Jennifer S. Hendricks[*]
The LGBT community may soon win a legal victory that has been decades in the making: passage of the Employment Non-Discrimination Act (ENDA). As passage of the bill becomes more likely, debates about how much to compromise for that victory have become increasingly important. The current version of ENDA represents a series of compromises that should now be reconsidered.
After years of failed attempts to add “sexual orientation” to Title VII of the Civil Rights Act of 1964, ENDA’s proponents decided they would have a better chance with a stand-alone bill stripped of several of Title VII’s protections: they gave up disparate impact claims and affirmative action as a remedy for proven discrimination. Most recently, Representative Barney Frank, the Act’s sponsor, agreed to remove protection for transgendered people in order to win passage in the House.
Less than a year later, a federal court showed Representative Frank the risks of compromise. In September 2008, the D.C. district court held in Schroer v. Billington that transgender people are already protected by Title VII’s ban on sex discrimination. This decision would have been less likely if ENDA had passed last year because the enacted version would have excluded gender identity claims. Moreover, the current version of ENDA has the potential to do real harm not only to transgendered plaintiffs but also to lesbian, gay, and even straight-identified plaintiffs whose claims are based on sex stereotyping.
This Essay proposes a revision of ENDA that would avoid those harms and at least leave the door open to gender identity claims. Instead of enacting ENDA as currently drafted, the next Congress should amend the “because of sex” provision of Title VII to prohibit discrimination “because of gender,” defined to include sexual orientation and gender identity. “Sex” refers to an anatomical classification as male or female, while “gender” refers to socially constructed norms associated with sex. As explained in Part I, the courts have already interpreted Title VII to allow some kinds of claims based on gender rather than sex but have refused to recognize sexual orientation as part of gender. Part II notes Congress’s response when the courts wrongly adopted a narrow interpretation of “sex” that excluded pregnancy: Congress responded effectively through Title VII instead of enacting a separate law about pregnancy. Part III argues that Congress should do the same for sexual orientation and gender identity.
I. The History of “Sex”
Title VII of the Civil Rights Act declares it unlawful to discriminate against a person in employment “because of such individual’s . . . sex.” Despite this clear language, sex discrimination law got off to a slow start. Many judges believed that “sex” had been added to the Civil Rights Act solely as an attempt to kill the bill and were reluctant to enforce the ban on sex discrimination.
Eventually, courts accepted that sex discrimination really was against the law. Having leaped that hurdle, they at times showed surprising willingness to innovate and redress subtle forms of sex discrimination. For example, the creation of sexual harassment law was a nearly wholesale adoption of feminist theory by the judiciary. The courts also recognized sex-plus and sex stereotyping claims by women who were punished for acting either too masculine or too feminine.
There were limits, however, to judicial creativity and insight into gender-based oppression. Courts floundered embarrassingly in cases involving sexual harassment of male plaintiffs, especially when the perpetrator was also male. They refused to recognize that the logic of sex-plus applied to sexual orientation, and until the D.C. district court’s decision in Schroer, had extended it to gender identity claims only when they fit the model of sex-stereotyping claims. Workers thus remain largely unprotected from discrimination based on core aspects of their gender.
II. The Pregnancy Precedent
Judicial resistance to sexual orientation and gender identity claims was not the first time the federal courts embraced a narrow definition of discrimination “because of sex.” In Geduldig v. Ailleo, the Supreme Court infamously declared that pregnancy discrimination was not sex discrimination because it merely distinguished between “pregnant women and non-pregnant persons.”
Congress responded not by adding “pregnancy” to the list of forbidden classifications but instead by defining “because of sex” to include “because of pregnancy.” By crafting the Pregnancy Discrimination Act (“PDA”) not as a separate discrimination law but as an amendment to Title VII, Congress not only insisted that pregnancy discrimination be illegal but also symbolically rejected the reasoning of Geduldig. Although the Court has never repudiated that reasoning, the PDA is nonetheless part of the ongoing dialogue between the courts and the legislature on the meaning of equality. This dialogue eventually led the Court to uphold the Family and Medical Leave Act (“FMLA”) as sex equality legislation, suggesting that Congress’s view has prevailed. A Court that can perceive sex discrimination in the denial of caretaking leave has come a long way since Geduldig.
