Cover of DSM-5 hardback editionWilliams v Kincaid, 2022 WL 3364824 (4th Cir. August 16, 2022) is not the first case to consider gender dysphoria as a disability, and marks the second time the Fourth Circuit has considered the difference between gender identity and gender dysphoria.¹ However, as a circuit court opinion holding that gender dysphoria can be a disability covered by the ADA it has special prominence, not only because of its possible effect on individuals suffering from gender dysphoria, but also because of what it teaches about interpreting the ADA in light of changing science and technology and the possible need for updating the ADA’s definitions in light of those changes.

The ADA has a specific exclusion from the definition of disability for homosexuality and bi-sexuality (42 U.S.C. §12211(a) and for “transvestism, transsexualism, pedophilia, exhibitionism, voyeurism, gender identity disorders not resulting from physical impairments, or other sexual behavior disorders” (42 U.S.C. §12211(b)(1)). Although these were treated differently in the statute, both exclusions appear to be based on the fear that employers and others would not be able to exclude LGBTQ people based on issues of sexuality.²

To understand the importance of Williams v. Kincaid as a guide to interpreting the ADA requires a little history. When the ADA was introduced in 1989 the Diagnostic and Statistical Manual of Mental Disorders III treated anything but traditional straight gender identity or behavior as a form of mental illness. The DSM is the standard reference for mental disorders, and because the ADA definition of disability begins with “a physical or mental disorder” it was not hard to see that gay, lesbian, bisexual, transsexual and other people might claim that the ADA prohibited discrimination against them based on questions of sexuality. This was morally abhorrent to many legislators and so the ADA was written with a specific exclusion from the definition of disability for homosexuality and bi-sexuality (42 U.S.C. §12211(a) and for “transvestism, transsexualism, pedophilia, exhibitionism, voyeurism, gender identity disorders not resulting from physical impairments, or other sexual behavior disorders” (42 U.S.C. §12211(b)(1)). Although these were treated differently in the statute, both exclusions appear to be based on the fear that employers and others would not be able to exclude LGBTQ people based on issues of sexuality² and a belief that these behaviors represented a moral failing rather than a mental disorder.³ The only issue with the exclusion appeared to be that it might exclude individuals with HIV/AIDS because of the association of that disease with homosexual behavior. The exclusion remained after different legislators pointed out that HIV/AIDS was not limited to homosexuals. What the ADA ended up doing was excluding conditions that were regarded as mental disorders in the scientific community but which most legislators thought were merely moral failings.

The treatment of LGTBQ sexuality as a mental disorder continued in DSM IV, but the current version of the DSM – version 5 – no longer treats questions of sexual identity as a mental disorder. It recognizes instead that “gender dysphoria” can be a mental disorder because individuals whose sexual identity does not match their gender at birth may suffer from severe mental distress as a result. In the thirty years since the ADA was passed science has advanced and the notions about sexual identity current in 1989 are no longer considered to have a scientific basis. Note that this is not an observation about social change. Societal attitudes about LGBTQ individuals have also dramatically changed in the last thirty years, but the DSM understanding of sexual identity is based on science, not social change.

This brings us to the analysis in Williams v. Kincaid. Williams is a transgender woman who had lived for fifteen years as a woman and was shown female on her drivers license. She was diagnosed as having gender dysphoria and received hormone therapy as part of her treatment. When arrested and jailed she was originally treated as a woman, but later transferred to male prisoner housing in the county jail and otherwise treated as a man. She was denied her hormone therapy for at least two weeks and only received it irregularly after that. Not surprisingly she was mocked by male prisoners and suffered at least some abuse from male guards. After her release she sued under the disability discrimination provisions of the ADA and Rehabilitation Act. Kincaid, the sheriff, defended on the exclusion of transsexuality from the definition of disability under the ADA.

The Fourth Circuit rejected this defense based only on principles of statutory interpretation. In its simplest form the argument went like this:

  • transsexuality as used in the ADA must have been based on the definitions of mental disorders in the DSM III (4) in which it was one of several “gender identity disorders.”
  • “gender dysphoria” was not mentioned as a mental disorder in the DSM III.
  • DSM 5 no longer recognizes gender identity disorders, but does recognize “gender dysphoria” as a mental disorder.
  • therefore Congress could not have intended to exclude gender dysphoria from the definition of disability in the ADA.

In other words, Congress only intended to exclude what it knew to be mental disorders in 1989 and the exclusion in the statute could not be re-written to encompass a newly recognized mental disorder even if that mental disorder arose out of transsexuality. As the Fourth Circuit wrote:

Thus, the ADA excludes from its protection anything falling within the plain meaning of “gender identity disorders,” as that term was understood “at the time of its enactment.” Bostock, 140 S. Ct. at 1738. But nothing in the ADA, then or now, compels the conclusion that gender dysphoria constitutes a “gender identity disorder” excluded from ADA protection. For these reasons, we agree with Williams that, as a matter of statutory construction, gender dysphoria is not a gender identity disorder.

