Each time I read a new article or interview with an American lawyer or legal scholar reacting to the recent decision by the Facebook Oversight Board (FOB) to invoke international human rights law in sustaining Facebook’s suspension of Donald Trump – I feel seized by the impulse to respond with an unsolicited public primer on what international human rights law (IHRL) is. It is not an unfamiliar feeling. On the contrary, the impulse (which I experience as uncomfortably paternalistic) has emerged repeatedly in the past, say, 20 years, during any one of countless exchanges with lawyers or academics who have voiced both certainty about what it is and confusion about why anyone not actually working in the field of IHRL would need to know it.
How do these conversations go? A few examples. Said one eminent international relations theorist: “You teach both constitutional law and international human rights law – aren’t those, like, opposites? I mean, the Constitution is quite clear but the other is just so ephemeral.” (Said I, to start: Actually, our Constitution leaves more questions open than you might expect.) Or this from a distinguished professor of law: “There’s no serious argument these treaties are actually binding on the United States, right?” (I tried naively to clarify: You mean, other than the argument that Article VI of the U.S. Constitution makes treaties part of the supreme law of the land?) Or chatting with a senior U.S. government intelligence community lawyer in the weeks following the 2015 decision of the Court of Justice of the European Union declaring inadequately protective of privacy rights the US-EU “Safe Harbor” agreement that had governed transfers of personal data between Europe and the United States, discussing why the government had found itself scrambling to react to the commercially cataclysmic decision: “I mean, we didn’t really know that was a court we needed to be following.”
I get it, I really do. International law is not a required course in the vast majority of American law schools, and it seems to have been long perceived as a hopelessly esoteric field, broadly irrelevant to most areas of legal practice in the United States. Compared to other fields of international law, IHRL in particular is in its relative infancy, a creature in most formal respects of the post-World War II world, with key treaties not entering into force until the 1970s. Although the United States played a leading role after the war in drafting the principles at the root of modern human rights law today), it wasn’t until the 1990s that the United States ratified the human rights treaties most discussed in contemporary debates – ratifying the International Covenant on Civil and Political Rights (discussed in the FOB decision) in 1992, and the Convention Against Torture (much at issue in debates surrounding post-9/11 U.S. interrogation practices) in 1994. The modern form of the European Court of Human Rights (ECtHR), which is today perhaps the world’s most influential specialty court of human rights with jurisdiction over Europe’s 47 member states comprising nearly a billion people, did not crystallize until it was restructured in 1998.
A lot has changed in a relatively short period of time – both in the development of formal rules of IHRL, and in the development of institutions capable of helping to “liquidate” (as James Madison once put it) those rules’ meaning. And while hurdles abound to litigating IHRL issues in U.S. courts (though IHRL-related litigation happens all the same, even here), the FOB decision has helped bring into sharper focus for many what international law scholars and lawyers have long understood: whatever the United States’ position, IHRL has today come to play an influential role in the legal landscape and actual behavior of most constitutional democracies on the planet. And because we Americans communicate, travel, trade, catch criminals, share intelligence, and cooperate militarily with these countries every day (to name just a few of our entanglements), it is simply unrealistic for U.S. lawyers to assume they can continue to ignore it safe in the belief that whatever “IHRL” is, it doesn’t really matter here.
Which brings us back to the current debate surrounding the FOB’s reliance on IHRL in its important decision about the former president’s account. How best to balance interests of open expression, public safety, and sustainable democracy on private social media platforms is a question of enormous complexity, one I will not pretend to answer here. Neither should anyone imagine that IHRL, or U.S. Constitutional Law, or any other off-the-shelf body of rules regulating human expression has remotely settled answers to all of the content regulation questions posed by today’s historically novel communications ecosystem.
But in voicing arguments in recent months about why the FOB’s repeated reliance on IHRL in its growing body of cases is problematic, critics I fear have at times merged concepts of U.S. and international law, conflated substantive law with the various courts and other institutions that interpret it, and otherwise struggled to discuss what remain, for most American lawyers, unfamiliar standards and institutions. It thus seems useful to clarify some basic details about how and why IHRL matters to the FOB; and what the Board’s reliance on IHRL principles means about the regulation of content on Facebook going forward.