Rather than enact a stand-alone ENDA or add “sexual orientation” to the list of prohibited classifications within Title VII, Congress should do what it did with the PDA. Title VII should be revised in two respects:
(1) The word sex in 42 U.S.C. § 2000e-2, which prohibits discriminatory employment practices, should be replaced by gender wherever it appears. For example, 42 U.S.C. § 2000e-2(a)(1) would say, “It shall be an unlawful employment practice for an employer to fail or refuse to hire or to discharge any individual, or otherwise discriminate against any individual with respect to his or her compensation, terms, conditions, or privileges of employment, because of such individual’s race, color, religion, gender, or national origin . . . .”
(2) The first sentence in the definition of “because of sex,” found at 42 U.S.C. § 2000e(k), should be revised to say, “The terms ‘because of gender’ or ‘on the basis of gender’ include, but are not limited to, because of or on the basis of sexual orientation; gender identity; and pregnancy, childbirth, or related medical conditions.”
On one hand, this amendment would tell the courts that they got it wrong when they failed to make the connections between sex, gender, and sexual orientation. On the other, by passing it, Congress would endorse much of what the courts have done in sex-plus and sex stereotyping claims as well as correct course on sexual orientation.
III. The Case for a Gender Amendment
As currently written, the freestanding version of ENDA is dangerous in its treatment of gender identity. Previous versions of ENDA, proposed over the years, referred only to sexual orientation. In 2007, the sponsors added gender identity to the bill. They took it back out to win passage in the House, thereby losing much of the bill’s support from advocacy groups. The D.C. district court’s decision in Schroer sheds timely light on the controversy over whether to push for gender-identity protection and also reveals two additional advantages of a gender amendment.
First, a stand-alone ENDA that excludes gender identity will undermine the progress that has been made for sex-plus and sex stereotyping claims. Admittedly, judge-made law on sex stereotyping has not been looking too good lately, especially when it comes to personal appearance as a manifestation of gender identity. The Ninth Circuit recently held that a casino grooming policy requiring women to give themselves makeovers every morning was equal to requiring men to trim their fingernails and comb their hair. Schroer is a welcome push in the opposite direction, with its long-overdue recognition that discrimination based on gender identity is literal sex discrimination as well as a form of stereotyping. The same is true for discrimination based on sexual orientation. However, Congress’s failure to enact ENDA has served as a signal to many courts not to pursue this logic. Passing ENDA as a freestanding law, without protection for gender identity, would likely be seen as a rebuff to the D.C. court and an implicit endorsement of a narrower view.
At a minimum, excluding gender identity from ENDA would likely lead to the overruling of Schroer’s holding that gender identity itself is a protected characteristic because it is part of “sex.” In addition, exclusion of gender identity has the potential to undermine other sex stereotyping claims, especially when brought by transgender, gay, or lesbian plaintiffs. The exclusion would invite the defense that an employer acted not on the basis of sex or sexual orientation but instead on inappropriate gender presentation. For these reasons, an ideal gender amendment to Title VII would define gender to include gender identity as well as sex and sexual orientation.
If that is not in the cards, however, a narrower gender amendment, leaving out the explicit reference to gender identity, has a second advantage over a freestanding ENDA. It would provide, frankly, the opportunity for Congress to punt on gender identity. A gender amendment would explicitly include sexual orientation and would endorse sex stereotyping theory. If “gender identity” cannot be explicitly included in the definition of “gender,” the statute would at least provide opportunities for litigants to continue to push the law in that direction. Given that the courts are now split on whether gender identity is already covered by Title VII, and a few have embraced claims by transgender plaintiffs under the rubric of sex stereotyping, Congress’s failure to speak to the issue would not definitively resolve it one way or the other. A general reference to gender is better than conspicuous exclusion of gender identity.
