Federal Government Improperly "Coerced" and "Significantly Encouraged" Certain Speech Restrictions by Social Media Platforms

In yesterday's decision in Missouri v. Biden, the Fifth Circuit (Judges Edith Clement, Jennifer Elrod, and Don Willett) held that the federal government violated the First Amendment by causing social media platforms to block posts on various topics (including "the COVID-19 lab-leak theory, pandemic lockdowns, vaccine side-effects, election fraud, and the Hunter Biden laptop story").

The court acknowledged that the First Amendment doesn't bar social media platforms from acting on their own to restrict user speech, since the First Amendment applies only to the government and not to private parties (including large corporations). But the court concluded that the First Amendment may be violated "when a private party is coerced or significantly encouraged by the government to such a degree that its 'choice'—which if made by the government would be unconstitutional—'must in law be deemed to be that of the State.' This is known as the close nexus test."

As to what constitutes "significant[] encouragement by the government" to restrict speech, the court held:

For encouragement, we read the law to require that a governmental actor exercise active, meaningful control over the private party's decision in order to constitute a state action. That reveals itself in (1) entanglement in a party's independent decision-making or (2) direct involvement in carrying out the decision itself. In any of those scenarios, the state has such a "close nexus" with the private party that the government actor is practically "responsible" for the decision, because it has necessarily encouraged the private party to act and, in turn, commandeered its independent judgment.

As to what constitutes "coerc[ion]," the court held:

For coercion, we ask if the government compelled the decision by, through threats or otherwise, intimating that some form of punishment will follow a failure to comply…. [T]o help distinguish permissible persuasion from impermissible coercion, we turn to the Second (and Ninth) Circuit's four-factor test. Again, honing in on whether the government "intimat[ed] that some form of punishment" will follow a "failure to accede," we parse the speaker's messages to assess the (1) word choice and tone, including the overall "tenor" of the parties' relationship; (2) the recipient's perception; (3) the presence of authority, which includes whether it is reasonable to fear retaliation; and (4) whether the speaker refers to adverse consequences.

Each factor, though, has important considerations to keep in mind. For word choice and tone, "[a]n interaction will tend to be more threatening if the official refuses to take 'no' for an answer and pesters the recipient until it succumbs." That is so because we consider the overall "tenor" of the parties' relationship. For authority, there is coercion even if the speaker lacks present ability to act so long as it can "reasonably be construed" as a threat worth heeding.

As for perception, it is not necessary that the recipient "admit that it bowed to government pressure," nor is it even "necessary for the recipient to have complied with the official's request"—"a credible threat may violate the First Amendment even if 'the victim ignores it, and the threatener folds his tent.'" Still, a message is more likely to be coercive if there is some indication that the party's decision resulted from the threat. Finally, as for adverse consequences, the government need not speak its threat aloud if, given the circumstances, it is fair to say that the message intimates some form of punishment.  If these factors weigh in favor of finding the government's message coercive, the coercion test is met, and the private party's resulting decision is a state action.

(Note that there is a good deal of caselaw on the coercion side, but much less on the significant encouragement side. Courts have suggested in the past that significant encouragement, even when it's not coercive, may implicate the government in the encouraged parties' decision. But the court cited few appellate cases to actually apply this to invalidate government action, and those struck me as quite different in the nature of the government action involved. In this respect, this case seems to set an important new precedent, unless it's overturned by the Supreme Court.)

Applying the tests, the court held "that the White House, acting in concert with the Surgeon General's office, likely … coerced the platforms to make their moderation decisions by way of intimidating messages and threats of adverse consequences":

Generally speaking, officials from the White House and the Surgeon General's office had extensive, organized communications with platforms. They met regularly, traded information and reports, and worked together on a wide range of efforts. That working relationship was, at times, sweeping. Still, those facts alone likely are not problematic from a First-Amendment perspective. But, the relationship between the officials and the platforms went beyond that. In their communications with the platforms, the officials went beyond advocating for policies, or making no-strings-attached requests to moderate content….

