Updated at 9:33 a.m. ET on January 27, 2021.
There is a rich historical irony to the fact that today, conservatives are the ones who argue most forcefully that the decisions by private companies to “deplatform” certain speakers threaten what President Donald Trump described in 2020 as the “bedrock” American right to freedom of speech. Until very recently, this was an argument made almost exclusively by those on the left.
The decision by Twitter, Facebook, and a host of other social-media outlets to ban Trump from their platforms after the January 6 attack on the Capitol intensified conservatives’ long-standing concerns that the powerful tech industry is violating their free-speech rights. Trump encouraged and amplified these arguments when he issued a (largely symbolic) executive order in May 2020 declaring that “free speech is the bedrock of American democracy,” and insisted that “in a country that has long cherished the freedom of expression, we cannot allow a limited number of online platforms to hand pick the speech that Americans may access and convey.”*
The deplatforming of the president appeared to many conservatives to offer vivid proof that these companies are just as dangerous to freedom of speech as Trump had claimed. Steve Daines, a Republican senator from Montana, took to Twitter to attack “Big Tech” for “censoring [Trump] and the free speech of American citizens.” Trump’s trade adviser, Peter Navarro, claimed that the platforms’ decision to restrict speech “threatened our democracy.” And on the floor of the Capitol building, newly sworn-in Representative Marjorie Taylor Greene of Georgia wore a mask bearing a single word—CENSORED—in stark white letters. Many liberals, meanwhile, insisted that the decision to deplatform the president had nothing to do with freedom of speech, at least not as protected by the First Amendment.
This is something of a reversal. Indeed, the idea that private actors, not just government officials, might threaten the freedom of speech guaranteed by the First Amendment, as well as the other rights protected by the Constitution, was first suggested by big-government liberals, whom contemporary conservatives love to hate. In the early 20th century, progressive legal scholars such as Felix Cohen and Robert Hale argued against the notion that the Constitution protects rights including freedom of speech from only government action. Private corporations wield tremendous power over individuals’ lives and fortunes, and to overlook that power when interpreting the meaning of constitutionally protected rights, Cohen and Hale believed, would make no sense.
This argument eventually found favor with progressive justices on the Supreme Court during the New Deal and led the court to conclude—as it did in the 1946 decision Marsh v. Alabama, for example—that the First Amendment could prevent private corporations from excluding speakers from property they owned and controlled when doing so was necessary to ensure that “the channels of communication remain free.” In later decades, although the Court struggled to define exactly when and under what circumstances the First Amendment applied to private actors, it continued to insist that it did sometimes apply. In 1968, for example, the great liberal lion, Justice Thurgood Marshall, wrote an opinion that held that a shopping mall’s private owner could not exclude protesters from the mall’s passageways without violating their First Amendment rights. Only after President Richard Nixon appointed four pro-business conservative justices did the Supreme Court reject this view of the First Amendment, and insist that private corporations have no constitutional obligation to grant access to their property to speakers they dislike, no matter how powerful those corporations might be.
When Trump and other conservatives complain that the decision to remove the president from popular platforms violates his freedom of speech, they place themselves in strange company. They acknowledge, albeit only implicitly and perhaps opportunistically, that early-20th-century progressives were correct to worry about private power’s threat to constitutionally protected liberties.
This recognition is welcome, if overdue. For decades now, nearly all of the important forums of mass communication in the United States (radio and television stations, newspapers and magazines, movies and, yes, social-media platforms) have been privately owned. Given this state of affairs, private companies’ decisions about what speech to allow or exclude from their property obviously have the capacity to limit the free and open debate that sustains American democracy. The difficult thing is figuring out what to do about it.
In recent weeks, some conservatives have suggested that courts should impose the same First Amendment duties on today’s social-media companies that the Marsh v. Alabama Court imposed on a private owner of a company town. In principle, this approach makes a lot of sense. Just like the company town in Marsh, Twitter and Facebook today provide an important forum for public conversation and debate. They represent, as Senator John Cornyn has argued, the new “public squares” of the internet age.
In practice, however, extending the rule from Marsh to social media would effectively make the nine justices on the Supreme Court (many of whom have, by all appearances, a poor grasp of the basic mechanisms of digital technology) the final arbiters of freedom of speech on social media for 330 million Americans. One might doubt whether the Court is best positioned to assess how free-speech principles translate to this new technological environment. Even if one doesn’t doubt that, the Supreme Court has evinced no desire to extend its holding in Marsh to new kinds of private property. If anything, the opposite.
Rather than imposing First Amendment duties on the powerful private companies that operate today’s virtual public squares, some on the left have argued, the best option for preserving freedom of speech on social media is to allow the companies to self-regulate, by creating internal speech policies that limit their ability to pick and choose what speech to allow on the platform. This is an idea that, unlike the probably doomed idea of reviving Marsh, is already being put into practice.
