that, on average, women tend to live longer than men.  22  The discrimination was formal in nature, targeting women as a group.

The problem is that sex discrimination looks very different today. Sex discrimination has become highly individualized. Modern sex discrimination does not target all men or all women, nor does it target subgroups of men or women - such as women who are aggressive and men who are effeminate.  23  The victims of modern sex discrimination are particular men and women who face discrimination because they do not or cannot conform to the norms of the workplace.  24  In addition to Darlene Jespersen, it is the male truck driver who wears women's clothing;  25  it is the bus driver who cannot find a bathroom to use while she is transitioning from male to female;  26  it is the effeminate man who sticks out like a sore thumb in a rural Wisconsin factory;  27  it is the new mother who needs extra breaks during the workday to pump milk for her newborn baby;  28  it is the hairstylist who is fired from her salon because she is a butch lesbian;  29  and it is the overweight telemarketer who is told she is not pretty enough for a face-to-face sales position.  30  Time and time again, these employees, often the most marginalized employees in their respective workplaces, have struggled to find their footing in a sex   [*896]   discrimination regime that still views the group as the focal point of discrimination.

This Article initiates a conversation about the future of sex discrimination. It lays the foundation for a new sex discrimination regime that is tailored to the lived experience of sex discrimination as it exists today rather than as it once did. The driving force behind my argument is the idea that, when it comes to developing antidiscrimination protections, the lived experience of discrimination should determine the doctrine and not the other way around.  31  I believe that this is what Oliver Wendell Holmes Jr. had in mind when he wrote, in The Common Law, that "the life of the law has not been logic; it has been experience."  32  The toughest obstacle facing victims of modern sex discrimination is the need to anchor their discrimination claims to a narrative of group subordination - to show, in other words, that the discrimination they faced in the workplace harms their group as a whole. My goal is to imagine a sex discrimination regime that is not tethered to group subordination.

Sex discrimination law needs to move in a new direction. And religious discrimination law provides a template for how to do that. Religious discrimination occupies a special place within Title VII thanks to two interrelated doctrinal features. The first is a dynamic conception of what constitutes a religion. Like other areas of Title VII, religious discrimination protects employees against status discrimination  33  - for instance, an employee being fired for being Jewish.  34  But what separates religious discrimination from the rest of Title VII is that it also protects employees against religious-practice discrimination  35  - for instance, an employee being fired for refusing to   [*897]   work on the Sabbath.  36  What is more, religious discrimination law embraces an attitude of liberal neutrality toward the particulars of a person's religion.  37  For a belief or practice to count as a religion, all an employee needs to show is that it is religious within her own scheme of things and that it is sincerely held. The result is a body of law that recognizes a vast universe of religious practices, each as distinctive as the next.  38

The second doctrinal feature of religious discrimination law flows immediately from the dynamic conception of religion as a protected trait. In addition to the standard prohibition against disparate treatment,  39  religious discrimination law also imposes on employers a duty to reasonably accommodate an employee's religious practice.  40  As a theory of discrimination, reasonable accommodation goes further than the standard disparate treatment protection, which   [*898]   only restricts employers from introducing bias into the workplace. Reasonable accommodation, by contrast, requires employers to redesign the workplace to account for the needs of their employees, provided that doing so will not create an "undue hardship" for the employer.  41  From an organizational perspective, the advantage of an accommodation framework is that it should promote collaboration between employers and employees, the hope being that when an employee's religious practice conflicts with an employer's policy, the employee will negotiate rather than litigate to find a solution to the problem.

Taken together, these doctrinal features create a legal environment that caters to individuals rather than groups. Sex discrimination law can - and I argue that it should - work this way, too. My proposal is that the new sex discrimination law should track the doctrinal structures of religious discrimination law. First, sex discrimination law should adopt a more dynamic conception of sex as a protected trait, one that affirmatively protects sex as both a status and a practice. Of all the proposed reforms in this Article, this one should be the easiest to implement, as the gender-stereotyping theory of sex discrimination has laid the groundwork for just such a robust protection. On top of that, discrimination claimants should have greater freedom to define the nature of their sex practice, just as claimants in religious discrimination cases do.

The second major doctrinal change is that sex discrimination law should supplement disparate treatment analysis with a reasonable accommodation protection. This is admittedly a more controversial proposal, though it nevertheless dovetails nicely with a more dynamic conception of sex as a protected trait. In making this proposal, I am mindful of Professor Kenji Yoshino's warning that we should not think of reasonable accommodation as legal penicillin.  42  Reasonable accommodation is not a panacea, and I do not think it will cure all the ills of sex discrimination. But I do think that reasonable accommodation offers a better way for the law to contend with the harms of modern sex discrimination than does the legal landscape currently available to a discrimination claimant. The defining characteristic of modern sex discrimination is that it is experienced individually, and reasonable accommodation is an individualized theory of discrimination. As we look to the future of sex   [*899]   discrimination law, we need a framework that treats discrimination claimants as individuals, not as members of a group. Reasonable accommodation provides this framework.

For these doctrinal changes to take hold, we will also need to recalibrate the vision of equality that undergirds sex discrimination law. For most of its history, sex discrimination law has conceived of equality in terms of treating men and women, as groups, in the same way.  43  According to this view, equality is cast in neutral terms, providing formal protections for the sexes as compared against each other. But the new sex discrimination demands a vision of equality that is rooted in difference rather than sameness, a vision of equality that understands that no two women (and no two men) are the same. Going forward, the central task of sex discrimination law should be to better recognize - and, in turn, protect - the distinctive ways in which employees express their maleness and femaleness. It is these differences, after all, that shape the way employees experience modern sex discrimination.

The Article proceeds in four parts. The purpose of Part I is to put my argument in context. To that end, it outlines the three main theories of discrimination currently available in employment discrimination law and contrasts two competing visions of equality. From there, Part II tells the story of why employees like Darlene Jespersen have such a hard time raising actionable sex discrimination claims. More specifically, this Part develops a wide-ranging critique of existing sex discrimination law, challenging the doctrinal rules, historical fictions, and normative values that anchor the law to narratives of group subordination. As a part of that discussion, I also propose a new normative baseline for Title VII's sex equality project.  44  The thrust of this claim is that Title VII should not define sex equality in terms of what is good for all or even most women (or men), but rather in terms of what individual men and women need to flourish in the workplace.

Part III sketches a new framework for sex discrimination law. Modeled on the protections currently available to employees in religious discrimination cases, this new framework relaxes sex   [*900]   discrimination law's conception of identity, making the law more responsive to the needs of employees as individuals. It also adds reasonable accommodation as a formal response to sex discrimination, which will force employers and employees to shoulder the burden, together, to accommodate differences in the workplace. Taken together, these doctrinal reforms will bring discrimination law more in line with the lived experience of discrimination as it exists today. Reimagining sex discrimination law in this way will also require that we rethink what equality means. Part III also argues that sex discrimination law needs to embrace a more holistic vision of equality, one that is sensitive to difference as well as sameness. Stepping back, Part IV considers the broader implications of my argument. The goal of this discussion is to jumpstart a conversation about the future of employment discrimination. Accordingly, Part IV raises some hard questions about where employment discrimination law has been, where it is now, and where it needs to go in the future.

I. Context

Before turning to the substance of my argument, this Part offers a brief tour of the doctrinal landscape of employment discrimination law. This discussion highlights two fundamental features of the law. First, it outlines the three main theories of discrimination that currently exist in employment discrimination law - disparate treatment, disparate impact, and reasonable accommodation. Such a discussion is helpful because my proposal, developed in detail below, is that sex discrimination law should shift from one theory of discrimination (disparate treatment) to another (reasonable accommodation). Second, it distinguishes between two competing visions of equality that undergird employment discrimination law. In terms of scope, the following discussion is purely descriptive. My goal in the immediate Section is to provide a doctrinal foundation for my normative claims about the past, present, and future of sex discrimination law.

A. Theories of Discrimination

As a body of law, employment discrimination is concerned with the "rights and responsibilities of employers and employees."  45  In   [*901]   particular, employment discrimination law seeks to regulate a small slice of employment decisions. Discrimination, in the broadest sense of the term, is essential to the organization of a workplace. Employers discriminate all the time - they make decisions about whom to hire and fire, whom to assign to certain shifts and special projects, and whom to promote and transfer. Employment discrimination law only comes into play when one of these employment decisions implicates a trait that is protected by statute.  46  The primary task of employment discrimination law, then, is to determine if an employment decision is based on one of these prohibited traits. 47

To facilitate this inquiry, employment discrimination law distinguishes between three theories of discrimination - disparate treatment, disparate impact, and reasonable accommodation. These theories are mutually exclusive, with each responding to a distinct experience of discrimination. Disparate treatment governs a situation in which an employer treats employees differently on account of a protected trait. Disparate impact addresses cases of unintentional discrimination, in which a nondiscriminatory rule has a discriminatory effect. And reasonable accommodation covers instances in which an employer fails to account for an employee's special needs. I consider these theories in turn.

1. Disparate Treatment. Disparate treatment is the analytical backbone of employment discrimination law. The theory rests on a principle - often referred to as an anticlassification principle - that an employer cannot make employment decisions that classify an employee (or a group of employees) on the basis of a protected trait, such as race, sex, or religion.  48  The critical feature of disparate   [*902]   treatment analysis is the causation requirement, which flows from the statutory command that prohibits discrimination "because of" a protected trait.  49  Consider an example. Say that two current employees - a man and a woman - apply for a promotion to be the new director of sales, and the employer promotes the male employee. To state a disparate treatment claim, the female employee must show that she was denied the promotion for an illegitimate reason - because she is a woman - and not for some legitimate reason, such as an inferior sales record or a spotty attendance record.

