The Limits of Law for Facebook’s Legitimacy

Facebook expects to set up its Oversight Board for content moderation by the end of the year. Soon the company will fill in the final details for the Board. One big question: who should serve on the Board? Facebook will initially select specific individuals for the Board; it’s likely that the initial choices will then select their successors. At this point, we might ask a more general question: what kind of individuals should be selected? By that I mean: what should be their background and qualifications for service?


Law professor Noah Feldman has advised Facebook about its Board, and he has some definite ideas about who should serve. Writing in 2018, Feldman outlined a system of appellate review for Facebook that looks a lot like the U.S. courts system. He foresaw a penultimate level of review (Tier 4) and something like a Supreme Court (Tier 5). He wrote, “Tier 4 judges…should be paid, independent contractors with careers of their own, generally in law and (perhaps) legal scholarship or (sometimes) other analogous disciplines.” Feldman says Tier 5 judges should

enjoy national or international public reputations within their professions. They could be of various ages and career stages, but they should bring prestige to the Facebook Supreme Court at least as much as they derive prestige from it. In some cases, they may be retired judges of constitutional or other courts or high‐​profile former government lawyers. They should reflect a broad political range and should come from a range of geographic and cultural backgrounds. (emphasis added)

Feldman assumes that most members of the Facebook Board will be lawyers. He is, after all, outlining a judicial model of content moderation review.


A judicial model staffed mostly by lawyers has some appeal. Most people recognize the importance of the rule of law and in this instance, contrast it favorably with “the rule of Mark.” The rule of law might extend from Facebook’s basic law (its Community Standards) down to its content moderation. Appeals from content decisions then go to a Board staffed by experts trained in understanding and applying law. This “legal” process might appropriate the legitimacy accorded to law and the courts in the offline world. And legal expertise might foster acceptance by “the governed.” Absent user acceptance, Facebook’s Board will likely fail.


Is process enough for legitimacy? Surely who is on the Board matters. Whom do they represent? U.S. courts can answer that question. The U.S. Supreme Court answers to “We, the People” and their text, the Constitution. True, it does not represent the will of the current electorate, nationally or in the states. To make the courts responsive to current majorities would make them less independent, a dangerous turn since courts are called upon to constrain current majorities speaking through representative institutions. But the Court can claim to represent the enduring commitments of American citizens.


The United States also has robust representative institutions. U.S. courts are independent of the legislative and executive branches of government, both of which are elected and have strong claims to represent various majorities. And those institutions (along with the courts) were created by a Constitution that grew out of a deliberative constitutional convention, ratification debates and votes, and an amendment process, all of which represent the considered will of “We, the People.”


Facebook lacks such robust representative institutions. Facebook policymakers created the Community Standards and the amendments thereto. Facebook’s users did not ratify either the basic rules or later amendments to them. No doubt the Community Standards and the amendments did take the concerns of users into account. Facebook policymakers did not simply impose their preferences on the users through the Community Standards. They sought to attract users to a multi‐​sided platform; the Community Standards were and are guesses about what users might want in the way of rules. Users who do not like the rules can refuse to consent to the rules. But then, of course, they cannot join Facebook.


The quality of consent matters here. At the proposal stage, Facebook’s Community Standards are much like the U.S. Constitution: an elite deliberated and proposed the document. But the U.S. basic law also involved deliberative state‐​by‐​state voting by the broadest franchise then on offer. Deliberating about whether to join Facebook and accept its rules is rather different. Do many people consider the obligation to obey the Community Standards when they join Facebook? Consent to the rules is not absent, but it is also not deliberative and ultimately, an ambiguous ratification of the rules.


None of this means Facebook should throw out its Community Standards, call a “constitutional convention” of users, and thereafter hold ratification votes. The more pragmatic question would be how Facebook’s rules and their application might evolve to gain legitimacy. How might Facebook users think of the rules as their own? The contrast with the U.S. Constitution indicates representation warrants Facebook’s attention. Its Board’s search for legitimacy requires balancing representation, the rule of law, and independence.


Up to this point, I have contrasted the rule of law and representation. But that contrast misleads. The rule of law itself is not just about judges and courts. Juries are drawn from the community at large. Why? A body of experts might well more accurately determine the guilt or innocence of the indicted. A jury (and grand juries) are supposed to represent the wider community both in determining guilt and innocence and in ensuring laws accord with the community’s norms and conception of justice. Such broader participation fosters wider legitimacy for the rule of law. Such participation fully accords with Facebook’s business model. Facebook users are active; they engage with others and thereby provide valuable information to the company. Given this “spirit of participation,” shouldn’t users also shape the rules of their community in some way? If not, can those rules and their application attain legitimacy?


Finally, a practical consideration favoring attention to the value of representation. Many on the left and the right will fault Facebook’s content moderation and the decisions of its Board. Facebook might reply that those decisions reflect the companies’ rules as applied by experts. In our world, Facebook’s critics would respond by condemning the “rule of elites” bent on suppressing the voice of the people. If the Board gives the correct weight to representation, Facebook can then reply: “Our process is not driven by elites. Our rules and their application represent our users.”


In sum, Professor Feldman overstates his case for lawyers on Facebook’s Board. Expertise in law, and the judicial model more generally, will increase the likelihood the Board will be accepted. But representation matters too. The risk now is not that the judicial model will be slighted in favor of pure democracy. Rather, the undoubted appeal of the rule of law may obscure the importance of representation of the users who, after all, will be governed by the Facebook Board. A subsequent post will take up ways the Board can honor the value of representation.