The Partial Facial Challenge Option in the Netchoice Cases

Much of the oral argument yesterday in Moody v. Netchoice focused on the fact that the respondents brought a facial challenge to Florida's law forbidding (as the Eleventh Circuit described it) "certain social-media companies from 'deplatforming' political candidates under any circumstances, prioritizing or deprioritizing any post or message 'by or about' a candidate, and, more broadly, removing anything posted by a 'journalistic enterprise' based on its content." Why is that significant? Because under the Court's precedents, a facial or overbreadth challenge to a law regulating speech can succeed only if the challengers can demonstrate that the law is "substantially" overbroad, as "judged in relation to [its] plainly legitimate sweep."

Neither the plaintiffs/respondents nor the defendants/petitioners thought that was a problem. Florida Solicitor General Henry Whitaker claimed that the state law is not overly broad because it doesn't target speech at all. Characterizing social media platforms as essentially passive conduits for third-party speech (except when they deviate to engage in what Whitaker tendentiously referred to as "censorship"), he thought that the law did not target any protected speech, much less that it did so in an overly broad manner.

At the other extreme, Netchoice's lawyer (and former U.S. Solicitor General) Paul Clement claimed that the law is invalid in all its applications. Thus it would be invalid even under the more demanding standard that the Court applies (or says it applies) in overbreadth cases outside the free speech context. 

At various points, each of Justices Alito, Kagan, and Barrett indicated that they thought the Florida law would be valid as applied to Gmail, DM apps, and other user-to-user communications that social media platforms do not curate (algorithmically or otherwise). Thus, these Justices suggested, Google as Gmail owner and other covered providers are not acting like newspaper editors in providing these services and hence can be subject to the Florida law. Therefore, they suggested, the Florida law has a plainly legitimate sweep, and the overbreadth challenge should fail. Perhaps Clement's clients could successfully sue to block the Florida law as applied to something like the Facebook news feed, but, in this view, the current challenge should fail.

Clement pushed back hard but not very successfully. His best argument appeared to be that dissolving the preliminary injunction would lead to disastrous results. The Florida law allows each and every Floridian to sue to collect $100,000 for each violation of the Act. Allowing it to go into effect would thus lead to a radical transformation of the social media landscape. Because of the difficulty of restricting content to one state, the companies' fear of crippling liability would lead them to adopt a nationwide strongly laissez-faire attitude towards content that doesn't fit into the categories of Section 230 of the Communications Decency Act (which the Florida law carves out). All social media sites would become even more awful cesspools than they already are.

Accordingly, by the end of the argument, I read the Justices as looking for a way to say that the Florida law violates the First Amendment as applied to curated feeds and the like but not as applied to user-to-user communictions like Gmail and DMs. The question is how to get to that result.

Solicitor General Prelogar suggested one route. Florida had not argued in the lower courts that its law should be upheld based on the existence of a plainly legitimate sweep in applications to Gmail and DMs. Rather, it had essentially waived that argument, instead contending that the core services of Facebook, Instagraram, X, and YouTube do not involve speech by the platforms. The Court could reject that argument, leaving Florida not arguing anything. Thus, the preliminary injunction would be preserved, but the Court could make clear in its ruling that (insofar as permitted by rules of preclusion), the Florida law could be applied to Gmail and DMs, and that certainly other state laws could be. (Although the argument in the Texas case covered some of the same ground, Clement stated early in the argument that the Texas law is narrower in not applying beyond conventional social media apps like Facebook's news feed.)

That approach might work in this case, but it won't work in every case. To my mind, that's a problem. If Florida hadn't waived the relevant argument, would we really want the Court to allow the Florida law to go into effect notwithstanding its unconstitutionality as applied to its core targets and notwithstanding the national consequences?

Consider a broader way out. In Doe v. Reed, the Court treated a challenge that was halfway between facial and as-applied as de facto facial because "the plaintiffs' claim and the relief that would follow . . . reach[ed] beyond the particular circumstances of these plaintiffs." Crucially, the Court said it was important to look past labels. Perhaps even more directly on point, in Ayotte v. Planned Parenthood, the Court treated the question of facial invalidity as calling for a solution at the remedial phase: If the law was invalid in some but not all applications, the Court said, then so long as the invalid applications are severable from the valid ones (a question it punted to the lower court on remand), the remedy should be to enjoin enforcement with respect to the invalid applications only.

Putting together the principles of those two cases ought to allow the Court in the Florida case to invalidate the state law facially as applied to conventional social media platforms but not to applications like Gmail and DMs--even without relying on the waiver theory SG Prelogar advanced. Here's hoping the Court sees its way clear to deciding the real issues in play rather than writing another confusing disquisition on the proper way to evaluate a facial challenge.