Moody v. NetChoice and the Unfinished Question of Who Owns the Feed

Conservatives are supposed to believe in two things that, in the matter of social media, pull in opposite directions. We believe that private property deserves protection from government interference, including a private company's right to run its own business its own way. And we believe that free speech is the lifeblood of a self-governing people, and that a handful of companies with the power to silence millions deserve scrutiny. Put those two commitments in a room with a social media platform and watch them argue.

That argument is the real subject of Moody v. NetChoice, decided by the Supreme Court on July 1, 2024. The Court did not settle it. It said some important things, sent the cases back and left a divided right to sort out what it believes. I think the sorting is worth doing carefully. So here is the plan: define the terms, state both conservative positions at their strongest, then argue which one better protects speech over time.

Terms

Content moderation. A platform's decisions about which posts to host, remove, label, demote or promote.

Curation. The assembly of a feed: choosing and ranking content for each user's screen. A feed is not a neutral list. It is a selection.

Common carrier. A business that holds itself out as serving the public indiscriminately and is therefore subject to a duty to serve all comers on reasonable terms. Historically: ferries, railroads, inns, telegraph and telephone companies.

Facial challenge. A lawsuit arguing that a law is unconstitutional in all or nearly all of its applications, as opposed to a challenge to the law as applied to the plaintiff. Facial challenges are hard to win, and that detail decides this case.

The two state laws

In 2021 both Florida and Texas passed laws regulating large social media platforms.

Florida's law, Senate Bill 7072, restricted platforms from banning candidates for political office, limited their treatment of content from "journalistic enterprises," and imposed requirements that platforms apply their rules consistently and explain certain decisions to users.

Texas's law, House Bill 20, applied to platforms above a certain size, defined as having a very large number of monthly active users in the United States, and forbade them from censoring users based on viewpoint, along with imposing disclosure and complaint-process requirements.

Trade associations representing the platforms, NetChoice and the Computer & Communications Industry Association, challenged both laws as violations of the First Amendment. The two federal appeals courts split. The Eleventh Circuit largely blocked Florida's law, reasoning that the platforms' content moderation is protected expression. The Fifth Circuit upheld the Texas law, reasoning that the platforms' moderation was not protected speech and that the state could treat the platforms as common carriers. The Supreme Court took both cases together.

What the Court held

Justice Elena Kagan wrote the majority opinion. Chief Justice Roberts and Justices Sotomayor, Kavanaugh and Barrett joined it in full, and Justice Jackson joined most of it. The decision was unanimous in vacating the lower court judgments and sending both cases back, but it was not unanimous in reasoning.

The core of the decision runs as follows.

1. The lower courts analyzed the wrong question. Both appeals courts had treated the cases as ordinary facial challenges but had not done the work that a facial challenge requires: identifying every function the laws cover, working out which functions raise First Amendment problems and weighing those against the laws' legitimate uses. The laws cover far more than Facebook's News Feed and YouTube's homepage. They might reach direct messaging, online marketplaces, ride-hailing and other services. The Court said that scope had not been analyzed, so it could not tell how much of the laws was unconstitutional.
2. Curation of a feed is expressive activity. The majority said that when a platform selects, orders and presents third-party content in a feed, it is engaged in an editorial activity protected by the First Amendment. It drew on a line of precedent: Miami Herald Publishing Co. v. Tornillo (1974), which struck down a Florida law forcing newspapers to print replies from candidates they had criticized; Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston (1995), which held that a parade organizer could not be forced to include a message it did not want to carry; and Turner Broadcasting System v. FCC (1994), on cable operators' editorial discretion.
3. The government's interest in "balancing" the marketplace is not enough. Texas had defended its law as a way to correct the platforms' supposed bias and to promote a diversity of views. The Court said that a state may not interfere with private editorial choices merely to change the mix of speech in the public square. The government cannot, in the Court's reasoning, justify itself by saying the private editor's choices skew the debate.
4. The cases return to the lower courts. The courts below must now do the facial-challenge analysis properly.