Why was the FOB even talking about IHRL rules in evaluating Facebook’s decision to bar Donald Trump from posting content on the platform? Doesn’t Facebook already have its own content moderation rules?
It certainly does. Facebook has its own, detailed body of rules called Community Standards, specifying what is and is not permitted on the platform. It also has its own set of interpretive “values,” broad principles it uses to help guide the way it applies its rules. Accordingly, the very first thing the FOB opinion did was to evaluate whether Facebook’s decision to block Trump complied with Facebook’s own Community Standards and values (in this case, most relevant was the standard prohibiting “content that praises, supports, or represents events that Facebook designates as terrorist attacks, hate events, … and [other] violating events”).
Why did the FOB then, separately, analyze whether Facebook’s decision was consistent with “human rights standards”? It took this additional step because the FOB’s Charter and Bylaws instruct it to do so. Article 2:2 of the FOB Charter provides: “When reviewing decisions, the board will pay particular attention to the impact of removing content in light of human rights norms protecting free expression.” Facebook reinforced this focus on IHRL when it announced, as a matter of corporate policy issued after the Jan. 6 insurrection, that it would “strive to respect” human rights “as defined in international law.” The analysis of IHRL was no surprise to close observers of the FOB. This three-part framework of applying community standards, its overarching “values,” and IHRL is common to all the decisions issued by the FOB.
If that’s all the FOB Charter says about human rights, how does the FOB know what IHRL “norms” to apply?
Just as “U.S. law” is contained in a variety of sources (statutes, regulations, judicial opinions, and more, depending on the field), so, too, is international law. First (and most relevant here) are a set of multilateral treaties – written agreements, between states, which have consented to be legally bound by the terms of the agreements. There are nearly two dozen major IHRL treaties in effect today (depending what one counts as “major”), including the two on which the FOB relied: the International Covenant on Civil and Political Rights (the ICCPR, to which 167 states are party, including the United States), and the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD, to which 182 states are party, including the United States).
Within some limits, states can agree to be bound by all, or just most, of the provisions of these agreements, and the United States (among others) qualified its consent to a few provisions of each of these treaties – including one provision relevant to expression that was not at issue in the FOB Trump decision. The treaty provision most central to the FOB Trump decision is Article 19 of the ICCPR, which states, in relevant part:
Everyone shall have the right to freedom of expression; this right shall include freedom to seek, receive and impart information and ideas of all kinds, regardless of frontiers, either orally, in writing or in print, in the form of art, or through any other media of his choice. The exercise of the[se] rights…. [may] be subject to certain restrictions, but these shall only be such as are provided by law and are necessary: (a) For respect of the rights or reputations of others; (b) For the protection of national security or of public order…or of public health or morals.
Ok, maybe that’s a bit more detailed than our own First Amendment, but it sure leaves a lot of questions open. In domestic law, courts function to fill in gaps and settle indeterminacies in the Constitution. IHRL has no such authoritative, institutional interpreter to settle its meaning.
Let’s set aside for present purposes what exactly is meant by “authoritative” (given our own courts have no independent enforcement power), and exactly what is meant by “settle” (given our own Supreme Court decides only the tiniest fraction of the constitutional disputes litigated in the United States every year, and that major constitutional questions – like whether the President can torture people – quite regularly never get anywhere near judicial resolution in any U.S. court on the merits). For a relatively young body of law, the ICCPR has had some useful “liquidation” already.
Take, for example, the General Comments issued by an international institution called the Human Rights Committee (HRC). The HRC is an expert body created by the ICCPR itself (Arts. 28-45), its membership comprising a rotating set of distinguished jurists from all over the world. The HRC, among other things, issues periodic “comments” setting forth its views of the meaning of various provisions of the ICCPR. The key HRC Comment relied on by the FOB made clear, among other things, that in determining the circumstances in which expression rights may be restricted under ICCPR Art. 19, a state’s purpose must be legitimate, and the means it chooses to achieve that purpose must be (broadly speaking) the least restrictive available.
