We’re back with the nightcap of today’s doubleheader, consisting of summaries of Moody v. NetChoice (No. 22-277), an important First Amendment case that turned into an important case on facial constitutional challenges; City of Grants Pass, Oregon v. Johnson (No. 23-175), in which the Court concluded that many western cities’ “camping bans” do not violate the Eighth Amendment by effectively criminalizing the status of being homeless; and Texas v. New Mexico (No. 141, Orig.), an original-jurisdiction water dispute in which a closely divided Court rejected a proposed consent decree between the involved states based on an objection of the United States.

In Moody v. NetChoice (No. 22-277) (decided with NetChoice v. Paxton (No. 22-555)), The Nine coalesced around an opinion on the standard for facial First Amendment challenges, while offering a diverse range of views on a number of substantive questions that will impact both these cases on remand and any future cases regarding the application of the First Amendment to speech regulated by private social-media platforms.

To set the stage, both Florida and Texas enacted statutes in 2021 regulating large social-media companies and other internet platforms. While the statutes differ in relevant ways, in each instance, conservative state legislatures (and governors) sought to limit the ability of platforms to engage in content-moderation that they believed disfavored conservative speech. (As described in last week’s decision in Murthy v. Missouri (2024), many conservatives believe that social-media platforms targeted and discriminated against conservative users during the pandemic and following the 2020 election.) In addition to limiting the platforms’ discretion over content moderation, each statute required platforms to provide individualized explanations for their content-moderation decisions, which would of course be an immense burden for platforms hosting hundreds of millions of users.

NetChoice, a trade association whose members include Facebook and YouTube, challenged both laws in federal court. Importantly, in each case, NetChoice brought facial challenges to the statutes, alleging that they violated the First Amendment in all possible applications and not just as applied to NetChoice or its members. District courts in Florida and Texas enjoined the laws, and the Eleventh and Fifth Circuit’s reached differing conclusions on appeal. The Eleventh Circuit upheld the injunction of Florida’s law, holding that it violated social-media platforms’ First Amendment right to “editorial discretion” in choosing what content to host or amplify. The Fifth Circuit, however, vacated the injunction of Texas’s law, holding that the platforms’ content-moderation activities were “not speech” and therefore did not implicate the First Amendment. For good measure, the Fifth Circuit concluded that, even if content moderation was protected speech, Texas had its own interest in “protecting a diversity of ideas” that trumped the rights of the platforms.

In a unanimous, but complicated, decision, the Supreme Court vacated both judgments and remanded for further proceedings, concluding that neither the Fifth Circuit nor the Eleventh had properly analyzed the cases as facial First Amendment challenges. Writing for the Court, Justice Kagan observed that “NetChoice chose to litigate these cases as facial challenges, and that decision comes at a cost.” By design, facial challenges are difficult to win. In the First Amendment context in particular, a plaintiff bringing a facial challenge must show that “a substantial number of [the law’s] applications are unconstitutional, judged in relation to the statute’s plainly legitimate sweep.” The lower courts did not pay much attention to the facial nature of NetChoice’s challenges, focusing instead only on the laws’ application to major platforms like Facebook and YouTube. Given that, the record on appeal was not adequate for the Supreme Court to assess the laws’ scope and determine which applications of the laws might violate the First Amendment and which might not. Because “this is a court of review, not of first view,” the Court unanimously decided to vacate the judgments of both the Fifth and Eleventh Circuits and remand for the lower courts to conduct a proper facial analysis.