We start with coercion. On multiple occasions, the officials coerced the platforms into direct action via urgent, uncompromising demands to moderate content. Privately, the officials were not shy in their requests—they asked the platforms to remove posts "ASAP" and accounts "immediately," and to "slow[] down" or "demote[]" content. In doing so, the officials were persistent and angry. When the platforms did not comply, officials followed up by asking why posts were "still up," stating (1) "how does something like [this] happen," (2) "what good is" flagging if it did not result in content moderation, (3) "I don't know why you guys can't figure this out," and (4) "you are hiding the ball," while demanding "assurances" that posts were being taken down.

And, more importantly, the officials threatened—both expressly and implicitly—to retaliate against inaction. Officials threw out the prospect of legal reforms and enforcement actions while subtly insinuating it would be in the platforms' best interests to comply. As one official put it, "removing bad information" is "one of the easy, low-bar things you guys [can] do to make people like me"—that is, White House officials—"think you're taking action."

That alone may be enough for us to find coercion. Like in Bantam Books v. Sullivan (1963), the officials here set about to force the platforms to remove metaphorical books from their shelves. It is uncontested that, between the White House and the Surgeon General's office, government officials asked the platforms to remove undesirable posts and users from their platforms, sent follow-up messages of condemnation when they did not, and publicly called on the platforms to act. When the officials' demands were not met, the platforms received promises of legal regime changes, enforcement actions, and other unspoken threats. That was likely coercive.

That being said, even though coercion may have been readily apparent here, we find it fitting to consult the Second Circuit's four-factor test for distinguishing coercion from persuasion. In asking whether the officials' messages can "reasonably be construed" as threats of adverse consequences, we look to (1) the officials' word choice and tone; (2) the recipient's perception; (3) the presence of authority; and (4) whether the speaker refers to adverse consequences.

First, the officials' demeanor. We find, like the district court, that the officials' communications—reading them in "context, not in isolation"—were on-the-whole intimidating. In private messages, the officials demanded "assurances" from the platforms that they were moderating content in compliance with the officials' requests, and used foreboding, inflammatory, and hyper-critical phraseology when they seemingly did not, like "you are hiding the ball," you are not "trying to solve the problem," and we are "gravely concerned" that you are "one of the top drivers of vaccine hesitancy." In public, they said that the platforms were irresponsible, let "misinformation [] poison" America, were "literally costing … lives," and were "killing people." While officials are entitled to "express their views and rally support for their positions," the "word choice and tone" applied here reveals something more than mere requests….

[M]any of the officials' asks were "phrased virtually as orders," like requests to remove content "ASAP" or "immediately." The threatening "tone" of the officials' commands, as well as of their "overall interaction" with the platforms, is made all the more evident when we consider the persistent nature of their messages. Generally speaking, "[a]n interaction will tend to be more threatening if the official refuses to take 'no' for an answer and pesters the recipient until it succumbs." Urgency can have the same effect. See Backpage.com v. Dart (7th Cir. 2015) (finding the "urgency" of a sheriff's letter, including a follow-up, "imposed another layer of coercion due to its strong suggestion that the companies could not simply ignore" the sheriff). Here, the officials' correspondences were both persistent and urgent. They sent repeated follow-up emails, whether to ask why a post or account was "still up" despite being flagged or to probe deeper into the platforms' internal policies. On the latter point, for example, one official asked at least twelve times for detailed information on Facebook's moderation practices and activities.

Admittedly, many of the officials' communications are not by themselves coercive. But, we do not take a speaker's communications "in isolation." Instead, we look to the "tenor" of the parties' relationship and the conduct of the government in context.  Given their treatment of the platforms as a whole, we find the officials' tone and demeanor was coercive, not merely persuasive.

Second, we ask how the platforms perceived the communications. Notably, "a credible threat may violate the First Amendment even if 'the victim ignores it, and the threatener folds his tent.'" Still, it is more likely to be coercive if there is some evidence that the recipient's subsequent conduct is linked to the official's message…. Here, there is plenty of evidence—both direct and circumstantial, considering the platforms' contemporaneous actions—that the platforms were influenced by the officials' demands.