The specifics of how a discrimination claimant proves a disparate treatment claim - what kind of evidence is needed to support a claim and what evidentiary framework a fact finder will use to assess the claim - are beyond the scope of the immediate discussion.  50  But it is important to note that the touchstone of disparate treatment is discriminatory intent or motive.  51  In the previous example, the employer's decision created a discriminatory result because a man was promoted over a woman. For the female employee to state a disparate treatment claim, however, she must show that the discriminatory result was motivated by a discriminatory purpose - for instance, that the employer thought women were not cut out for managerial positions.  52

 [*903]   Under a disparate treatment rule, an employer cannot inject bias into the workplace.  53  This is what separates disparate treatment from reasonable accommodation. In the latter case, discrimination is defined in terms of whether the employer has adequately adjusted the job to satisfy the needs of individual employees. By contrast, disparate treatment does not question the way an employer organizes its workplace. In the words of Professors Pam Karlan and George Rutherglen, disparate treatment "essentially takes jobs as it finds them."  54

2. Disparate Impact. If discriminatory intent is the touchstone of disparate treatment, then discriminatory effect is the touchstone of disparate impact. Disparate impact captures unintentional discrimination, cases in which an employment policy is fair on its face but harms one group of employees more than another.  55  The least intuitive of the three theories of discrimination, disparate impact is best explained in the context of an example.

Alabama had a statute that set a minimum height and weight requirement to work in a correctional facility (five feet two, 120 pounds).  56  The female plaintiff applied for a position as a correctional counselor but was rejected because she fell short of the statute's weight requirement.  57  She challenged the policy under a disparate impact theory. At the time of the suit, women fourteen years of age or older made up 52.7 percent of the Alabama population but held only 12.9 percent of the correctional counselor positions.  58  Taken together, the height and weight requirements created a disparate impact: the rule excluded from consideration 41.1 percent of women and less than 1 percent of men.  59  Importantly, the plaintiff's claim did not allege purposeful sex segregation. Even though Alabama adopted   [*904]   the height and weight requirements because it was concerned for the safety of its correctional officers, the rule still violated Title VII.  60

The full story of disparate impact is worthy of serious consideration as we think about the future of employment discrimination law. It is not, however, integral to my goal in this Article. For this reason, the remainder of this Article is focused on disparate treatment and the third main theory of discrimination, reasonable accommodation.

3. Reasonable Accommodation. The newest of the three theories of discrimination, reasonable accommodation, emerged on the scene as part of the Rehabilitation Act of 1973,  61  the statutory precursor to the Americans with Disabilities Act (ADA).  62  Today, reasonable accommodation is a part of only a small handful of antidiscrimination statutes,  63  most notably the ADA and Title VII's religious discrimination provision.  64

The basic thrust of reasonable accommodation is that the employer must take steps to modify a job to fit the needs of a particular employee.  65  Take an example from Title VII's religious   [*905]   discrimination jurisprudence. Say that an employee, a devout Jew, is unable to work her scheduled shift on a Saturday afternoon. She explains the conflict to her employer, noting that she cannot abide by her religious beliefs if she is required to work on the Sabbath. At this point, the employer is faced with a choice: adjust the work schedule if possible, or face the threat of a lawsuit. Say that the employer cannot make the requested change to the schedule, perhaps because doing so would greatly disrupt the whole work schedule and inconvenience many employees. In that situation, the requested accommodation would pose an "undue hardship," thereby relieving the employer of any obligation to accommodate the employee.  66  By contrast, if the employer can make the change without great difficulty, the law requires the employer to reschedule the employee so as to eliminate the conflict between her work obligations and her religious practice.

Reasonable accommodation provides a more individualized protection than disparate treatment. Under a disparate treatment regime, the employer is prohibited from introducing bias into the workplace but bears no responsibility to tailor the job to the needs of its employees. Reasonable accommodation, by contrast, seeks to mold the job around the needs of the employee, when possible. Negotiation is integral to reasonable accommodation: employers and employees must come together to discuss their respective needs and expectations.  67 By fostering collaboration in this way, reasonable accommodation turns conventional antidiscrimination discourse on its head. If, as Professor Christine Jolls describes it, "the canonical idea of "antidiscrimination' in the United States condemns the differential treatment of otherwise similarly situated individuals on the basis of   [*906]   race, sex, national origin, or other protected characteristics,"  68  then reasonable accommodation marks a significant shift in the American antidiscrimination project. Specifically, it expands the idea of antidiscrimination to cover the unique ways in which individuals express their identities.

B. Visions of Equality

The fundamental goal of employment discrimination law is to promote equality in employment. Though it lies at the heart of American antidiscrimination law, equality is not a self-defining concept. Consider the following question: What does it mean to treat employees equally? How we answer this question reveals a great deal about our normative expectations of what wrongs employment discrimination law is supposed to remedy and how it should go about doing so.

1. Equality as Sameness. The dominant view of equality in employment discrimination law is based on the idea of sameness. The sameness model provides that employers should treat like people in a like manner.  69  To satisfy this equality command, an employer must act as if it cannot - or at least does not - appreciate its employees' identity traits.  70  The idea that employers should blind themselves to their employees' traits is deeply engrained in American antidiscrimination law.  71  In his celebrated work Prejudicial Appearances, Professor Robert Post seeks to uncover the hidden logic of American antidiscrimination law.  72  Post discovers that trait blindness is crucial to the inner workings of our antidiscrimination project: "Blindness," he argues, "renders forbidden characteristics   [*907]   invisible; it requires employers to base their judgments instead on the deeper and more fundamental ground of "individual merit' or "intrinsic worth.'"  73

Certain traits are designated off-limits as bases for an employment decision. The idea is that these traits - such as race, sex, and religion - are culturally salient and often trigger deeply held stereotypes and prejudices.  74  Employment law takes these traits off the table, in the hope that restricting their influence in individual cases will, in the long run, lead to wholesale social change.

2. Equality as Difference. We can contrast the sameness model of equality with the difference model of equality. Whereas the sameness model seeks to create formal equality, the difference model is interested in creating substantive equality.  75  It mandates that some people need to be treated differently to be treated equally.  76  The difference model is the animating force behind reasonable accommodation. When an employer adjusts the workplace to meet the needs of a given employee, the employer is accounting for the employee's difference - the characteristics that set the employee apart from her coworkers. Writing in the context of disability law, Professor Anna Kirkland argues that the ADA charts a new course for antidiscrimination law, specifically with respect to its conception of equality: "What is new is the turn to accommodations for difference and the acceptance that difference may be insoluble, and that ignoring it may be the height of oppression rather than the best hope for seeing past it."  77

I want to break down the difference model into three interrelated ideas. First, employees are different. There is no such thing as a truly homogenous group: no two members of a group   [*908]   express a protected trait in exactly the same way. Second, these differences matter. The experience of discrimination is all about the salience of an employee's identity traits. Those who stand out - that is, those who are different from their peers - tend to be the primary targets of employment discrimination. At the same time, the thing that marks an employee as different is often integral to the employee's sense of self. Finally, the law should protect employees because of their differences, not despite them.

C. The Claim

At this point, the necessary pieces are in place to preview my claim. In terms of restructuring sex discrimination law, I propose that sex discrimination law should take a page out of religious discrimination law's playbook. Unlike the other kinds of employment discrimination prohibited by Title VII, religious discrimination adopts a vision of equality that incorporates notions of sameness and difference, blending them together to create a more holistic equality command. It can do this because it allows employees to raise reasonable accommodation as well as disparate treatment claims. Such a framework would likewise make sense in the context of sex discrimination. Like their peers in religious discrimination cases, employees experience sex discrimination both because of who they are (sex as status) and because of how they behave (sex as practice).

This last point is important. My proposal rests on the idea that sex discrimination doctrine should reflect the experience of sex discrimination as it exists today rather than how it used to be. Modern sex discrimination is not easy. It is a highly subjective experience, often targeting a single employee. It cuts across identity traits, defying traditional notions of causation. And it is a product of work culture, burdening employees who do not fit in with their coworkers. The simple fact is that existing sex discrimination doctrine is not equipped to deal with these sorts of cases. It is time for a new sex discrimination regime.

II. Sex Discrimination and Group Narratives

When Darlene Jespersen challenged Harrah's new makeup policy, she faced an uncertain landscape. Because Jespersen's claim focused on her appearance rather than her status as a woman, it did not fit easily among existing sex discrimination norms. The canonical case of sex discrimination is formal disparate treatment, in which an   [*909]   employment policy subjects women to worse treatment than men (or vice versa). Jespersen could not allege that she was treated differently because she is a woman, as the makeup policy applied to all female employees. Nor would it have been easy for her to prove that women fared worse than men under Harrah's "Personal Best" policy, as the policy also had requirements that applied exclusively to men (for example, no makeup, no ponytails). Ultimately, what Jespersen wanted - to be free to work without having to wear makeup - is not the sort of remedy that existing sex discrimination norms can provide.

This Part tells the story of why employees like Jespersen have such a difficult time raising actionable sex discrimination claims. It starts by discussing the complicated history of Title VII's "sex" provision. Although this history is interesting in its own right, it is especially important for my purposes because courts have long used the history of the "sex" provision as a reason to narrow the reach of Title VII's sex equality project. From there, this Part considers the issue of subgroup discrimination, an experience of discrimination that has confounded the courts since the early days of Title VII. In particular, this discussion zeroes in on what has proven to be at once the most elusive and most transformative theory of subgroup discrimination: the gender-stereotyping theory of sex discrimination. I use the gender-stereotyping theory as a catalyst to rethink Title VII's sex equality project. The thrust of my argument is that, as a normative matter, Title VII should not define sex equality in terms of what is good for all or even most women (or men), but rather in terms of what individual men and women need to flourish in the workplace.

This Part concludes by considering another important gap in existing sex discrimination norms: employees who face discrimination along multiple axes of bias. As a regulatory force, sex discrimination law tends to flatten identity, forcing employees to shed part of themselves so they can squeeze into discrete identity categories. I argue that sex discrimination law should instead seek to promote a vibrant conception of identity, one that is driven not by the strictures of doctrine but by the needs of individual employees.

