On October 20, 2023, the Knight Institute will host a closed convening to explore the question of jawboning: informal government efforts to persuade, cajole, or strong-arm private platforms to change their content-moderation policies. Participants in that workshop have written short notes to outline their thinking on this complex topic, which the Knight Institute is publishing in the weeks leading up to the convening. This blog post is part of that series.


In assessing the Fifth Circuit’s historic decision in Missouri v. Biden, it would be easy but mistaken to be distracted by line-drawing and fact-categorizing difficulties. Where the line falls between impermissible government coercion and permissible cooperation, or between significant government encouragement and mere recommendation, is not directly pertinent to the disposition of this case. Its facts present a fusion of private and governmental action so plain as to blunt any protection that the state action and government speech doctrines might have provided. That there is a dawn and a dusk, as the old proverb goes, does not mean there is no noon or no midnight. The government’s unapologetic disdain for an independent, decentralized marketplace of ideas, and the social media platforms’ fawning obsequiousness in kowtowing to its “suggestions”—detailed exhaustively in the district court’s opinion and distilled by the Fifth Circuit—leave no doubt as to the side of the line on which their joint enterprise falls.

More important—and this will be the point of these remarks—are the largely unstated but crucial legal premises underpinning those courts’ decisions: that the marketplace of ideas exists today in a public square occupied mainly by social media platforms; that protecting an independent, decentralized marketplace requires safeguarding the speech freedom not only of the government and social media platforms but also of listeners and the general public; that the machinery of democracy depends upon keeping open those channels of political communication; and that when they are blocked, it is the duty of the courts to re-open them. (I elaborate in a recent article and forthcoming book, from which this post draws.)

To appreciate the seminal importance of these decisions, it’s useful to step back from line-drawing and categorization issues and recognize the starkness of the clash of competing legal visions. Begin with the government’s position, which has been the same in this case and related cases now before the Supreme Court involving Texas and Florida statutes. Its position is that only its speech and the speech of the social media platforms are constitutionally protected. Its speech, it argues, is protected by the so-called government speech doctrine, which permits it to say what it wishes free of First Amendment constraints, and the platforms’ speech is protected by the state action doctrine, which leaves private actors free to speak as they wish (or not to speak, by “editorially” suppressing users’ posts). Thus, the First Amendment, as construed by the government, would empower it to make known its wishes freely to the companies and would empower the companies to suppress any content by any user for any reason—whereas the speech interests of users and the general public would count for naught.

Now, accept for purposes of argument that choices involved in curating material constitute a form of speech. There are powerful reasons to reject that argument, as did the Fifth Circuit in NetChoice v. Paxton—pointing out, among other things, that the platforms themselves have repeatedly denied engaging in editorial judgment, and that Congress, in granting them immunity, explicitly rejected the proposition that they are publishers or speakers. Even if the companies are engaged in a kind of speech, however, other participants in the marketplace of ideas also have speech interests that warrant protection. The importance of the Fifth Circuit’s and the district court’s decisions in Missouri v. Biden lies in their unequivocal affirmation that protection of social media listeners’ interests is essential to the operation of the modern marketplace of ideas—that the First Amendment cannot mean that only the speech of the government and the companies is entitled to constitutional protection, without regard to the free speech interests of social media users and the general public.


Why protect listeners? Because public policy cannot be forged intelligently and democracy cannot function effectively unless people are free to hear what others have to say and to modify their own views accordingly. Preventing us from listening to others impedes our ability to find common ground, injuring the entire polity. This insight is hardly new. “To suppress speech is a double wrong,” Frederick Douglass said. “It violates the rights of the hearer as well as the speaker.”

The Fifth Circuit in Missouri understood this. It quoted one of several state officials who testified to the importance of listening to constituents on social media:

[M]ask and vaccine mandates for students have been a very important source of concern and public discussion by Louisiana citizens over the last year. It is very important for me to have access to free public discourse on social media on these issues so I can understand what our constituents are actually thinking, feeling, and expressing about such issues, and so I can communicate properly with them.