That unanimous holding, however, came on the 18th of 96 pages of opinions in the case. Joined by the Chief and Justices Sotomayor, Kagan, Barrett, and (in part) Jackson, Justice Kagan went on to explain at considerable length why the Fifth Circuit’s decision was based on a “serious misunderstanding of First Amendment precedent and principle.” The Court has repeatedly held that requiring a speaker to provide a forum for another’s views implicates the First Amendment when the regulated party is engaged in its own expressive activity, as in the case of a newspaper or parade. Thus, the government cannot require newspapers to give political candidates a “right to reply to critical coverage,” require cable companies to allocate certain channels to local broadcast stations, or require a parade to admit participants who wish to convey a message the parade organizers disagree with. These principles are just as applicable in the context of “ever-advancing technology,” as represented by large-scale social media companies. Facebook’s News Feed and YouTube’s homepage reflect the platforms’ preferences for what speech to highlight and what speech to include. The Fifth Circuit, Kagan mused, was wrong to ignore the platforms’ editorial rights, focusing instead on the fact that the companies’ algorithms typically highlight speech that their algorithms predict users will engage with, as opposed to the platforms’ considered views. And it ignored precedent in concluding that Texas’s purported interest in ensuring a diversity of views somehow trumped the social media companies’ interests in exercising editorial control. Thus, in an epic dictum, the Court warned the Fifth Circuit not to make the same mistakes on remand.

Justice Barrett filed a solo concurrence to flag the “dangers of bringing a facial challenge,” particularly given the range of speech-adjacent, but potentially unprotected, functions that social-media platforms engage in. For example, platforms primarily rely on algorithms to prioritize and remove content on their feeds. But algorithms don’t have free-speech rights. In some cases, the First Amendment would still be implicated, such as where a human being provides the instructions on which speech to favor and disfavor. But it’s conceivable that an algorithm could operate entirely independent of human instruction, particularly given the rapid evolution of Artificial Intelligence. At an even more basic level, the degree of First Amendment protection a social-media platform enjoys may depend on where it is incorporated. Foreign corporations (like other foreign persons) don’t enjoy First Amendment rights. “So a social-media platform’s foreign ownership and control over its content-moderation decisions might affect whether laws overriding those decisions trigger First Amendment scrutiny.” (Psst, she might be looking at you, TikTok….) These and other questions, Barrett noted, show why as-applied challenges are preferable in the context of the First Amendment and social media.

Justice Jackson also penned a solo concurrence, urging the lower courts on remand to “address these cases at the right level of specificity.” “The question,” Jackson said, “is not whether an entire category of corporations (like social media companies) or a particular entity (like Facebook) is generally engaged in expression. Nor is it enough to say that a given activity (say, content moderation) for a particular service ([Facebook’s] News Feed, for example) seems roughly analogous to a more familiar example from our precedent.” Rather, in evaluating a broad facial challenge, courts “must make sure they carefully parse not only what entities are regulated, but how the regulated activities actually function before deciding if the activity in question constitutes expression and therefore comes within the First Amendment’s ambit.” For that reason, Jackson noted that “further factual development” may be necessary before the courts can adequately address NetChoice’s facial challenges.

Next came Justice Thomas, who concurred only in the judgment. He agreed with the Court’s decision to vacate in light of the lower courts’ failure to conduct the proper facial-challenge analysis but disagreed with the majority’s “decision to opine on certain applications of th[e] statutes.” Having criticized the majority’s dictum, Thomas went on to provide some advice of his own, including that the “common carrier doctrine should continue to guide the lower courts’ examination of the trade associations’ claims on remand.” For good measure, Thomas added some thoughts on how the federal courts lack authority to deem statutes “facially unconstitutional” in the first place.

Justice Alito (joined by Thomas and Gorsuch) likewise joined in the judgment vacating the decision below, insisting that “[e]verything else in the opinion of the Court is nonbinding dicta.” Alito criticized the majority for inaccurately describing the Florida and Texas laws, as well as the litigation below, and for providing an “incomplete” summary of the Court’s precedents. The majority’s dicta, Alito insisted, “compelled [him] to provide a more complete discussion of those matters than is customary in an opinion that concurs only in the judgment. But since dictum criticizing dictum is really a bit much for July reading, we’ll leave it to those who may be interested to work their own way through Alito’s 30-page retort. In sum, he took issue with the majority’s view that it did not matter whether a “compiler” of speech like Facebook let in most speech and only excluded a minority of disfavored views. While the majority seemed content to apply existing precedents to new areas, Justice Alito (like Justice Jackson, though perhaps with different priors) seems poised to adopt new rules that take account of the vast influence and power these platforms have over public discourse (a theme of his Murthy dissent, as well.)