As a whole, this Part offers a wide-ranging critique of existing sex discrimination law, challenging the doctrinal rules, historical fictions, and normative values that anchor the law to narratives of group subordination. To develop this critique, this Part highlights the stories of victims of modern sex discrimination. Along the way, we meet transgender men and women, pregnant women, working mothers, gender-nonconforming men and women, crossdressers,   [*910]   unmarried women, lesbians, and gay men, among others. These stories draw attention to a set of doctrinal pitfalls that, taken together, substantially limit the reach of Title VII's sex equality project. These stories also show that sex discrimination law's conception of identity leaves much to be desired. More than anything, though, these stories remind us that antidiscrimination law is about righting wrongs and providing relief, both in the legal and psychological sense of the term. These employees have been fired, harassed, humiliated, and otherwise disrespected. Some struggle just to get through the day, while others lose their livelihood altogether. Given what is at stake, we owe it to them, as well as to other employees who face similar circumstances, to do the work of antidiscrimination law as best we can.

A. A Troubled History

When Congress enacted Title VII, it embarked on an unprecedented journey with respect to sex discrimination. Not only was the sex provision the first of its kind, but it was seemingly boundless.  78  After all, the text of Title VII never defines "sex" or "discrimination," and the legislative history of the sex provision leaves unanswered the central question of all of sex discrimination law:  79  What constitutes unlawful sex discrimination under Title VII?

In the story that haunts employment discrimination law, Title VII's sex provision is a quirk of history. The principal evil that the Civil Rights Act sought to remedy was, of course, race discrimination.  80  Sex did not become a part of the Civil Rights Act until late in the legislative process, when Congressman Howard Smith proposed an amendment to add "sex" as a protected trait.  81  Because the amendment came so late in the process, Congress did not hold hearings about sex equality in American society, nor did any   [*911]   committee issue a report about the scope of Smith's proposed sex amendment. Except for a few hours of floor debate about the new provision, Congress was otherwise silent on the issue of sex discrimination. The prevailing view is that Smith, a staunch opponent of civil rights,  82  proposed the amendment in the hope of killing the Civil Rights Act before it could become the law of the land.  83  Although the country may have been ambivalent about racial equality, there was no way people were going to support the Civil Rights Act if it also applied to sex, or so he thought. In the end, Smith's plan backfired. When the Civil Rights Act became law, sex was right there alongside the other protected traits.

This story is a trap. Like most good traps, it is attractive, luring unsuspecting victims into its clutches. It is hard not to be drawn into the idea that Title VII's ban on sex discrimination was never meant to be. Indeed, courts have taken the bait time and time again, relying on this story as a reason to narrow the reach of Title VII.  84  As Professor Cary Franklin writes in a recent article, "Claims about the narrow mindset and goals of the Eighty-Eighth Congress have exerted a powerful regulative influence over the interpretation of Title VII's sex provision."  85  Franklin is part of a growing chorus of scholars who have sought to correct the record on sex discrimination law's history.  86    [*912]   Though their weapons are academic in nature, their fight has considerable real-world implications. Sex discrimination law is hamstrung by the story of a political ploy gone wrong. That Smith may have introduced the amendment to thwart civil rights does not tell us anything about why legislators ultimately voted in favor of the provision.  87  And yet the story endures, fostering a narrow vision of sex equality that constrains Title VII's capacity to respond to emerging forms of sex discrimination.

This Section considers three instances in which sex discrimination's history stands in the way of sex discrimination's future. The first is discrimination against transgender employees. By their very existence, transgender employees pose a challenge to historical conceptions of sex and sex discrimination. Their identities demonstrate that sex is not immutable, and their discrimination claims suggest that sex discrimination is not as simple as treating all men and women equally. In this regard, they are the poster children for the new sex discrimination. I also consider two other instances in which sex discrimination's past inhibits its future - discrimination against pregnant women and women who are breastfeeding. Technically distinct experiences of discrimination, the pregnancy and breastfeeding cases nevertheless belong together. Not only do both revolve around motherhood, but in both instances the doctrine rests on a suspect claim about the history of sex discrimination.

1. Transgender Employees. In 1981, Eastern Airlines fired a pilot who had been with the company for over a decade.  88  The reason for the termination was that the employee had undergone sex reassignment surgery without the company's knowledge.  89  Hired in   [*913]   1968 as Kenneth Ulane, a decorated Army pilot who served in Vietnam, the pilot was fired in 1981 as Karen Frances Ulane.  90  The airline could not argue that Karen lacked the requisite experience or qualifications to fly because, despite her new appearance, Karen was the same pilot she was before the surgery. At the time, transgender discrimination was still a new issue for the courts. Yet Ulane's case presented a unique opportunity to fold transgender discrimination into sex discrimination law. After all, Ulane was her own comparator. What better example of sex discrimination than an employer who is willing to employ a man but not a woman, when the man and the woman are the same person?  91

The court did not see it that way. The Seventh Circuit held that Title VII's prohibition on sex discrimination did not reach a claim brought by a transgender person.  92  History was not on Ulane's side. According to the court, "Our responsibility is to interpret this congressional legislation and determine what Congress intended when it decided to outlaw discrimination based on sex."  93  From there, the court surveyed the thin legislative history of Title VII's sex provision.  94  It called the sex provision a "gambit" and a "ploy" designed to "scuttle the adoption of the Civil Rights Act," which ultimately led to sex being "abruptly added to the statute's prohibition against race discrimination."  95  Taken together, the "circumstances of the amendment's adoption" and its "total lack of legislative history" led the court to the following conclusion: "Congress never considered nor intended that this 1964 legislation apply to anything other than the traditional concept of sex."  96

Let us step back for a moment to consider what the court meant by the "traditional concept" of sex, as well as what it means for the   [*914]   contours of sex discrimination law.  97  Once again, Professor Cary Franklin's writing on the history of sex discrimination is instructive. Franklin argues that the idea that there is a traditional concept of sex discrimination is problematic for two reasons. First, the traditional concept of sex discrimination is an invented tradition, not one that is deeply rooted in sex discrimination law.  98  In this regard, Ulane v. Eastern Airlines, Inc.  99  is part of a long string of cases that not only refer to but also directly rely on a history that never actually occurred. Second, because it appeals to an invented tradition, the traditional concept of sex discrimination serves to narrow the scope of Title VII's sex equality project.  100  We can see this at work in Ulane's case. The court rejected an emergent manifestation of sex discrimination because it did not conform to what the court thought of as the classic manifestation of sex discrimination, namely, formal rules that treat men and women differently.  101  Thus we can think of Ulane as part of a sustained effort to weaken the normative force of sex discrimination, an effort that is as old as Title VII itself and continues to the present day.

Now turn back to the court's reasoning in Ulane. In rejecting Ulane's claim, the court cited a handful of transgender discrimination cases that likewise rejected employees' sex discrimination claims on largely historical grounds.  102  Together with Ulane, these cases form the transgender discrimination canon. The thread unifying these cases is a concern about group narratives. The reason that these employees lost their cases was because they could not map their claims onto a narrative of group subordination - they could not, in other words,   [*915]   show that the discrimination they faced was bad for all men or all women.  103  In the eyes of their respective courts, these employees faced a form of niche discrimination: most men and women do not take hormones and undergo sex reassignment surgery, nor do they dress and self-identify as a sex different than their birth sex. For these courts, transgender employees cease to be men and women; they are first and foremost transgender persons, wholly defined by their most marginalizing trait.  104 Once ascribed with the transgender label, these employees find themselves shut out of sex discrimination law.  105

There are signs that judicial attitudes toward transgender discrimination may be softening, however. In recent years, two new theories of sex discrimination have emerged for transgender employees. The first allows transgender plaintiffs to raise actionable claims based on their gender nonconformity. For instance, in one prominent case, Smith v. City of Salem,  106  a preoperative male-to-female transsexual was suspended from her job as a firefighter after she began appearing in the workplace as a woman.  107 The court held that the employee could sustain a sex claim based on a gender-stereotyping theory  108  - a theory I discuss in great detail below.  109    [*916]   According to the court, "Employers who discriminate against men because they do wear dresses and makeup, or otherwise act femininely, are also engaging in sex discrimination, because the discrimination would not occur but for the victim's sex."  110

The obvious upside of the court's reasoning is that it provides a viable route to recovery, one that can be followed by transgender employees in the future. But it comes at a cost. The court's reasoning effectively erases transgenderism as an identity. Although the employee was asserting a female identity in the workplace, to avail herself of the gender-stereotyping theory she had to take on a male identity, namely, as a man who wanted to participate in the workplace dressing and looking like a woman. As one commentator notes of the case, "Transsexuals or transgender people per se do not really exist in the Smith opinion; there just happen to be some men out there who want to wear dresses."  111

A second case worth mentioning involves an employee who applied for a research position with the Library of Congress.  112  The employee applied in her capacity as a man and received an offer,  113  but the offer was rescinded after the employer learned that the employee planned to join the workplace as a woman.  114  The court in this case compared the employee's situation to that of an employee who is in the process of converting from one religion to another.  115  As the court noted, Title VII's prohibition on religious discrimination would certainly protect an employee who faces discrimination because of a religious conversion.  116  Like a religious convert, the transgender employee in this case was transitioning from one extant identity to another,  117  both of which fall within Title VII's prohibition against sex discrimination.  118  It thus follows that discrimination aimed   [*917]   at a transgender employee constitutes sex discrimination under Title VII.  119

It is important to remember that these cases are exceptions to the prevailing rule, at once a vision of a brighter future and a reminder of how far we are from realizing this vision. A 2007 case in Utah drives home the point. The plaintiff in the case, a male-to-female transsexual, worked as a city bus driver in Salt Lake City.  120  If they needed to use the bathroom during their shift, bus drivers had to use public facilities on their route,  121  and the Utah Transit Authority made arrangements with businesses for drivers to use their restrooms.  122  The plaintiff was fired after her supervisor discovered that she was using female public bathrooms while wearing a work uniform.  123  Siding with the employer, the court framed the case in terms of group harms: the city's policy was permissible because it did not disadvantage one sex more than the other.  124  And, once again, history proved to be an obstacle, with the court explicitly citing Ulane and other decisions in the transgender discrimination canon as controlling authority for the rule that Title VII does not protect transsexuals as a class.  125

2. Pregnant Women and New Mothers. The thinness of the sex provision's legislative history also played a major role in the outcome of the pregnancy cases in the 1970s. In the span of a few short years, sex discrimination law was turned on its head, and then back again, with respect to pregnancy discrimination. The battle over pregnancy discrimination started with the Supreme Court's 1976 decision in   [*918]   General Electric Co. v. Gilbert.  126  Gilbert involved a challenge to General Electric's (GE) disability benefits program.  127  As a general matter, GE's disability program covered male and female employees equally.  128  The plan did not, however, cover pregnancy and related conditions.  129  As the Court saw it, GE's disability plan divided the workforce into two groups: "pregnant women and nonpregnant persons."  130  Although the former group was made up entirely of women, the latter group included both men and women, thereby making it hard to compare how the two groups fared under the disability plan.  131

Ultimately, the Supreme Court concluded that GE's disability plan did not run afoul of Title VII's prohibition on sex discrimination.  132  What matters most is how the Court reached this decision. It started by looking at the legislative history of Title VII's sex provision, observing that the history is "notable primarily for its brevity."  133  Without a strong statement of legislative purpose on which to rely, the Court looked for guidance in its race discrimination jurisprudence.  134 "Discrimination," according to the Court, means   [*919]   something specific, having been developed by the courts for almost a century.  135  Reluctant to break new ground, the Court refused to endorse any definition of discrimination that was "different from what the concept of discrimination has traditionally meant."  136  But what did "discrimination" traditionally mean? The answer, of course, is that "discrimination" means formal discrimination, where employers divide employees into discrete groups.