2022 WL 3364824, at *6 (emphasis added).
This is interesting enough, but does it have broader ramifications in interpreting the ADA? It should. After all, advances in science and technology since 1989 have not been limited to the understanding of mental illness. Courts arguing one way or the other concerning the application of the ADA to websites and internet based applications almost always assert that Congress never thought about the internet when the ADA passed because it was relatively insignificant at the time. The Fourth Circuit approach – what Congress didn’t think about it didn’t include – supports the law in the various circuits that find stand-alone websites are not places of public accommodation.(5)
This approach may also require re-thinking the meaning of phrases like “architectural barriers” or “communication barriers.” Is the definition of an architectural or communication barrier frozen in terms of assistive technologies available in 1989 or will improvements in technology change the way these phrases are understood? The same may become true of technologies that ameliorate a disability. Congress explicitly stated that the ameliorative effects of various physical devices would not be considered in determining whether a physical disability substantially limits a major life activity(6), but there is no similar provision for gene therapy, another technology not available in 1989. There may be a time when some mental or physical disorders are not reasonably considered disabling because of the effectiveness of new therapies that don’t fit the definition of “medication” in the list of things not to be considered when determining whether a disability exists.
The very high profile controversy about websites as public accommodation has prompted efforts at legislative change(7).The decision in Williams v. Kincaid tells us that this may not be the only statutory reform needed. In Williams v. Kincaid the Fourth Circuit was able to bring the ADA in line with modern medical practice because of a change in vocabulary, but that may not always be true. Congress went to a lot of much trouble to specifically exclude the ameliorative effect of different technologies without considering whether those technologies might, in the future, become the equivalent of “ordinary eyeglasses” whose effects can be considered. Courts can only do so much to bring an old law into a new millennium. Congress may need to do some work as well.


¹ See my blog for a discussion of earlier district court cases, and see Grimm v. Gloucester County Sch. Bd., 972 F.3d 586, 611 (4th Cir. 2020), as amended (Aug. 28, 2020), cert. denied, 141 S. Ct. 2878 (2021) for an earlier discussion of gender dysphoria discrimination.

² In the Congressional Record a poll showing widespread opposition to laws that would forbid discrimination against homosexuals was entered in the record on the same day the first draft of the ADA was presented. (C.R. 101 – May 9, 1989 at pp. 39, 138). The Fourth Circuit certainly viewed the ADA exclusion as evidence of discrimination in Grimm v. Gloucester County Sch. Bd. writing:

First, take historical discrimination. Discrimination against transgender people takes many forms. Like the district court, we provide but a few examples to illustrate the broader picture. See Grimm, 302 F. Supp. 3d at 749 (“[T]here is no doubt that transgender individuals historically have been subjected to discrimination on the basis of their gender identity, including high rates of violence and discrimination in education, employment, housing, and healthcare access.” (collecting cases)). As explained in the Brief of the Medical Amici, being transgender was pathologized for many years. As recently as the DSM-3 and DSM-4, one could receive a diagnosis of “transsexualism” or “gender identity disorder,” “indicat[ing] that the clinical problem was the discordant gender identity.” See John W. Barnhill, Introduction, in DSM-5 Clinical Cases 237–38 (John W. Barnhill ed., 2014). Whereas “homosexuality” was removed from the DSM in 1973, “gender identity disorder” was not removed until the DSM-5 was published in 2013. See Kevin M. Barry et al., A Bare Desire to Harm: Transgender People and the Equal Protection Clause, 57 B.C. L. Rev. 507, 509–10, 517 (2016). What is more, even though being transgender was marked as a mental illness, coverage for transgender persons was excluded from the Americans with Disabilities Act of 1990 (ADA) after a floor debate in which two senators referred to these diagnoses as “sexual behavior disorders.” See Barry et al., supra, at 510; see also 42 U.S.C. § 12211(b)(1). The following year, Congress added an identical exclusion to the Rehabilitation Act of 1973, “stripping transgender people of civil rights protections they had enjoyed for nearly twenty years.” Barry et al., supra, at 556; see also H.R. Rep. No. 102-973, at 158 (1992).

972 F.3d 586 at 611.

³ I found these in the materials collected at The ADA Law Project

(4) Although it did not survive into the final statute the DSM III was explicitly included in one of the early versions as part of the description of what was excluded from the meaning of disability.

(5) The Fourth Circuit has not ruled on this issue, but District Courts in the Circuit have held that stand-alone websites are places of public accommodation. The Ninth and EleventhCircuits say that stand-alone websites are not places of public accommodation and the Third and Sixth have excluded non-physical places of business in other contexts. The First, Second and Seventh have suggested in non-website cases they would extend the ADA to definition of public accommodation in ways that include websites. District courts in all but the Ninth and Eleventh Circuits have taken varying views of the question.

(6) 42 U.S.C. §12102.

(7) see my blog The Online Accessibility Act of 2020 – does it do what it needs to do?