By Daniel Lyons
AEI recently published a thought-provoking report by Professor Richard A. Epstein addressing censorship of conservative viewpoints online. Building on initial comments offered last year in the Wall Street Journal, Epstein identifies the problem as a systemic progressive bias among dominant social media companies, coupled with steep barriers to entry that reduce competition as a potential disciplining force in the short term. The solution, he posits, is a common carriage regime that would prevent digital platforms from abusing their positions in ways that distort public debate.
I am sympathetic to Epstein’s concerns. As I’ve written elsewhere, social media is a volatile battleground, and any gatekeeper’s perception of particular content is likely to be informed at least subconsciously by one’s priors. While evidence of systemic bias remains unclear, high-profile anecdotal missteps (such as the Hunter Biden laptop story) certainly reinforce conservative grievances. But it’s not clear to me that social media platforms fit the common carriage paradigm, and even if they do, common carriage treatment likely violates the First Amendment.
Social Media and Common Carriage
Epstein argues that common carriage developed as a solution to natural monopolies. Setting aside whether Facebook, Twitter, and the like exercise this kind of market power, Christopher Yoo has shown that this historical justification for common carriage treatment is muddled at best. Market power is neither a necessary nor sufficient condition for common carriage treatment. For example, until 1992, cable providers often held exclusive franchises but were statutorily exempt from common carriage treatment, while wireless companies were saddled with common carriage obligations in a competitive market.
Surveying the regulatory history, the District of Columbia Circuit Court defined common carriers as companies that hold themselves out to serve the public indiscriminately, without making individual business decisions regarding with whom to deal and on what terms. Under this definition, common carriage is a poor fit for social media. Unlike a telephone company or the postal service, which carry communications between users without regard to the underlying message, social media companies’ terms of service explicitly reserve the right to treat customers differently by moderating individual user content to offer users a personalized, curated experience.
Common Carriage and the First Amendment
This curation function raises a more significant obstacle to common carrier treatment, however justified: the First Amendment. In Miami Herald v. Tornillo, the Court recognized that companies engaged in the publication and dissemination of speech possess a First Amendment right of editorial control that protects their judgments about what content to carry and how. Tornillo struck down a Florida right-of-reply statute that required newspapers to carry political candidates’ responses to critical editorials. Like Epstein, Florida argued that compelled access to the platform was necessary to prevent the platform’s bias from distorting public debate. But the Court found this insufficient to overcome the newspaper’s First Amendment rights.
The Court has not recognized a common carriage exception to this right of editorial control. If anything, Tornillo itself implies the opposite. The Court recognized that the newspaper had significant market power over dissemination of political speech, and barriers to entry made alternative distribution unlikely. But it rejected Florida’s argument that this economic reality justified infringing the newspaper’s rights. Epstein correctly notes that earlier cases rejected First Amendment challenges to newspapers’ judgments about employee hiring and to anticompetitive withholding of stories from competitors. But the newspaper still decided what those employees said in print and which stories it would carry.
And this makes sense, as curation is how these platforms compete for user attention. Different platforms draw lines in different places, thereby cultivating different types of communities that appeal to different groups. Facebook aggressively removes pornographic and violent content, Twitter is more permissible but puts questionable content behind warning labels, and other platforms are free-for-all cesspools. Through millions of micro-level editorial judgments each day, platforms reveal their values, views, and community standards. In this way, the First Amendment not only protects the companies’ freedom of expression but also allows for richer and more dynamic competition among platforms.
We should be wary of vesting this editorial power in the government instead. Under the state action doctrine, the First Amendment prohibits only governmental abridgment of speech. Legislatures and courts have a poor track record when taking it upon themselves to decide which private spaces are “public” enough to be saddled with government-like duties. The fairness doctrine illustrated how government-compelled access to platforms could become a tool to reward political allies and punish enemies while chilling the very speech the doctrine was supposed to protect. First Amendment doctrine recognizes private editorial control rights not as an unalloyed good but as the lesser of two evils. In the long run, private regulation of censorship is less threatening than government regulation of censorship. Common carriage is a helpful tool to discipline less competitive markets, but it becomes more complicated when applied to markets for speech.
Does the Law of Common Carriage Override the First Amendment?
Recent Developments Tend to Undermine the Status of Media Platforms as Independent Social Actors.
By Richard A. Epstein
My thanks to Daniel Lyons for writing this brief response to my report, Should Platforms Be Treated as Common Carriers? It Depends. The last two words exhibit a clear note of caution about platforms’ status, as I am genuinely uneasy whether the common carrier rules on taking all comers are either sufficient or necessary. The gist of my argument rests on what Lyons calls the “volatile” status of common carriers in today’s novel and technologically dense environment.
Normally, the prospects of new entry is an effective reason to avoid imposing take-all-comers rules—under fair, reasonable, and nondiscriminatory (FRAND) terms. But that conclusion is uneasy in a network industry, where entry has to be successful on a large scale for it to happen at all. The current players have held their dominant positions now for several years, with little sign of erosion. So my proposal sought to force them to bear these obligations until new entrants arrive, at which point the competitive model could take over.
In making this argument, I assume with ample historical evidence that the common carrier rules had imposing service obligations of persons with monopoly power as one of their major objections. Lyons is right to note the complications about whether this definition is broad enough to cover all the cases. He notes that the District of Columbia Circuit Court defined common carriers as those that hold themselves out to serve the public indiscriminately, but that definition is wrong in two key ways.
First, it makes it appear that common carrier status depends on the willingness of firms to serve the public indiscriminately. But any railroad or public utility was, under traditional law, bound to offer the service on FRAND terms. So voluntary acceptance is not necessary for that status, nor was it sufficient. Every restaurant in town could easily agree to accept all comers, but so long as there are multiple outlets, that is just a promise and not evidence of common carrier status. Thus, while common carriers can withdraw from the market only by giving notice or meeting certain regulatory standards, the self-anointed common carrier can just pull down that sign.
But if we confine ourselves to the narrower definition of common carriers, do social media platforms fall within the definition? Here the weakest test is that they just have a huge market share. But there are other related reasons that might justify regulation that cannot be ignored. Collusion among the network platforms could easily trigger an antitrust inquiry for illegal horizontal cooperation. And evidence, if commonplace, that the Biden administration pressured Twitter to ban Alex Berenson would tend to subject Twitter to common carrier obligations as an agent of the federal government.
Choosing the right kind of regulation does matter. But in this setting, I think that Lyons is incorrect to deny the strong evidence of systematic bias, for I cannot think of a single high-profile liberal commentator who has ever been forced off Twitter. At this point, therefore, it is hard to ignore the remedial question, on which I remain convinced, that elaborate reporting obligations—largely silly paperwork—do count as an unreasonable restraint on free speech. But insisting on the usual rules of nondiscrimination by content (by way of nondiscrimination in prices) strikes me as a sensible and workable accommodation if, as now seems the case, the simplistic view of some strong separation of media platforms from each other and from the state seems, regrettably, to be less likely with each passing day.
Daniel Lyons is a nonresident senior fellow at the American Enterprise Institute, where he focuses on telecommunications and internet regulation. Richard A. Epstein is the inaugural Laurence A. Tisch Professor of Law at New York University School of Law, serving as a director of the Classical Liberal Institute.
The post Social Media, Common Carriage, and the First Amendment appeared first on American Enterprise Institute – AEI.