The story gets more complicated from there. Gilbert is that rare case that confirms that Congress listens when the Court speaks. Just two years after the Court handed down Gilbert, Congress passed the Pregnancy Discrimination Act (PDA).  137  The PDA explicitly overturned Gilbert by amending Title VII to "prohibit sex discrimination on the basis of pregnancy."  138  The statute further defines its scope as reaching discrimination on the basis of "pregnancy, childbirth, or related medical conditions."  139  Most important for my purposes, however, was the Court's explanation - in a later case interpreting the PDA - that Congress also disapproved of Gilbert's "test of discrimination."  140  This, of course, is a reference to the Gilbert Court's conclusion that sex discrimination be understood as the formal division of the sexes into two groups.

By all accounts, the PDA should have put Gilbert to rest. Yet the decision continues to play a major role in sex discrimination law, preventing Title VII from intervening in current controversies. Gilbert operates as what Professor Deborah Widiss calls a "shadow precedent," a case that courts continue to follow even though it is no longer good law.  141  Widiss demonstrates that lower courts act as if   [*920]   Congress never enacted the PDA, sometimes even citing Gilbert as controlling authority.  142

We can see this dynamic at work in cases involving discrimination against female employees who are breastfeeding. Like pregnancy, breastfeeding is a uniquely female experience. And also like pregnancy, it does not affect all female employees, which means that it creates a subclass of female employees. The breastfeeding example is especially useful because it highlights the individualized nature of modern sex discrimination, as the employee is seeking to mold the workplace around her unique needs.

Consider a prominent example.  143  The plaintiff, a news producer for MSNBC, returned to work shortly after giving birth to her son.  144  Because she was breastfeeding, the plaintiff wanted a private space where she could pump.  145  Initially, MSNBC let her use an empty editing room.  146  But her coworkers, who did not know the room was occupied, eventually tried to enter the room while she was using it.  147 Despite her requests, MSNBC was unable to provide the plaintiff with a suitable alternative space to pump.  148  This led to a larger conflict about scheduling, and the plaintiff ultimately resigned her position.  149

She sued and lost. The district court concluded that the "drawing of distinctions among persons of one gender on the basis of criteria that are immaterial to the other … is not the sort of behavior covered by Title VII."  150  This tracks the Supreme Court's reasoning in Gilbert, limiting the reach of sex discrimination law to employment policies that divide men and women into distinct groups. The court did not stop there, however. As a means of lending historical force to its ruling, the court explained that this rule was "was made clear more than twenty years ago in General Electric Co. v. Gilbert."  151  It is as if the PDA never existed.

 [*921]   Stepping back, the larger point I want to make about history is that sex discrimination law is haunted by the ghosts of its early cases. These cases settled on a specific idea about what does and does not count as sex discrimination, and that idea has become so commonplace that courts act as if it were embedded in the DNA of sex discrimination law. I want to push back against the idea that the traditional conception of sex discrimination should continue to undergird Title VII. However entrenched it may seem, this idea is not essential to Title VII's sex equality project. There is nothing in the language of the statute - nor, for that matter, is there anything in the actual legislative history of the law - that limits Title VII to such a narrow conception of discrimination. And yet the idea persists because it is tied to a historical account of the sex provision that is suspect at best.

It has been almost fifty years since the Civil Rights Act became law, and in that time sex discrimination - and by this I mean the lived experience of sex discrimination - has evolved into something new. Sex discrimination law has not kept pace with the changing nature of sex discrimination. At its most fundamental level, antidiscrimination law is in the business of creating change - to change people's attitudes about each other, to change the way employers organize their workplaces, and to change the social structure of society at large. To bring about the kinds of change needed today, sex discrimination law must also change. It can start by moving past its own troubled history.

B. Subgroups and Stereotypes

Since the early days of Title VII, courts have been confounded by the issue of subgroup discrimination. Some of the earliest sex discrimination cases were brought by flight attendants - or stewardesses, as the job was called back then - who were challenging the airline industry's discriminatory practices toward women.  152  To cultivate the dual image of flight attendants as both potential bride and sex object,  153  the airlines went to great lengths to control their   [*922]   employees' private lives. Marriage 154  and pregnancy  155  were grounds for dismissal, and the airlines forced flight attendants to retire when they reached a certain age, usually between thirty and thirty-five.  156  The airlines were not engaging in wholesale discrimination against all women. Rather, they were discriminating against some women. Within the subgroup of women who work as flight attendants, the rules drew distinctions between single women and married women, as well as between older women and younger women.

Subgroup discrimination is not always defined by job categories, however. Another prominent form of subgroup discrimination occurs when an employer targets some women (or men) because they have - or do not have - a particular trait. For instance, say an employer will hire men with school-aged children but not women with school-aged children.  157  Even though the employer may not have a policy against hiring women as a general matter, this specific rule disadvantages a subgroup of women (women with school-aged children). 158  A leading antidiscrimination scholar, Kimberly Yuracko, has coined the phrase "trait discrimination" to describe the experience of discrimination in these subgroup cases.  159  The label is especially valuable because it captures the essence of subgroup   [*923]   discrimination: discrimination based on sex plus some other trait.  160  The other trait can be another protected trait, like race,  161  or an unprotected trait, like whether the employee has short or long hair.  162

Within subgroup discrimination, the thorniest issue concerns the scope of the gender-stereotyping theory of sex discrimination. The gender-stereotyping theory has its roots in a 1989 Supreme Court decision, Price Waterhouse v. Hopkins. The case revolved around Ann Hopkins's unsuccessful partnership bid at the consulting firm Price Waterhouse. Hopkins had worked for the firm for five years before applying for the promotion to partner.  163  Of the eighty-eight employees up for partnership that year, Hopkins was the only woman. 164  In fact, had she been successful, Hopkins would have been only the eighth female partner at the firm, out of 662 partners then affiliated with the firm.  165  As part of its review process, Price Waterhouse solicited feedback about the candidates from partners across the country, even from partners who had little to no contact with the applicant.  166

The feedback on Hopkins revealed that the partners were conflicted about her candidacy. On the one hand, the partners viewed her work product favorably, citing in particular her work landing, almost singlehandedly, a lucrative contract with the federal   [*924]   government.  167  On the other hand, the partners voiced concerns about Hopkins's interactions with coworkers. The partners depicted Hopkins as difficult to work with and rude to support staff.  168  It was these latter comments, about Hopkins's so-called "interpersonal skills,"  169  that raised the specter of sex discrimination. The partners said that she was "macho," that she "overcompensated for being a woman," and that she needed "a course at charm school."  170 Perhaps the most critical fact, however, was the advice given to Hopkins about how she could improve her chances for making partner in the future.  171  She was told to "walk more femininely, talk more femininely, dress more femininely, wear make-up, have her hair styled, and wear jewelry."  172

Hopkins sued Price Waterhouse under Title VII, alleging sex discrimination. The case made its way to the Supreme Court, where the Court ultimately sided with Hopkins. In many ways, the case was easy for the Court. As Justice Brennan wrote for a plurality of the Court, "If an employee's flawed "interpersonal skills' can be corrected by a soft-hued suit or a new shade of lipstick, perhaps it is the employee's sex and not her interpersonal skills that has drawn the criticism."  173  But the Court went deeper than that, offering a seemingly broad theory of sex discrimination. "In the specific context of sex stereotyping," Justice Brennan wrote, "an employer who acts on the basis of a belief that a woman cannot be aggressive, or that she must not be, has acted on the basis of gender."  174  This marked a significant change in the nature of sex discrimination jurisprudence, which up until Price Waterhouse had been focused primarily on status discrimination, that is, discrimination against an employee in her capacity as a woman (or his capacity as a man). Recognizing that it was breathing new life into sex discrimination law, the Court bolstered its legal rule by declaring that "we are beyond the day when   [*925]   an employer could evaluate employees by assuming or insisting that they matched the stereotype associated with their group."  175

Note how the Court's reasoning recast the relationship between the employee and the employee's group. Rather than formally comparing men and women as groups, the Court's definition of discrimination focused on how the particular employee failed to live up to the employer's idea about how men or women are supposed to look and act. The discriminatory comparison is therefore between the employee and a stereotypical employee, a heuristic rather than a real person. Moreover, Price Waterhouse pushes Title VII beyond the realm of biological sex to capture the performative aspects of an employee's identity. In this regard, the decision echoes the work of feminist scholars who sought to disaggregate sex and gender  176  - the former refers to biological differences between men and women, whereas the latter describes the cultural expressions of masculinity and femininity.  177  Although Hopkins was the only woman up for partnership that year, she stumbled not because she was a woman but because of how she performed her womanhood in the workplace - what she wore, how she walked, how she talked. Price Waterhouse was indeed a watershed moment in the arc of sex discrimination law. After Price Waterhouse, writes Professor Katherine Franke, "bodies have dropped out of the equation."  178

1. Contested Terrain. In practice, the gender-stereotyping theory has proven to be a mixed blessing. On the one hand, the theory has been a source of novel victories for outsider employees. For instance, lesbian and gay employees,  179  transgender employees, 180  and working   [*926]   mothers  181  have all used the gender-stereotyping theory as a means to remedy sex discrimination. On the other hand, other outsider employees, sometimes the very same groups of outsider employees who have had success with the theory, have found that the theory frustrated their cases, with courts concluding that such employees are trying to "bootstrap" protection for unprotected traits.  182  The doctrinal confusion over the theory - the struggle over whether it is a pioneering step for sex discrimination law or the basis for a suspicious litigation tactic - exists because the normative underpinnings of the theory remain elusive. When the Supreme Court declared that sex stereotyping violated Title VII, it never really explained why.  183

Stepping into this void, scholars have offered different ways of thinking about the gender-stereotyping theory in particular and subgroup discrimination more generally. I want to highlight two such scholarly perspectives as a means of distinguishing my own account, which I develop in the following Section. These scholars offer competing visions of how and why the gender-stereotyping theory works, and they do so by considering the analogue to Ann Hopkins's masculine woman: a man in a dress.

