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 the wake of the Fifth Circuit’s extraordinary decision in Missouri v. Biden, there has been significant criticism of the court’s assertion that the jawboning of private speech intermediaries violates the First Amendment not only when it “coerces” intermediaries to act, but also when it “significantly encourages” them to do so. Commentators have warned that such a holding imperils important free speech interests, by rendering virtually any government speech about the social media platforms—or television networks, or newspapers, etc.—potentially unconstitutional. It threatens government officials’ own freedom of speech.

In this post, I argue that these criticisms of the significant encouragement test are correct as applied to the Fifth Circuit decision, but fundamentally misguided as a general proposition. It is true that the Fifth Circuit applied the test in the Missouri decision too broadly, to sweep in informal communications between government actors and the social media companies that do not pose a serious threat to the independence of the marketplace of ideas. But the idea that jawboning can violate the First Amendment not only when it threatens its target with legal harm (what “coercion” in jawboning cases is usually understood to mean) but also when it incentivizes them to act in a particular way is obviously correct, given fundamental features of the speech marketplace—namely, the economic incentives that underpin that marketplace, and how sensitive these incentives make speech intermediaries to even soft government pressure. It would be a grave mistake as a result for First Amendment scholars or courts to respond to the Fifth Circuit’s unduly broad application of the significant encouragement test by embracing the extremely narrow test of jawboning that, until Missouri v. Biden came along, tended to dominate in the cases. That test, by failing to take account of the inescapable fact of intermediary sensitivity to government pressure, fails to adequately protect free speech interests, including the interest of us all in a speech marketplace which the government cannot secretly manipulate by means of carrots as well as sticks.

Instead, courts and scholars should focus not on a particular doctrinal test, but on the harm that the First Amendment rule against jawboning is meant to guard against. This harm is not the harm of coercion per se (in fact, the government coerces speech all the time). As the Supreme Court made clear in the one and only jawboning case it has ever decided, it is the harm of constitutional evasion: the harm of a system of speech regulation outside the reach of the courts and, consequently, the First Amendment. Rather than fighting over the formal terms of the jawboning test—formulations that, as the Fifth Circuit decision makes clear, can easily be manipulated to mean very different things than originally intended—courts and scholars should analyze the constitutionality of government jawboning campaigns with this in mind. This means that both negative and positive inducements by government actors (what usually get classified in the cases as “coercion” and “encouragement”) should be understood as capable of violating the First Amendment, but only when they pose an undue risk of this kind of constitutional subversion. The task for scholars today is not to debate whether government encouragement can violate the First Amendment, but to help courts figure out what kinds of government encouragement, or coercion, or other kinds of action pose this kind of risk.

The Significant Encouragement Test in Missouri v. Biden

In Missouri v. Biden, the Fifth Circuit held that a significant number of officials in the Biden administration likely violated the First Amendment rule against jawboning when they either coerced or significantly encouraged social media platforms to suppress speech about the COVID-19 pandemic, Hunter Biden’s laptop and other politically contentious matters. In interpreting the First Amendment to prohibit informal government actions both when they coerced and when they significantly encouraged private actors to suppress speech, the Fifth Circuit relied upon a decades-old Supreme Court decision, Blum v. Yaretsky.

Blum involved a due process challenge to the decision by a number of private nursing homes to discharge or transfer Medicaid-funded patients without a hearing. The plaintiffs argued that because the clinics were encouraged to make the transfer or discharge decisions by Medicaid policies, they were entitled to the same procedural rights as patients discharged by government facilities. The Court rejected this claim because it found the private clinics possessed some degree of private discretion in the choice to transfer or discharge. In his majority opinion, Justice Rehnquist explained that “a State normally can be held responsible for a private decision only when it has exercised coercive power or has provided such significant encouragement, either overt or covert, that the choice must in law be deemed to be that of the State.” The fact that the clinics could have chosen not to transfer or discharge the patients meant, the Court concluded, that their choice could not “be deemed to be that of the State.” Private discretion broke the chain of responsibility, and therefore, constitutional constraint.

Blum set out, in other words, what appeared at the time a very narrow test of state action that meant that, in most cases in which government actors encouraged private actors to act, constitutional constraints simply did not apply. This was certainly how most lower courts interpreted it subsequently. The Ninth Circuit, for example, interpreted Blum to require evidence that a government official “overwhelm[ed] the private party and essentially compel[led] the party to act in a certain way” in order for the private party’s decision to be deemed “that of the State.”

