Some of those changes have been obvious, as when the Court overruled Roe v. Wade and deprived half the country of a right they had counted on for almost half a century. But overturning precedent is costly to an institution whose legitimacy turns significantly on its commitment to be bound by precedent, and so it must be done sparingly.
In the 2024–2025 term, the Court did not explicitly reverse any precedents, but it certainly made new law, by the less obvious path of misreading or simply refusing to follow prior cases. In three of its most significant decisions—involving gender-affirming care for minors, Internet pornography, and religious objections to public school curricula—the Court divided 6–3 along partisan lines and made up new rules that seemed designed to reach a conservative result that would have been foreclosed had it followed precedent. In a fourth major 6–3 decision, involving Trump’s attempt to deny constitutionally guaranteed birthright citizenship to the children of certain immigrants, the Court yet again rewrote the rules, this time to restrict lower courts’ authority to issue nationwide injunctions.
The Court also had an unusually active “emergency docket,” with many cases involving requests to stay lower court rulings against the Trump administration. With a few notable exceptions, it granted most of those requests, generally without explanation and often by the same 6–3 margin.
Taken together, these decisions reveal a conservative majority that is willing to rewrite the rules to reach its desired results and that at least on initial review appears less skeptical about many of Trump’s initiatives than numerous lower courts have been. The Court will be called upon over the next three years to review many of those initiatives on the merits. It is too early to write off the Court; after all, as Noah Feldman recently argued, what is the alternative?1 But the early signs are not encouraging.
Constitutional law in the United States is governed by a set of meta-rules about the “standards of review” that courts should apply to particular kinds of government actions. These rules are designed to focus the courts’ attention on the sorts of actions that warrant the most skepticism, while calling for deference to the political branches on other kinds of decisions.
Thus, for example, when a law distinguishes on the basis of race, the state must satisfy “strict scrutiny,” which requires it to show that the law is the most narrowly tailored means of furthering an extremely important, or “compelling,” interest. The same strict scrutiny applies to laws that restrict speech based on its content. Laws that discriminate on the basis of sex must satisfy “heightened scrutiny,” only slightly less demanding than strict scrutiny; sex-based laws will be upheld only if the state can show they are closely tied to an important state interest. By contrast, most economic or social legislation is given “rational basis” review and upheld as long as any rational justification can be imagined for it.
These rules impose some discipline on the courts by specifying the applicable standard of review rather than leaving it to judicial discretion. But they don’t impose discipline if a court can simply refuse to apply them when they would produce a result the majority doesn’t prefer in a particular case. The Supreme Court did exactly that this term in United States v. Skrmetti, which challenged a Tennessee law that banned the provision of hormones and puberty-blocking medication to minors if prescribed for the purpose of “enabling a minor to identify with, or live as, a purported identity inconsistent with the minor’s sex.” The law defined “sex” as sex designated at birth. Thus a child born male can take testosterone or puberty blockers to conform to male norms, but a child born female cannot.2
Tennessee’s law classifies on the basis of sex on its face. If a law forbade minors from pursuing careers or wearing makeup “inconsistent with the minor’s sex,” we would all understand it as sex discrimination. Such laws enforce stereotypes about gender roles. And ever since the Court recognized in the 1970s that the Equal Protection Clause forbids sex discrimination, it has applied heightened scrutiny to any law that classifies on the basis of sex, in part to flush out laws based on sexual stereotypes.
Had the Court applied heightened scrutiny to Tennessee’s law, it almost certainly would have struck it down, as every lower court judge to apply such scrutiny had. The state defended the law on the grounds that it protects teens, especially the small percentage who later conclude that they do not want to transition, from the medical risks of gender-affirming care. The drugs can cause infertility, among other things. But for many minors, they offer invaluable relief and avoid the more intrusive measures needed if one transitions after puberty.
So why a categorical ban, and why only on this one form of medical treatment? All medical interventions pose risks. Nonetheless, when, as in this case, the medical profession deems a treatment appropriate or necessary in some instances, the state generally leaves the decision about whether to pursue the treatment to the family and their doctors. Tennessee noted that several European countries have recently cut back on access to gender-affirming care, citing concerns about the evidentiary support for the treatment. But none of those countries bans the treatment altogether, regardless of a particular person’s need for it.
