Social media, government, and coercion: A partisan case raises big questions

A Law Dork Q&A with Alex Abdo, the litigation director at the Knight First Amendment Institute at Columbia University, about Monday's upcoming SCOTUS arguments.

This is the most significant term — yet — at the U.S. Supreme Court for social media.

Yes, there are many other very important topics in front of the justices this term, but they should not erase the similarly important ways in which the justices are potentially molding the future of the internet — and democracy — with this term’s social media cases.

Specifically, it is a term about limits — limits on public officials blocking members of the public; limits on state officials regulating the companies; and, finally, limits on the federal government influencing the companies. The cases and decisions can have overlapping effects in other areas, but those are how the cases were brought to the justices.

On Monday, the U.S. Supreme Court is set to hear arguments in the third of those: Murthy v. Missouri, a case in which Republican state leaders and people who, primarily, were dissatisfied with the Biden administration’s efforts to push the major social media platforms to stop Covid and election disinformation from spreading on their platforms.

In July, U.S. District Judge Terry Doughty issued an unexpected July Fourth ruling and injunction in the case, concluding that the Biden administration had likely gone too far — “arguably involv[ing] the most massive attack against free speech in United States’ history,” his opinion posited — and purporting to limit many people throughout the administration from even talking with or contacting social media companies for any of a number of broad purposes while the case proceeded.

It was a dramatic ruling that could have had dramatic consequences. The U.S. Court of Appeals for the Fifth Circuit, however, substantially pared it back in a September decision. Under the appeals court’s ruling, much of the most expansive language in the district court injunction was tossed aside. Nonetheless, the Fifth Circuit still held that the plaintiffs had standing; that the White House, the Surgeon General, the CDC, and the FBI “likely coerced or significantly encouraged social-media platforms to moderate content,” which “render[ed] those decisions state actions“; and that a more narrow injunction was appropriate.

That led the case to the Supreme Court — which then led to some really weird procedural shenanigans at the Fifth Circuit and a slightly revised opinion — and, ultimately, the Supreme Court granted review of the case on October 20, 2023. The high court also put the injunction, as modified by the Fifth Circuit, on hold pending Supreme Court review. Notably, Justice Sam Alito, joined by Justices Clarence Thomas and Neil Gorsuch, dissented from the decision to grant the stay of the injunction during the appeal.

On Monday, Principal Deputy Solicitor General Brian Fletcher will defend the government’s efforts, and Louisiana Solicitor General Ben Aguiñaga will represent the challengers. Notably, the social media companies at the center of the dispute — the speech intermediaries — won’t have an advocate at the podium. It will be Fletcher’s third argument at the court this term (and 23rd Supreme Court argument overall), and Aguiñaga’s first argument before the justices — although he previously clerked for Alito.(Before becoming Louisiana’s solicitor general in January for newly elected Louisiana Attorney General Liz Murrill, Aguiñaga was an associate at Jones Day.)

The issues involved are complicated, and the particularly partisan lens that the case has been seen through — given the plaintiffs, the administration, and the topics, as questionable as the underlying case is¹ — make it difficult at times to focus on the stakes and consequences of the broader issues underlying the case.

To help sort through those issues, Law Dork spoke with Alex Abdo, the litigation director at the Knight First Amendment Institute at Columbia University. From spyware to social media, Abdo is involved in litigation surrounding countless aspects of how the law is responding to and dealing with the changes being brought about by and on the internet.

In the Murthy case, the Knight First Amendment Institute, with Abdo as counsel of record, submitted an amicus curiae brief “in support of neither party.” Now, sometimes, while that is the proper label for a brief for a technical reason, it still rather clearly supports one side. Here, though, Knight submitted a brief that lays out a distinct perspective for analyzing and judging the case.

In the decision being appealed, the Fifth Circuit primarily relied on the 1982 Blum v. Yaretsky case — that set forth a test for finding state action — while only nodding at the 1963 Bantam Books v. Sullivan case addressing when government action directed at a speech intermediary amounts to a what the court called an unconstitutional “system of informal censorship.” The Supreme Court briefing and questions presented continue that state action focus — a focus that the Knight First Amendment Institute brief challenges.

---

This interview has been condensed and edited for clarity.

LAW DORK: This case started for many of us — probably a little earlier for you — but for many of us on the Fourth of July, when we unexpectedly got the district court decision, which upended things rather quickly. What is what is this case about? Like, how did we get a case like this in the first place?