The first comes from Professor Mary Anne Case, who, a few years after the Court handed down Price Waterhouse, wrote an influential article on effeminate men in sex discrimination law. Case argues, simply but elegantly, that "the world will not be safe for women in frilly pink dresses - they will not, for example, generally be as respected as either men or women in gray flannel suits - unless and until it is made safe for men in dresses as well."  184  Case's perspective is rooted in the goal of protecting "the stereotypically feminine" from   [*927]   becoming enshrined in law and culture as being less worthy of respect than traits traditionally coded as masculine.  185  Importantly, Case does not endorse a sameness model of sex equality, nor, for that matter, does she see her intervention as making a difference-based argument.  186  Instead, Case argues for melding the two approaches, with respect to equality of the sexes, by disaggregating gender from sex and by allowing both men and women to express both masculine and feminine behaviors and identities.  187  As she describes it, the goal of Case's project is "to make the world safe for us all, norms and exceptions, men and women, masculine and feminine, and every shade in between."  188

We can contrast Case's view with that of Yuracko's, who articulates a different vision of the theory in her work on trait discrimination.  189  Whereas Case sees value in protecting men in "frilly pink dresses," Yuracko questions whether sex discrimination should protect such behaviors. She argues that it means something different when a woman wears a dress than when a man does.  190  "All gender norms … are not created equal,"  191  Yuracko argues. Therefore, she continues, "employers may recognize some norms without impeding [sex] equality in the workplace."  192  The problem with trait neutrality, Yuracko argues, is that we lose sight of the real harms of sex discrimination - norms that prevent men or women from being full and active participants in the workplace.  193  This is why she is reluctant   [*928]   to protect the man in the dress against discrimination. "Men are not disadvantaged in the work world by being forced to mimic the clothing style of the ideal male worker," she writes.  194  Though such a rule will disadvantage "some men,"  195  - namely, men who wear women's clothing - Yuracko is more concerned with the "substantive sex equality that is Title VII's goal,"  196  from which she seems to adopt a group-centric conception of equality.

2. Rethinking Sex Equality. The example of the man in a dress is not just some abstract hypothetical. In Oiler v. Winn-Dixie Louisiana, Inc.,  197  Peter Oiler lost his job as a truck driver for the grocery chain Winn-Dixie after his boss discovered that Oiler was cross-dressing during his off-hours.  198  Oiler is a good case with which to rethink Title VII's commitment to sex equality, specifically with respect to Title VII's antistereotyping principle. The thrust of my argument is that, as a normative matter, Title VII should not define sex equality in terms of what is good for all or even most women (or men), but rather in terms of what individual men and women need to flourish in the workplace. To make this point, I borrow an idea from Case's account of sex equality, and I disagree with one from Yuracko's account of sex equality.

Let us start, however, with the nature of the discrimination faced by Peter Oiler. Winn-Dixie did not have a formal rule prohibiting men from wearing women's clothing.  199  Instead, the company applied an ad hoc rule once it discovered Oiler's secret life as a cross-dresser.  200  In this sense, Oiler was in a class all by himself; he was targeted because of the way he, as an individual, performed his manhood. The company was, no doubt, operating on the basis of a stereotype: men should not wear women's clothing. Prescriptive in   [*929]   nature,  201  the stereotype not only dictates appropriate behavior - a "real" man wears men's clothing - but also punishes those who fail to conform to this behavior. This highlights the defining characteristic of modern discrimination. Whereas earlier forms of discrimination focused on an employee's status as a member of a group, modern discrimination has more to do with work-culture norms and the ways in which an employee's behavior violates these norms. After all, not all outsiders face discrimination in the workplace. Discriminators target particular victims because of who they are and how they act. There is no question that an employee's group membership factors into this calculus; this is unavoidable so long as marginalized traits are salient in our culture. But the critical point is that, by and large, the days of top-down discrimination are behind us.  202  Modern discrimination is the product of a complex web of work-culture norms,  203  stereotypes,  204  and unconscious biases,  205  which work together to make discrimination subtle, messy, and more personal than ever before.  206  Discrimination is no longer just about who you are; it is also about how you express yourself and whether this self-expression is welcome in your workplace.

Central to Mary Anne Case's view of Title VII's antistereotyping principle is the idea that we should look to the margins as a guide for determining the substance of sex equality.  207  When she was writing in the mid-1990s, the effeminate man resided at the margins of society.  208    [*930]   Cultural norms have since shifted and the effeminate man, once a challenge for courts, is now an easy case.  209  But he has been replaced at the margins by a cast of characters as diverse as they are stigmatized. Given its remedial nature, Title VII should steer toward the needs of these employees, the outsiders of today. This is where I disagree with Kimberly Yuracko's vision of sex equality. She argues that Title VII should prohibit only discrimination "based on gender norms or scripts that inhibit the ability of individuals of a particular sex to participate successfully in the work world."  210  Under this view, Title VII would provide relief for Ann Hopkins, the abrasive business consultant,  211  but not Peter Oiler, the man in a dress.  212  Both violated conventional gender norms, but only Hopkins faced a norm that inhibits all or most women from flourishing in the workplace. "If employers were permitted to act upon the gender script equating aggressiveness in women with bitchiness," Yuracko writes, "all women would be undermined in their ability to participate fully and successfully in the workplace."  213  Oiler, by contrast, violated a gender norm that affected, at most, a tiny population of men. "Eradicating this particular gender norm is not necessary for the substantive [sex] equality of women and men in the work world," she concludes.  214

Aside from biological attributes like being pregnant, we should be cautious about enshrining in law the idea that there are certain things that women do and certain things that men do. It may very well be that, as Yuracko points out, women who wear dresses conform to social norms while men who wear dresses buck them.  215  But the question should not be whether a particular gender norm is harmful to all or most women (or men); it should be whether the norm harms an individual woman (or man). Cross-dressing is probably not important to most men and certainly not all men. But it was very   [*931]   important to Peter Oiler's sense of his identity, just as not wearing makeup was very important to Darlene Jespersen's identity. Sex discrimination law should foster an ethic of self-definition; it should shield employees against gender norms that seek to dampen constitutive elements of an employee's identity. After all, the universe of gender performances is vast, if not infinite, and one person's femaleness (or maleness) is no more authentic than another person's.  216

At this point, some may question whether sex discrimination law is capable of personalizing antidiscrimination protections in this way. How, for instance, will the law distinguish between norms that strike at constitutive elements of a person's identity and other less harmful norms? I address these sorts of issues in greater detail in the following Part, in which I argue that we should remake sex discrimination law in the image of religious discrimination law. For now, however, the short answer is that it is up to each individual employee to determine the constitutive elements of her identity. The doctrinal structures of religious discrimination law - the very structures I want to import into sex discrimination law - facilitate this process of self-definition. In a religious discrimination case, the central question is not whether a particular trait or behavior is, from an objective standpoint, religious in nature. Instead, the inquiry is whether the employee, in her subjective capacity, believes the trait or behavior to be religious within her own worldview.  217  Because each employee is given room to define the contours of her own religious beliefs and practices, religious discrimination law makes accommodation, rather than disparate treatment, the centerpiece of its analysis. In doing so, religious discrimination law accomplishes something that Title VII has otherwise been unable to do: define discrimination from the   [*932]   perspective of the victim rather than the discriminator.  218  Because an employee gets to determine what behaviors are central to her identity, the central task for the law is to determine whether the employer discriminated against the employee by not making accommodations for the employee's behavior.

C. Multiples

One of the hallmarks of antidiscrimination analysis is the assumption that we can organize people into discrete categories.  219  All people have a race or a sex, for instance, so the law sets out to categorize people along these lines. The problem, of course, is that modern discrimination is a messy enterprise that defies neat categorization. In its attempt to impose order on something disorderly, employment discrimination law neglects the needs of employees who face discrimination aimed at multiple parts of their identity. Writing in the late 1980s, Professor Kimberle Crenshaw sparked a discussion about the problem of intersectionality in antidiscrimination law,  220  which prompted scholars to develop a robust conception of identity for purposes of proving discrimination.  221  Using discrimination against a black woman as her jumping off point, Crenshaw argued that the "intersectional experience is greater than   [*933]   the sum of racism and sexism."  222  A black woman's experience cannot be compared to the experience of either a black man or a white woman. Neither of these latter examples captures the full range of stereotypes and prejudices that attach uniquely to a black woman's experience.

Because it is stuck in what Crenshaw calls a "single-axis framework,"  223  antidiscrimination law stumbles in the face of discrimination that cuts across multiple identity traits.  224  Ultimately, this is a failure of group-based analysis. The dominant method of proving discrimination today is to consider how the employer treated the claimant as compared to a similarly situated employee.  225  The closer the similarity, the easier it is to isolate the reason for the adverse employment action and, in turn, identify whether the employer was motivated by an illegitimate purpose.  226  This is discrimination by algebraic equation: cancel out the traits in common, and the trait that motivated the employer's action is left remaining.  227  Although helpful in theory, such analysis only works if the claimant has someone to compare herself to.

