The quintessential public forum, such as a town square, is a place for citizens to gather and discuss issues of the day and ideas related to the polity. Under the public forum doctrine, spaces that the government has opened to the public for expressive activities cannot be subject to content-based or viewpoint-based discrimination. These designated public fora can be physical spaces or virtual spaces where people engage in speech online; in fact, civic republican discourse is more likely today to rely on such virtual spaces.
Recently, in Knight First Amendment Institute at Columbia University v. Trump, the U.S. District Court for the Southern District of New York found that the “‘interactive space’ where Twitter users may directly engage with the content of the President’s tweets” constituted a designated public forum under the First Amendment, and that consequently, the President could not block the users from engaging with his tweets in that space. The Knight court’s approach raises questions about the role of Twitter’s own free speech rights, under which it and any other social media platform presumably can control who may use their platform and what may be posted. In finding that a portion of a private platform can be a designated public forum—and in therefore imposing the obligations of the First Amendment—the case exposes the incoherence of the public forum doctrine; the doctrine provides no good answer for what to do when the rights of participants in that forum potentially clash with the speech rights of the private-forum owners themselves, and raises doubt as to the control that a private actor may exercise on its own platform.
Twitter is an online platform that “allows users to post short messages . . . called tweets.” These are posted both to Twitter and to a user’s “timeline” of videos and links to which others can reply. Users can also tweet replies to other tweets, which appear as responses. Finally, users can “retweet,” or repost the message on their own timeline, or “like” tweets to convey approval or acknowledge receipt. Each tweet can be thought of as creating its own “interactive space . . . in which other users may directly interact with the content of the tweets by, for example, replying to, retweeting, or liking the tweet.” Users can also block other Twitter users, which stops the blocked individuals from being able to see the blocking user’s tweets. More generally, as it is a private platform, Twitter is able to remove any user temporarily or permanently from the platform for any (or no) reason.
President Trump blocked seven users in response to critical replies to his tweets. The users brought suit against President Trump and White House Social Media Director Daniel Scavino, alleging a First Amendment violation and requesting declaratory and injunctive relief. The Knight First Amendment Institute joined as a plaintiff, arguing that it suffered injury by being denied the opportunity to read replies the blocked users might have posted. The district court granted the plaintiffs’ claim for declaratory relief, finding that the speech at issue — the users’ criticism of President Trump — was core political speech, and that the “interactive space” created by each of President Trump’s tweets created a limited public forum in which the President could not viewpoint discriminate.
First, after finding that the plaintiffs had standing, Judge Buchwald addressed the “applicability of forum doctrine” to four discrete portions of Twitter related to the @realDonaldTrump account: “the content of the tweets sent, the timeline composed of those tweets, the comment threads initiated by each of those tweets, and the ‘interactive space’ associated with each tweet.”
The public forum doctrine seeks to designate a “right of access” to certain “public places for free speech purposes.” Those spaces found to be public fora have the protections of the First Amendment attached to them, such that content- or viewpoint-based discrimination, for example, is not permitted.
The court indicated that, to be a privately owned public forum, a space must be somehow “owned or controlled by the government.” Judge Buchwald found that three spaces — the tweets, the timeline, and the interactive space of each tweet — were controlled by the government because President Trump and Scavino managed various features of them for governmental purposes. However, the tweets and the timeline resembled government speech rather than a government-created forum for public discourse because the government had total discretion in those areas to make “content-based judgments in deciding what . . . to make available to the public.” In contrast, the decision to retweet or reply (in other words, to engage with the “interactive space” around the tweets) rests solely with other users — the public at large.
As a result, the court proceeded with the threshold analysis to determine if forum doctrine was applicable only to the interactive space of each tweet because the space’s “essential function” was “to allow private speakers to engage with the content of the tweet, which supports the application of forum analysis.” Accordingly, Judge Buchwald considered what type of forum the interactive space was. A given forum is typically classified as a traditional public forum, a designated (or limited) public forum, or a nonpublic forum; strict scrutiny applies to viewpoint- or content-based discrimination in the first two. The court held that the interactive space was not a traditional public forum, because there was “no historical practice of the interactive space of a tweet being used for public speech and debate,” in large part because “there is simply no extended historical practice as to the medium of Twitter” generally. Rather, “the interactive space [was] a designated public forum” based on the “governmental intent” inferred from “policy and past practice,” the nature of the space, and the space’s “compatibility with expressive activity.” Since @realDonaldTrump’s tweets are “generally accessible” and anyone who “has not been blocked may participate in the interactive space,” the factors counseled in favor of a designated public forum.
Judge Buchwald then turned to the question of “whether the blocking of the individual plaintiffs [was] permissible in [such] a designated public forum.” Content- or speaker-based restrictions on speech in a designated public forum “are permissible ‘only if they are narrowly drawn to achieve a compelling state interest.’” Moreover, “[v]iewpoint discrimination . . . is presumed impermissible when directed against speech otherwise within the forum’s limitations.” Judge Buchwald noted that the defendants never challenged the fact that the plaintiffs were blocked from @realDonaldTrump because of their views on the policies and ideas discussed in the tweets; they “were indisputably blocked as a result of viewpoint discrimination.” She rejected the argument that the blocking was an exercise of the President’s associational interests. As the blocked user is unable to see or respond to any of @realDonaldTrump’s tweets, blocking goes further than does choosing not to engage and actively restricts the user’s “right to speak in a discrete, measurable way.” This is especially true since it is understood that “[t]he audience for a reply extends more broadly than the sender of the tweet being replied to, and blocking restricts the ability of a blocked user to speak to that audience.” In this way, the court concluded that the President’s “blocking [of] . . . individual plaintiffs as a result of the political views they have expressed is impermissible under the First Amendment.”