“Indeed,” the court observed, “the CDC’s own witness explained that if content were censored and removed from social-media platforms, government communicators would not ‘have the full picture’ of what their citizens’ true concerns are.”

Individual plaintiffs who were denied the right to listen suffered similar injury. The plaintiffs identified more than 150 other speakers whom they followed on social media and who were specifically targeted and silenced. The plaintiffs, it is well to recall, included prominent physicians the soundness of whose medical judgment rested on the ability to keep abreast of current scientific opinion—opinion that was suppressed by the government and social media. Though the Missouri courts did not explicitly consider the individual plaintiffs’ interest as listeners, the Fifth Circuit seemingly recognized the validity of those claims. “The [government] officials have engaged in a broad pressure campaign designed to coerce social-media companies into suppressing speakers, viewpoints, and content disfavored by the government,” it held. “The harms that radiate from such conduct extend far beyond just the Plaintiffs; it impacts every social-media user.” Listeners count.

The government, however, would have none of it. It objected to weighing any listeners’ rights, whether social media users or the general public. Such a right extends only to “intended recipients with some connection to the speaker,” the government insisted. Neither the state plaintiffs nor the individual plaintiffs had a right, in its view, to hear or listen to any of the suppressed speech, even speech that was concededly true.

Under the government’s own standard, this argument fails. The plaintiffs obviously were intended recipients of the suppressed materials; as social media users, each was specifically intended to receive the content posted by other social media users. It is, indeed, reasonable to infer that the authors of that content further intended to give it the widest possible circulation (including circulation to those of us who have not signed on to those platforms and have no wish to support them). Such recipients, moreover, have a manifest connection with the author—indeed, as the plaintiffs pointed out, the most fundamental of connections: “but for the government’s censorship, the listener would otherwise hear the speaker’s message.” The whole point of writing and speaking and posting on social media is, after all, to connect with a listener.

The Supreme Court addressed listeners’ rights in Lamont v. Postmaster General (1965), a case overlooked in the government’s briefs. In it, the Court confronted a statute that required persons who received “communist political propaganda” sent through the mail by foreign governments to explicitly request delivery by placing their names on a post office registry. The Court struck down the law even though, as Burt Neuborne has noted, it found that the “speaker”—the foreign government—had no First Amendment rights whatsoever. The Court ruled that the recipient had an independent right to receive the literature. “[W]here a speaker exists,” the Court later elaborated, “the protection afforded is to the communication, to its source and to its recipients both.” The right to hear or read that communication is, it said, “an inherent corollary of the rights to free speech and press that are explicitly guaranteed by the Constitution.” The Court added subsequently that “the right to receive ideas follows ineluctably from the sender’s First Amendment right to send them.” Justice William Brennan’s concurrence in Lamont burnished the majority opinion. The right to receive information, he argued, is a fundamental right. “The dissemination of ideas can accomplish nothing if otherwise willing addressees are not free to receive and consider them,” he wrote. “It would be a barren marketplace of ideas that had only sellers and no buyers.” And, we might add today, it would be a barren digital marketplace that protected only the speech interests of the government and the social media platforms but not those of listening users or the general public.

To the contrary, if the aim is to protect the marketplace of ideas, it’s imperative to examine the speech interests of all participants, to identify all rights that are potentially impinged, and to protect those rights directly, rather than safeguarding only those that happen adventitiously to be represented by a particular litigant. The government rejects doing so. That, the government argued in Missouri, would necessitate adopting “boundless theor[ies] of standing” of the sort the Court has “consistently rejected.” The government did not refer explicitly to the “generalized grievances” doctrine, but it seems to have had such justiciability limits in mind. That doctrine precludes standing “when the asserted harm is a generalized grievance shared in a substantially equal measure by all or a large class of citizens.” Thus plaintiffs relying upon the Constitution’s requirement of a regular statement and accounting of all governmental expenditures, for example, or its prohibition against an officer of the United States simultaneously holding congressional office, have no standing to challenge members of Congress serving in the military reserves, or the secrecy of the CIA’s budget. The theory is that such limits serve separation of powers purposes by insulating the political branches from judicial encroachment; similar problems, the government argued in Missouri, would arise if standing were found in a suit such as this. Existing case law gives the Missouri plaintiffs good and sufficient answers to such objections, as the Fifth Circuit found, but in different circumstances such concerns could take on greater weight.