Another important reason to unite ENDA with “because of sex” is to avoid perpetuating the problem of intersectionality that is already inherent in Title VII. As Kimberlé Crenshaw first demonstrated, many Title VII plaintiffs are confounded by the statute’s category-by-category approach. When a Black woman is discriminated against, she is asked to separate her race from her sex and tell the court which one is the basis for her claim. If she claims it is both, she must prove them separately. If white women and black men have fared tolerably well in her workplace, she will lose on both counts.
The same fate awaits lesbian plaintiffs if ENDA is passed as a stand-alone statute rather than as a gender amendment to Title VII: if straight women and gay men have fared well, the lesbian plaintiff will lose on both ENDA and Title VII counts. Moreover, having separate statutes with separate remedial structures will make it even more important for the factfinder to isolate the claims, parse the evidence more finely, and ignore intersectionality. The difference is that Congress has a clear opportunity with ENDA to avoid the problem by instead passing a gender amendment. As a plaintiffs’ lawyer, I can attest anecdotally that a nontrivial number of sex discrimination claims are “really” about actual or perceived—often closeted—sexual orientation. Such claims would include, for example, cases in which a woman was treated differently from a man, but the driving force behind the employer’s hostility was her sexual orientation. Those claims will become, in some ways, more difficult to win if sex discrimination and sexual orientation discrimination are deemed so distinct that they do not even belong in the statute.
Finally, tying ENDA to sex discrimination with a gender amendment to Title VII instead of a stand-alone statute could have important consequences for the scope of Congress’s power to prohibit discrimination on the basis of sexual orientation and gender identity. If ENDA is to be fully enforceable against the states, it must be backed up not just by the commerce power but also by Congress’s power to enforce the Fourteenth Amendment. So far, the Supreme Court has found the Fourteenth Amendment power applicable to civil rights laws built around established suspect classes such as race and sex but not other classifications such as age and disability. A freestanding ENDA virtually guarantees that sexual orientation will fall into the “other” category.
Congressional action, however, can sway the Court’s view of Congress’s Fourteenth Amendment power. The Court takes umbrage when Congress tries to flex its muscle where the Court has already marked a limit. But if Congress acts first by defining sex to discrimination to include gender identity, as with the FMLA, the congressional understanding of the Fourteenth Amendment where the recognized suspect class of sex includes gender can prevail. Indeed, in light of the decision upholding the FMLA, even the PDA almost surely falls within the Fourteenth Amendment umbrella. Congress should similarly insist that discrimination based on sexual orientation and gender identity is of a piece with sex inequality. If it does, there is a better chance that the Court will decide that equal opportunity regardless of gender “occupies a more protected position in our constitutional system than does the movement of cattle, fruit, steel and coal across state lines.”
A gender amendment might be unsatisfactory to some supporters of ENDA if it is perceived as failing to give adequate symbolic recognition to the LGBT community. The LGBT movement is not a footnote to feminism. I do not think that the gender amendment would diminish the moral significance of an act banning sexual orientation and gender identity discrimination; to the contrary, those claims will be strengthened. In a legal sense, all claims to nondiscrimination in this country are footnotes to the struggle for racial equality because the principle equality entered the Constitution through the Fourteenth Amendment. Having a separate law would be a lesser symbolic achievement than joining the constitutional fabric of which the Civil Rights Act of 1964 is a central piece.
The gender alternative to ENDA would still be
far from perfect. Even the most
inclusive version of ENDA proposed so far was startling in its embrace of stark
gender binarism, protecting transgender employees only to the extent they
clearly declared themselves to be one sex or the other. The political process may continue to demand
caveats on affirmative action and disparate impact claims. Judicial treatment of the new law could go in
several directions. Judicial enthusiasm
for the PDA, for example, has been mixed. Courts have sliced and diced sex, pregnancy, sex-plus, and sex
stereotyping claims in defiance of Congress’s view that all are of a piece. There
is no guarantee that their enforcement of a gender amendment would be more
enthusiastic. A gender amendment would,
however, create at least a chance for some progress, and a chance to avoid
creating new problems for intersectional and other non-archetypal plaintiffs.
*. Associate Professor, University of Tennessee College of Law.