Over the past few years, social-media companies have expended considerable effort developing internal policies that they claim are designed to ensure that “all people can participate in the public conversation [on the platforms] freely and safely.” Instead of exercising the unbounded freedom that post-Nixon First Amendment cases give them to exclude whomever they like from their platforms, companies such as Twitter and Facebook have declared themselves bound by a principled, though not legally mandated, duty to promote freedom of speech on their platforms, and have developed policies that allow speech to be removed, flagged, or hidden only when it satisfies certain conditions. These policies also provide limited due-process rights to those regulated by them.
When they banned Trump, the platforms took care to justify the decision by reference to these policies. Twitter, for example, provided a detailed explanation of why Trump’s speech violated its policy against glorifying violence and therefore could be removed despite the company’s general preference for “the public [to] hear from elected officials and world leaders directly.” Mark Zuckerberg made a similar argument to explain why Facebook was banning Trump until the end of his presidency, and just recently Facebook asked its court-like Oversight Board, established a few months ago to provide independent oversight of its speech-regulating decisions, to evaluate whether the ban violated the company’s policies.
These efforts to justify Trump’s deplatforming by reference to social-media companies’ internal speech policies—and in particular, Facebook’s willingness to have that decision reviewed by an independent, quasi-judicial Oversight Board—suggest that the project of platform self-regulation is gaining traction. The important question facing internet users in the United States and around the world is whether the platforms’ self-regulation will be sufficient to protect the important democratic and expressive freedoms that the American free-speech tradition cares about.
There are reasons to be skeptical that self-regulation will be enough. Perhaps the primary reason is the fact that, notwithstanding their presumably sincere commitment to freedom of speech, social-media companies are, in the end, for-profit entities that offer a forum for speech in order to make money. Will they protect expressive freedom even when it conflicts with corporate profits? Conversely, outside the extraordinary circumstances of the Capitol invasion, will they take down genuinely harmful speech that brings readers to their platforms? Past history suggests that the answer to both of these questions will be no. Certainly the often–ad hoc and inconsistent decision making that the platforms demonstrated during the 2020 election campaign is alone concerning.
Given all of this, it is worth considering a third option that has been used in the past, and could once again be used, to protect expressive freedom from private power: laws requiring that the private media companies governing the mass public sphere abide by basic nondiscrimination and, often, due-process obligations. Even when the First Amendment intruded further into the private sphere than it does today, statutory nondiscrimination and due-process requirements were lawmakers’ primary tools to ensure that the private companies that controlled the telegraph and telephone wires, the radio and television airwaves, and the cable networks did not use their power to discriminate in favor of certain political viewpoints, or otherwise undermine the vitality of public debate. The most famous, and controversial, example of these laws was the Fairness Doctrine, which imposed extensive, if vague, nondiscrimination duties on radio and television broadcasters, and to an extent, cable-television companies, from the 1930s until the late ’80s, when Ronald Reagan’s FCC repealed it. But the Fairness Doctrine is only one example of a much wider array of media nondiscrimination laws, many of which continue to ensure, to this day, that, as one senator put it in 1926, the “few men” who control the “great publicity vehicles” of radio and television do not limit the range of ideas and viewpoints that the public can hear.
In this context as well, a significant shift in political attitudes has occurred. For much of the 20th century, conservatives were the ones who railed against the constraints that federal laws like the Fairness Doctrine imposed on private media companies, and liberals and progressives defended these policies against attack. Today, however, many conservatives argue for the need to impose statutory nondiscrimination duties on social-media companies, while many liberals express alarm about the constraints such bills would impose on the freedom of private companies.
Although some of the bills that have been proposed to rein in social-media companies’ power are certainly poorly drafted and could easily be abused by self-interested politicians, advocates on the left should not give up on the possibility of using regulation to protect freedom of speech on the platforms. Designing nondiscrimination rules that can work effectively in social media’s new technological environment will be no easy feat. But that does not mean it cannot be done. There is no reason Congress could not impose minimum procedural requirements on the platforms when they act to remove their users’ speech.
All of which is to say that the debate about free speech on social media should not be viewed primarily as a debate about whether the social-media companies violated Trump’s freedom of speech when they banned him, or whether they violate anyone else’s freedom of speech when they make thousands of similar decisions every day. Instead, it should be viewed primarily as a debate about what freedom of speech means on social media, and, perhaps most importantly, about who gets to decide—courts, corporations, or legislatures. That liberals and conservatives have switched perspectives on these questions in recent years reflects the extraordinary political fluidity, and perhaps possibility, of the current moment.
However the political alignments work out, Trump’s deplatforming illuminated a basic insight worth keeping in mind: Private companies not only participate in the marketplace of ideas but also determine to a significant extent who else can participate in it. We should not take comfort in the fact that the speech-regulating decisions by Big Tech companies do not and cannot violate the First Amendment as it is currently understood. Conservatives are correct to be worried about the threat that the private platforms pose to freedom of speech, even if this makes them more like big-government liberals than they might be willing to acknowledge. Those big-government liberals should realize as much, and act accordingly.
* This article previously misstated the date when Trump signed the executive order concerning free speech. It was May 2020, not May 2019.