Does Social Media Have a “Democracy Deficit”?

The United States Supreme Court has it easy. No doubt their job is difficult, but at least they know what they should be doing: interpreting and applying the Constitution and other laws. Social media content moderators have a harder task. No one wrote and ratified a Constitution, no legislature passed Community Standards. Content moderators created Community Standards in resolving cases as judges did with the common law.

But content most American moderators lack the legitimacy accorded common law judges.

For example, most Americans think content moderators are biased. If content moderators were thought to be applying rules independent of their employers’ interests and their own political commitments, the task of judging online speech would become easier. International human rights law (IHR) seems to offer such independent rules. But can IHR be legitimate for billions of social media users? Legal professors Ilya Somin and John McGinnes raise interesting questions about the legitimacy of international law.

McGinnis and Somin distinguish between "domesticated" and "raw" international law. Congress and the president have expressly made "domesticated" international law "part of our law." In contrast, "raw" international law "has not been endorsed by the domestic political process." The authors see "domesticated" law as "largely unproblematic" because it has been accepted by Congress and the president through statute or treaty. "Raw" international law is different, and McGinnis and Somin are concerned that U.S. courts might nonetheless incorporate such rules into the domestic legal order.

McGinnis and Somin focus on the process leading to international law rather than the ensuing content. They find a "democracy deficit" in the processes that create international law. The deficit in question can be measured by the difference between the relatively democratic path to domestic law (and thus to “domesticated international law”) and the undemocratic process culminating in “raw” international law. They favor a presumption in the United States against incorporating the latter into the former because of the "democracy deficit." Of course, social media standards are not domestic law. But if IHR suffers from a “democracy deficit,” it may be less likely to confer legitimacy on social media standards.

Customary international law is the conventional term for what McGinnis and Somin call "raw" international law. They argue "nation-states do not explicitly agree on many principles that are deemed customary international law. Instead, these rules are inferred from state actions by publicists—such as international law professors—and international courts." In particular, publicists also examine the "reasonableness" of a rule to establish how much state practice is required to recognize it as "law."

Elites thus create (not just interpret) customary or "raw" international law. The authors argue:

The problem with lodging discretion in publicists is that it has high agency costs. The “people who care” are self- appointed and cannot be considered the faithful agents of anyone but themselves. They certainly are not likely to be agents of their fellow citizens, because they are not accountable to them and do not represent popular views about what counts as “reasonable.”

The publicists of international law, on this account, are like a foreign administrative state imposing their will on domestic populations lacking even an "intelligible principle" to link elite decisions to laws made by elected officials. And, of course, the world lacks a legislature to oversee in theory the work of the publicists. Given all this, “raw” international suffers from a “democracy deficit.”

McGinnis and Somin's argument matters beyond "raw" international law. As noted in an earlier post, Article 20(2) of the International Covenant on Civil and Political Rights (ICCPR) requires ratifying nations to enact "hate speech" laws. The United States ratified the ICCPR but did not endorse Article 20(2). Social media platforms have rules banning "hate speech." But those rules cannot claim democratic legitimacy from Article 20(2) at least in the United States. Social media face their own “democracy deficit” since their content moderators apply “hate speech” prohibitions.

Yet this critique may be misplaced. I am most concerned about speech and content moderation. That's not just a subjective preference. The charter of Facebook's new Oversight Board begins "The purpose of the Oversight Board is to protect freedom of expression..." Accountability to majorities may not protect freedom of expression as effectively as elites who on average are more likely to support free speech. Majorities often oppose free speech in concrete cases. Some evidence suggests elites may be more protective of expression than most voters. Insofar as "raw" international law concerns protecting expression, elites might do a better job than more populist alternatives.

Yet elites may also threaten expression by interpreting IHR to favor one party or cause. Content moderation would become a partisan censor cloaked in human rights rhetoric. Assuming the need for one set of rules, this danger can be mitigated by making sure adherents of one party or cause do not dominate the institutions creating and interpreting social media standards. McGuiness and Somin think publicists represent one political mind in creating “raw” international law. Social media managers could enhance both the liberal and democratic credentials of their content moderation by seeking views rarely found among international law publicists.

The content of social media standards also matters. Here I see an opportunity to reconcile democracy and voice. Recall that McGinnis and Somin recognize that much international law has been domesticated and thus is "unproblematic." Article 19 of the ICCPR was ratified by the United States and most other nations. Other parts of the ICCPR do not enjoy such broad support.

Content moderators could simply ignore parts of the ICCPR that do not have democratic legitimacy, such as Article 20(2)'s "hate speech" requirement. International law might then confer to online standards both liberal and democratic legitimacy. Of course, that reconciliation of democracy and liberty would mean current social media "hate speech" rules could not appeal to international law for their legitimacy. They might even turn out to be invalid for the social media space.