So, at the end of the day, all nine Justices agreed that the lower courts needed to take a mulligan on these cases. But while several Justices offered competing advice on how those courts should address NetChoice’s facial challenges on remand, it seems much more likely that NetChoice will follow Justice Barrett’s advice and bring as-applied challenges to the laws.

City of Grants Pass, Oregon v. Johnson (No. 23-175) was perhaps the most peculiar case on the OT23 docket in a couple respects. First, it involves a public-policy problem the Supreme Court rarely confronts: homelessness, a particularly acute crisis in many Western states. And it turns on a philosophical question the Court hasn’t really addressed since Robinson v. California (1962): Can states criminalize a person’s status, as opposed to the person’s conduct, and if not, how do courts draw a line between the two? Writing for the Court’s six conservative justices, Justice Gorsuch concluded that an Oregon city’s laws criminalizing sleeping on public property do not constitute cruel and unusual punishment under the Eighth Amendment, rejecting the Ninth Circuit’s conclusion that those laws were unconstitutional because they effectively made it a crime to be homeless even for those whose homelessness was involuntary.

Grants Pass begins with the problem of homelessness, which some have called the “defining public health and safety crisis in the Western United States,” where most of the U.S. homeless population lives. As a result of both internal and external factors, homeless individuals in these states have increasingly come to reside in communal encampments, many of which are found in or around parks or other public land. One measure many Western cities have taken to address this problem is enacting so-called camping bans, which (depending on the jurisdiction involved) may prohibit everything from setting up long-term encampments with tents and fires to simply sleeping in a public space.

In 2019, the Ninth Circuit seriously limited these laws, holding in Martin v. Boise that Boise, Idaho violated the Eighth Amendment’s Cruel and Unusual Punishments Clause by enforcing its camping ordinance against individuals who lacked “access to alternative shelters,” that is, those who were involuntarily homeless because they had no place to live and no viable means of obtaining temporary shelter. Martin led to wave of litigation in the Ninth Circuit, with homeless advocates and homeless individuals themselves challenging various similar measures enacted by cities within the Ninth Circuit’s jurisdiction. One of those suits targeted the City of Grants Pass, a 38,000 person city in southern Oregon where roughly 600 people experience homelessness on any given day. But Grants Pass’s chosen laws were particularly harsh, imposing substantial fines and imprisonment (for repeat violators) on those who set up a “campsite,” defined as basically any material used for bedding—even a blanket—or “camp” (that is, sleep) in public space, including within a vehicle. Following Martin, an Oregon District Court certified a class of homeless plaintiffs potentially affected by these laws, concluded that many homeless persons in Grants Pass were “involuntarily homeless” because the City’s homeless population far outstrips available shelter beds, and enjoined the City from enforcing its laws against this population. A divided Ninth Circuit affirmed (though narrowing the injunction in some respects), and an en banc Ninth Circuit declined to rehear the case over the dissent of no fewer than 17 judges. The Court granted cert.

Writing for a majority of six, Justice Gorsuch reversed the Ninth Circuit, concluding that the enforcement of generally applicable laws regulating camping in public is not “cruel and unusual punishment” within the meaning of the Eighth Amendment. He began by observing that the Cruel and Unusual Punishment Clause is primarily directed at what kinds of punishment the government can after there has been a violation of its criminal laws, not what conduct those laws can criminalize in the first place, making it an ill fit for the local ordinances at issue here. But he then turned to an alleged “notable exception” to this general rule about the Clause’s scope—the one that was the basis for the Ninth Circuit’s Martin decision and the dissent—Robinson v. California (1962). In that case, the Warren-era Court held that a California statute violated the Eighth Amendment by prohibiting any person from “be[ing] addicted to the use of narcotics” in the State of California. That was so because the law effectively criminalized a “status”—being a drug addict—rather than a particular act. But in the six decades since Robinson was decided, the Court has hardly mentioned it and never relied on it to invalidate any other criminal law. So is it even right?