This is not how the Fifth Circuit read the Blum test, however. The Fifth Circuit interpreted Blum to mean that government officials could be held responsible for acts of speech suppression carried out by private parties—in this case, the social media platforms—not only when there was no evidence that the platform’s will was “overwhelmed” but no evidence that the platform felt pressured or incentivized by the government officials to act at all. Instead, the Fifth Circuit found that the mere fact that government officials (White House officials, representatives from the CDC and the FBI) communicated frequently with platform employees about their content moderation decisions, and that these communications influenced the decisions the platforms made, was sufficient to deem those decisions “that of the State,” and therefore unconstitutional. In other words, in a remarkable act of judicial chutzpah, the Fifth Circuit turned what had been an extremely narrow test of state action into a very broad test of unconstitutional jawboning and one that appeared to prohibit government officials from communicating with private speech intermediaries in any way that ends up having an impact on their decision making about speech.

As commentators have noted, such a broad interpretation of the prohibition against jawboning is hard to justify by reference to the values and interests that the First Amendment protects. By making it constitutionally perilous for government actors to communicate with private speech intermediaries about the speech they carry in almost any form whatsoever, it helps safeguard the independence of the marketplace of ideas from informal, behind-the-scenes government influence and control, to be sure. But it also makes it very difficult, if not impossible, for officials to share information with speech intermediaries about the harms and benefits of speech even when that is information that the intermediaries would have no way to obtain on their own. It also makes it difficult for officials to push speech intermediaries to make speech-regulating decisions that are more public-oriented than their private economic incentives might otherwise lead them to make.

The Supreme Court has made clear that the First Amendment does not require such a costly rule. In the only jawboning case it has ever decided, Bantam Books v. Sullivan, the Court held that informal communications between government actors and private speech intermediaries are permissible so long as those communications are “purely advisory” rather than “coercive in purport.” In such cases—where government officials merely communicate information to private actors—the Court concluded that the government’s actions “need not retard the full enjoyment of First Amendment freedoms.” This is because they leave the private party with complete autonomy to make whatever choice they like about the speech they disseminate, even if the exercise of that autonomy is ultimately influenced by what the government officials had to say.

Commentators are therefore correct that the Fifth Circuit test of significant encouragement sweeps too broadly both as a matter of principle and precedent. This does not mean, however, that it is only when government officials threaten private speech intermediaries or speakers with harm that jawboning poses a constitutional problem. To understand why, it is necessary first to understand what the distinctive harm of jawboning is, and how it manifests in practice.

Why Threats of Legal Harm Are Not the Only Kinds of Coercive Pressure That Threaten Free Speech Values

As the language quoted above makes clear, in Bantam Books, the Court held that informal communication between government officials and private actors violates the First Amendment when it is coercive in nature: that is to say, when it is intended, and has the effect (or, as later courts added, could have the effect) of pressuring a private actor into doing something she would not otherwise have done voluntarily. This kind of coercive government communication is constitutionally problematic because it creates, or risks creating, what Justice Brennan described in his majority opinion as a “system of informal censorship.” It allows government officials to informally suppress speech or boost it without having to comply with the procedural and constitutional constraints that would apply if they used formal means (regulations, statutes, executive orders) to do so. This is why the Court concluded that government jawboning of this type possesses a “capacity for suppression of constitutionally protected publications [that] is far in excess of that of the typical licensing scheme held constitutionally invalid by this Court.” It poses a greater threat to free speech values than the systems of administrative censorship that have long been associated with the worst abuses of the repressive state, because it is informal and often secret—and thereby less easily subject to judicial review, and constitutional control.

Because the jawboning scheme at issue in Bantam Books involved threats of criminal obscenity prosecutions against booksellers who refused to stop selling books and magazines that government officials believed unsuitable for teenage readers, some lower courts interpretedthe decision to mean that government jawboning can violate the First Amendment only when government actors threaten private speakers or speech intermediaries with criminal prosecution, or similarly serious legal sanctions. Indeed, this has been the dominant interpretation of Bantam Books in the federal courts.

This is a poor interpretation of the opinion, however. If the threat that government jawboning poses to the First Amendment is the threat of constitutional evasion—as Bantam Books tells us it is—that threat is realized whenever a government official succeeds in pressuring a private actor into regulating speech differently than she would have voluntarily chosen to do. Every time a government official manages this feat, they do what the First Amendment says they may not: they interfere in the operation of the marketplace of ideas by attempting to “give one side of a debatable public question an advantage in expressing its views to the people.” And what the long history of jawboning in the United States makes clear is that it is not only when government officials threaten private companies with criminal prosecution or some other equally serious legal harm that they pull off this feat.

Consider for example the Penthouse v. Meese case. In that case, a commission within the Department of Justice sent a letter to 7-Eleven, warning it that if it did not hear back, the commission would include the retailer in a list of “identified distributor[s]” of pornography in a forthcoming report. The letter contained no threat of legal sanctions. Pornography is legal in the U.S., thanks to the First Amendment. Nevertheless, in response to the letter, 7-Eleven immediately stopped selling not only Playboy magazines but also temporarily suspended the sale of Cosmopolitan, for fear that the half-naked perfume ads in the magazine would attract the DOJ’s ire. The company’s fear of reputational harm prompted immediate and eager compliance with the commission’s unstated and unspecified request to take magazines off its shelves.

