The use and misuse of medicine in Skrmetti

This is part of SCOTUSblog’s term in review series, in which scholars analyze some of the most significant cases of the 2024-25 Supreme Court term. For a contrasting view on United States v. Skrmetti, please see this piece by Erin Hawley.

To uphold Tennessee’s ban on pediatric gender-affirming care, the court in United States v. Skrmetti hid from the specter of discrimination. The court’s opinion repeatedly highlighted statutory language (which appeared only once in the statute) that targeted certain diagnoses. Yet it all but ignored language that appeared several times in the statute that explicitly targeted minors based on sex. Thus, the court concluded that the law did not discriminate against people – it only discriminated against medical diagnoses. Such reasoning mischaracterized the arguments at issue, the court’s prior doctrine, and the medicine involved.

The Case

Skrmetti involved a challenge to Tennessee’s law that banned minors from receiving medical care that enabled them “to identify with … a purported identity inconsistent with the minor’s sex” or that was “discordan[t]” with their sex. The law defines sex as “a person’s immutable characteristics … as determined by anatomy and genetics existing at the time of birth.” The statute repeats the prohibition on sex inconsistent/discordant treatment several times – once in the statutory definitions and no fewer than three times in its operative provisions.

The discrimination is obvious. The statute facially classifies based on sex: sex attributed at birth determines drug access. Cisgender girls can grow unwanted body hair (diagnosed among them as hirsutism), which may cause distress by clashing with their gender identity. So do many transgender girls. But through an exception targeted at cisgender conditions, Tennessee’s ban allows the cisgender girl to access medical treatment, but not the transgender girl. Cisgender and transgender boys would similarly experience unequal treatment for analogous conditions.

Under existing Supreme Court doctrine, laws that discriminate based on certain categories like race and sex receive closer judicial scrutiny. But sex discrimination can sometimes be justified as serving important goals. The litigants – both the Biden administration and the private plaintiffs – asked the court to send the case back to the lower court to assess whether such goals existed here. Such an assessment would have involved a detailed exploration of district court findings and the medical evidence, where the state would have had to justify its discrimination.

But this the court refused to allow.

Rather, to avoid a remand, the court circumvented the law’s language prohibiting treatment inconsistent or discordant with sex. To do this, it took two approaches: one that discounted the sex classification, and another that made it disappear.

Approach 1: discount the sex classification

The court’s first move was to discount the classification. The court held that that the classification only counted as (a) a “mere reference to sex,” (b) “[i]n the medical context.” Together, this meant that “heightened scrutiny” was inapposite.  

The suggestion that the classification was a “mere reference” to sex was a straw man. Sure, the mere use of the word “sex” in any context does not invite heightened scrutiny. To draw an example from the main dissent: A law that requires certain action “regardless of sex” does not classify and therefore does not invite elevated scrutiny. But the plaintiffs never made that argument. Rather, their point was not just that the statute used the word “sex” but that it used the word to classify: to create differing categories of children based on sex assigned at birth. And sex classifications invite heightened scrutiny.

This was apparent from the single case the court cited for this argument: 2001’s Nguyen v. INS, which upheld a law that automatically gave citizenship to children whose mothers were unmarried U.S. citizens, but not to those whose fathers were unmarried U.S. citizens. Nguyen characterized the government provision not as simply a reference to sex, but as “a gender-based classification.” It then applied heightened scrutiny, seeking “important governmental objectives … substantially related to” the means used.

Skrmetti went out of its way to avoid acknowledging Nguyen’s application of heightened scrutiny. Nguyen is infamous for claiming that “basic biological differences” satisfied heightened scrutiny and justified the sex classification. Skrmetti also emphasized “biological differences between men and women.” But rather than quote from Nguyen, and apply heightened scrutiny, it pulled its “biological differences” language from a random FDA webpage last updated a few weeks into the Trump administration. Skrmetti invoked biological differences, not to satisfy heightened scrutiny (as in Nguyen) but to justify a new “medical context” exception to the application of such scrutiny. What role the “medical context” exception will play in future doctrine, the reader is left to tease out.

Approach 2: make the sex classification disappear

The second move the court made was to simply pretend the sex classification didn’t exist. The court avoided quoting the statute’s prohibition on providing care “inconsistent” or “discordan[t]” with birth-assigned sex until more than halfway through its sex discrimination analysis. But the court explained away this explicit sex classification by holding that the language was essentially equivalent to another statutory provision: one that prohibited the exception that allows treatment for cisgender people from applying to “gender dysphoria, gender identity disorders, and gender incongruence” diagnoses. The issue, then, was medical discrimination, not sex discrimination.

