In mid-October, Twitter took the unprecedented step of preventing its users from sharing a link to a New York Post story reporting allegations about the Biden family’s foreign financial interests. Whenever controversies about social media censorship arise, a mantra is repeated in their defence, usually by progressives: these are private companies and, as such, they can regulate speech on their platforms however they wish. Notice that this claim does not simply maintain that companies can restrict information in pursuit of specific policy aims (like thwarting criminal activity such as human trafficking efforts or combating terrorist networks). Instead, they can restrict expression on their platforms simply because those platforms are company property.
Among the most familiar responses to this argument is the view that, even though these tech giants are private entities, they function like public forums in which free expression should be upheld. Here, I spell out a different approach: that while social media companies are indeed private entities, they should nevertheless be unable to regulate non-criminal activity on their platforms for a variety of moral and legal reasons.
Do social media platforms have a moral responsibility to limit expression?
Two justifications are usually presented for enforcing speech codes on social media, both of which implicate policy aims. The first has to do with combating crime and terrorism. The second involves censoring what has become widely known as “misinformation.”
Of these two policy aims, combatting crime seems the more palatable, not least because online activity that amounts to criminal or terrorist behavior is easier to define and identify than misinformation. If an activity is criminal in the real world, combatting online behavior that facilitates it (or virtual conduct that is itself criminal, like the distribution of child pornography) is a legitimate pursuit. It remains debatable whether social media companies should have the ability to enforce criminal laws through censorship or termination of accounts or be required to act in concert with government enforcement agencies to combat criminality. Nevertheless, it is broadly uncontroversial that criminal behavior should not be exempt from enforcement simply because it’s facilitated by digital communications technologies.
More controversial is the emerging pattern of regulating legal behavior and expression, like the dissemination of “misinformation” and speech we find disputable, uncomfortable, or politically distasteful. Tech giants have assumed responsibility for curbing this type of expression, presumably to protect factual information and to cultivate an informed citizenry. But should they be endowed with this weighty responsibility? Consider the following LinkedIn post which demonstrates the complexity of the task at hand (typos in the original post have been retained):
Pompeo’s West Bank trip would be unthinkable for any other US Secretary of State. Pompeo is a joke internationally, the stress in the Middle East goes way beyond anti-semitism, one factor was the Brits bringing in 350,000 immigrants between 1920 and 1947 into the region of Palestine. After the Ottoman Empire fell, the Arabs thought the land would go back to the them, but no, a secret treaty (sikes Picot) by the Europeans divided the land between the UK, France, with ascent of Italy, and Russia as colonial powers. The UK was pushed out of the region in 1947 and the US was left with much of the mess.
At first glance, there doesn’t appear to be anything wrong with this post. Yet, many of its author’s assertions are incorrect. For instance, the British didn’t “bring” 350,000 (Jewish) immigrants into Palestine between 1920 and 1947. A British White Paper, published on May 17th, 1939, stipulated that, despite the persecution of European Jews, Britain would cap Jewish immigration into Palestine at 15,000 per year for five years, a total from which illegal immigrants would be deducted. This was done to pursue the letter of the Balfour Declaration (which sought to balance the interests of European Jewry with those of the Palestinian Arabs). The British government, in other words, attempted to curb Jewish immigration into post-Ottoman Palestine and was harshly criticized by Zionist leaders at the time for doing so.
Errors like these may not be readily identified by readers (including social media personnel) unfamiliar with the subject. So does this post amount to “misinformation”? The post was composed and uploaded to an online platform by its author. It was also written in a way that portrays its claims as indisputable and, based on the numerous “likes” it received, one can assume that it influenced or at least reinforced others’ opinions on a politically charged topic. When does the willful publication of factually disputable content become sanctionable as misinformation rather than simply being wrong?
The answer is we don’t know—and neither do the tech companies themselves. The history of Mandatory Palestine is just one of innumerable sensitive political issues prone to controversy, even among experts. Does this mean there aren’t better and worse (or even right and wrong) answers to such questions? No, but it is clear that social media companies are ill-equipped to arbitrate on questions of fact and shouldn’t impose half-baked orthodoxies onto the public by dismissing unpopular or disputable content as “misinformation.”
Even if a piece of information or an argument enjoys widespread assent, strong ideas do not need an authority to bolster them—these ideas and arguments must stand on their own and justify themselves on the basis of their own evidentiary support and logical coherence. The most reliable bulwark against an uninformed populace isn’t top-down information control undertaken by private companies, but the protection of the free exchange of all ideas and questions among curious citizens.
Do social media platforms have a legal right to limit expression?
For the sake of this discussion, let us grant that social media platforms are in fact private companies rather than “functionally public places.” Are these companies therefore able to do whatever they like with their private property? In other property contexts, such a suggestion would be absurd.
Owners of real property know that both contractual covenants and statutes can restrict freedom of use. Just because property is privately held, that doesn’t mean the owner can use it in any way he likes. Restrictive covenants and zoning laws, for example, limit the use of property in a number of ways—a regulation may prevent the owner using land for commercial purposes; a local statute may cap the height of new buildings; a covenant may prohibit the construction of more than one residential unit on a piece of land. There are also common law limitations on the use of private property when the rights of other property owners or the general public are encroached upon, like when a property owner maintains a nuisance. Finally, the free use of private property is restricted by laws governing landlord-tenant relationships—renters’ rights laws have broadened in recent decades across the US, and rental agreements prevent landlords from evicting tenants arbitrarily.
Why should limitations to the otherwise free use of private property not extend to private property owned by tech giants? Congress could, for example, enact a statute prohibiting social media companies from tampering with non-criminal speech on their platforms. The federal legislature could also conceivably impose special penalties for censoring non-criminal speech on the basis of political affiliation. Such statutes would apply, for example, to Twitter’s selective enforcement of internal speech codes on the basis of its users’ apparent political views. The New York Post article was suppressed, ostensibly on the basis of concerns about unverified and illegally released materials, while a New York Times tweet which cited a report about Trump’s illegally leaked tax returns was left untouched.
Further, the fact that users exchange valuable data for what is essentially a possessory interest in using a platform could also justify imposing statutory limits on tech companies’ internal enforcement mechanisms, perhaps informed by the landlord-tenant framework. While lease agreements are flexible in terms of covenants they can include, contractual provisions are rendered unenforceable if they contravene law. Absent criminal activity (the enforcement of which should arguably be left to government anyway), statutes akin to renter’s rights laws could ensure that social media companies wouldn’t be able to, for example, arbitrarily “lock” accounts until users comply with a company’s protean speech codes—a policy defended by Twitter CEO Jack Dorsey in a recent Senatorial hearing. At any rate, it is unlikely progressives would be as strident in their support for social media companies’ unfettered use of their property if speech codes were being enforced in the other direction.
We learn by actively conversing with others, and arguments are strengthened (and also revealed to be unpersuasive) when pressed and exposed to conflicting information. Even beyond the realm of factual dispute, adult members of a multicultural society will invariably differ in terms of their values, which inform and are informed by political, moral, and religious commitments. Tech companies are not equipped to rule on messy and complex disputes over truth. But, more importantly, they have neither a moral nor a necessary legal imperative to do so—even on their own platforms.
Join the newsletter to receive the latest updates in your inbox.