Is this interpretation of Art. 19 legally binding? No. It is, however, deeply informed, thoughtful, and often relied upon in the absence of more controlling authority. Such a status is entirely familiar in U.S. law. Federal courts of appeals commonly look to their sister appellate courts’ non-controlling opinions on analogous matters for guidance. And while opinions of the Justice Department Office of Legal Counsel are only formally “controlling” within the Executive Branch, they regularly feature prominently in judicial opinions undertaking to reach more broadly binding conclusions about the meaning of constitutional law. That courts and other interpretive bodies should view HRC comments this way is likewise unsurprising. After all, the states that agreed to be bound by the ICCPR also created the HRC to offer precisely this kind of guidance in applying it. Countries like the United States, which joined the ICCPR relatively late, were aware of the HRC’s track record in issuing these legal opinions before ratifying the treaty.
In any case, the HRC is not the only institutional body that interprets the ICCPR (or other parts of IHRL). Domestic and international courts of varied jurisdictions all over the world interpret aspects of the ICCPR (or analogous provisions in regionally binding treaties) when it is relevant in cases before them. (Even our own Supreme Court has interpreted provisions of the ICCPR from time to time.) In addition to each treaty party’s domestic courts, international human rights courts with limited jurisdiction over particular regions of the world may opine on treaty provisions, as may the International Court of Justice (the international court created by a separate treaty annexed to the treaty chartering the United Nations – a treaty to which every state in the world is party).
Beyond courts, there are additional non-binding but nonetheless often influential efforts by more ad hoc groups of distinguished experts to elaborate on the application of ICCPR standards. In this category, one can put the Rabat Plan of Action cited on several occasions by the FOB – a consensus report produced after a series of international meetings with legal experts convened by the UN High Commissioner for Human Rights, which asked a range of scholars to begin to sketch out, in a way that might one day achieve more binding international approval, how ICCPR freedom of expression principles should be understood to apply to hate speech and violent incitement in the digital universe. Indeed, it was just such an influential effort by an American law professor, David Kaye, while in his former role as an independent human rights expert retained by the UN, that helped lead Facebook to embrace human rights law to the degree it has.
Such reports, and voluntary statements of standards like the UN Guiding Principles on Business and Human Rights (UNGPs) to which Facebook now adheres, are no more legally binding than, say, a particularly opinionated Restatement of Law prepared by the American Law Institute – attempting to synthesize and at times extend a best approximation of what current law provides. Yet such writings have the virtue of availability, and often persuasiveness, and it is thus commonly in such “softer” forms that norms destined eventually to become binding rules of law are born.
Ok, so maybe the problem with IHRL is not a total institutional inability to fill in substantive indeterminacies in application. The problem remains that, unlike in U.S. domestic law, in IHRL there’s no singular institution that can decide for all purposes which interpretation should prevail when the views of different bodies conflict.
The existence of conflicting legal interpretations by bodies of different degrees of authority and legal jurisdiction neither distinguishes IHRL from U.S. domestic law, nor does it (necessarily) matter in the present context. In the United States, lower federal and state courts disagree about constitutional meaning all the time, and because the Supreme Court hears such a tiny fraction of constitutional disputes each term it sits, the vast majority of these conflicting interpretations go unresolved in any given year. The result is that different regions of the country may be governed by different interpretations of the U.S. Constitution and federal statutes at certain times (sometimes, over long periods of time). On occasion, such differences have benefits. It allows different populations to “test out” the effect of different interpretations in practice – experience that can help lawmakers at different levels recognize when laws need to be supplemented or revised. As long as it is clear to governments and individuals which interpretive view governs their affairs (what the law is, for example, for the states subject to the jurisdiction of the U.S. Court of Appeals for the Fifth Circuit as opposed to the Second Circuit), individuals can act accordingly and the legal system functions. More complicated problems arise when, for example, more than one interpretation of the same legal rule (or two conflicting legal rules) apply to the same individual at the same time – and the field of Conflict of Laws was born to help make just such determinations.