1. LGBT stands for lesbian, gay, bisexual, and transgender. “Transgender” describes people whose social identity differs from the biological sex to which they were assigned at birth, regardless of whether they have undergone sex reassignment surgery or other treatment. “Transsexual” refers to a transgender person who has undergone such treatment or would like to do so. Discrimination against transgender people is referred to as discrimination on the basis of “gender identity.”
2. Examples of failed versions of ENDA are S. 2056, 104th Cong. (1996); H.R. 1863, 104th Cong. (1995); H.R. 4636, 103d Cong. (1994); H.R. 2074, 96th Cong. (1979); H.R. 8269, 95th Cong. (1977); H.R. 166, 94th Cong. (1975).
3. See 153 Cong. Rec. H13228, H13252 (daily ed. Nov. 7, 2007). The text of the bill that passed the House, H.R. 3685, 110th Cong. (1st Sess. 2007), is at pages H13241-43. The earlier version of the bill, with protection for gender identity, was H.R. 2015, 110th Cong. (1st Sess. 2007).
4. See Schroer v. Billington, No. 05-1090, 2008 WL 4287388 (D.D.C. Sept. 19, 2008). The plaintiff in Schroer planned to begin her new job with a female gender identity and to undergo male-to-female sex reassignment surgery a year later. The Schroer court agreed with a growing trend in the Courts of Appeals in holding that discrimination against a transgender employee can be cognizable as a sex stereotyping claim, where the discrimination occurs because the employee is insufficiently masculine or feminine. Schroer, however, went further, holding that discrimination because of transsexuality itself was literally discrimination because of sex.
5. 42 U.S.C. § 2000e-2.
6. See Michael Evan Gold, A Tale of Two Amendments: The Reasons Congress Added Sex to Title VII and Their Implications for the Issue of Comparable Worth, 19 Duq. L. Rev. 453, 453–54 (1981) (describing scholarly and judicial attitudes toward ban on sex discrimination and rebutting that view based on legislative record).
7. See generally Catharine A. MacKinnon, The Sexual Harassment of Working Women: A Case of Sex Discrimination (1979).
8. A sex-plus claim is one in which the basis for discrimination is sex combined with some other characteristic. For example, an employer might hire both women and men but refuse to hire women with young children. See Phillips v. Martin Marietta Corp., 400 U.S. 542 (1971) (holding that employer’s refusal to hire women with young children is sex discrimination under Title VII). A sex stereotyping claim is a special case of sex-plus, in which the “plus” is a characteristic stereotypically associated with sex. For example, an employer might fault women but not men for acting assertively or for not being pretty enough. See Price Waterhouse v. Hopkins, 490 U.S. 228 (1989) (finding sex discrimination where employee was told to wear makeup and to talk and walk more femininely).
9. See Oncale v. Sundower Offshore Servs., Inc., 523 U.S. 75, 79 (1998).
10. For decisions rejecting sexual orientation as an aspect of “sex” under Title VII, see, for example, Simonton v. Runyon, 232 F.3d 33, 35 (2d Cir. 2000); Higgins v. New Balance Athletic Shoe, Inc., 194 F.3d 252, 259 (1st Cir. 1999); Williamson v. A.G. Edwards & Sons, Inc. 876 F.2d 69, 70 (8th Cir. 1989). Several early cases also rejected claims by transsexual plaintiffs. However, since Price Waterhouse established sex stereotyping as a legitimate basis for a claim, some courts have reconsidered and accepted those claims if the discrimination can be characterized as based on stereotypes about the plaintiff’s masculinity or femininity. See Smith v. Salem, 378 F.3d 566, 574–75 (6th Cir. 2004) (reviewing cases). Schroer is the first case to hold that gender identity itself is “sex” under Title VII.
11. 417 U.S. 484 (1974).
12. Id.at 496 n.20. Geduldig was a constitutional holding. Its logic was extended to Title VII in Gilbert v. GE, 429 U.S. 125 (1976).
13. 42 U.S.C. §§ 2000e2000e-17.