Notice the structure. The majority's statements about feeds were strong, but the formal judgment was only to vacate and remand. The Court did not strike down the Texas or Florida laws. It said the law must be examined function by function, and that, as to the central feed functions, the platforms had a strong First Amendment claim.

The separate writings

Several justices wrote separately, and they matter.

Justice Barrett concurred, cautioning that facial challenges to laws governing complex platforms are difficult and noting that questions arise about algorithms and about the role of artificial intelligence in curation, and about platforms controlled by foreign owners. She was raising questions the majority left open.

Justice Jackson joined part of the majority opinion and concurred in part and in the judgment, arguing that the lower courts should specify the level of analysis, that more factual development may be needed, and that the Court should avoid previewing merits rulings.

Justice Alito wrote an opinion concurring in the judgment, joined by Justices Thomas and Gorsuch. He agreed with Justices Thomas and Jackson that the Court's guidance to the lower courts on remand was unnecessary, and cautioned that when a constitutional requirement meets new technology, courts should proceed with caution rather than compare social media too readily to traditional media.

Justice Thomas also concurred in the judgment, objecting that the majority's discussion of specific applications was unnecessary. He has separately developed a different line of thinking about whether digital platforms might be regulated as common carriers: in an April 2021 concurrence in a different case, Biden v. Knight First Amendment Institute, he suggested that Twitter, Facebook and similar companies might be regulated as common carriers.

So the Court's position is a mixture: a strong majority view that feed curation is expressive, a clear refusal to decide the laws' fate, and a visible minority that wanted to leave more room for state regulation.

The steelman for the common-carrier position

Here is the best case for conservatives who want the Texas approach, as its defenders would put it.

One. Some platforms are not ordinary publishers. They are the infrastructure of public conversation. When a handful of companies host most online political speech, their decisions to remove a speaker are not like a newspaper's choice of op-eds. They are more like a telephone company hanging up on a caller.

Two. The common-law tradition does recognize that businesses that hold themselves out to the public and that are essential to commerce carry duties to serve all. Inns could not turn away travelers, ferries could not refuse passengers on a whim. The telegraph and telephone were regulated this way. There is no principled reason a digital equivalent could not be.

Three. Concentration changes the picture. A speaker banned from a small forum can go elsewhere. A speaker banned from a dominant platform may lose most of his audience. Network effects mean that "go build your own platform" is a poor answer when the network is the product.

Four. Platforms have, in practice, been inconsistent in how they apply their rules, and their decisions have appeared to lean in one political direction. A law requiring viewpoint neutrality would restrain that.

That is a serious argument, made by serious people, and conservatives who make it are not being inconsistent. They are applying an old legal category to a new technology.

The steelman for the private-property position

And here is the best case for the other side of the conservative house.

One. A feed is an editorial product. Choosing what to show is what editors do. The Court has said so for fifty years, in Tornillo, where it held that a government cannot force a newspaper to carry a reply it does not wish to print. If the government can force a platform to carry speech it does not want to host, it can force a newspaper, a bookstore, a church bulletin or a talk show to do the same.

Two. The principle that protects the platforms today protects conservative publishers tomorrow. Suppose a different political majority passes a "fairness" law requiring every large platform, and every large media company, to carry viewpoints the owners find abhorrent. Conservatives would be the first to object. A rule that depends on who is in power is not a rule.

Three. The common-carrier label does not fit. A common carrier carries messages without choosing among them. A platform's entire product is its selection, ranking and recommendation. Calling it a carrier by statute does not change what it does, and forcing it to carry everything would produce a product no user wants: an unfiltered stream of spam, scams, harassment and worse.

Four. The remedy for bad moderation is competition and persuasion, not compulsion. Users can leave. Alternatives exist and can be built. Pressure from users and advertisers can change a platform's policy. A government mandate would freeze the arrangement in law.

The verdict

I come down on the private-property side, and I want to say why in terms that take the other side's concerns seriously.