A case such as Missouri presents a welcome opportunity to restrict the scope of the generalized grievances doctrine, in recognition that its sweeping application renders important parts of the Constitution judicially unenforceable. The constitutional provisions involved, Erwin Chemerinsky has written, “could be blatantly disregarded, and yet the courts would be powerless to halt the violations.” One possible solution, I’ve suggested, would be to carve out an exception for situations where no plaintiff can be hypothesized who would meet standing requirements, or where, as here, the alleged harm lies in blockage of the channels of communication on which the operation of democratic political processes depends. In the face of a systemic breakdown, the alternative is a complete absence of any legal remedy and continued constitutional violation. “The very essence of civil liberty certainly consists in the right of any individual to claim the protection of the laws, whenever he receives an injury,” Chief Justice John Marshall wrote in Marbury v. Madison. “One of the first duties of government is to afford that protection.” That duty surely is at its highest when the very functioning of the democratic system at risk, as Owen Fiss observed:

The duty of the state is to preserve the integrity of public debate…not to indoctrinate, not to advance the “Truth,” but to safeguard the conditions for true and free collective self-determination. It should constantly act to…make certain that the status quo is embraced because we believe it is the best, not because it is the only thing we know or are allowed to know.


Vindicating the rights of listeners, then, is one step toward adapting the First Amendment to the digital era. But more is needed. Like over-enforcement of the generalized grievances doctrine, over-enforcement of the state action doctrine and the government speech doctrine has led to under-protection of the marketplace of ideas. Unless the courts are to become bystanders, as the Missouri judges understood, the potential reach of those doctrines will need to be retracted. Doing so may entail adopting something like Tim Wu’s proposed “accomplice liability” doctrine, under which “an official who spurs private censorial mobs to attack a disfavored speaker might … be subject to a court injunction or even damages.” It may entail viewing major social media platforms as public utilities or common carriers subject to viewpoint neutrality requirements, as Justice Thomas suggested. Historically, he pointed out, businesses known as common carriers, such as railroad companies, insurers, and telegraph companies, have long been subject to special regulations, including a general requirement to serve all comers. “In many ways,” he wrote, “digital platforms that hold themselves out to the public resemble traditional common carriers.” It may entail requiring public disclosure of content moderation communications between the government and the platforms. It may entail requiring the platforms to make available an optional “channel” filtered under First Amendment principles, giving users the choice to decide what content to view. Most importantly, to avoid the pitfalls of ad hoc, case-by-case balancing, it may entail integrating such elements in a new definitional category of speech in which special rules apply, an approach outlined by Melville Zimmer.

Weighing competing interests ad hoc is sensible in resolving policy conflicts, Zimmer wrote, but not conflicts that arise in the delicate realm of freedom of expression. With ad hoc balancing, “there is no rule to be applied, but only interests to be weighed.” This results in no standard to measure one interest against another. The resulting uncertainty is chilling, because it “deters all but the most courageous…from entering the marketplace of ideas.” Moreover, in nearly all major cases in which case-by-case, ad hoc balancing has been applied, free speech has lost. Under ad hoc balancing, Zimmer wrote, the First Amendment “becomes a dead letter.”

The alternative, he suggested, is definitional balancing: employing balancing not to determine which litigant wins in a particular case, but to define which forms of speech are constitutionally protected and under what circumstances. Unlike fact-dependent ad hoc balancing, which creates no meaningful precedent, “a rule emerges from definitional balancing which can be employed in future cases without the occasion for further weighing of interests.” Some commentators have referred to this approach as “category balancing,” because the rule created by balancing interests creates a category that precludes ad hoc balancing in future cases that fall within it.