When officials asked for content to be removed, the platforms took it down. And, when they asked for the platforms to be more aggressive, "interven[e]" more often, take quicker actions, and modify their "internal policies," the platforms did—and they sent emails and assurances confirming as much. For example, as was common after public critiques, one platform assured the officials they were "committed to addressing the [] misinformation that you've called on us to address" after the White House issued a public statement.

Another time, one company promised to make an employee "available on a regular basis" so that the platform could "automatically prioritize" the officials' requests after criticism of the platform's response time. Yet another time, a platform said it was going to "adjust [its] policies" to include "specific recommendations for improvement" from the officials, and emailed as much because they "want[ed] to make sure to keep you informed of our work on each" change. Those are just a few of many examples of the platforms changing—and acknowledging as much—their course as a direct result of the officials' messages.

Third, we turn to whether the speaker has "authority over the recipient." Here, that is clearly the case. As an initial matter, the White House wields significant power in this Nation's constitutional landscape. It enforces the laws of our country, and—as the head of the executive branch—directs an army of federal agencies that create, modify, and enforce federal regulations…. At the very least, as agents of the executive branch, the officials' powers track somewhere closer to those of the commission in Bantam Books—they were legislatively given the power to "investigate violations[] and recommend prosecutions."

But, authority over the recipient does not have to be a clearly-defined ability to act under the close nexus test. Instead, a generalized, non-descript means to punish the recipient may suffice depending on the circumstances…. [A] message may be "inherently coercive" if, for example, it was conveyed by a "law enforcement officer" or "penned by an executive official with unilateral power." In other words, a speaker's power may stem from an inherent authority over the recipient. That reasoning is likely applicable here, too, given the officials' executive status.

It is not even necessary that an official have direct power over the recipient. Even if the officials "lack[ed] direct authority" over the platforms, the cloak of authority may still satisfy the authority prong….

True, the government can "appeal[]" to a private party's "interest in avoiding liability" so long as that reference is not meant to intimidate or compel. But here, the officials' demands that the platforms remove content and change their practices were backed by the officials' unilateral power to act or, at the very least, their ability to inflict "some form of punishment" against the platforms. Therefore, the authority factor weighs in favor of finding the officials' messages coercive.

Finally, and "perhaps most important[ly]," we ask whether the speaker "refers to adverse consequences that will follow if the recipient does not accede to the request." Explicit and subtle threats both work—"an official does not need to say 'or else' if a threat is clear from the context." Again, this factor is met.

Here, the officials made express threats and, at the very least, leaned into the inherent authority of the President's office. The officials made inflammatory accusations, such as saying that the platforms were "poison[ing]" the public, and "killing people." The platforms were told they needed to take greater responsibility and action. Then, they followed their statements with threats of "fundamental reforms" like regulatory changes and increased enforcement actions that would ensure the platforms were "held accountable." But, beyond express threats, there was always an "unspoken 'or else.' After all, as the executive of the Nation, the President wields awesome power. The officials were not shy to allude to that understanding native to every American—when the platforms  faltered,  the  officials  warned  them  that  they  were "[i]nternally … considering our options on what to do," their "concern[s] [were] shared at the highest (and I mean highest) levels of the [White House]," and the "President has long been concerned about the power of large social media platforms." …

Given all of the above, we are left only with the conclusion that the officials' statements were coercive….

And the court held that the White House and the Surgeon General's office "also significantly encouraged the platforms to moderate content by exercising active, meaningful control over those decisions" by "entangl[ing] themselves in the platforms' decision-making processes, namely their moderation policies"—an independent basis, in the court's view, for treating the government's action as state action, even apart from coercion:

The officials had consistent and consequential interaction with the platforms and constantly monitored their moderation activities. In doing so, they repeatedly communicated their concerns, thoughts, and desires to the platforms. The platforms responded with cooperation—they invited the officials to meetings, roundups, and policy discussions. And, more importantly, they complied with the officials' requests, including making changes to their policies.

The officials began with simple enough asks of the platforms—"can you share more about your framework here" or "do you have data on the actual number" of removed posts? But, the tenor later changed. When the platforms' policies were not performing to the officials' liking, they pressed for more, persistently asking what "interventions" were being taken, "how much content [was] being demoted," and why certain posts were not being removed.