Moreover, the Tennessee law expressly states that its goals include “encouraging minors to appreciate their sex” designated at birth. And that is nothing less than enforcing a stereotype. The vast majority of children born male will identify as men throughout their lives, but some will not. And a generalization that is true of most but not all is the very definition of a stereotype.
How, then, did the six conservative justices uphold the law? By declining to apply heightened scrutiny. They asserted that the law classified on the basis of medical use, not sex. Taking hormones and puberty blockers for purposes “inconsistent with sex,” the Court reasoned, is a different medical treatment than taking the same drugs for a purpose “consistent with sex.” But that is circular; the Court simply built the sex distinction into its conception of medical treatment. And the law banned any use “inconsistent with sex,” not just gender-affirming care. Tennessee chose to define the prohibited conduct in sex-specific terms; the Court simply chose to look the other way.
Some doubts have been raised recently about the scientific basis for gender-affirming care, but those doubts do not support a categorical ban under all circumstances. In any event, that uncertainty was not the basis for the Court’s ruling. Instead it simply applied the law’s most deferential standard, rational basis review, which allowed it to uphold the law regardless of what the evidence showed. In a case where calling balls and strikes under existing rules would have required the Court to call this one a strike, it changed the rules.
In Free Speech Coalition v. Paxton, the Court used the same ploy to affirm, by the same vote, a Texas law mandating that websites with material of which more than one third is “harmful to minors” verify that visitors are adults by requiring them to submit either government-issued identification or “transaction data” such as mortgage records. The Court has long ruled that “obscenity”—namely, sexually explicit material that appeals to the “prurient interest,” is “patently offensive,” and lacks any redeeming social value—is unprotected by the First Amendment. But most sexually explicit material is not obscene under contemporary standards, as is reflected by the virtual absence of any obscenity prosecutions nationwide. So most pornography is protected speech for adults, even though states can prohibit minors from accessing it.
On four prior occasions the Supreme Court had ruled that where laws restrict adults’ access to sexually explicit material in order to protect minors, the laws must satisfy strict scrutiny. That’s because such laws regulate speech on the basis of its (sexually explicit) content, and content-based laws require skeptical review out of concern that the government may be seeking to suppress disfavored views. The prior cases involved limits on access to such material on the telephone, on cable television, and on the Internet. Indeed the Texas law is virtually identical to a federal statute, the Child Online Protection Act, that the Court subjected to strict scrutiny in 2004 and that the lower courts eventually struck down on that basis.
This time, though, the Court applied only a much more deferential “intermediate scrutiny” test. It had previously applied that standard only to content-neutral restrictions on speech. The Texas law is obviously not content-neutral, but the Court said intermediate scrutiny was appropriate because the law imposed only an “incidental” burden on adult speech. As Justice Elena Kagan pointed out in a dissent that eviscerated the majority’s reasoning, there is nothing “incidental” about the Texas law’s effect on adult speech. Under First Amendment doctrine, to say that a law “incidentally” burdens speech is just another way of saying the law is neutral as to content. A law barring all public burning, for example, would have only an incidental effect on someone who sought to burn the American flag, because it regulates the conduct of public burning without regard to what, if anything, the burning communicates. But Texas’s law regulates only speech, and precisely because of its sexually explicit content. The law also requires websites to post warnings that “pornography is potentially biologically addictive” and “proven to harm human brain development,” making clear that the state disfavors the speech for all persons, not just minors.
The majority openly worried that the Texas law might not survive strict scrutiny, but that concern is no justification for failing to apply the well-established standard for content-based laws that the Court had applied four times previously to similar restrictions. In any event, as Kagan pointed out, a properly tailored age verification requirement could well satisfy strict scrutiny, as long as the state could show there is no less restrictive way to further the interest of protecting kids. Texas’s law was anything but narrowly tailored; it applied to all of a website’s material, even if only one third was “harmful to minors”; it allowed unlimited access to the same images through Google searches, the mechanism kids use most; and it imposed onerous verification requirements without even trying to pursue content-filtering alternatives. Under intermediate scrutiny, however, careful tailoring is not required. Even if there are less onerous ways to protect children, through more identity-protective age verification methods, for example, the Texas law survives under the majority’s deferential approach. Just as the Court justified its result in Skrmetti by declining to apply heightened scrutiny to a sex-based law, so in Free Speech Coalition it refused to apply strict scrutiny to a content-based statute. Change the rules, change the result.