Justice Gorsuch found it unnecessary to answer that question—with no party asking the Court to reconsider Robinson—because even if it was, it didn’t support the Ninth Circuit’s approach in Martin. That’s because rather than criminalizing a status—e.g., being a homeless person—Grants Pass’s ordinances forbid actions—like creating a campsite or camping in public. And they do so in a generally applicable way: They draw no distinction between people who violate them because they are homeless and those who violate them for other reasons, like vacationing backpackers or protesting college students. The dissent countered that the camping bans of places like Grants Pass and Boise really aren’t that different from Robinson, because they prescribe acts a homeless person (at least an involuntary one) can’t help but take: Sleep is a biological necessity, so a person with no place else to go will necessarily have to sometimes sleep in public. But Gorsuch rejected that argument, finding it foreclosed by Powell v. Texas (1968), where the Court upheld a Texas law that made it a crime to “get drunk or be found in a state of intoxication in any public place” against a challenge brought by an alcoholic who, not unlike the homeless plaintiffs in this case, argued that his violation of the law—being drunk in public—was an “involuntary” result of his status as an alcoholic. Having already declined to extend Robinson’s rationale to similar situations, Gorsuch saw no reason why the Court should do so now.

Gorsuch concluded with a lengthy discussion of two policy points, reinforcing his conclusion that the Eighth Amendment wasn’t the right tool, and courts not the right actors, for addressing the problems of chronic homelessness. First, he recognized that the involuntary nature of homelessness may have some bearing on whether Grants Pass could properly charge and convict homeless people under its laws. But the tools for raising those arguments weren’t an Eighth Amendment-based injunction barring enforcement of the ordinances; rather, it was for defendants charged with violating the ordinance to raise defenses like necessity, diminished capacity, or duress in criminal proceedings. Second, he noted the enormous practical problems created by the Ninth Circuit’s approach in Martin: It is practically impossible for courts to clearly define or determine when a person is involuntarily homeless (so as to preclude enforcement of camping bans and other laws). And courts’ constitutionalizing of this area of public policy prevents more representative government entities (like states and cities) from crafting more tailored and perhaps effective approaches to the homelessness crisis. For these reasons, as well as several others, Gorsuch concluded the better path was to leave this area of law to local legislatures.

Justice Thomas joined the Court’s opinion in full. But he also concurred, solely for himself, arguing that the Ninth Circuit’s decision was wrong because Robinson was too. In his view, that case departed from the Eighth Amendment’s original understanding, which merely precluded certain forms of punishment judged cruel and unusual at the time of the Amendment’s ratification.

Justice Sotomayor, joined by Justices Kagan and Jackson, dissented. She began by discussing at length the problem of homelessness, focusing particularly on two facts downplayed by the majority: that a significant percentage of homeless individuals are homeless due to events outside their control and that studies have consistently shown that punishing homelessness (whether through fines or imprisonment) is at best ineffective and at worst counterproductive.

Sotomayor then turned to Robinson, criticizing the majority’s understanding of the case as “plainly wrong” and asserting that “[e]very shred of evidence” shows that Grants Pass’s laws were intended to and did in fact criminalize the status of being a homeless person. While the majority described these laws as ones of general applicability, citing backpackers and students as examples of other who could be ensnared in them, the record made clear that Grants Pass’s City Council passed these ordinances specifically to combat “vagrancy problems” and that it designed the ordinances to be so excessively punitive that they would simply drive homeless people away from the region. The laws’ text points to the same place: Contrary to the majority, they don’t criminalize setting up a campsite; they bar doing so “for the purpose of maintaining a temporary place to live,” which makes them applicable really only to homeless people. And that is particularly so given that the ordinances target a necessary bodily function—sleeping—with the result that anyone who has the forbidden status—being homeless—cannot help but avoid violating it. If it was cruel and unusual for California to criminalize “being a drug addict,” then it was equally cruel and unusual for it to criminalize “being a drug addict and breathing.” Finally, Grants Pass has enforced its ordinances precisely as intended, systematically applying them against homeless people in order to force them away from the City.