Take Dawn Dawson as an example. Dawson worked as an assistant and stylist-in-training at Bumble & Bumble, a high-end salon   [*934]   in New York City.  228  A self-described gender-nonconforming lesbian woman,  229 Dawson was fired from her assistant position and kicked out of the stylist training program.  230  The salon's reason for these actions was that Dawson had little chance of finding a stylist position except in New York City because her demeanor and appearance would frighten most people.  231  On top of that, Dawson also alleged that several of her coworkers harassed her on a regular basis, subjecting her to a steady stream of demeaning comments, often in front in clients.  232  For instance, they called her "Donald."  233  They said she should act less like a man and more like a woman.  234  They said she wore her sexuality "like a costume."  235 And they said she "needed to have sex with a man."  236  Dawson responded by bringing a sex discrimination claim under Title VII, which she lost handily.  237

As Dawson's case demonstrates, intersectionality is a conceptual blind spot for antidiscrimination law.  238  In the eyes of the court, Dawson was first and foremost a lesbian.  239  That Dawson was also a masculine woman, and that she faced discrimination both as a masculine woman and as a lesbian, did not factor into the court's thinking. In this sense, Dawson's case confirms Professor Crenshaw's central insight into the limits of existing antidiscrimination discourse: an employee cannot inhabit more than one identity at a time. Because it saw her as a lesbian - and as nothing but a lesbian - the court viewed Dawson's sex discrimination claim as an attempt to bootstrap   [*935]   protection for sexual orientation into Title VII.  240  Though Bumble & Bumble employed outsiders of various stripes,  241  there was no one who could serve as a comparator for Dawson. And without a comparator in sight, the court was able to recast Dawson's claim, transforming it from a hard case about gender norms into an easy case about sexual orientation.

Taking a step back, intersectionality theory highlights an important point about the regulatory force of employment discrimination law. By channeling discrimination claims into a single-axis framework, employment discrimination doctrine effectively shapes an employee's identity. We see this in Dawn Dawson's case, as the doctrine marked her as a lesbian and, in the process, erased any sense of her female masculinity. We see it in the transgender discrimination cases, as the doctrine forces a male-to-female transsexual to self-identify as a man in a dress and thereby give up her hard-fought female identity. And we see it in discrimination against a black woman - the paradigm case of intersectionality - as the law pits her blackness and femaleness against each other, forcing her to claim one identity and forego the other. As a regulatory force, employment discrimination law tends to mute difference. The time has come for a new sex discrimination regime, one that amplifies difference rather than dampens it. The next Part sets out to imagine just such a regime.

III. The New Sex Discrimination

In this Part, I sketch a new framework for sex discrimination law that is modeled on the protections currently available to employees in religious discrimination cases. The defining characteristic of religious discrimination law is its elasticity. Through various doctrinal mechanisms, the law bends to the needs of employees, seeking to empower employees to practice their faith without having to sacrifice their position at work. I want to be clear at the outset, however, that I am not proposing this shift because I think it will automatically translate into more victories for employees in sex discrimination cases. Rather, I propose it because religious discrimination law offers a more sophisticated way of thinking about difference and discrimination.

 [*936]   The prevailing discourse in religious discrimination law is built around twin goals of neutrality and balance. The law takes a neutral position on what counts as religious in belief and practice, thereby giving employees plenty of room to determine for themselves what is or is not required as part of their faith.  242  At the same time, the law seeks to strike a balance between an employer's needs in organizing its workplace and an employee's needs in furtherance of her faith. On this latter point, employers clearly have the upper hand. In an at-will environment,  243  antidiscrimination protections can only go so far, and the law reaches its limit once an employer offers a reasonable accommodation or shows that no such accommodation is possible. Yet the inquiry is what really matters. The ultimate question in a religious discrimination case is whether an employer can adapt its workplace to the needs of its employee and not the other way around. Sex discrimination should follow suit. By reorganizing sex discrimination law in this way, we can advance the needs of men and women at the margins of our society while at the same time advancing the work of antidiscrimination law more generally. This, in turn, would bring us closer to the ideal of a workplace culture in which employers make decisions on the basis of merit rather than identity.

I develop this argument in three sections. The first articulates a doctrinal shift for sex discrimination, tracking the doctrinal framework currently used in religious discrimination cases. The next Section argues for a new vision of equality for sex discrimination law, one that is steeped in difference rather than sameness. Finally, the last Section sharpens my argument by responding to potential critiques.

A. A New Framework

To begin, consider a garden-variety example of religious discrimination.  Abercrombie & Fitch  hired Lakettra Bennett to work   [*937]   as a sales person - or, as Abercrombie & Fitch calls it, "model" - in one of its stores.  244  Soon after that she was promoted to a manager-in-training position and transferred to one of the company's Hollister stores.  245  At the time, Bennett followed the company's "Look Policy," which required employees to wear clothes that were consistent with the Hollister brand, namely, "ripped-up jeans, a little revealing, sporty, California beach style, laid back."  246  Female employees in particular were supposed to make themselves look sexy by wearing tight clothing that accentuated their bodies.  247

The Hollister dress code soon became a problem for Bennett. After converting to the Apostolic religion, Bennett wanted to dress more modestly. She exchanged her short skirts for long skirts that fell below the knee, her low cut shirts for long sleeve shirts that did not show any cleavage. 248  On her first day back to work after her conversion, she wore an ankle-length denim skirt, which was unlike anything Hollister had ever sold.  249  Her new attire clearly violated Hollister policy, though the company was willing to work with her. Over the course of several meetings, the company offered a number of alternatives to resolve the conflict between Hollister's dress code and Bennett's clothing preferences. One option was to wear jeans instead of skirts.  250  Another was to wear short skirts with leggings underneath to cover her legs.  251  The final option was to look in other stores for skirts that would be consistent with both the dress code and her religious beliefs.  252

Bennett rejected all three proposals.  253  The only accommodation she was willing to settle for was an exception to the company's dress   [*938]   code so that she could wear long skirts.  254  Unwilling to make this concession, Hollister gave Bennett two weeks to make a decision: adhere to the policy or resign.  255  She opted to resign.  256  Soon afterward she brought a religious discrimination claim under Title VII, alleging that Hollister made no effort to reasonably accommodate her religious practice.  257

The first step in evaluating a claim of religious discrimination is to determine whether an employee's beliefs constitute a "bona fide" religious belief under Title VII.  258  This is a notably relaxed standard. There is no list of permitted religions. Nor is there any expectation that an employee's beliefs be in the mainstream of her faith. Instead, the court considers whether the belief is religious within the given employee's worldview.  259  Piercings have been held to be religious under this standard,  260  as have tattoos,  261  veganism,  262  witchcraft,  263  and even atheism.  264  In addition, an employee must show that the religious belief is sincerely held.  265  It was the sincerity requirement that may have posed the biggest problem for Lakettra Bennett in her suit against Abercrombie & Fitch. Although she claimed that her faith did not permit form-fitting clothing, she showed up to her deposition wearing a tight shirt, which she herself described as "body conscious."  266  The purpose of the sincerity requirement is to make   [*939]   sure that employees do not disingenuously appeal to religion so they can avoid work obligations.  267  In one particularly memorable case, a plaintiff was unable to convince the court that his practice of eating Kozy Kitten Cat Food was a sincerely held religious belief.  268

Once an employee satisfies these definitional requirements - and it is worth noting that, contrary to the cat-food eater, most employees do in fact satisfy these requirements - the next step in the framework is to look at an employer's response to the conflict.  269  The bulk of religious discrimination litigation focuses on this step. Title VII imposes on employers a duty to make reasonable accommodations for an employee's religious beliefs or practices. An employer is not bound to accept an employee's requested accommodation.  270  Nor is the obligation limitless. An employer does not have to make an accommodation if doing so would pose an "undue hardship."  271  In an important case, the Supreme Court defined undue hardship as a requested accommodation that would make an employer "bear more than a de minimis cost."  272  This standard is a boon for employers, and it also helps reign in the reasonable accommodation theory, which has the potential to cripple business if left unrestrained.

Lakettra Bennett's case highlights the important role that negotiation plays in reasonable accommodation. Rather than rejecting her requests out of hand, the company offered Bennett three different options to accommodate her faith. This is a hallmark of reasonable accommodation; it facilitates an interactive process between employer and employee, who must work together to try to fix the problem in such a way that the employee can get back to work and the employer's business will not suffer too much on account of   [*940]   the changes.  273  In this regard, reasonable accommodation upends traditional antidiscrimination discourse, and in a good way. It transforms the employment relationship into a kind of partnership. Not an equal partnership, to be sure, but these partnerships make it possible for employers and employees to have an honest conversation about difference - which differences matter, why they matter, and whether there is space in the workplace for these differences.  274  A workplace brimming with these sorts of conversations is a breeding ground for real social change, the kind of change that is felt beyond the walls of the workplace.  275

1. Sex as Practice. The first step in my proposal is that sex discrimination law track the causation standard currently in use in religion cases. This would shift the analysis away from the conventional "because of sex" standard that dominates sex discrimination law today and move it more toward an individualistic standard that focuses on the employee's lived experience of discrimination. For this to work, sex discrimination law will have to embrace the idea that sex is both a status and a practice. Employees face sex discrimination both because of who they are (status) and because of how they act (practice). The new sex discrimination is all about sex as a practice, capturing the performative side of a person's identity. The critical question will be one of self-definition: Does the employee sincerely believe that the practice in question is constitutive of her identity as a woman (or his identity as a man)?

Consider two examples. Start with Darlene Jespersen and Harrah's no-makeup policy.  276  For Jespersen, wearing makeup conflicted with her own sense of her womanhood. She found it   [*941]   degrading, so much so that she could not do her job - a job she had done exceedingly well for over twenty years - effectively. And she sincerely felt this way. It was not the case that Jespersen immediately developed her preference when the company announced its new policy; she had not worn makeup at any time in her many years with the company, nor did she wear makeup outside work. Her stance against makeup was a deeply held conviction.