ALEX ABDO: There are a lot of ways to answer that question. Let me let me try a couple. On the surface, what the case is about is the efforts by the Biden administration to reduce the spread of disinformation online, mostly on the social media platforms, and mostly disinformation about COVID vaccines and the 2020 election. And the question in the case is whether any of those efforts crossed the constitutional line between what the government is allowed to do when it comes to constitutionally protected speech and what the government is not allowed to do.

At a higher level of generality, now that the cases before the Supreme Court and the Supreme Court hasn't dealt with this set of questions in 60 years, the case has the potential to be a referendum on what limits the First Amendment imposes on the ability of the government to try to persuade our speech intermediaries — like the social media platforms — to take down speech. There's a possibility that the court provides a lot of clarity in answering that question and the possibility that we'll end up with a kind of messy doctrine, and maybe somewhere in between.

LAW DORK: I wanted to talk to you both because of your broad expertise in the area and also because your brief looked a little different than then both the petitioner and respondents. I guess, can we start from there: Whether this case is about state action, like the Blum case focused on in the briefs, or something else.

ABDO: Maybe the clearest place to start would be the Bantam Books decision in 1963. So in 1963, the Supreme Court decided an enormously important case that basically held that the First Amendment prohibits the government, not only from directly censoring protected speech, but from indirectly doing so by trying to coerce intermediaries for speech — today it’s the platforms, in 1963 was book distributors — into taking books that are constitutionally protected off of their shelves. The court said that coercion in that context was wrong.

Most courts since then have understood that case to stand for the proposition that it's OK for the government to engage in persuasion, to try to convince the public, to try to convince even speech intermediaries to agree with the government's understanding of what's socially harmful and what's socially beneficial, but that it is unconstitutional for the government to coerce private actors into censoring protected speech. And that test is a test that most courts apply in cases like this one, what people in the First Amendment community call jawboning cases — cases where the government tries to jawbone or pressure others into doing its censorship for it.

But some courts, including the Fifth Circuit, in the Murthy case, applied a very different test — the test from Blum v. Yaretsky. It's a test for state action. It wasn't a First Amendment case. Blum was a case about whether patients who had been discharged by private nursing homes could sue the government based on the theory that the private nursing homes were acting as the government — on the theory that they were state actors because they were so entangled with the government based on the regulations of nursing homes. And the Supreme Court articulated its state action test — one of many state action tests — in that case, holding that you could sue the government for the conduct of the private actor if you could show that the private actor was so coerced or substantially encouraged by the government that the choice in law must be deemed to be that of the state, which is a very high standard. And it is an understandably high standard, because the consequence of a finding of state action is that the plaintiff can sue not just the government for the action of the private actor — the plaintiff could even sue the private actor and hold the private actor to constitutional standards. And in fact, in the lower court litigation in the Blum case, the plaintiffs had sought an injunction directly against the private nursing homes.

So some courts have applied that in the context of jawboning cases.

LAW DORK: And that the that's the main focus of both DOJ and Missouri's briefs at the Supreme Court because — basically because that's what the Fifth Circuit did, that's just what they're fighting over. And I sort of got the impression that — not in a bad way, but — that your brief was basically jumping up and down being like, “Wait, why? Why is this the fight that we’re having?”

ABDO:  Yeah, as a First Amendment litigator, it was totally perplexing to see them fighting over this generic state action case, as opposed to Bantam Books, which is the only Supreme Court case to really set out the standard for exactly that circumstance.

LAW DORK: And yet, it ultimately makes sense that DOJ is going with that route and fine with it because it's a much higher standard. And it it's much easier to make it look like the Fifth Circuit screwed up and the district court screwed up on finding state action than it would be to find that they had gone too far under Bantam Books, at least as you all interpret it.

ABDO: I think Blum is a very permissive standard in terms of the government. If the state action tests from Blum were the standard in jawboning cases, then the government could do quite a lot before it actually reached the level of coercing a private actor within the meaning of the Blum test. And that is not a very speech-protective test. It's understandable for that reason that DOJ is fine with it, and it's strange for that reason that the Fifth Circuit relied on that test, given that the direction the Fifth Circuit clearly wanted to go, when it clearly thought the government had crossed the line. I think it would have been easier for the Fifth Circuit to reach that conclusion if it had relied on what I think of as the right test, the Bantam Books test.

