Twitter Isn’t a Shopping Mall for First Amendment Purposes (Duh)–Johnson v. Twitter

Multitudinous lawsuits–mostly by political “conservatives”–seek to deprive social media services of their property rights to exclude unwanted customers. Or, more precisely, these lawsuits seek to censoriously restrict the social media services’ First Amendment-protected right to free speech/press by forcing them to publish content they, in their editorial discretion, would choose not to publish. Fortunately, these lawsuits are likely to fail (as happened with PragerU v. YouTube), though they will remain an embarrassing and permanent stain on the reputation of the folks who advanced them–plus an expensive lesson in the law for those plaintiffs who lose anti-SLAPP motions.

This case involves Chuck Johnson, who Twitter’s CEO called a “troll.” (We know this because Buzzfeed published some Twitter executives’ emails saying that). Twitter perma-banned him for his allegedly trollish behavior, including a tweet that apparently urged followers to support “taking out” a civil rights activist. Johnson sued Twitter. In a “tentative’ ruling, the lower court grants Twitter’s anti-SLAPP motion to strike.

[Procedural note: state trial court opinions are often short and not especially well-drafted. The judges rarely have clerks to help with the drafting, they don’t get a lot of experience writing complex opinions because many cases involve routine legal matters that don’t require detailed or sophisticated explanations, and the judges often have extremely busy dockets that don’t leave time to write in-depth opinions. Plus, many state court judges come from a criminal law background and spend a majority of their courtroom time on criminal matters, so they rarely have experience with counter-intuitive federal immunities like Section 230. As a result, state trial court opinions often are a little disjointed and underdeveloped. This particular opinion is sound, but its analysis is thin, which makes it a challenge to blog.]

The court says the lawsuit arises from protected activity. The court says the parties apparently agree that Twitter is a public forum for anti-SLAPP purposes and Twitter’s control of its platform is an issue of public interest. The court says the lawsuit is premised on Twitter exercising its “traditional editorial functions.”

The court then explains why Johnson failed to make a sufficient prima facie showing that he can win the case. His unpersuasive arguments included: that Twitter is the equivalent of the old town square; that Section 230 only authorizes removal of “illicit” speech (a bastardization of Section 230(c)(2)’s safe harbor for removal of objectionable content); that Twitter contravened its marketing representations as a public free speech forum; and that it’s inconsistent for Twitter to claim First Amendment protection when it also claims Section 230 immunity as a neutral public forum. [Tip: one sure-fire way to tell if someone grossly misunderstands Section 230 is if they claim it only applies to neutral public forums.] Among other responses, the court notes that Twitter relies on Section 230(c)(1), not (c)(2).

The most interesting part of the opinion discusses the potential that Twitter is a state actor for First Amendment purposes, in which case its removal decisions would need to comply with the First Amendment’s free speech requirements. The court says Twitter is private sector entity that provides services to customers only if they abide by its TOS. The court says Twitter has the “First Amendment right to exercise independent editorial control over the content on its platform,” and terminating Johnson’s account for allegedly bad behavior “is an editorial decision regarding how to present content.” The court rejects the analogy to the Pruneyard decision, which obligated a private shopping mall to tolerate protesters. The court explains:

Plaintiff’s tweet reasonably could be, and in fact was, interpreted as threatening and harassing, unlike [the protesters’] activity that “would not markedly dilute defendant’s property right.” Moreover, Defendant’s rules were adopted to ensure that Defendant is able to maintain control over its site and to protect the experience and safety of its users.

This is a nice judicial repudiation of the UGC site/shopping center analogy, though the pro-censorship crowd surely will keep trying the analogy.

The court’s latter point reinforces why I’m so baffled about the end game of the anti-free speech forces. The lawsuits seek to let bad actors have unrestricted rights to spread their garbage on social media–a scenario that would destroy social media completely. Fortunately, the court recognizes that OF COURSE social media services need to patrol and control their virtual premises to preserve the service’s utility to other users, and the First Amendment gives them the discretion to do so. This is another demonstration of how the Constitution’s framers were amazingly far-sighted about how censorious impulses would manifest themselves, even a quarter-millennium later.

Case citation: Johnson v. Twitter, Inc., No. 18CECG00078 (Cal. Superior Ct. June 6, 2018)


The Berkeley Technology Law Journal recently published an article, The “Sovereigns of Cyberspace” and State Action: The First Amendment’s Application–or Lack Thereof–to Third-Party Platforms, by Jonathan Peters. From the abstract:

This Article analyzes the state action doctrine as it exists today, examining: (1) how it distinguishes the public and private spheres, and (2) whether it forecloses the First Amendment’s application to nongovernmental Internet companies, specifically third–party platforms like Facebook and Twitter. The Article concludes that the state action doctrine does foreclose such an application.

The article itself is even more emphatic: “there is no doubt that the state action doctrine does not permit the First Amendment’s application to such platforms” (emphasis added).

After a ruling like this, Prof. Peters deserves a victory lap.

See also my short essay, Of Course the First Amendment Protects Google and Facebook (and It’s Not a Close Question)