Now consider Krystal Etsitty, the transgender city bus driver who was fired for using women's restrooms along her route.  277  The critical question in the case would be to determine Etsitty's sex. Existing law defines Etsitty's sex according to her birth sex, which was male. This explains why Etsitty had to resort to a theory of sex discrimination that defined her not as a woman but as a man who faced discrimination because he wanted to wear a dress. By contrast, my proposal allows Etsitty to define her sex consistent with her transition, thereby allowing her to be a female. The sex practice in question was Etsitty's use of a women's restroom. She would have to argue - and I am confident she could establish - that using a women's restroom was constitutive of her identity as a woman. And such a belief would, no doubt, be about as sincere as they come.

2. Accommodating Sex. The first step is the easy part. The second step - determining whether the employer must accommodate the sex practice - is, as it should be, a harder question. To avail oneself of the reasonable accommodation protections, an employee would have to engage her employer, alerting the employer to the conflict - if the employer were not already aware of it - and opening a line of communication to try to remedy the situation. An employer's duty to accommodate would not mean that an employer has to accept any proposal put forth by one of its employees; the standard would be one of reasonableness. If an employer offers a reasonable solution, an employee could not reject it in favor of her own preferred solution and still state a claim under Title VII. And if no accommodation were possible, I would likewise adopt the undue-hardship safety valve. A surprisingly low standard, the undue-hardship test is far more favorable to an employer's interests than an employee's.

Two more points bear mentioning before considering some examples. The first is that reasonable accommodation is a fact-intensive inquiry, which makes sense given that reasonable   [*942]   accommodation seeks to craft individualized remedies. As such, every case of reasonable accommodation is tailored to the needs of the employee alleging discrimination. The second point is more of a reminder: the real value of reasonable accommodation is that it facilitates a conversation between employers and employees about difference.

Now return to Darlene Jespersen's case. Jespersen wanted an exception to Harrah's makeup policy. Harrah's, by contrast, sought uniformity; it was trying to foster an image in its casinos, and makeup was an important part of that image. Jespersen's history with the company undercuts the argument that female bartenders needed to wear makeup to do their job well. Although Harrah's had long encouraged women to wear makeup,  278  Jespersen never did. And her work performance did not suffer because of it. It is simply hard to argue that her job now calls for wearing makeup. Under a reasonable accommodation standard, Harrah's would have to identify the costs of honoring Jespersen's request, as a means of arguing that accommodating Jespersen would amount to an undue hardship. This would be an uphill battle.

Krystal Etsitty's case would be harder. The employer never inquired whether certain businesses along her route would have allowed Etsitty to use their bathrooms,  279  though it is not clear that the employer would have been required to look into this possibility. If that were possible, however, that might have satisfied everyone's wishes. Another possibility is that Etsitty could have waited to use the bathroom until after completing her route, possibly until she returned to the bus station. If that worked, Etsitty would have had to bear the brunt of the costs, as she would have had to wait to go the bathroom until she finished her shift. Yet it is also possible that there are no good solutions here. Let me stress that my proposal does not mean that every claimant has to be accommodated. My hope is that Etsitty and her employer would be able to work things out without having to resort to litigation. Having an accommodation framework in place incentivizes these sorts of negotiations, making litigation a last resort when compromise is not possible.

 [*943] 

B. A New Vision of Equality

The backbone of the new sex discrimination regime is the idea that difference is universal. In his new book Far From the Tree, Andrew Solomon makes this point better than I can.  280  Solomon is writing in the context of parents who have children with different identities than their own - straight parents who have a gay child, able-bodied parents who have a disabled child, and parents of average intelligence who have a prodigy, to name a few examples.  281  Yet his claim is no less applicable to my project: "Difference unites us. While each of these experiences can isolate those who are affected, together they compose an aggregate of millions whose struggles connect them profoundly. The exceptional is ubiquitous; to be entirely typical is the rare and lonely state."  282  What Solomon has identified - and it is a point I want to stress - is that each person's search for identity is hers alone. And this search does not get put on hold when an employee enters the workplace. The idea that our work and private lives occupy separate spheres is more metaphor than reality. Employees bring into the workplace their preferences and biases, their relationships and associations, their identities and senses of self.  283  The workplace is an artificial environment, assembling an increasingly diverse set of individuals who would otherwise not interact with each other in their daily lives.  284  If anything, work magnifies the social pressures facing outsider employees, making it all the more important for outsiders to fit in among their coworkers.  285

For the new sex discrimination regime to take hold, we need to recalibrate the vision of equality that undergirds sex discrimination law. Put simply, we need to start thinking about equality in terms of cultivating difference among employees. A helpful way to facilitate this shift would be to expand the way we talk about Title VII's protections. Title VII lists prohibited bases for employment - race, sex, religion, among others. I have made the conscious choice throughout this Article to refer to these as "protected traits" rather   [*944]   than "protected groups." In my experience, the latter formulation is far more prevalent than my rendering.  286  Though it may seem like a minor matter of semantics, the difference between these two formulations is actually quite meaningful, cutting to the heart of what is at stake in employment discrimination law. When we couch Title VII in the language of groups, we not only anchor the statute to the needs of groups, but we also marginalize those employees who do not fit neatly into recognized identity groups, whether because of the complexity of their identity - think intersectionality - or because they are the outsiders among other outsiders, the employees who have the hardest time fitting in. My position is that sex discrimination law needs to make a better effort to reach these employees.

I am not arguing that we should completely abandon the sameness model. What we need is a holistic vision of equality, one that is capable of pivoting between sameness and difference as the case calls for it. Group-based discrimination still occurs. Take the recent Wal-Mart litigation. 287  The plaintiffs in that case charged, among other things, that Wal-Mart discriminated against women as a group in its promotion decisions.  288  Contrast their claims with Darlene Jespersen's, Ann Hopkins's, and Karen Ulane's claims. These latter cases are all about particular women who sought a personalized remedy. Whereas the women of Wal-Mart wanted to be treated the same as men, Jespersen, Hopkins, and Ulane wanted to be treated differently than all their coworkers, men and women alike, because they had different needs than their coworkers. Each practiced her womanhood on her own terms - as a woman who refuses to wear makeup, as a woman who curses and acts aggressively, and as a woman who used to be a man. Their search for identity has carried them beyond the bounds of what our culture currently deems appropriate for women, and they suffered the consequences at work because of it. Equality is a remarkably pliable concept. In its prevailing form it works to constrain identity, discouraging marginalized employees from embracing the full range of their identities. Yet it also has the capacity to liberate identity, providing a blank canvas on which employees can sketch their own concept of   [*945]   who they are. It is possible for sex discrimination law to speak of equality and difference in the same breath as part of a common project. My hope is that sex discrimination law will move in that direction.

C. Concerns

My argument is sure to have raised red flags along the way. In this Section, I respond to three major objections, all of which I have faced at one point or another as I have explored this project.

1. Groups Matter. The first objection is that my argument has overlooked the value of group membership, especially in the realm of civil-rights practice. People benefit immensely from joining and identifying as a member of a group. Members of a group share a common history. For many outsiders, the first step in asserting a new identity is to walk the path of those who came before them. The promise of group membership is a community based on shared experience. It makes no difference where you are or where you have been, even if you are the only person around who has your identity. You can always identify with others like you. This is what makes Dan Savage's "It Gets Better Project" so powerful.  289  Savage, an influential sex-advice columnist, sought to create a vehicle to reach gay youth who are victims of bullying.  290  The immediate goal of the project was to urge gay kids not to commit suicide in the face of harassment and social ostracism. The broader aim, however, is to reassure these kids, many of whom have no gay role models in their day-to-day lives, that they are already part of a larger community of people just like them.  291

Groups matter. I would not want to suggest otherwise. My critique of sex discrimination law is that it has lost sight of the individual at a time when the individual is more important than ever. This is not to say that sex discrimination can or should stop thinking about groups. That is why I am envisioning a two-tiered sex discrimination regime. The first tier is sex discrimination as we have always known it, taking aim at discrimination for being male or female. Such status discrimination cases revolve around an   [*946]   employee's group membership. The second tier is focused on individualized harms. These are the cases in which the employee faces discrimination not because of who she (or he) is, but because of how she (or he) acts in the workplace. Whereas the first tier is concerned with being, the second tier is concerned with doing. The reforms I have presented would apply only to the second tier of cases. The purpose of the new sex discrimination is to supplement the old sex discrimination, not to replace it. Together, they will work in concert to strike at the full spectrum of sex discrimination as it exists today.

2. Flood and Stretch. The next major concern is that my proposal broadens the concept of sex discrimination far beyond anything we have seen before in American law. Consider a hypothetical to sharpen the critique.  292  Say that an employee wants to be excused from work two days a week so he can practice karate. When the employer resists - as the employer most certainly would - the employee demands an accommodation because his karate practice is an essential part of his manhood. How is it that we can still call this sex discrimination? There are traits that, even if you do not think should be protected, are clearly linked to sex in some way - refusing to wear makeup, cross-dressing, and transitioning from one sex to another are prime examples. But practicing karate is something else entirely. If the employee gets to define what practices are constitutive of identity, then the employee has a lot of room - arguably too much room - to determine the substance of sex discrimination law.

There are two ways of thinking about this objection - flood and stretch. Let me start with flood. The argument is that my proposal will lead to a deluge of sex discrimination cases, which will drain resources that would be better spent on other, more serious forms of discrimination.  293  For instance, one may argue that neither being forced to wear makeup nor being denied time to practice karate can compare to a case in which an employer discriminates on the basis of race. Though I am sympathetic to this concern, I am uncomfortable with any conception of workplace equality that subscribes to a strict hierarchy of harms. Title VII is not like the Equal Protection Clause; there are no levels of scrutiny, no differing standards depending on   [*947]   the trait at issue. Although others might see makeup as a relatively minor matter, it was not to Jespersen, and that is what matters most to the vision of sex equality I have put forth in this Article. Moreover, the flood concern assumes that a more lenient standard automatically leads to more litigation. The time, money, and inconvenience of litigating, not to mention how it disrupts their employment relationship, will dissuade most people from wasting the effort of bringing a sex discrimination case. Only those who feel strongly about the employer's decision will resort to litigation. So if the employee truly believes that his manhood is tied up in his karate practice, then sex discrimination law should care about it, even if it ultimately cannot provide the man with a remedy.