LAW DORK: In your brief, you talk about why the Bantam Books test can work when we're not talking about booksellers, when we're talking about platforms, and we're talking about this era where so much important public speech is happening on privately owned platforms. Can you lay out what those interests are?

ABDO: Yeah, I think there are really three main interests. Obviously, the First Amendment is implicated when the government is trying to pressure private actors to suppress speech. But there are a couple of different interests that are at stake in those interactions, and we argue that the court should have these interests in mind when it tries to flesh out the Bantam test.

The first one is the obvious one, which is: People who are speaking on these platforms, and the platforms themselves, have a First Amendment right to a forum for speech that has not been unduly distorted by the government. This is background First Amendment law. We have a right to speak without the government telling us directly or coercing us indirectly into speaking consistently with the government's message.

The second interest that we pointed to is the interest of the public in being able to elect a government that is empowered to try to shape public opinion — not through coercion, but through persuasion. The government reflects the views of the majority that put it into power and is uniquely situated to help steer public attention toward the most important issues of the day. The government has an important role to play in that regard.

And then the final interest that we pointed to is really a kind of a procedural one, because the danger of jawboning is most acute when the government is acting behind closed doors, when it is trying to either informally or surreptitiously convince or coerce — depending on your perspective — speech intermediaries like the platforms into suppressing speech.

And so, whatever the rule is that the court comes up with for applying the tests from Bantam Books, it needs to be one that is sensitive to all of those interests — to preventing coercion, to enabling the government to express its views and try to steer public attention, and also one that can in fact meaningfully be enforced to prevent evasions of those core First Amendment limits on government censorship.

LAW DORK: What's the harm if they decide this on the basis of the Blum test?

ABDO: I think the harm is that the test would not be nearly protective enough of the First Amendment rights of platform users and the platforms themselves because there's quite a bit of distance between the Bantam test and the Blum test. And again, the Bantam test drew a line between coercion and persuasion, and the Blum test is a kind of super-coercion test — coercion to the point that we should treat the private actor as though it was the government. If Blum were the right test, that would mean the government could engage in a lot of coercion — just not that level of coercion — in trying to suppress protected speech, giving the government a lot of leeway to not just steer public attention, but to effectively mandate content moderation policies on the platforms. And I think that would be risky. We don't expect the government to have that kind of power over public discourse.

LAW DORK: And then you have the added implications for the platforms then being subject to litigation under government standards. There's this issue that a few of the briefs talk about, this idea that, in order to affirm the Fifth Circuit under Blum, you effectively have to water down the Blum test enough to get all of these actions to fit under it as government action. But then when you do that, you're effectively making platforms less free for speech, because you're subjecting these platforms now to potential litigation as government actors.

ABDO: Yeah, that's absolutely right. That’s exactly what the Fifth Circuit did. It diluted the test for state action under Blum because it clearly thought that the administration was engaged in coercion, and whatever your views on that, it would have been, again, easier for the court to reach that conclusion under the Bantam test.

Now, there may be jawboning cases where the plaintiff can satisfy not just the Bantam test for coercion, but also the Blum test for coercion. And in those cases, there might be a subset where it is appropriate to treat the private speech intermediary as a state actor. But, if you do so, there are added First Amendment complications. For one, if you really think the platform was coerced, it is strange to then subject the platform to liability for doing what it was coerced into doing. That might be a due process problem to hold the platform liable.

LAW DORK: The worse the government's behavior was, the more likely it is that the platform is now liable.

ABDO: That's right. It's strange. The theory under Blum, basically, is the government has a gun to your head, and you're doing the government's bidding, and now potentially, you could be subject to damages liability for responding in the way that anyone would respond if you’re actually being coerced to that extent. But the other complication is that whatever remedy the plaintiff gets in that kind of the case might interfere with the First Amendment rights of the platforms and their users. You can imagine in the Murthy case, if the plaintiffs met the higher state action test, they might be entitled to an injunction directing, say Facebook or Twitter, to reinstate their accounts, or reinstate their posts, or even change their content moderation policies. And that has implications for the First Amendment.

Alex Abdo

LAW DORK: It was rather direct, I thought, for an amicus brief to basically say, “You can't solve this problem, don't try to solve everything with this case.” And you didn't just hint at that but directly talked about other ways in which these issues should be dealt with, including through Congress. Is that just like a reminder to the court? Or was it more of a nudge to Congress? Or both?