Having concluded that Robinson resolved the case, Sotomayor addressed some points that supported most of the majority’s analysis. First, the majority’s concern that courts were inserting themselves into public-policy debates better left to representative governments was overblown because Martin and the lower courts’ injunction in this case were narrow. None of those decisions barred cities from enforcing ordinances barring tents or camping gear, regulating the time and place of sleeping outside, or imposing fines and jail time on homeless people who decline accessible shelter options. And while it’s true that courts will have to draw lines as to when homelessness is truly involuntary, in the years since Martin, the Ninth Circuit has shown that it can sensibly do so. Second, Sotomayor found Powell irrelevant because it had no majority decision: Four justices in that case concluded that Texas’s law did violate Robinson, four justices disagreed, and the ninth found it unnecessary to decide who was right because the defendant failed to show he genuinely had no choice but to be intoxicated in public. But even if one were to take Powell’s result as binding precedent, the law at issue in that case was much different the one here: Even an alcoholic can get drunk at home. She concluded by agreeing with the majority that other constitutional provisions and doctrines, like the Excessive Fines Clause and the Due Process Clause, may bar Grants Pass from enforcing its ordinances, even if the Cruel and Unusual Punishment Clause does not.

Finally, continuing with the Western theme, Texas v. New Mexico (No. 141, Orig.) is one of the Court’s occasional original-jurisdiction forays into Western riparian disputes. This time, the Court rejected a proposed consent decree between Texas, New Mexico, and Colorado that would have resolved their water-rights dispute. Why? Because granting it would have disposed of the claims of the United States, which intervened years ago to advance its “unique federal interests.”

The water rights here concern the Rio Grande and the jurisdictions it touches. At the highest level, those jurisdictions are the United States and Mexico, which entered into a treaty in 1906 in which the United States promised to provide Mexico a certain amount of water each year as part of the Rio Grande Project, an irrigation system implemented by the United States Bureau of Reclamation. Later, in a series of “Downstream Contracts,” the United States agreed to (after allocating Mexico’s share), deliver apportionments of water to two irrigation districts, the Elephant Butte Irrigation District in New Mexico and the El Paso County Water Improvement District. Then come the states, Colorado, New Mexico, and Texas. In 1938, they resolved their competing water-rights claims with the Rio Grande Compact, which facilitated an “equitable apportionment” of the waters among them. That apportionment (and in particular, Texas’s share of the waters) depends upon the Bureau of Reclamation operating the Elephant Butte Reservoir under the Downstream Contracts. Pursuant to the Constitution’s Compact Clause, Congress approved the Rio Grande Compact. But as the decades passed, the Rio Grande’s waters diminished. As a result, entities in southern New Mexico increased their pumping of groundwater to support the state’s agriculture. This had downriver consequences, as it intercepted water bound for Texas and accordingly meant the Bureau of Reclamation had to release more water from the Elephant Butte Reservoir to comply with its various obligations. The Bureau of Reclamation addressed this issue by adopting a linear regression model based on data from 1951 to 1978 to help it predict how much water would be available to its irrigation districts based on a specific release of water from the Elephant Butte Reservoir.

