Earlier this month, a federal judge in Louisiana issued a ruling that restricted various government agencies from communicating with social-media companies. The plaintiffs, which include the attorneys general of Missouri and Louisiana, argued that the federal government was coercing social-media companies into limiting speech on topics such as vaccine skepticism. The judge wrote, in a preliminary injunction, “If the allegations made by plaintiffs are true, the present case arguably involves the most massive attack against free speech in United States’ history. The plaintiffs are likely to succeed on the merits in establishing that the government has used its power to silence the opposition.” The injunction prevented agencies such as the Department of Health and Human Services and the F.B.I. from communicating with Facebook, Twitter, or other platforms about removing or censoring content. (The Biden Administration appealed the injunction and, on Friday, the Fifth Circuit paused it. A three-judge panel will soon decide whether it will be reinstated as the case proceeds.) Critics have expressed concern that such orders will limit the ability of the government to fight disinformation.
To better understand the issues at stake, I recently spoke by phone with Genevieve Lakier, a professor of law at the University of Chicago Law School who focusses on issues of social media and free speech. (We spoke before Friday’s pause.) During our conversation, which has been edited for length and clarity, we discussed why the ruling was such a radical departure from the way that courts generally handle these issues, how to apply concepts like free speech to government actors, and why some of the communication between the government and social-media companies was problematic.
In a very basic sense, what does this decision actually do?
Well, in practical terms, it prevents a huge swath of the executive branch of the federal government from essentially talking to social-media platforms about what they consider to be bad or harmful speech on the platforms.
There’s an injunction and then there’s an order, and both are important. The order is the justification for the injunction, but the injunction itself is what actually has effects on the world. And the injunction is incredibly broad. It says all of these defendants—and we’re talking about the President, the Surgeon General, the White House press secretary, the State Department, the F.B.I.—may not urge, encourage, pressure, or induce in any manner the companies to do something different than what they might otherwise do about harmful speech. This is incredibly broad language. It suggests, and I think is likely to be interpreted to mean, that, basically, if you’re a member of one of the agencies or if you’re named in this injunction, you just cannot speak to the platforms about harmful speech on the platform until, or unless, the injunction ends.
But one of the puzzling things about the injunction is that there are these very significant carve-outs. For example, my favorite is that the injunction says, basically, “On the other hand, you may communicate with the platforms about threats to public safety or security of the United States.” Now, of course, the defendants in the lawsuit would say, “That’s all we’ve been doing. When we talk to you, when we talk to the platforms about election misinformation or health misinformation, we are alerting them to threats to the safety and security of the United States.”
So, read one way, the injunction chills an enormous amount of speech. Read another way, it doesn’t really change anything at all. But, of course, when you get an injunction like this from a federal court, it’s better to be safe than sorry. I imagine that all of the agencies and government officials listed in the injunction are going to think, We’d better shut up.
And the reason that specific people, jobs, and agencies are listed in the injunction is because the plaintiffs say that these entities were communicating with social-media companies, correct?
Correct. And communicating in these coercive or harmful, unconstitutional ways. The presumption of the injunction is that if they’ve been doing it in the past, they’re probably going to keep doing it in the future. And let’s stop continuing violations of the First Amendment.
As someone who’s not an expert on this issue, I find the idea that you could tell the White House press secretary that he or she cannot get up at the White House podium and say that Twitter should take down COVID misinformation—
Right.
Does this injunction raise issues on two fronts: freedom of speech and separation of powers?
Technically, when the press secretary is operating as the press secretary, she’s not a First Amendment-rights holder. The First Amendment limits the government, constrains the government, but protects private people. And so when she’s a private citizen, she has all her ordinary-citizen rights. Government officials technically don’t have First Amendment rights.
That said, it’s absolutely true that, when thinking about the scope of the First Amendment, courts take very seriously the important democratic and expressive interests in government speech. And so government speakers don’t have First Amendment rights, but they have a lot of interests that courts consider. A First Amendment advocate would say that this injunction constrains and has negative effects on really important government speech interests.
More colloquially, I would just say the irony of this injunction is that in the name of freedom of speech it is chilling a hell of a lot of speech. That is how complicated these issues are. Government officials using their bully pulpit can have really powerful speech-oppressive effects. They can chill a lot of important speech. But one of the problems with the way the district court approaches the analysis is that it doesn’t seem to be taking into account the interest on the other side. Just as we think that the government can go too far, we also think it’s really important for the government to be able to speak.
And what about separation-of-powers issues? Or is that not relevant here?