ABDO: It was a little bit of both. Here's the concern that motivated that section of our brief, it’s the last section of our brief. There's something underlying, I think, all of the recent free speech cases that have come before the Supreme Court — the ones that relate to the social media platforms. It’s the five or six cases, depending on how you count, this term. It's the two cases last term, the Gonzalez case and the Taamneh v. Twitter case. And I think what underlies all those cases, why the Supreme Court is taking them, is that there is a sense that our democracy has lost control over these very important spaces for public discourse. And they are now controlled by a handful of people in this country who can decide what can be said and what will be heard.

The sense that we got is that that's in part what is motivating the Court’s interest in taking these cases on. Last term, for example, you had, prior to the Supreme Court taking the Gonzalez case, you had Justice Thomas expressing concern that Section 230 had been interpreted too broadly and immunized too much of the platforms’ conduct. And then when they took the case, you could tell they had some buyer's remorse and actually punted on it.

What we were trying to do in the last section of our brief is say, “The concentration of power of private power of public discourse is, in fact, a huge problem.” It's a problem specifically for jawboning because the fewer speech intermediaries there are, the fewer major platforms there are, the easier it is for the government to jawbone because they can just go through a very small Rolodex if they want to actively jawbone disfavored speech off of the internet. But you shouldn't expect jawboning doctrine to solve more than that narrow jawboning problem.

The core of the problem is concentration, not any one of these more specific doctrinal areas. So we were trying to caution the court against trying to prophylactically stretch jawboning doctrine to solve this broader problem and to say there, “There are alternatives. We have not yet lost control of the digital public sphere. There are other tools available to our democracy to rein in what I think of as the underlying problem of concentration.” That's what we're trying to do. I don't know whether it was effective, but as we were trying to do.

LAW DORK: What are you going to be watching at the arguments?

ABDO: I think at the end of the day, I'm going to be watching for how the court balances the Bantam rule of preventing coercion with the expectation in a democracy that we can have a government that can talk freely, not coerce, but talk freely about what it what it thinks of as really important societal issues.

And I will be especially interested to see how the argument or how the decision may affect the ability of researchers to study the platforms. One thing that is not discussed enough in the parties’ briefs is the role of private researchers in this case. Because private researchers study the platforms, they collaborate with the government, they coordinate their efforts sometimes with the government, and the researchers generally have a constitutional right to do all of that so long as they're not essentially coerced pass-throughs for government censorship, which none of the research, to my mind, seems like. It's important that researchers be able to do that work, because we know what we know about these platforms and their effect on society because of their work. And so I'd be very worried about a constitutional rule that made it too difficult for the researchers to collaborate with the government in all the ways that we might expect them to for the public good and for understanding these digital spaces.

That's one of the things that will be front of mind for me, in part because we're into an election year and a lot of the research efforts that were the subject of this litigation have been cut back. And that makes me worried that we're heading into this election with a blindfold on, without the research —

LAW DORK: Even more of a black box.

ABDO: — to understand what role social media has, and what role, say Russian state interference has with our electoral politics, which is really important for us to know whether you want more or less regulation, or more or less effort by the platforms to regulate this stuff. It's important for us to know what the scale of the problem is.

LAW DORK: Anything else?

ABDO: I'm cautiously optimistic that the court will approach this legal question carefully. The NetChoice cases, which were argued just a couple of weeks ago, had a similar dynamic where there were these two diametrically opposed paths that are open to the court to take on a really important First Amendment question.

At the beginning of the argument, I had the sense anyway, that it felt very similar to the beginning of the Gonzalez argument last year, the court maybe had a little bit of buyer's remorse. But by the time that Paul Clement took the podium, the court seemed to be coming around to understanding the complexity of the question before it and seem to be leaning toward a narrow ruling that would recognize the First Amendment rights of the platforms and their users against the most onerous of the requirements in Texas’s and Florida’s social media laws — but that still allowed space for reasonable and well-drafted regulation that would serve the public good.

I hope I'm accurate in that assessment of what was taking place in that argument, and I hope the court approaches this argument in the same way.

1. This excellent piece from Tech Policy Press’s Justin Hendrix and Just Security’s Ryan Goodman also published on March 17 focuses in on the troubling aspects of the so-called facts underlying the case. I added a link in here so that people can go over and read that as well. This was added at 1:10 p.m.