Eventually, the officials pressed for outright change to the platforms' moderation policies. They did so privately and publicly. One official emailed a list of proposed changes and said, "this is circulating around the building and informing thinking." The White House Press Secretary called on the platforms to adopt "proposed changes" that would create a more "robust enforcement strategy." And the Surgeon General published an advisory calling on the platforms to "[e]valuate the effectiveness of [their] internal policies" and implement changes. Beyond that, they relentlessly asked the platforms to remove content, even giving reasons as to why such content should be taken down. They also followed up to ensure compliance and, when met with a response, asked how the internal decision was made.

And, the officials' campaign succeeded. The platforms, in capitulation to state-sponsored pressure, changed their moderation policies. The platforms explicitly recognized that. For example, one platform told the White House it was "making a number of changes"—which aligned with the officials' demands—as it knew its "position on [misinformation] continues to be a particular concern" for the White House. The platform noted that, in line with the officials' requests, it would "make sure that these additional [changes] show results—the stronger demotions in particular should deliver real impact." Similarly, one platform emailed a list of "commitments" after a meeting with the White House which included policy "changes" "focused on reducing the virality" of anti-vaccine content even when it "does not contain actionable misinformation." Relatedly, one platform told the Surgeon General that it was "committed to addressing the [] misinformation that you've called on us to address," including by implementing a set of jointly proposed policy changes from the White House and the Surgeon General.

Consequently, it is apparent that the officials exercised meaningful control—via changes to the platforms' independent processes—over the platforms' moderation decisions. By pushing changes to the platforms' policies through their expansive relationship with and informal oversight over the platforms, the officials imparted a lasting influence on the platforms' moderation decisions without the need for any further input. In doing so, the officials ensured that any moderation decisions were not made in accordance with independent judgments guided by independent standards. Instead, they were encouraged by the officials' imposed standards.

In sum, we find that the White House officials, in conjunction with the Surgeon General's office, coerced and significantly encouraged the platforms to moderate content. As a result, the platforms' actions "must in law be deemed to be that of the State."

The court also found impermissible coercion and significant encouragement as to certain FBI requests:

We start with coercion. Similar to the White House, Surgeon General, and CDC officials, the FBI regularly met with the platforms, shared "strategic information," frequently alerted the social media companies to misinformation spreading on their platforms, and monitored their content moderation policies. But, the FBI went beyond that—they urged the platforms to take down content. Turning to the Second Circuit's four-factor test, we find that those requests were coercive. [Details omitted. -EV] …

We also find that the FBI likely significantly encouraged the platforms to moderate content by entangling themselves in the platforms' decision-making processes. Beyond taking down posts, the platforms also changed their terms of service in concert with recommendations from the FBI. For example, several platforms "adjusted" their moderation policies to capture "hack-and-leak" content after the FBI asked them to do so (and followed up on that request). Consequently, when the platforms subsequently moderated content that violated their newly modified terms of service (e.g., the results of hack-and-leaks), they did not do so via independent standards. Instead, those decisions were made subject to commandeered moderation policies.

In short, when the platforms acted, they did so in response to the FBI's inherent authority and based on internal policies influenced by FBI officials. Taking those facts together, we find the platforms' decisions were significantly encouraged and coerced by the FBI.

As to the CDC, the court held that, "although not plainly coercive, the CDC officials likely significantly encouraged the platforms' moderation decisions, meaning they violated the First Amendment":

We start with coercion. Here, like the other officials, the CDC regularly met with the platforms and frequently flagged content for removal. But, unlike the others, the CDC's requests for removal were not coercive—they did not ask the platforms in an intimidating or threatening manner, do not possess any clear authority over the platforms, and did not allude to any adverse consequences. Consequently, we cannot say the platforms' moderation decisions were coerced by CDC officials.

The same, however, cannot be said for significant encouragement. Ultimately, the CDC was entangled in the platforms' decision-making processes.