The good news for Facebook, at least for now, is that no such complexity exists. There is currently only one body of relevant jurisdiction that has decided whether Facebook’s Trump suspension is consistent with its corporate human rights obligations as informed by ICCPR Art. 19: the FOB, with authoritative jurisdiction over Facebook in decisions on whether to retain or remove certain content.
Look, IHRL rules were designed to regulate the conduct of states. Facebook is not a state. It doesn’t make sense for IHRL to apply to Facebook’s content decisions.
A number of scholars have made this point, and I have at times struggled to understand why exactly they believe it matters. One argument critics might be making is that IHRL by its terms cannot be understood to be legally binding on private actors, much like the First Amendment by its terms has been interpreted not to apply as a matter of law to non-state action. Of course, no one is claiming that IHRL is legally binding on Facebook as a function of IHRL’s direct application; the FOB applied IHRL principles in its ruling because Facebook, through private contract law (the FOB’s Charter), told it to – rather like any private parties might decide to incorporate by reference some other existing rules or definitions into their own agreement – a state of affairs that would also seem to obviate any worries about party consent or fairness.
Perhaps, then, critics believe that ICCPR Art. 19 (for example) only makes conceptual sense when applied to a state, that its meaning is simply indiscernible when applied to a private party. This might indeed be a credible argument if the relevant IHRL provisions were written in the same terms as the First Amendment itself, which famously begins: “Congress shall make no law … abridging the freedom of speech.” In the absence of Congress, or any state actor, what or who exactly does this restrict? But IHRL is as famously not written around particular state institutions, neither are its provisions framed solely in terms of state action. As a textual matter, many IHRL rights are most comfortably read not merely to restrict governments, but to acknowledge and protect the dignity of individuals. So begins Article 19, for example: “Everyone shall have the right to freedom of expression….” Again, in a poor attempt to encapsulate entire libraries of scholarship on this point, it is at least conceptually plausible that Art. 19-type provisions might be read to apply to both non-state actors and states.
No, the issue isn’t really some categorical concern about international laws or institutions, state actors or not. The real worry is about the normative content of IHRL rules regarding free expression.
There are plenty of important arguments to be made for or against the application of a particular IHRL rule to a particular case (or issue) on the ground that it is likely to produce a normatively undesirable outcome. A few scholars have, valuably, begun to do just that, suggesting that a putative IHRL rule is (variously) either insufficiently protective of free expression or vastly overprotective in that it would effectively disable any social media company from any algorithmic regulation of content. Remarkably, however, these analyses often manage to advance normative worries about IHRL rules without actually engaging what any IHRL source (like those on which the FOB relied) says about the content of those rules. Nate Persily, for example, argues that direct application of IHRL would effectively invalidate the vast majority of Facebook’s existing content regulations, including those on “hate speech, obscenity, self-harm and disinformation.” How does he know IHRL rules would require throwing out so much of Facebook’s existing playbook? Because, he points out, these existing content regulations, regularly applied pre-emptively through algorithmic selection, would certainly violate the First Amendment’s prohibition on prior restraints.
But while IHRL and the U.S. Constitution’s First Amendment have many substantive ideas in common, they are not the same law. Indeed, thus far, different IHRL institutions interpreting different provisions of IHRL (not only ICCPR Article 19 but analogous protections of free expression under European and Inter-American treaties) have taken different approaches on the acceptability of what U.S. law calls prior restraints. Our own Supreme Court has yet to contemplate at all the application of prior restraint rules to the universe of algorithmic speech. Neither has the FOB. There are thus plenty of questions yet to be answered. But that, too, is a feature of all legal systems’ efforts to apply old law to new cases. It is a failing of legal development. Not a failing unique to IHRL. And not fatal to the project of moderated speech.