In finding that a portion of a private platform can be designated as a public forum by separate government action, Knight complicates the public forum doctrine: the overall conflict between Twitter’s speech rights and the constraints of the public forum is undefined, and may be severe. Twitter has speech rights, and by creating a limited public forum, the government may be causing the First Amendment to apply to Twitter, as if Twitter were also the state. As a result, Twitter’s rights may be burdened because the government has chosen to use Twitter as a public forum.
While Twitter may not be a state actor, a limited public forum designation may nonetheless burden it with increased obligations and restrictions because the government-ownership-or-control question is largely intertwined with the state action question in an unclear way with respect to platforms like Twitter that are used by public officials. The Supreme Court has explained in previous cases that “First Amendment protection extends to corporations.” However, by using Twitter as a government forum, the state creates free speech rights for private users on the platform where none previously existed, burdening Twitter via the government’s choice to act in a way that invokes burdensome public forum doctrine.
The Court has noted that the government may not limit corporate speech in the name of antidistortion — or bringing balance to discourse — even if the corporate speech dominates and has “no correlation to the public’s support” or sentiment around an issue. This holding suggests that prior to Knight, if Twitter decided simply to block all members of a certain political party and “distort” discourse, it could do so because its platform is its speech and can weigh heavily one way.
That ability seems to conflict with Knight’s ruling — which effectively imposes an antidistortion rule onto the portions of the platform the court found to be public fora. Presumably this ruling could stop Twitter from blocking users engaged in those spaces outright, even if it otherwise could have — effectively allowing the government to “tak[e] the right to speak from some and giv[e] it to others,” or impermissibly regulate speech on an antidistortion theory.
As a result, though the court did not directly address how Twitter’s speech rights are impacted by the finding, there seems to be a direct conflict between the speech rights of Twitter and the speech rights of citizens engaging in the public forum created by @realDonaldTrump’s tweets on Twitter. Never in the court’s analysis did Judge Buchwald raise the issues of Twitter’s speech rights, whether there was an infringement of those rights, or how to handle such a conflict.
Arguments stretching back to Buckley v. Valeo indicate that “the concept that government may restrict the speech of some elements of our society in order to enhance the relative voice of others is wholly foreign to the First Amendment;” yet, in holding that the interactive spaces were public fora, Knight seemingly restricts Twitter’s rights to impose viewpoint or content controls—in effect, disfavoring its speech rights in favor of others’.
It is worth noting that the Supreme Court has addressed similar issues in the context of private physical spaces. In PruneYard Shopping Center v. Robins, the Court found that a state law extending the state’s constitutional free speech right to private malls did not abridge malls’ First Amendment rights. Professor Noah Feldman notes that PruneYard’s reasoning predated some of the development around corporations’ speech rights: “Today, the mall could probably argue that it did not want . . . any noncommercial politics of any kind.” Other space cases, such as Marsh v. Alabama, often involved company-owned towns and balanced property rights against speech rights rather than inquiring into the speech rights of the space owners. In other words, the novel case of public space on private online platforms—a problem that will only appear more as discourse moves online—creates a direct clash between speech rights that will require a similarly novel approach: looking at those clashes case-by-case evaluation against underlying First Amendment goals.
Social media platforms represent a fundamental shift in how speech and civic engagement happens, and novel conflicts of speech rights can arise as a result. As the dissent in a similar social media suit, Davison v. Randall, suggests, courts are failing to recognize serious problems about its ruling in light of these novel issues. The framework of civic republicanism, a justification for the First Amendment holding that the First Amendment should focus on creating conditions ideal for people to participate in governance and democratic discourse, is helpful in understanding these shortfalls.
Civic republicanism requires, on the one hand, protecting any core political speech that relates to governance and ideas about the polity, and on the other hand, actually curtailing certain speech that is incendiary, chilling, or otherwise disruptive in order to create good conditions for political participation. “[T]he final end of the State [is] to make men free to develop their faculties,” and civic republican theory’s view is that this occurs through political deliberation and the exercise of governing.
For instance, the fact that an interactive space attributed to the President is designated as a forum might have different civic republican consequences than if the space were attributed to state or local officials with smaller constituencies, where it is more important that every additional voice be heard. In the context of this particular case, the civic benefits of bolstering the speech rights of the individuals over Twitter are strong. Protecting their participation through public forum designation “ha[s] the potential to initiate public discourse among citizens who might otherwise never interact” with each other or with the government and increases the likelihood they will engage in other political participation. Indeed, the forum concept itself originates in the idea that spaces should exist for the public to come together as participants in democracy and share and respond to ideas and questions about how to govern. Finding a public forum here allows any and all to participate in those critical conversations online.
However, there are also significant civic republican interests served by prioritizing Twitter’s rights to moderate content. If speech is policed for a lack of civility, individuals are unable to out-yell or chill each other’s speech through hateful rhetoric. A designated public forum cannot control for these distortions, but Twitter—if able to control all parts of its platform as its own speech—certainly can through enforcement of its community guidelines, potentially creating better conditions for political engagement because a small number of voices will not intimidate or dominate discussion.
The court should have acknowledged the implications for Twitter’s First Amendment rights in its holding. If its finding that the interactive space is a public forum prevents Twitter from blocking users of its platform, then we face a clash between Twitter’s free speech rights and the users’ free speech rights.