14. See Nev. Dep’t of Human Res. v. Hibbs, 538 U.S. 721 (2003).
15. The “or her” is also a change.
16. See 153 Cong. Rec. H11383 (daily ed. Oct. 9, 2007) (statement of Rep. Frank) (explaining decision to remove gender identity from bill); Equality Federation Statement on the House Vote on the Employment Nondiscrimination Act (Nov. 7, 2007), available at www.ENDAunited.org (ad hoc umbrella organization of LGBT advocacy groups pledged to oppose any version of ENDA that fails to cover gender identity) (link).
17. See Jespersen v. Harrah’s Op. Co., Inc., 444 F.3d 1104 (9th Cir. 2006) (en banc). The court analyzed the case as if the problem were whether the employer’s policy imposed a greater burden on women or men in terms of money or time. On this point, the court absurdly refused to take judicial notice of the fact that the requirements for women were far more costly and time-intensive than those for men. The court suggested that the case might have come out differently if the plaintiff had put her receipts in evidence. This analysis ignored the true nature of the complaint. The plaintiff’s objection was not that it took her longer and cost her more to dress for work but that she was forced to conform to a stereotyped, sexualized image of femininity.
18. See, e.g., Simonton v. Runyon, 232 F.3d 33, 35 (9th Cir. 2000) (“Admittedly, we have little legislative history to guide us in interpreting the Act’s prohibition against discrimination based on sex. . . . But we are informed by Congress’s rejection, on numerous occasions, of bills that would have extended Title VII’s protection to people based on their sexual preferences.”) (internal quotation marks and citation omitted); Bibby v. Phila. Coca Cola Bottling Co., 260 F.3d 257, 261, (3d Cir. 2001) (following similar reasoning).
19. This argument is not meant to imply that it would be proper for a court to rely on Congress’s failure to act on gender identity in 2009 as a guide for interpreting a law passed in 1964, but merely to predict that courts will do so—just as they have relied on the nonpassage of ENDA to carve sexual orientation out of Title VII.
20. See Kimberlé Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics, 1989 U. Chi. Legal F. 139.
21. See Seminole Tribe of Fla. v. Florida, 517 U.S. 44 (1996) (holding that pre-Civil War enumerated powers cannot abrogate state sovereign immunity).
22. See Nev. Dep’t. of Human Resources v. Hibbs, 538 U.S. 721 (2003) (sex); Bd. of Trs. of the Univ. of Ala. v. Garrett, 531 U.S. 356 (2001) (disability); Kimel v. Fla. Bd. of Regents, 528 U.S. 62 (2000) (age); Katzenbach v. Morgan, 384 U.S. 641 (1966) (race).
23. See, e.g., City of Boerne v. Flores, 521 U.S. 507 (1997).
24. This optimistic prediction about the PDA would not have been possible before Hibbs. See Jennifer S. Hendricks, Women and the Promise of Equal Citizenship, 8 Tex. J. Women & L. 51, 63–70 (2000) (discussing likely invalidity of PDA as applied to states under then-existing law); Joan C. Williams, Hibbs as a Federaliam Case, Hibbs as a Maternal Wall Case, 73 U. Cin. L. Rev. 365, 365 (2004) (noting unexpectedness of result in Hibbs).
25. Edwards v. California, 314 U.S. 160, 177 (1941) (Douglas, J., concurring).
26. Compare UAW v. Johnson Controls, 499 U.S. 187, 198–99 (1991) (relying in part on PDA to invalidate fetal protection policy), with Puente v. Ridge, 2005 WL 1653017, *4 (S.D. Tex. July 6, 2005) (dismissing claim by woman forbidden to take breaks to express milk, while other employees were permitted to take similar breaks to smoke, on grounds that lactation was not “related to pregnancy”).
27. See, e.g., Troupe v. May Dep’t Stores Co., 20 F.3d 734, 737–38 (7th Cir. 1994) (holding that firing a pregnant employee because her supervisor thought she would not come back to work after having the baby was not actionable under the PDA; similar action would clearly be actionable under sex-plus theory).
Copyright 2008 Northwestern University
Cite as: 103 Nw. U. L. Rev. Colloquy 209 (2008).
Persistent URL: http://www.law.northwestern.edu/lawreview/colloquy/2008/43