That’s enough math; let’s get to the litigation. In 2013, Texas took New Mexico straight to the Supreme Court, alleging New Mexico was violating the Compact by allowing local entities to pump groundwater at levels higher than those contemplated in 1938. Soon after, the United States sought to intervene and filed its own complaint against New Mexico, claiming New Mexico’s groundwater pumping threatened its ability to fulfill its obligations to Mexico and the irrigation districts. As true Court Fans know, when the Supreme Court exercises original jurisdiction (generally in suits brought by one state against another), it acts as a trial court, though typically it delegates the real work to a special master. Here, the Special Master recommended dismissing the United States’ complaint, but the Supreme Court, in a 2018 opinion, disagreed, allowing the United States to intervene based on its interest “in seeing that water is deposited in the [Elephant Butte] Reservoir consistent with the Compact’s terms,” in order to fulfill its obligations under the Downstream Contracts. After the case returned to the Special Master, Texas and New Mexico negotiated a proposed consent decree that would establish a new method of determining the allotment of water New Mexico must deliver to Texas based on groundwater pumping levels “reflected in the 1951–1978 timeframe rather than [requiring] a strict return to a pumping condition as existed in 1938.” The United States objected to the consent decree, arguing it would wrongly dispose of its claims. That takes us up to the present.

In a 5-4 opinion authored by Justice Jackson and joined by the Chief and Justices Sotomayor, Kagan, and Kavanaugh, the Court agreed with the United States and rejected the consent decree. In doing so, the Court had to determine “whether the United States has valid Compact claims” and “whether the proposed consent decree would dispose of those claims.”

First, Justice Jackson concluded that the United States has valid claims. In her view, the Court had effectively decided this issue back in 2018 when it allowed the United States to intervene to “defend” its “distinctively federal interests.” Recounting the Court’s rationale for intervention, Justice Jackson explained the Compact is “inextricably intertwined” with the Rio Grande Projects and the Downstream Contracts, meaning the United States serves as an “agent” of the Compact or, put differently, the Compact incorporates the Downstream Contracts by reference. She also noted New Mexico had conceded the United States plays a key role in the Compact’s operation. And since the United States’ ability to deliver water to Mexico depends on New Mexico complying with its Compact obligations, a breach of the Compact could result in a breach of the United States’ treaty.

Second, the consent decree would dispose of the United States’ otherwise valid claims. As Jackson explained, the United States claims New Mexico’s groundwater pumping breaches its Compact duty not to interfere with the Rio Grande Project and seeks an injunction prohibiting such interference. But the consent decree would dispose of that claim without imposing a “duty of noninterference” or prohibiting New Mexico “from allowing groundwater pumping beyond 1938 levels.” To the contrary, the consent decree would incorporate New Mexico’s pumping into the Compact by adopting a new method for apportioning the Rio Grande’s waters, which would allow pumping based on the 1951 to 1978 conditions that led to New Mexico’s pumping increase.

Justice Jackson made quick work of the states’ competing arguments (echoed by the dissenters). First, she determined the Court’s 2018 intervention decision is “all but dispositive” of the arguments that the United States lacks Compact claims. Second, she brushed aside the idea that rejecting the consent decree would “unjustly expand the scope” of the case and that the United States should just bring its claims to another forum. Whether the United States’ and Texas’s interests have now diverged is irrelevant, she maintained; just because Texas wants to compromise does not mean the United States is expanding the case. And the United States cannot just take their dispute to another court, as the consent decree would settle the issue the United States cares about: New Mexico’s levels of groundwater pumping.

Justice Gorsuch penned a dissent, joined by Thomas, Alito, and Barrett. To the dissenters, the consent decree gives the United States exactly what it asked for when it intervened: “the protection of its existing federal reclamation operations.” Indeed, when the United States intervened, it claimed that Texas’s view that the Compact required a 1938 baseline would imperil its use of the 1951–1978 data when performing under the Downstream Contracts and the consent decree essentially would maintain business as usual. But in now abandoning this position (while not amending their complaint) and taking Texas’s original position, the United States was expanding the litigation even though it could simply litigate its claims in the lower courts.

Gorsuch also concluded there was no reason to reject the consent decree. It is consistent with the Compact, which only guarantees Texas “some minimum amount of Rio Grande water each year,” and it would impose no new duties on the United States or restrict it from pursuing any claims it may still have. What’s more, the consent decree would not dispose of any of the United States’ claims, as they could be dismissed without prejudice to the United States pursuing them to the lower courts.

So, back the case goes to the Special Master, where we’ll see if the third time is the charm.