I think the way that the First Amendment is interpreted in this area is an attempt to protect some separation of powers. Government actors may not have First Amendment rights, but they’re doing important business, and it’s important to give them a lot of freedom to do that business, including to do things like express opinions about what private citizens are doing or not doing. Courts generally recognize that government actors, legislators, and executive-branch officials are doing important business. The courts do not want to second-guess everything that they’re doing.
So what exactly does this order say was illegal?
The lawsuit was very ambitious. It claimed that government officials in a variety of positions violated the First Amendment by inducing or encouraging or incentivizing the platforms to take down protected speech. And by coercing or threatening them into taking down protected speech. And by collaborating with them to take down protected speech. These are the three prongs that you can use in a First Amendment case to show that the decision to take down speech that looks like it’s directly from a private actor is actually the responsibility of the government. The plaintiffs claimed all three. What’s interesting about that district-court order is that it agreed with all three. It says, Yeah, there was encouragement, there was coercion, and there was joint action or collaboration.
And what sort of examples are they providing? What would be an example of the meat of what the plaintiffs argued, and what the judge found to violate the First Amendment?
A huge range of activities—some that I find troubling and some that don’t seem to be troubling. Public statements by members of the White House or the executive branch expressing dissatisfaction with what the platforms are doing. For instance, President Biden’s famous statement that the platforms are killing people. Or the Surgeon General’s warning that there is a health crisis caused by misinformation, and his urging the platforms to do something about it. That’s one bucket.
There is another bucket in which the platforms were going to agencies like the C.D.C. to ask them for information about the COVID pandemic and the vaccine—what’s true and what’s false, or what’s good and what’s bad information—and then using that to inform their content-moderation rules.
Very different and much more troubling, I think, are these e-mails that they found in discovery between White House officials and the platforms in which the officials more or less demand that the platforms take down speech. There is one e-mail from someone in the White House who asked Twitter to remove a parody account that was linked to President Biden’s granddaughter, and said that he “cannot stress the degree to which this needs to be resolved immediately”—and within forty-five minutes, Twitter takes it down. That’s a very different thing than President Biden saying, “Hey, platforms, you’re doing a bad job with COVID misinformation.”
The second bucket seems full of the normal give-and-take you’d expect between the government and private actors in a democratic society, right?
Yeah. Threats and government coercion on private platforms seem the most troubling from a First Amendment perspective. And traditionally that is the kind of behavior that these cases have been most worried about.
This is not the first case to make claims of this kind. This is actually one of dozens of cases that have been filed in federal court over the last years alleging that the Biden Administration or members of the government had put pressure on or encouraged platforms to take down vaccine-skeptical speech and speech about election misinformation. What is unusual about this case is the way that the district court responded to these claims. Before this case, courts had, for the most part, thrown these cases out. I think this was largely because they thought that there was insufficient evidence of coercion, and coercion is what we’re mostly worried about. They have found that this kind of behavior only violates the First Amendment if there is some kind of explicit threat, such as “If you don’t do X, we will do Y,” or if the government actors have been directly involved in the decision to take down the speech.
In this case, the court rejects that and has a much broader test, where it says, basically, that government officials violate the First Amendment if they significantly encourage the platforms to act. And that may mean just putting pressure on them through rhetoric or through e-mails on multiple occasions—there’s a campaign of pressure, and that’s enough to violate the First Amendment. I cannot stress enough how significant a departure that is from the way courts have looked at the issue before.
So, in this case, you’re saying that the underlying behavior may constitute something bad that the Biden Administration did, that voters should know about it and judge them on it, but that it doesn’t rise to the level of being a First Amendment issue?
Yes. I think that this opinion goes too far. It’s insufficiently attentive to the interests on the other side. But I think the prior cases have been too stingy. They’ve been too unwilling to find a problem—they don’t want to get involved because of this concern with separation of powers.
The platforms are incredibly powerful speech regulators. We have largely handed over control of the digital public sphere to these private companies. I think there is this recognition that when the government criticizes the platforms or puts pressure on the platforms to change their policies, that’s some form of political or democratic oversight, a way to promote public welfare. And those kinds of democratic and public-welfare concerns are pretty significant. The courts have wanted to give the government a lot of room to move.
But you think that, in the past, the courts have been too willing to give the government space? How could they develop a better approach?
Yeah. So, for example, the e-mails that are identified in this complaint—I think that’s the kind of pressure that is inappropriate for government actors in a democracy to be employing against private-speech platforms. I’m not at all convinced that, if this had come up in a different court, those would have been found to be a violation of the First Amendment. But there need to be some rules of the road.