The CDC's relationship with the platforms began by defining—in "Be On the Lookout" meetings—what was (and was not) "misinformation" for the platforms. Specifically, CDC officials issued "advisories" to the platforms warning them about misinformation "hot topics" to be wary of. From there, CDC officials instructed the platforms to label disfavored posts with "contextual information," and asked for "amplification" of approved content. That led to CDC officials becoming intimately involved in the various platforms' day-to-day moderation decisions. For example, they communicated about how a platform's "moderation team" reached a certain decision, how it was "approach[ing] adding labels" to particular content, and how it was deploying manpower. Consequently, the CDC garnered an extensive relationship with the platforms.

From that relationship, the CDC, through authoritative guidance, directed changes to the platforms' moderation policies. At first, the platforms asked CDC officials to decide whether certain claims were misinformation. In response, CDC officials told the platforms whether such claims were true or false, and whether information was "misleading" or needed to be addressed via CDC-backed labels. That back-and-forth then led to "[s]omething more."

Specifically, CDC officials directly impacted the platforms' moderation policies. For example, in meetings with the CDC, the platforms actively sought to "get into [] policy stuff" and run their moderation policies by the CDC to determine whether the platforms' standards were "in the right place." Ultimately, the platforms came to heavily rely on the CDC. They adopted rule changes meant to implement the CDC's guidance. As one platform said, they "were able to make [changes to the 'misinfo policies'] based on the conversation [they] had last week with the CDC," and they "immediately updated [their] policies globally" following another meeting. And, those adoptions led the platforms to make moderation decisions based entirely on the CDC's say-so—"[t]here are several claims that we will be able to remove as soon as the CDC debunks them; until then, we are unable to remove them." That dependence, at times, was total. For example, one platform asked the CDC how it should approach certain content and even asked the CDC to double check and proofread its proposed labels.

Viewing these facts, we are left with no choice but to conclude that the CDC significantly encouraged the platforms' moderation decisions. Unlike in Blum v. Yaretsky (1982), the platforms' decisions were not made by independent standards, but instead were marred by modification from CDC officials. Thus, the resulting content moderation, "while not compelled by the state, was so significantly encouraged, both overtly and covertly" by CDC officials that those decisions "must in law be deemed to be that of the state."

But the court concluded that, as to the National Institute of Allergy and Infectious Diseases, the State Department, and the Cybersecurity and Infrastructure Security Agency, "there was not, at this stage, sufficient evidence to find that it was likely these groups coerced or significantly encouragement the platforms":

For the NIAID officials, it is not apparent that they ever communicated with the social-media platforms. Instead, the record shows, at most, that public statements by Director Anthony Fauci and other NIAID officials promoted the government's scientific and policy views and attempted to discredit opposing ones—quintessential examples of government speech that do not run afoul of the First Amendment….

As for the State Department, while it did communicate directly with the platforms, so far there is no evidence these communications went beyond educating the platforms on "tools and techniques" used by foreign actors. There is no indication that State Department officials flagged specific content for censorship, suggested policy changes to the platforms, or engaged in any similar actions that would reasonably bring their conduct within the scope of the First Amendment's prohibitions. After all, their messages do not appear coercive in tone, did not refer to adverse consequences, and were not backed by any apparent authority. And, per this record, those officials were not involved to any meaningful extent with the platforms' moderation decisions or standards.

Finally, although CISA flagged content for social-media platforms as part of its switchboarding operations, based on this record, its conduct falls on the "attempts to convince," not "attempts to coerce," side of the line. There is not sufficient evidence that CISA made threats of adverse consequences—explicit or implicit—to the platforms for refusing to act on the content it flagged. Nor is there any indication CISA had power over the platforms in any capacity, or that their requests were threatening in tone or manner. Similarly, on this record, their requests—although certainly amounting to a non-trivial level of involvement—do not equate to meaningful control. There is no plain evidence that content was actually moderated per CISA's requests or that any such moderation was done subject to non-independent standards….

