The decision by a half dozen of the largest social media companies in the United States to ban President Trump and thousands of his supporters after their posts instigated the violent invasion of the Capitol building on January 6 raised all kinds of difficult questions about the meaning of freedom of speech in the United States. Was the deplatforming a condemnable act of censorship or a necessary means of preventing future attacks? In what circumstances could incendiary and duplicitous speakers like President Trump justifiably be excluded from the platforms? And who should be responsible for determining the speech rules that govern the social media public sphere? These questions pervaded public debate about what some began to call the Great Deplatforming and they generated a wide range of answers.
Despite the vigorous debate about freedom of speech raised by the deplatforming, there was wide agreement—at least among lawyers—that the First Amendment provided no help in answering them. Almost all litigators and law professors who weighed in the Great Deplatforming agreed that, while it raised difficult policy questions, it raised no serious First Amendment questions because the social media companies were private actors to whom the First Amendment did not apply. Jameel Jaffer, the Executive Director of the Knight First Amendment Institute at Columbia University, eloquently expressed this view when he tweeted that “The First Amendment question is easy. All the other questions are hard.”
The idea that the First Amendment was irrelevant here shocked many outside the law world. Republican politicians in particular continued to insist that the social media companies had violated the First Amendment when they banned the President because of his speech.
The First Amendment’s irrelevance flows almost ineluctably, however, from the rigid framework that the contemporary cases apply to determine its scope. Under existing rules, the Constitution limits only state actors. Private persons confront the Speech Clause in very few circumstances: mostly when they exercise a function that has traditionally been the exclusive prerogative of the government, or when their conduct is coerced or very strongly encouraged by government actors.
Doctrinally, then, the conclusion that the social media companies were private rather than state actors was easy to reach. Running a social media platform is not a public function that has traditionally been performed by the government. Nor did government actors dictate the companies’ deplatforming decisions or even strongly encourage it. Twitter and Facebook came to those decisions on their own.
Most scholars and litigators who assessed the merits of the Great Deplatforming also endorsed the view that the social media companies ideally should not be subject to the Speech Clause, given the significant constraints that it would impose on their ability to regulate user speech that flowed through their platforms. Under existing interpretations, the First Amendment prohibits state actors from almost ever discriminating against speech because of its viewpoint or message. If they were subject to that rule, social media platforms would find it extremely difficult to exclude lies or misinformation, to ban encouragement of violence that did not amount to explicit incitement, or to prohibit the repetition of odious racist stereotypes. Were courts to analogize the platforms to “quintessential public forums” like public streets, parks, and sidewalks, they would impose even greater constraints on the companies’ ability to regulate. Because the government is generally prohibited from discriminating against speech because of its content when it regulates the traditional public sphere, the extension of these duties to the social media companies would prevent them from being able to engage in the content-based acts of filtering, recommending, and ranking speech that play such an important role at present in organizing the digital public sphere.
It is not difficult, therefore, to understand why even enthusiastic advocates of free speech were unwilling to argue that the social media companies should be subjected to those rules. Doing so appeared very likely to make the platforms, as Jack Balkin put it, “far less valuable to end users, if not wholly ungovernable,” and no one wanted that.
The result was nevertheless to lead most scholars and litigators to agree that the most important legal guarantee of freedom of speech in the United States provided no protection whatsoever against what the ACLU rightly described as the “unchecked power” of the social media companies “to remove people from platforms that have become indispensable to the speech of billions.” And not only that.
In recent years, the Court has interpreted the First Amendment not only to not prohibit private acts of speech discrimination but also to actively protect the right of private actors to discriminate free from legislative control. So in concluding that the social media companies were private actors, those who weighed in on the Great Deplatforming at least implicitly endorsed a view of the First Amendment that could be used to render those companies immune not only from judicial checks on their decisionmaking, but from legislative checks as well. The ACLU’s public statement certainly appeared to recognize as much. After warning darkly of the great power that the social media companies possessed, the statement identified voluntary transparency as the only mechanism for checking that power—the ACLU said it could only “hope that these companies will apply their rules transparently to everyone.”
We think it is an impoverished view of freedom of speech that leaves such weak means for protecting the expressive and democratic freedom of so many people. The Great Deplatforming may have illuminated many other things, but one thing it clearly spotlighted is how poorly equipped contemporary First Amendment doctrine is for conceptualizing, let alone doing anything about, the threats to freedom of speech that result from private control of the mass public sphere. Because virtually every forum of mass communication in the United States is privately owned and controlled, the First Amendment today sharply limits the ability of the democratic government to oversee the operation of its most important forums.
We do not want to have to choose between John Roberts and Mark Zuckerberg as the guardians of democracy, though that is what current doctrine seems to require. Luckily, the contemporary framework is not the only one available to us.