As for stretch, the argument is that if anything and everything counts as sex, then there is no limit on sex discrimination. I want to respond in two ways. The first is to emphasize that there is a limiting principle built into my proposal. Just like in a religious discrimination case, an employment decision would be sex based provided two conditions are met. First, the identity trait in question must be sex based within the employee's worldview. This may seem far too open-ended, but it will work for the same reason it already works in religious discrimination cases: individuals are in the best position to define the terms of their own identity. Second, the employee's belief that the trait under attack is sex based must be sincerely held. As is the case with religious discrimination, the purpose of the sincerity requirement is to safeguard against employees using sex discrimination as an excuse to avoid work obligations.

The second point in response to the stretch argument is that we should not think of antidiscrimination law only as a means to right wrongs. In addition to remedying specific cases of discrimination, antidiscrimination law also facilitates a critical conversation about identity and difference - a conversation that takes place in workplaces, in courts, in the media, and in people's daily lives. Although he ultimately leans away from the law as a tool for civil rights,  294  Kenji Yoshino celebrates the virtues of a "reason-forcing conversation," a conversation in which those who seek to impose a burden on an outsider must justify their reason for doing so.  295  As   [*948]   Yoshino writes, these conversations "reveal the true dimension of civil rights."  296

Stretch is not a bad thing. Outside of biological differences, sex does not have natural boundaries. The problem with existing sex discrimination doctrine is that it tries to draw firm lines around sex. By contrast, my proposal takes a hands-off approach, leaving it up to the individual to set the boundaries for what constitutes sex. None of us is in a good position to tell the man that practicing karate is not essential to his manhood; it is something only he can decide for himself. What the law can do, however, is facilitate a conversation between the man and his employer, in the hope that they can come to some resolution of the situation. In this case, the requested accommodation seems unlikely. I do not expect that any court would second-guess the employer's decision not to let the man take off work to practice karate. But that is beside the point I want to make here. The conversation is what matters. At every turn, my proposal points toward conversation as an important means of bringing about social change. It gives the employee an outlet to express his identity. It lets employers explain what they need from their employees to maintain good order in the workplace. And, most important of all, it harnesses the power to change people's minds about difference.

3. Religion Is Special. Religion holds a special status in our legal culture. Under the First Amendment, the state can neither prefer nor inhibit religious practice.  297  Religion receives this special status because of the critical role that religious freedom played in the founding of our country. Yet this special status also stems from the fact that we tend to think of religion as being somehow different than the other traits protected by antidiscrimination law. This helps to explain the so-called "ministerial exception" to Title VII, a judicially created rule that allows religious institutions to engage in overt discrimination for positions related to the institution's religious mission.  298  And it also helps to explain why claimants in religious discrimination cases have access to a separate antidiscrimination regime than claimants alleging other forms of discrimination. Thus it   [*949]   is only fair to question whether sex discrimination likewise deserves accommodation.

This is an important objection, raising a roadblock to my proposal that is both legal and ideological in nature. Let me start with a concession: I cannot overcome the point that religion deserves special treatment for historical reasons. Employment discrimination law is structured so as to give employees a reasonable amount of space to practice their faith, space that it is not available to employees to explore other aspects of their identity. Set aside the historical and textual reasons for treating religious discrimination differently. Are there other reasons to explain why religion is different than sex (or any other protected trait under Title VII)?

The obvious point to consider is that religion is mutable and the other traits are immutable. The theory here is that religion requires special treatment because it is not a fixed identity like a race or sex. It comes down to a person's control over her identity. If a person can change her religion - or even abandon religion altogether - then the person has a greater say in who she is and how she lives her life. We do not, by contrast, get to choose our race, race being biologically determined. Sex is another story, though. Although the vast majority of people do not exercise choice over their sex, some people do, going to great lengths to change their birth sex. Indeed, the great lesson of the transgender cases is that, where there is a will, there is a way to change one's sex. At least as far as immutability is concerned, religion and sex are more alike than different.

Maybe we need a softer definition of immutability. Rather than thinking of traits as locked identities, we can define immutability as a trait that is so central to our sense of self that it would be extremely difficult to change. We see this view of immutability used in asylum law, in which courts have ruled, for instance, that sexuality and gender identity "are immutable; they are so fundamental to one's identity that a person should not be required to abandon them."  299  Under this view, immutability is more about the effect of changing one's identity rather than the ability to change it. Yet this new definition would not resolve the problem at hand. Once again, the transgender example suggests that changeability is not a useful dividing line. Converting one's sex is no less fraught than converting one's religion; that a person wants to change his or her sex, and is   [*950]   willing to spend the time, money, and emotional energy to do so, says a lot about how important the change is to the person's sense of self. We could say the same about converting from gay to straight, forcing oneself not to cross-dress, or wearing makeup even though it makes one's skin crawl. Forcing or expecting people to change traits that are constitutive of their identity is like asking them to be somebody else.  300  In this regard, religion is not all that different than any other important identity trait that is capable of change.

IV. The Future of Employment Discrimination

Before concluding, I want to take a step back to consider some of the broader implications of my argument. My goal in this Part is to lay the groundwork for future discussions. To that end, I raise two questions about the future of employment discrimination law in light of my proposals for sex discrimination law. Because these questions provoke ideas that are each capable of sustaining an article unto itself, I simply cannot answer them in this space. Yet they are worth considering, if fleetingly, in the hope that they will stimulate future conversation about how antidiscrimination law should adapt to the changing nature of employment discrimination.

A. Beyond Sex

Sex discrimination is only a small sliver of employment discrimination law. What about the other traits protected by Title VII? Surely one could make a parallel argument about, say, race.  301  Like sex, race is both a status and a practice, a marker of both who an employee is and how the employee presents herself in the workplace. And like sex discrimination, modern race discrimination is primarily about the performative aspects of a person's racial identity. Consider a prominent case.  302  American Airlines refused to allow Renee Rogers, an African-American woman, to wear her hair in cornrows.  303  Rogers argued that the all-braided hairstyle held a special significance   [*951]   for her as a black woman in America.  304  It was, in other words, critical to her sense of her own racial identity.

Rogers ultimately lost her case, with the court drawing a distinction between a natural hairstyle (such as an Afro) and one based on artifice.  305  Braids were of the latter variety, according to the court, and therefore beyond the reach of Title VII.  306  Scholars have written strong critiques of the court's decision.  307  My interest in the case lies elsewhere. Later on in its opinion, the court notes that American Airlines would have allowed Rogers to put her hair in a bun and wrap a hairpiece around the bun.  308  After trying it out, though, Rogers rejected this option because the hairpiece gave her severe headaches.  309  The court does not frame the inquiry this way, but this is textbook accommodation analysis. The employer had a rule that prevented Rogers from wearing her hair in a manner that was critical to her sense of self as a black woman. The employer proposed a solution to the conflict - wearing her braids in a bun and hairpiece. It is hard to say whether the proposed accommodation was reasonable, but that is something that could be addressed through litigation.

Renee Rogers's case further demonstrates that accommodation is a natural way of thinking about identity and difference in the workplace. Moreover, it also suggests that existing conceptions of racial identity - like existing conceptions of sex and gender - do not capture the full range of how employees perform their identity in the workplace. Thus, as we look to the future of employment discrimination law, we need to consider whether other kinds of discrimination - race, national origin, and age, among others - should likewise move toward the antidiscrimination model I have proposed for sex discrimination.

 [*952] 

B. Beyond Title VII

There is a growing sense among scholars that Title VII may not be up to the task of eradicating employment discrimination as it exists today. Suzanne Goldberg writes that employment discrimination law "is in the midst of a crisis."  310  Elizabeth Glazer argues that employment discrimination law "needs help."  311  Marcia McCormick notes that employment discrimination law has effectively stalled, accomplishing little since the 1980s.  312  And Nancy Levit has shown that employment discrimination law is on the wrong side of changing workforce demographics.  313  I am likewise concerned that Title VII is ill-equipped to face the challenges raised by modern discrimination. The reforms I have proposed in this Article only go so far. Though they may improve the law's approach to sex discrimination, they do not alter the architecture of employment discrimination law as a whole.

Perhaps the time has come to adopt a new regulatory scheme for employment discrimination. The idea has intuitive appeal. After all, modern discrimination bears little resemblance to what discrimination looked like when Title VII became law. In her work on trait discrimination, Kimberly Yuracko offers Justice  Sandra Day O'Connor 's inability to find legal work as an attorney as a good example of what discrimination used to look like.  314  Despite graduating third in her class at Stanford Law School, the only law firm job Justice O'Connor could get was as a legal secretary.  315  Discrimination used to be about formal segregation and exclusion. Today, however, it is about not fitting in at work. The Justice O'Connors of the world have been replaced by the likes of Darlene Jespersen, Peter Oiler, and Dawn Dawson, men and women whose identities mark them as different from their coworkers. If discrimination has changed so much, why has discrimination law changed so little?

 [*953]   Whether we should adopt a new regime and what that regime might look like are important questions that warrant deeper consideration than I can offer in this space. Though I may not have broken Title VII's mold, I have offered a new way of thinking about sex discrimination - and possibly other forms of discrimination, too. My hope is that, at the very least, this Article will stimulate further discussion about whether and how we can make employment discrimination law more attuned to the needs of employees as it finds them today.

Conclusion

In an influential sexual harassment case, Justice  Scalia  once wrote that "statutory prohibitions often go beyond the principal evil to cover reasonably comparable evils, and it is ultimately the provisions of our laws rather than the principal concerns of our legislators by which we are governed."  316  Though he was writing about a particular issue in sex discrimination law,  317  I take Justice  Scalia 's words to be a broader statement about the scope and ambitions of employment discrimination law in general. He recognized both that discrimination will change over time and that, for it to be effective, the law must change, too. In that spirit, this Article has proposed significant structural changes to sex discrimination law. My hope is that these changes will make the law more effective in its ongoing and ever-changing fight against sex discrimination.

Duke Law Journal Copyright (c) 2014 Duke Law Journal Duke Law Journal

Footnotes