The third case in which the Court adjusted the standard of review to reach a desired result, this time applying a more demanding test than precedent called for, was Mahmoud v. Taylor. Religiously observant parents claimed that the “free exercise” of religion required public schools in Montgomery County, Maryland, to provide them with notice and an opportunity to pull their children out of class whenever the teachers read aloud from books the parents object to on religious grounds. In this case the offending books portrayed same-sex couples and gay and transgender individuals positively, but followers of another religion could object to books that feature women working outside the home or interracial marriages.
Once again the Court rewrote the applicable standard of review, this time applying strict scrutiny where precedent called for only rational basis review. As a result, public schools across the country must now provide notice to parents whenever their lessons might include any material that any religious adherent might find offensive and accommodate any objecting parent by allowing their child to opt out of the lesson. (Montgomery County initially tried that approach and changed only when the administrative costs proved too much; now every school district will have to shoulder those burdens.)
The school system argued that merely exposing children to such books did not infringe on their parents’ free exercise rights. The Court’s precedent supported that view. In 1990 it ruled that where a law or policy does not single out religion but imposes a “neutral law of general applicability,” the fact that it has the (here, accurately “incidental”) effect of burdening someone’s religious practice does not require heightened scrutiny. In that case, two Native Americans objected to a law prohibiting all uses of peyote, which had the effect of barring them from using it in a religious ritual. The law undeniably burdened their religious practice, but because it prohibited peyote across the board and did not single out religious uses, the Court ruled, the law called for only rational basis review. Justice Antonin Scalia, writing for the Court, explained that because there is no limit to the religious beliefs people can hold, ruling otherwise would make “every citizen…a law unto himself.”
Under that precedent, the school policy in Mahmoud was plainly valid. It applied equally to all and did not single out religion, and it furthered the school system’s legitimate interest in fostering mutual respect and avoiding the serious administrative costs of managing individual opt-outs on an ad hoc basis.
But the majority applied strict scrutiny. It likened the case to a 1972 decision, Wisconsin v. Yoder, in which the Court ruled that the Amish had a right not to be compelled to send their children to high school. Amish practice requires that children be in the Amish community full-time during the high school years. The Court stressed that the law “affirmatively compel[led]” the parents, “under threat of criminal sanction, to perform acts undeniably at odds with fundamental tenets of their religious beliefs.” Compelling them to attend high school would take their children “away from their community” when they “must acquire…the specific skills needed to perform the adult role of an Amish farmer or housewife.”
It is one thing to be compelled to send a child to school when one’s faith requires, in effect, homeschooling. It is quite another to choose to send one’s child to public school and then insist on a right to opt out of any individual lesson deemed objectionable, even if it involves nothing but listening to a book read aloud that features gay people. Yet the majority treated them as the same in order to avoid applying the standard of review that would have produced a different ruling. As a result, the Court interpreted the Constitution as requiring religious intolerance to prevail over mutual respect for all. And the broader pattern that emerges from all three cases, in which the conservative majority simply eschews the applicable rules when they command a result it does not like, is deeply corrosive of the Court’s—and the law’s—legitimacy.
The fourth case in which the Court, again by the same 6–3 vote, declined to follow precedent was Trump v. CASA. It involved a challenge to one of Trump’s first executive orders, which announced that his administration would refuse to honor the citizenship of people born on US soil to certain noncitizen parents. The order flies in the face of the text of the Fourteenth Amendment, Supreme Court precedent, and long-standing executive practice, all of which recognize that those born here are citizens, unless their parents are immune from US jurisdiction at the time of birth, as foreign diplomats are but foreign nationals are not. With one exception, every court that has addressed Trump’s order has declared it unconstitutional. The cases before the Supreme Court were brought by several states and CASA, an organization suing on behalf of its immigrant members. Once the lower courts concluded that the policy was unconstitutional in all its applications, they enjoined it nationwide.
