With the presidential election nearing, calls for government regulations that prohibit social media platforms from censoring speech based on political biases have become more frequent and urgent. Many of those calling for restrictions on platforms’ censorship authority have presumed that government regulations will not infringe on the platforms’ First Amendment rights because they are “platforms,” not publishers.
But the U.S. Supreme Court’s recent decision in Manhattan Community Access Corp. v. Halleck (June 17, 2019) makes clear that private media platforms have the same First Amendment rights to curate the speech on their platforms that private publishers enjoy. This means any government regulation that restricts an platform’s authority to censor speech will, in fact, amount to a form of viewpoint censorship that will violate the platforms’ First Amendment rights.
This does not mean, however, that platforms’ censorship powers are beyond the reach of any government regulation. Instead of trying to adopt regulations that restrict platforms’ ability to censor based on political bias, the government could adopt due process regulations that require platforms to adopt clear rules regarding what speech is not allowed on the platform and protect users from the arbitrary enforcement of those clear rules. Unlike regulations that restrict platforms’ authority to censor, these due process regulations would be able to withstand a First Amendment challenge because due process regulations are “viewpoint neutral,” meaning that they don’t prevent platforms from controlling the types of viewpoints expressed on these platforms.
The recent revelations from Project Veritas and the high-profile censoring of Steven Crowder suggest to many conservatives that some form of regulation is needed to ensure that conservatives can fairly compete in the 2020 elections. But if those pushing for regulation fail to take into account the platforms’ First Amendment rights and push forward with legislation aimed at eliminating political bias on these platforms—legislation like that currently proposed by Sen. Josh Hawley—that legislative effort will be blocked by the courts, leaving the problem unaddressed.
Social Media Platforms Have the Right to Censor Speech
The quintessential protection provided by the First Amendment is a protection against government efforts to limit your right to express your viewpoint. While its protections limit the government’s ability to enact overt forms of censorship, it also protects against many types of “soft” censorship, such as speech restrictions embedded as conditions for grants (Agency for Int’l Dev. v. Alliance for Open Soc’y Int’l) and most forms of compelled speech (Wooley v. Maynard).
The Supreme Court has recognized that compelled speech is an especially pernicious form of viewpoint censorship because rules that force you to say certain things—for example, rules compelling you to recite the Pledge of Allegiance (Board of Education v. Barnette)—can influence your freedom of thought.
As the decision in Manhattan Community Access makes clear, media platforms do not forfeit their First Amendment rights simply because they have opened their platform to other users. The publisher/platform dichotomy is a red herring. Any regulation that compels the platform to host content that the platform does not want to host will amount to government censorship of the platform through compelled speech.
The fact that a platform like Twitter has a dominant market presence cannot justify government restrictions on that platform’s censorship authority. The Supreme Court, in fact, rejected a very similar argument when it held that newspapers could not be compelled to provide a “right to respond” to individuals who were attacked in that newspaper (Miami Herald Pub. Co. v. Tornillo). On the issue of market dominance, the court ruled that compelling the newspaper to carry an article could not be justified by the massive the imbalance between a newspaper’s ability to reach the public and an individual’s ability to do so.
Another claim some commentators have made to justify restrictions on platforms’ authority to censor speech is to analogize platforms to utilities, which are routinely subject to significant government regulation. But this argument will also fail because the Supreme Court has already held that heavily regulated utilities cannot be compelled to convey speech to their customers that the utilities do not want to convey (Pacific Gas and Elec. Co. v. Public Utilities Com’n of California).
Nor can the government fall back on anti-monopoly laws to justify restricting platforms’ right to censor based on viewpoint. Even if platforms were monopolies (remember Myspace?) the Supreme Court has already ruled that anti-monopoly laws cannot be used to justify a restriction on core First Amendment rights (Eastern R. R. Presidents Conference v. Noerr Motor Freight, Inc.).
Finally, proposals like Hawley’s—that would tie censorship restrictions to the acceptance of the government-granted immunities in Section 230—are also destined to fail. While the government can and often does condition the grant of a government benefit on the recipient agreeing to certain conditions, the Supreme Court has made it clear that those conditions cannot include requirements that the recipient convey or not convey any particular message. For example, the court ruled that nongovernmental organizations that accept U.S. funds to fight the spread of AIDs in Africa could not be compelled to advocate against prostitution as a condition for accepting those funds (Agency for Int’l Dev. v. Alliance for Open Soc’y Int’l, Inc.).
Due Process Regulations Are Fair Game, Though
While First Amendment protections may prevent the government from enacting viewpoint-based speech regulations, this does not mean that the government can do nothing to address the harms of biased platforms. Regulations that do not require platforms to alter their message—like due process regulations—can withstand First Amendment challenges, especially when those regulations are tied to an optional government benefit like the immunity provided by Section 230. Moreover, due process protections will address most of the concerns being raised.
For example, the government could require companies relying on Section 230 only to censor speech if that speech violates clear restrictions in the platforms’ terms of service. The government could also require companies to provide users notice of any censorship and an opportunity to challenge the claimed violation.
These regulations would allow users to know the rules ahead of time so they can create content that complies with those rules. The regulations would also effectively prohibit stealth censorship like shadow-banning and throttling. Finally, the appeal right would prevent arbitrary enforcement.
Would mere due process protections really make a difference? Consider how the recent actions of two of the leading platforms for crafters and DIYers: Ravelry (knitting) and Pinterest (general DIY). Ravelry took an approach that would be consistent with due process protections when it openly announced to all of its users that anyone who makes posts in support of President Trump would be banned.
In contrast, after being caught shadow-banning a conservative pro-life website (by labeling it as pornography) before outright banning it from the platform, Pinterest has continued to insist that it has no biases against conservatives. Ravelry’s upfront announcement of its censorship policy allows potential users to make an informed choice on whether they want to make content for that site.
It also allows potential customers who don’t mind users knitting Trump-themed hats and blankets to market their competing site to likeminded yarn enthusiasts. Pinterest’s denials and use of covert methods of censorship like shadow banning are designed to trick users into believing no such biases exist.
Any platform that claims to be free of biases should be carefully scrutinized because biases are inevitable. The entire purpose of the “media” is to serve a curating function for the infinite amount of information available today. The types of decisions that are necessary to serve that curating function will always involve the curator’s individual biases.
Thus, in the end, it is not even possible to adopt legislation that will make platforms free from all biases, even if such legislation could withstand a First Amendment challenge. But speech-restricting regulations are not necessary because much of the perceived harms from censorship can be prevented by due process regulations that do not unconstitutionally infringe on First Amendment rights.