The Missouri courts seemed tacitly to appreciate the importance of this distinction. Both the district court and the Fifth Circuit eschewed ad hoc balancing. Nowhere in either opinion, for example, does one find a weighing of the government’s interest in suppressing evidence of the inefficacy of lockdowns against Dr. Jay Bhattacharya’s concrete, personal, individuated interest in publishing that evidence. To the contrary, the courts weigh broadly conflicting institutional interests. On one side of the scales, as the Fifth Circuit put it, is the need to preserve “a robust sphere of individual liberty” and to protect the speech of government officials who “were permissibly exercising government speech, ‘carrying out [their] responsibilities,’ or merely ‘engaging in [a] legitimate…action.’” On the other side of the scales is “a fundamental aspect of American life” that was jeopardized by their “coordinated campaign” of “suppressing millions of protected free speech postings by American citizens.” Weighing these structural values against each other makes for the rudiments of definitional balancing. By doing so, these courts took a laudable first step toward creating a new definitional rule: that within the category of digital speech on major social media, the interests of all participants, not simply those of the government and the platforms, are constitutionally protected.


It is now left for the Supreme Court to refine their work by defining more fully the rules that apply within that category and in the process, perhaps, by tackling deeper questions not directly presented in this case—such as how, in the future, the marketplace of ideas is to be protected from censorship by platforms operating in the face of governmental silence. The underlying problem, after all, is not a horsehead in the platforms’ bed; if it is, in the end, the vitality of an independent, decentralized marketplace that concerns us rather than merely government coercion, and if social media today are the most important places in the public square for the exchange of views, as the Supreme Court has said, why should it matter whether the delete key is pressed by an officious Facebook content moderator or some DHS “mal-information expert”? Suppose social media “voluntarily” join forces with the government not because it intimidates them, but because the two have a shared interest in shuttering the marketplace of ideas to all content they mutually dislike? And suppose they succeed?

The government seems to believe that would be fine. In such circumstances, it contends, the plaintiffs would lack standing. Symbiosis or “pervasive entwinement” is irrelevant, contrary case law notwithstanding. The government insists that the plaintiffs must show that but for the government’s urging, they would not have been censored. This is, of course, an all but impossible burden to meet. Requiring but-for causation in these circumstances would immunize the government from all responsibility for “voluntarily” intermediated censorship, for human behavior is almost always a function of multiple variables, many unconscious and unquantifiable. The Fifth Circuit rightly rejected this simplistic standing theory (though the government’s musclebound Silicon Valley partners will no doubt assure us that the government never frightened them).

No doubt the government would contend that preserving an inclusive, independent, decentralized marketplace of ideas is not the role of the courts, that their task is to address only specific harms presented in factually delimited cases and controversies. Fair enough; in simpler times, a minimalist judicial role was in many respects appropriate. But this, the Missouri courts seemed to realize, is a different era, an era in which a mix of technological advance and ethical regression led to “‘unrelenting pressure’ from certain government officials [that] likely ‘had the intended result of suppressing millions of protected free speech postings by American citizens.’” The courts cannot turn a blind eye to the corrosive systemic effects of that suppression. It stifles scientific progress and political reform. It creates martyrs or heroes out of suppressed authors. It gives their squelched messaging a “forbidden fruit” quality, making it more credible than it otherwise would be. It drives those messages underground where they are less open to rational rebuttal. It alienates the public and generates distrust. It fosters social division and political instability. It is, in a word, unsafe—Justice Oliver Wendell Holmes’s point in his landmark dissent in Abrams v. United States (1919), laying the groundwork for the modern First Amendment. Yes, these systemic distortions are the sorts of second- and third-order effects that judges normally look to the political branches to address. But if the judiciary does not now act remedially, who will? A Congress chosen by an electorate that censorship has blinded to its blindness? The same executive that orchestrated that censorship? Justice Brennan gave the right answer in Lamont: “It is the duty of courts to be watchful for the constitutional rights of the citizens, and against any stealthy encroachments thereon.” The Missouri courts honorably fulfilled that duty. It now falls to the Supreme Court to do the same.