A Spectrum of Public Duties
If we look to an earlier epoch of First Amendment jurisprudence—the era of the 1940s, 1950s, and 1960s—what we find is a far less formalistic approach to the delimitation of the First Amendment’s boundaries, one that is much more attentive to the problem of private power than the current framework tends to be, and far less antagonistic to the possibility of democratic supervision of the mass public sphere. This is a model that would better equip courts, lawyers and legislatures to think through the difficult free speech questions raised by events like the Great Deplatforming.
To get a sense of how this earlier approach operated, one need only look at one of the most famous but also perhaps most misunderstood cases in the First Amendment canon—the 1946 decision in Marsh v. Alabama. Marsh is almost always misremembered today as a decision that held that private corporations can be considered state actors when they perform a traditional government function, like running a town. But that is not what Marsh held. Neither the majority nor the dissent ever suggested the possibility that the corporation involved in the case—the Gulf Shipbuilding Corporation—was a state actor. (Gulf wasn’t in fact a party to the case.) All nine members of the Court instead agreed that the relevant state actors were the three Alabama state courts that either convicted or affirmed the trespass conviction of a Jehovah’s Witness who Gulf tried to kick out of the town it owned because it disliked her religious proselytizing but who refused to leave. The Marsh majority reversed the trespass conviction because it concluded that otherwise the corporation could use state property law to dictate what views could be expressed in the town’s public spaces, thus denying the town’s residents of a full opportunity to exercise their First Amendment rights.
The Marsh Court concluded, in other words, that a private corporation—a private actor!—had a duty to respect the free speech interests of other private persons because of the role it played in the regulation of the public sphere. More specifically, the Court held that the Gulf Shipbuilding Corporation had a duty to make sure the channels of communication in its town remained open because it had voluntarily, and for its own profit, turned its property into a part of the public realm. This was because, the Court noted, “[w]hether a corporation or a municipality owns or possesses the town, the public . . . has an identical interest in the functioning of the community in such manner that the channels of communication remain free.” It followed from this that “the managers appointed by the corporation [could not] curtail the liberty of press and religion of these people” any more than the government could.
This is an incredibly foreign way of thinking about the scope of the First Amendment when viewed from the vantage point of contemporary doctrine. Rather than presuming that the First Amendment imposes duties only on government actors, it envisages something like a continuum of constitutional obligations to match what we might think of as a continuum of public power. Private actors who open up their property more, and thereby acquire more of the attributes of government control, have a concomitantly greater responsibility to honor the statutory and constitutional interests of the public who use their property than private actors who open up their property less. But in all cases, the constitutional rights of members of the public extend beyond, or properly work to define, the public/private divide.
The result is to capture in doctrinal form an important feature of our contemporary reality that current First Amendment law almost entirely ignores: namely, the fact that private actors do not simply participate in the marketplace of ideas but also regulate it, and consequently possess quasigovernmental power to manipulate, or limit the scope of public debate. Yet, the Marsh framework does not require courts, or for that matter, anyone else, to treat private actors such as social media companies exactly like the government.
Indeed, the view of the First Amendment outlined in Marsh is entirely consistent with the state action requirement. As the opinion makes clear, private corporations may owe a duty to respect the constitutional rights of those who live, work, or speak on their property, but those duties are not directly enforceable against them. It is the responsibility of state actors to ensure that private property owners do not deny other private citizens important, constitutionally protected liberties, such as the right to access a diverse range of views on matters of public debate. This is more or less the view of state action that the Court employed in the Reconstruction Era cases in which it first defined the term. In its 1883 opinion in the Civil Rights Cases, for example, the Court insisted that a government that failed to punish those who “by force or fraud, interfere[d] with the enjoyment of the rights” guaranteed to private citizens by the Fourteenth Amendment satisfied the state action requirement—and violated the Constitution. Marsh holds essentially the same to be true, when it comes to the rights protected by the First Amendment.
The approach does not, in other words, deny that there are important differences in the constitutional status of government actors and even very powerful private corporations, like Twitter and Facebook and YouTube. However powerful corporations may be, they do not exercise the monopoly on the means of legitimate force that the government does; notwithstanding network effects, it remains much easier to leave Facebook than to leave the United States. The approach does not as a result threaten to create the ungovernable hell that Balkin fears.
But it also does not presume that the differences in the constitutional status of government and private actors transform the First Amendment into a grant of near-total immunity to private property owners to do with their property whatever they will. To the contrary: Marsh construed the First Amendment to require state actors to interfere with the autonomy of private property owners when their actions threaten others’ freedom of speech by, for example, refusing to enforce their property rights in such cases.