First, the common-carrier argument proves too much. The very features that make a platform dominant, its curation and recommendation, are the features a carrier regime would suppress or regulate. And the argument depends on a distinction between "carrying" and "editing" that the platform's actual product collapses. To the extent a function is a pure conduit, such as plain direct messaging, there may be a case for neutral-carriage rules. The Court's remand leaves room for that. But the feed, the thing that most upsets people, is the platform's editorial product.

Second, the long-run risk of the common-carrier approach is capture. A law that tells platforms what they must carry creates a regulator with the power to decide what counts as viewpoint discrimination. That regulator will be appointed by whoever holds power. The same office that polices "neutrality" today will police it differently under a different administration. The structure invites exactly the government pressure that I criticize in NRA v. Vullo and the Unanimous Rule Against Government Jawboning and Murthy v. Missouri Lost on Standing, Yet the Record It Built Still Stands. You cannot consistently oppose government pressure on platforms to remove speech and support government pressure on platforms to carry it. The principle in both cases is that editorial judgment belongs to the editor.

Third, the strongest conservative interest is in a world where private institutions, including ones we build, can hold and express their own viewpoints without government permission. A church, a school, a publisher, a platform founded by a conservative: each of them depends on the same right of association and editorial control. A rule that lets government dictate the content mix of large private speech platforms will not stay confined to the platforms we dislike.

I will concede what the common-carrier side gets right. Concentration is a real problem, and the answer is not to pretend it is not. But the right tools are the ones that preserve editorial freedom: enforcing the antitrust laws against genuine monopoly conduct, requiring transparency about how rules are applied, supporting interoperability and data portability so that users can leave and take their connections with them, and opening the market to competitors. Those measures attack the power without commandeering the editorial function. They fit the Founders' preference for dispersing power over concentrating it in one more regulator.

Notice, too, that the Founding is on the side of editorial freedom. The First Amendment forbids Congress to abridge "the freedom of speech, or of the press." Printers in the founding era chose what to print. A press that was compelled to print whatever the government or the public demanded would not have been free. The framers feared the state as censor, and they would have feared the state as compulsory publisher just as much.

What is still unfinished

The cases are back in the lower courts, and several questions remain open.

1. Which functions of the platforms are expressive, and which are not? Direct messages, marketplaces and other services may be treated differently from curated feeds.
2. Do algorithmic recommendations count as editorial judgment when no human decides each placement? Justice Barrett flagged the issue.
3. What can states require in the way of transparency, such as explaining moderation decisions or offering appeals, without compelling speech?
4. What obligations, if any, could attach to a platform that functions as a mere conduit?

The Court has handed the lower courts a map and left most of the territory unexplored. Expect years of litigation. In the meantime, the political work continues, and conservatives would be wise to fight about means as well as ends. For the tech companies' legal status more broadly, see Publisher or Platform, Eight Myths About Big Tech and the Law and The 1995 Prodigy Lawsuit That Accidentally Created Section 230. For how foreign law handles the same questions, see Europe's Digital Services Act and the First Amendment, Side by Side.

The case in brief

1. Moody v. NetChoice (July 1, 2024) vacated the Eleventh and Fifth Circuit rulings on Florida's and Texas's social media laws because the lower courts had not done the facial-challenge analysis.
2. The majority opinion by Justice Kagan said that curating a feed is protected editorial activity, resting on Tornillo, Hurley and Turner, and that a state cannot rebalance private speech just to change the mix.
3. Conservatives divide between a common-carrier view, which stresses concentration and the history of public-service duties, and a private-property view, which stresses editorial freedom and the risk of a captured regulator.
4. The private-property position better protects speech over time, because the same rule that shields platforms from compelled hosting shields every conservative publisher, church and school.
5. Real concentration problems should be answered with antitrust enforcement, transparency, interoperability and competition, not compelled carriage.

What you can do: read the syllabus of Moody v. NetChoice on the Supreme Court's website, then ask your state representative which functions of a platform a proposed law in your state would actually cover.