This move justifies the rest of the court’s reasoning. Comparing cisgender and transgender kids made no sense, held the majority: giving a cisgender girl medication to avoid the distress of unwanted hair is simply “a different medical treatment” from doing the same for a transgender girl, because the underlying medical condition is different.

Similarly, 2020’s Bostock v. Clayton County, which held that anti-transgender discrimination violates federal employment-discrimination laws, was inapposite. Bostock held that sex is a but-for cause in anti-transgender discrimination because such discrimination was based, in part, on the discrepancy between the employee’s sex assigned at birth and their actual sex. But, held the court, Bostock’s reasoning did not apply to Skrmetti, because medical categories rather than sex were the but-for cause of any discrimination.

Finally, claims of sex-based stereotyping also failed, according to the court, because the Tennessee legislature was motivated by medical reasons, rather than ones sounding in stereotype: it had found that the treatments are “experimental, can lead to later regret,” and involve significant “risk.” This argument is reinforced in the court’s assessment of the putatively rational basis underlying the law. 

But pretending that the discrimination was solely medical, and not sex-based was simply wrong.

First, it was incorrect to read away the sex inconsistent/discordant treatment prohibition and subsume it into the gender dysphoria treatment prohibition. As a trivial matter, such a reading violates the rule against surplusage – different language in statutes usually means different things. Here, that is indeed the case. All transgender people experience identity that is “inconsistent” with birth-assigned sex. Not all transgender people who seek medication are diagnosed with “gender dysphoria” and the other diagnoses listed. The two prohibitions do not operate coterminously.

Even if the provisions are collapsed, the court did it the wrong way round. The statute mentions prohibition on treatment “inconsistent”/“discordan[t]” with “sex” assigned at birth three times in its operative provisions and once in the findings. The medical condition prohibition, which lists gender dysphoria and related conditions and which the opinion recites with far greater frequency, was mentioned only once. If one provision had to be read using another, it would thus have made more sense to read the medical condition prohibition using the sex-explicit provision rather than the reverse.

But even if the court was right that the question was one of diagnostic discrimination rather than sex discrimination, problems would remain. 

First, medical diagnoses can incorporate sex-discriminatory criteria. Even in cases where diagnosis codes differ, it makes no sense to rely on medical discrimination to validate legal discrimination. Some doctors think of cardiac events differently for men and women, but sex is just one of many other characteristics they take into account. But medicine has its own history of sex stereotyping. If doctors suddenly elevated sex as a key factor and used different diagnosis codes for the same cardiac events in men and women, it wouldn’t automatically justify discrimination in the law.

Second, it is not always the case that medical encounters involve different diagnostic terms if the child is trans rather than cis. Doctors may sometimes use the same diagnosis code no matter how the patient identifies. In general, medical diagnosis line-drawing is a difficult process; its application to a specific patient is even harder. The exact treatment plan will depend on a range of factors – whether the child is trans is just one such factor. 

Looking forward

It is important to acknowledge that had the court not dodged the discrimination claim, the outcome could have been even more troubling. If the majority had squarely held that sex discrimination did exist, but heightened scrutiny did not apply in medical contexts, or that Bostock’s reasoning did not apply to the Constitution, or that transgender discrimination existed but did not merit heightened scrutiny, it would have upended the decisions of several lower courts, including appellate courts, that have held otherwise, and made the landscape for transgender rights litigation much bleaker. 

But the court has also made challenges to future anti-trans provisions fraught. When they can, policymakers will target transgender people using medical language. For example, in May, the court let the Trump Administration’s ban on transgender servicemembers move forward. The ban, the Department of Defense argued, did not discriminate based on transgender status, but rather, based on gender dysphoria, and related conditions, and that eliminating such conditions from the military served the goals of military readiness. While the court did not issue a reasoned opinion, it reversed a lower court’s stay that prevented the ban from going into effect. After Skrmetti, other provisions will be worded to target gender dysphoria, and will likely be upheld.

By splitting the baby, the court has allowed some constitutional trans-rights litigation to move forward. But the opinion’s reasoning will make equal protection challenges much harder. And the cost is an incoherent opinion that ignores the statute’s text, mischaracterizes the arguments of the plaintiffs, and creates a doctrinal muddle that will plague sex discrimination litigation for years to come.