On the one hand, I was suggesting that there are important democratic interests in not having too broad a rule. But, on the other hand, I think part of what’s going on here—part of what the facts that we see in this complaint are revealing—is that, in the past, we’ve thought about this kind of government pressure on private platforms, which is sometimes called jawboning, as episodic. There’s a local sheriff or there’s an agency head who doesn’t like a particular policy, and they put pressure on the television station, or the local bookseller, to do something about it. Today, what we’re seeing is that there’s just this pervasive, increasingly bureaucratized communication between the government and the platforms. The digital public theatre has fewer gatekeepers; journalists are not playing the role of leading and determining the news that is fit to print or not fit to print. And so there’s a lot of stuff, for good or for ill, that is circulating in public. You can understand why government officials and expert agencies want to be playing a more significant role in informing, influencing, and persuading the platforms to operate one way or the other. But it does raise the possibility of abuse, and I’m worried about that.
That was a fascinating response, but you didn’t totally answer the question. How should a court step in here without going too far?
The traditional approach that courts have taken, until now, has been to say that there’s only going to be a First Amendment violation if the coercion, encouragement, or collaboration is so strong that, essentially, the platform had no choice but to act. It had no alternatives; there was no private discretion. Because then we can say, Oh, yes, it was the government actor, not the platform, that ultimately was responsible for the decision.
I think that that is too restrictive a standard. Platforms are vulnerable to pressure from the government that’s a lot less severe. They’re in the business of making money by disseminating a lot of speech. They don’t particularly care about any particular tweet or post or speech act. And their economic incentives will often mean that they want to curry favor with the government and with advertisers by being able to continue to circulate a lot of speech. If that means that they have to break some eggs, that they have to suppress particular kinds of posts or tweets, they will do that. It’s economically rational for them to do so.
The challenge for courts is to develop rules of the road for how government officials can interact with platforms. It has to be the case that some forms of communication are protected, constitutionally O.K., and even democratically good. I want expert agencies such as the C.D.C. to be able to communicate to the platforms. And I want that kind of expert information to be constitutionally unproblematic to deliver. On the other hand, I don’t think that White House officials should be writing to platforms and saying, “Hey, take this down immediately.”
I never thought about threatening companies as a free-speech issue that courts would get involved with. Let me give you an example. If you had told me four years ago that the White House press secretary had got up and said, “I have a message from President Trump. If CNN airs one more criticism of me, I am going to try and block its next merger,” I would’ve imagined that there would be a lot of outrage about that. What I could not have imagined was a judge releasing an injunction saying that people who worked for President Trump were not allowed to pass on the President’s message from the White House podium. It would be an issue for voters to decide. Or, I suppose, CNN, during the merger decision, could raise the issue and say, “See, we didn’t get fair treatment because of what President Trump said,” and courts could take that into account. But the idea of blocking the White House press secretary from saying anything seems inconceivable to me.
I’ll say two things in response. One is that there is a history of this kind of First Amendment litigation, but it’s usually about private speech. We might think that public speech has a different status because there is more political accountability. I don’t know. I find this question really tricky, because I think that the easiest cases from a First Amendment perspective, and the easiest reason for courts to get involved, is when the communication is secret, because there isn’t political accountability.
You mentioned the White House press secretary saying something in public. O.K., that’s one thing. But what about if she says it in private? We might think, Well, then the platforms are going to complain. But often regulated parties do not want to say that they have been coerced by the government into doing something against their interests, or that they were threatened. There’s often a conspiracy of silence.
In those cases, it doesn’t seem to me as if there’s democratic accountability. But, even when it is public, we’ve seen over the past year that government officials are writing letters to the platforms: public letters criticizing them, asking for information, badgering them, pestering them about their content-moderation policies. And we might think, Sure, people know that that’s happening. Maybe the government officials will face political accountability if it’s no good. But we might worry that, even then, if the behavior is sufficiently serious, if it’s repeated, it might give the officials too much power to shape the content-moderation policies of the platforms. From a First Amendment perspective, I don’t know why that’s off the table.
Now, from a practical perspective, you’re absolutely right. Courts have not wanted to get involved. But that’s really worrying. I think this desire to just let the political branches work it out has meant that, certainly with the social-media platforms, it’s been like the Wild West. There are no rules of the road. We have no idea what’s O.K. or not for someone in the White House to e-mail to a platform. One of the benefits of the order and the injunction is that it’s opening up this debate about what’s O.K. and what’s not. It might be the case that the way to establish rules of the road will not be through First Amendment-case litigation. Maybe we need Congress to step in and write the rules, or there needs to be some kind of agency self-regulation. But I think it’s all going to have to ultimately be viewed through a First Amendment lens. This order and injunction go way too far, but I think the case is at least useful in starting a debate. Because up until now we’ve been stuck in this arena where there are important free-speech values that are at stake and no one is really doing much to protect them. ♦