Legislative Obligations to Protect Freedom of Speech
Although Marsh dealt only with judicial checks on rights-violating private action, in later cases, the Court, relying on essentially the same view of the First Amendment, made clear that Congress also had a responsibility to act, to ensure that the channels of communication ran free. One can see as much if one looks at the opinion in Associated Press v. United States, which held that the Sherman Antitrust Act could require the powerful new-gathering organization the Associated Press to open up its membership rolls—and therefore share its news with—newspapers that competed with existing members. The AP argued that enforcement of the Act violated its First Amendment rights but the Court vehemently disagreed. “The First Amendment,” Justice Black wrote, again on behalf of the Court, “far from providing an argument against application of the Sherman Act, provides powerful reasons to the contrary. . . Freedom of the press from governmental interference under the First Amendment does not sanction repression of that freedom by private interests. The First Amendment affords not the slightest support for the contention that a combination to restrain trade in news and views has any constitutional immunity.”
Two decades later, in Red Lion Broadcasting v. FCC, the Court similarly suggested that the right of reply obligations imposed on radio and television broadcasters by the Federal Communications Commission under the authority given to it by the 1934 Communications Act might be constitutionally compelled, not just constitutionally permitted. “It is the purpose of the First Amendment to preserve an uninhibited marketplace of ideas in which truth will ultimately prevail,” the Court declared, “rather than to countenance monopolization of that market, whether it be by the Government itself or a private licensee. . . . It is the right of the public to receive suitable access to social, political, esthetic, moral, and other ideas and experiences which is crucial here.” This right, the Court insisted, “may not constitutionally be abridged either by Congress or by the FCC” by failing to take action to prevent the private actors who controlled the radio spectrum from limiting the range of views that the public could access.
Red Lion is a complicated case, but the basic point for our purposes is this: in multiple cases, and over many decades, the Court articulated a view of the First Amendment that suggests that legislatures, as well as courts, have a constitutional duty to ensure that the social media companies do not use their obvious and significant power over the digital public sphere to unduly inhibit the marketplace of ideas, or to otherwise interfere with the First Amendment rights of those who use their platforms to speak, to associate, and to listen. Certainly, lawmakers cannot be constitutionally prohibited from ensuring the health of the speech environment, on this conception.
Democratic Control of Democracy’s Channels
This is a far more attractive view of the First Amendment than the one the Court tends to hold today because it does not pit freedom of speech against democratically-authorized regulation but views regulation as a mechanism of guaranteeing freedom of speech. It makes it possible to imagine a world in which the only avenue available for protecting expressive and associational freedom on the platforms is not to place all one’s hopes in the good will of corporate actor. Indeed, it makes it possible to imagine a world in which legislatures might be constitutionally authorized, even constitutionally required, to ensure that users of the platforms enjoy minimal due process protections against removal, or to require some degree of transparency by the social media companies in their operations, or to take other steps to ensure the vitality of speech in the digital public sphere.
Of course, as we noted earlier, the framework outlined in Marsh and Associated Press and Red Lion does not specify precisely what, in such a world, such a constitutional duty would require lawmakers to demand from private power. It does not, in other words, answer any of the difficult questions raised by the Great Deplatforming. But it makes it possible to view those questions as legal, and in fact constitutional, questions, not just questions of corporate self-regulation.
None of this presumes that legislatures will not be subject to economic and political forces that separate their interests from those of the people they represent. Nor does it ignore the gridlock that has hamstrung Congress. Without romanticizing lawmakers, however, it recognizes their role in contesting the sometimes-countervailing politics of courts and corporations.
Some might resist this approach for another reason. A more expansive First Amendment would likely include courts as arbiters of freedom of expression on the platforms and in recent years, a number of free speech scholars have argued that this is something to be avoided at all costs. Mary Anne Franks, for example, has argued that the social media companies should do all they can to escape the “black hole” of the First Amendment in order to avoid being trapped by the overly laissez faire view of freedom of speech employed by the contemporary First Amendment cases.
We completely agree that there are deep problems with how the First Amendment has been interpreted in recent decades. Indeed, we have both written extensively about them. But we think that there is still good reason to prefer a public battle that includes the courts and legislatures and corporations over the meaning and scope of democratic speech on the platforms than to put our faith solely in non-transparent, unelected, and profit-motivated corporate actors. In fact, we think it would be a calamity of significant proportions if the cramped view of the First Amendment that currently dominates the doctrine was taken to mean that debate about these questions had to take private, non-transparent and undemocratic channels, rather than to proceed vigorously via our democratic institutions.
Another way to put it is that the First Amendment questions raised by the Great Deplatforming are in fact very difficult ones, but recognizing that fact is better than believing that those questions are as easy to answer as current doctrine suggests. Freedom of speech can be broader and more complicated than formal First Amendment doctrine—the digital platforms are showing us that. That does not mean, however, that democratic forces need shape that conception solely as customers or critics. Citizens too ought to be able to shape the meaning of freedom of speech on massive corporate platforms, acting through the countervailing power of legislatures and regulators. But enabling them to do so will likely require us to reconceive what the First Amendment means, and where it applies.