The Trump administration asked the Supreme Court for a partial stay of those injunctions. It did not raise the merits of its birthright citizenship policy, presumably because even it knows that the Court is almost certain to strike down the policy once it addresses it. Instead the administration challenged only the nationwide scope of the rulings and asked that the Court limit the injunctions to the particular plaintiffs who brought the cases. That would mean the policy could take effect in some places and against some individuals, even though the courts found it unconstitutional in all applications.
The Supreme Court held that courts have power to provide relief only to the parties directly before them. In doing so, as Justice Sonia Sotomayor pointed out in dissent, the majority effectively reversed more than a century of practice in which federal courts, including the Supreme Court, issued or affirmed nationwide or statewide injunctions. The practice dates back to the English courts of equity, which the Court’s precedent identifies as a guide to the scope of power Congress gave US federal courts. Nationwide relief also makes sense here, Sotomayor argued: if denying birthright citizenship is unconstitutional anywhere, it is unconstitutional everywhere.
However, despite headlines suggesting otherwise, the Court did not rule out nationwide injunctions altogether. Instead it directed that they are permissible only where necessary to provide “complete relief” to the parties before the court. That rule leaves open several routes to nationwide relief, for example if plaintiffs file a class action on behalf of all persons nationwide who would be subject to the Trump policy, as the ACLU did the same day the Court ruled. In that case, a district court has now indeed found that a nationwide class action is appropriate and has accordingly awarded nationwide relief. The administration will likely appeal, but the CASA decision does not bar the relief now afforded, because only nationwide relief can provide relief to the nationwide class.
The Court also left open the possibility that, even in the absence of a class action, the states may need a nationwide injunction in order to obtain complete relief for themselves. Absent a uniform rule, the states argued, it will be extraordinarily burdensome for them to determine to whom they must provide services owed to citizens. The Court left that argument to be considered on remand by the lower courts.
Justice Brett Kavanaugh noted further in a concurrence that in challenges to federal agency action (as opposed to executive orders), Congress has expressly authorized courts to set aside a federal rule when it is unlawful, thereby permitting nationwide relief in such cases. And he and the majority explained that when the Supreme Court resolves the validity of a federal policy or law, even with regard to a single individual, its decisions are binding nationwide. Indeed, Barrett, in the majority opinion, cited Trump’s lawyer’s concession at oral argument that the administration would abide nationwide by a Supreme Court ruling on birthright citizenship in a single case. So the consequences of the nationwide injunction decision may be less dire than some have predicted.
Finally, the Court this term considered twenty requests for emergency relief in cases involving Trump policies. Such requests, either to stay injunctions granted by lower courts or to issue injunctions denied by lower courts, are generally resolved without oral argument and often without the Court providing any reasoning. The Court ruled for Trump in the vast majority of these emergency matters, but it did rule against him in some significant cases, including disputes involving the freezing of USAID funding, the mistaken deportation of Kilmar Armando Abrego Garcia, and the administration’s unprecedented invocation of the Alien Enemies Act to deport individuals said to be associated with a Venezuelan gang.3
But otherwise the Court ruled for the administration, and often along 6–3 partisan lines. These largely unexplained decisions have had tremendous consequences for those affected—including literally hundreds of thousands of foreign nationals now vulnerable to deportation under Trump policies the lower courts had enjoined and many thousands of federal employees Trump has threatened to fire. The Court’s emergency rulings do not resolve the merits, only the interim status of executive actions while cases challenging them make their way through the appeals process. In coming years the Court will confront the merits of many of these initiatives. Its emergency decisions are not binding on those later appeals, but they allow countless individuals to suffer from potentially illegal acts in the meantime. And these delphic interim rulings technically required the Court to conclude that the administration is likely to prevail when it does take up the merits, so they are not encouraging. The federal judiciary is the only part of the federal government that has demonstrated any willingness to check Trump’s many abuses. But those checks will mean nothing if the Supreme Court declares them void.