The court "emphasize[d] the limited reach of [its] decision":

We do not uphold the injunction against all the officials named in the complaint. Indeed, many of those officials were permissibly exercising government speech, "carrying out [their] responsibilities," or merely "engaging in [a] legitimate [] action." That distinction is important because the state-action doctrine is vitally important to our Nation's operation—by distinguishing between the state and the People, it promotes "a robust sphere of individual liberty." … If just any relationship with the government "sufficed to transform a private entity into a state actor, a large swath of private entities in America would suddenly be turned into state actors and be subject to a variety of constitutional constraints on their activities." So, we do not take our decision today lightly.

But, the Supreme Court has rarely been faced with a coordinated campaign of this magnitude orchestrated by federal officials that jeopardized a fundamental aspect of American life. Therefore, the district court was correct in its assessment—"unrelenting pressure" from certain government officials likely "had the intended result of suppressing millions of protected free speech postings by American citizens."

And the court held that the district court injunction was overbroad:

[Parts of the injunction] prohibit the officials from engaging in, essentially, any action "for the purpose of urging, encouraging, pressuring, or inducing" content moderation. But "urging, encouraging, pressuring" or even "inducing" action does not violate the Constitution unless and until such conduct crosses the line into coercion or significant encouragement….

[Certain other] provisions likewise may be unnecessary to ensure Plaintiffs' relief. A government actor generally does not violate the First Amendment by simply "following up with social-media companies" about content-moderation, "requesting content reports from social-media companies" concerning their content-moderation, or asking social media companies to "Be on The Lookout" for certain posts….

These provisions are vague as well. There would be no way for a federal official to know exactly when his or her actions cross the line from permissibly communicating with a social-media company to impermissibly "urging, encouraging, pressuring, or inducing" them "in any way." …

Finally, [one other] prohibition—which bars the officials from "collaborating, coordinating, partnering, switchboarding, and/or jointly working with the Election Integrity Partnership, the Virality Project, the Stanford Internet Observatory, or any like project or group" to engage in the same activities the officials are proscribed from doing on their own—may implicate private, third-party actors that are not parties in this case and that may be entitled to their own First Amendment protections. Because the provision fails to identify the specific parties that are subject to the prohibitions, and "exceeds the scope of the parties' presentation." …

That leaves [one remaining provision], which bars the officials from "threatening, pressuring, or coercing social-media companies in any manner to remove, delete, suppress, or reduce posted content of postings containing protected free speech." But, those terms could also capture otherwise legal speech. So, the injunction's language must be further tailored to exclusively target illegal conduct and provide the officials with additional guidance or instruction on what behavior is prohibited….[It] is MODIFIED to state:

Defendants, and their employees and agents, shall take no actions, formal or informal, directly or indirectly, to coerce or significantly encourage social-media companies to remove, delete, suppress, or reduce, including through altering their algorithms, posted social-media content containing protected free speech. That includes, but is not limited to, compelling the platforms to act, such as by intimating that some form of punishment will follow a failure to comply with any request, or supervising, directing, or otherwise meaningfully controlling the social-media companies' decision-making processes.

Under the modified injunction, the enjoined Defendants cannot coerce or significantly encourage a platform's content-moderation decisions. Such conduct includes threats of adverse consequences—even if those threats are not verbalized and never materialize—so long as a reasonable person would construe a government's message as alluding to some form of punishment. That, of course, is informed by context (e.g., persistent pressure, perceived or actual ability to make good on a threat). The government cannot subject the platforms to legal, regulatory, or economic consequences (beyond reputational harms) if they do not comply with a given request. The enjoined Defendants also cannot supervise a platform's content moderation decisions or directly involve themselves in the decision itself. Social-media platforms' content-moderation decisions must be theirs and theirs alone. This approach captures illicit conduct, regardless of its form….

Note that, when a court of appeals strikes down a federal statute, and the federal government then asks the Supreme Court to review the matter, the Court is very likely to say yes. The Court's view is that the judiciary may properly tell Congress that it can't do something—but if that's done, that should be the province of the Supreme Court, and not one of the lower courts. I expect the Justices would take the same view of an injunction that orders the President not to do things; if the Solicitor General seeks review by the Court, the Court is likely to agree to hear the matter.