Or consider R.C. Maxwell v. New Hope. In that case, a bank immediately terminated the lease of a billboard that a city manager identified as “objectionable” after the manager hinted that, if the bank didn’t take the billboard down, the city would look with disfavor on a development project the bank was planning. Here too, there was no threat of legal sanctions. There was instead the promise of a carrot if the bank complied: fast track passage of a valuable development project. But in this case also, the pressure was sufficient to prompt immediate and unequivocal compliance.

These cases illustrate a fundamental feature of the political economy of the speech marketplace: namely, that it will often be far more rational for private companies to comply with even relatively soft government pressure to suppress or keep up speech than it will be for them to contest it. This is because bookstores, banks, and platforms profit from serving a great deal of speakers, not from keeping up any particular acts of speech. The cost they face, as a result, of removing particular speakers or speech acts from their platforms will often be extremely minimal. Meanwhile, as the Maxwell example demonstrates, government officials often possess a great deal of discretionary power they can wield to the benefit or the detriment of the companies they regulate.

The Bantam Books Court was well aware of these features of the speech marketplace. It is why it rejected claims that because booksellers could always prove that the books they sold were not obscene if and when they were prosecuted—or, alternatively, could seek a declaratory judgment in advance of prosecution—the informal pressure campaign against them did not violate their First Amendment rights. The Court recognized that even if booksellers could defend their First Amendment rights in court, they were unlikely to do so. As Justice Brennan noted in his majority opinion, “[t]he distributor who is prevented from selling a few titles is not likely to sustain sufficient economic injury to induce him to seek judicial vindication of his rights.” The Court therefore insisted that courts should establish what it described as “adequate bulwarks” against the erosion of First Amendment rights. They should interpret the First Amendment to protect against the “gravely damaging yet barely visible encroachments” on freedoms of expression that the informal exercise of government power permitted.

The existing law of jawboning, prior to Missouri v. Biden, failed to live up to this promise. By equating coercion with threats of legal harm, and by insisting that the government unconstitutionally jawboned only when it coerced, the cases left government officials a great deal of room to use the soft powers granted them by virtue of their office—including here the power of the bully pulpit and power over the bureaucratic apparatus of the state—to pressure private companies into suppressing or boosting speech that those companies would never otherwise choose to suppress or to boost.

Jawboning as a Problem of Constitutional Evasion

The great advantage of the Fifth Circuit’s decision in Missouri v. Biden is that it recognizes, however imperfectly, that government jawboning can violate the First Amendment not only when it involves threats of legal harm but also when it involves other, more amorphous threats and promises of benefit. Indeed, the decision is remarkable not only for how broadly it construes the “significant encouragement” language from Blum but also in how broadly it interprets the category of government coercion.

The great problem with the Fifth Circuit’s decision, however, is its failure to recognize that these terms—coercion, significant encouragement—should be defined by reference to the harm they are meant to rout out: namely, the harm of constitutional evasion. As Justice Brennan put it in Bantam Books, the task for courts in jawboning cases is to “look through [the] forms [of government action] to substance” and determine when informal actions are “sufficiently [likely to] inhibit the circulation of publications to warrant injunctive relief.” The task for courts, in other words, is to determine whether the government’s actions are sufficiently likely to lead private parties to act, not because they believe on the merits it is the right thing to do, but because the government wants them to, to pose a serious threat to the independence of the marketplace of ideas and the system of free expression.

This is not a task that the Fifth Circuit performed, in any meaningful respect. But, to be fair to the Fifth Circuit, it is not a task that the other circuits have tended to perform very well in their jawboning cases. Instead, by embracing an overly narrow test of jawboning-as-threats-of-legal-harm, these other courts have similarly failed to understand what the role of the First Amendment anti-jawboning rule really is. The result has been an unduly simple and formalistic test of jawboning that focuses on the form of government action, not its substance—in direct contradiction of the command from the Court in Bantam Books.

Figuring out when government action is “sufficiently likely” to lead private parties to act involuntarily will be no easy task. It will not always be a simple matter to tell when a private actor makes a decision because a government official convinced her it was the right thing to do and when she makes that decision because she wants to stay in the official’s good graces, or thinks it is easiest just to do what the official says, or fears retribution etc. Nevertheless, it is the task that Bantam Books, and the First Amendment principles it vindicates, requires courts to attempt. The question is not whether the First Amendment disallows the use by government officials of informal significant encouragement. The question is what kinds of significant encouragement it disallows.