Introduction
What may and what should courts and legislatures do about private restrictions on speech? This has long been a critically significant and contested question, both before the Court and elsewhere. This Term’s cases — Moody v. NetChoice, LLC,1 Murthy v. Missouri,2 and, to a lesser degree, Lindke v. Freed3 — all considered the issue, but left it mostly undecided.
Professors Evelyn Douek and Genevieve Lakier have written important recent works on this general question,4 and their Lochner.com?5 is yet another major contribution on their part. I’d like to elaborate on some of the broader points that they discuss, and then turn to what the cases they focus on (Moody, Murthy, and Lindke) practically mean for the free speech/private power question going forward.
My general theme is that the question has defied, and continues to defy, any simple generalization — perhaps unsurprisingly, given that private power over speech is both dangerous and valuable. Our legal system has not concluded either that all private entities should be constrained by free speech principles, or that all private entities should be free to restrain speech on their property (or by their employees or students) as they prefer. Perhaps this is a problem that is, to borrow the late Professor Fred Schauer’s words from a different First Amendment context, “too hard” for a comprehensive rule.6 At least, if a rule is indeed possible, it has yet to be discovered.
The new cases add to this complexity, by largely acting cautiously and incrementally, perhaps because of the difficulty of the private power problem and not just because of the novelty of social media technology. I hope the analysis below helps us consider how private power questions might be considered in light of those cases. Where Douek and Lakier evaluate and critique the new precedents, I generally ask a complementary question: What might judges, lawyers, legislators, and academics do in the world those cases create?7
I. Restraints on Private Power Following Moody
A. Right to “Present[] a Curated Compilation of Speech”
1. The Majority. — To begin with, as Douek and Lakier note,8 the Moody majority strongly reaffirmed private entities’ power to exclude speech from their “curated compilation[s]”9 that make up “a single speech product,”10 such as news feeds, parades, and newspapers.11 That remains true even when the private entities have a great deal of influence over the public sphere.12
And this makes sense, partly because we rely on private entities to provide us as readers some valuable services that the First Amendment disables the government from providing. For instance, the government’s power to restrict misinformation is sharply limited. But we of course count on newspapers and other publishers to avoid misinforming their readers, including by screening third-party submissions (such as op-eds) for accuracy.
Indeed, it would be hard to have effective democratic self-government or search for truth without some private entities — newspapers, scientific journals, book publishers — that help us sort the true from the false and good ideas from bad ones. The Court concluded that the same principles that protect newspaper publishers, parade organizers, and the like also protect social media platforms.13 A magazine might want to present a conservative view or a liberal view. A parade organizer might want to organize a parade that conveys a particular theme and not other messages that the organizer views as inconsistent with the theme.14 Likewise for social media platforms striving to create particular “curated speech products” for their users.15
Private entities can also help promote useful discussions by trying to shape a pleasant environment for participants, readers, and listeners. Historically, many newspapers have had editorial policies aimed at satisfying what was seen as editors’ and readers’ preference for decency and propriety.16 Likewise, the moderator of an online discussion group may want to block people or posts that are unduly vulgar, menacing, or otherwise offensive, and that risk leading potential participants to leave. Indeed, without this, useful conversation might become difficult for all but the thickest-skinned.
A social media platform might similarly try to block similar material from comments posted on users’ pages, or the items that it includes in its news feed, in order to keep those pages and news feeds valuable to its users. It is especially important for such platforms to block spam, or else their products would become unusable. But even blocking offensive ideas may help them create a speech product that more readers will want to consume.
Douek and Lakier suggest the Court spoke too categorically in foreclosing the future viability of even modest right-of-access mandates such as “relatively modest nondiscrimination obligations” that mandate some degree of equal treatment of the speech of political candidates.17 (Think the narrow and precise obligations imposed on broadcasters by the candidate equal opportunity and noncensorship rule,18 rather than the broad, vague, and discretionarily applied obligations imposed by the old fairness doctrine.19) Perhaps there should be some more latitude for narrow laws that aim to limit “the capacity of the powerful tech companies to, for example, sway an election if they desire to do so.”20 But on balance, I think, the Court was right to conclude that, as to their curated feeds, platforms have the same broad curatorial power that newspapers do.
2. Possible Departures? — But there is a complication: One of the five Justices who joined the majority in full, Justice Barrett, filed a concurrence flagging questions for the future, and suggesting that certain kinds of “curation” might not be fully protected by the First Amendment after all. These might include:
- platforms’ using algorithms that “just present[] automatically to each user whatever the algorithm thinks the user will like — e.g., content similar to posts with which the user previously engaged”;21
- platforms’ “hand[ing] the reins to an AI tool and ask[ing] it simply to remove ‘hateful’ content,” based “on large language models” that “determine what is ‘hateful’”; and
- “foreign . . . corporations” making decisions “at the direction of foreign executives.”22
Justices Alito, Thomas, and Gorsuch were even more broadly open to certain kinds of restrictions on platforms.23 While some of their analysis was squarely rejected by the majority, some of it fits Justice Barrett’s reservations: They, too, expressly noted that “when AI algorithms make a decision, ‘even the researchers and programmers creating them don’t really understand why the models they have built make the decisions they make,’” and asked, “[a]re such decisions equally expressive as the decisions made by humans?”24
Finally, Justice Jackson’s minimalist concurrence suggested she may be saving such questions (among others) for a later day: “Faced with difficult constitutional issues arising in new contexts on undeveloped records, this Court should strive to avoid deciding more than is necessary.”25 There thus may be at least five Justices who are open to some limitations even on “[a] private party’s collection of third-party content into a single speech product.”26
To be sure, the majority rejected one possible basis for such limitations: the claim that such collection loses First Amendment protection “just because a compiler includes most items and excludes just a few.”27 But, as noted above, other bases may yet be available.
B. No Categorical Right to Control Property
1. In Moody. — While Moody protected some platform rights, it rejected the view that a property owner has a categorical First Amendment right to make sure that its property is not used to convey messages of which it disapproves.28 Such an argument was made, based on Wooley v. Maynard,29 and it could have derived some support from cases such as Janus v. AFSCME, Council 31,30 which held that the government may not require people to contribute money for ideological uses that they reject.31
But the Court made clear that “ordering a party to provide a forum for someone else’s views implicates the First Amendment” “if, though only if, the regulated party is engaged in its own expressive activity, which the mandated access would alter or disrupt.”32 And the Court acknowledged that PruneYard Shopping Center v. Robins33 and Rumsfeld v. FAIR34 upheld compulsions that an entity host speech on its property, “because in those cases the compelled access did not affect the complaining party’s own expression.”35
In the Moody oral argument, Chief Justice Roberts suggested that Rumsfeld might have been limited to conditions on government funding,36 though his opinion in Rumsfeld expressly disclaimed any such limitation.37 But the Moody opinion treated Rumsfeld as good law with regard to compelled access rules38 and declined an amicus’s call to overrule PruneYard.39
The Court thus echoed what Justice Breyer wrote in a 2020 dissent:
Requiring someone to host another person’s speech is often a perfectly legitimate thing for the Government to do. See, e.g., FAIR (holding that the Government may require law schools to host speech from military recruiters); PruneYard (holding that the Government may require the owner of a private shopping mall to host speech from politically minded pamphleteers).40
And this too makes sense, if one looks by analogy at the many property owners that are indeed disabled from excluding certain third-party views. Closest to platforms, phone companies — both landline monopolies and the famously competitive cellular companies — are “common carriers,” which can’t cancel, say, the Klan’s or the Communists’ phone lines, even if they don’t want their property used by such groups. One historical reason for this common carrier rule, as Lakier has documented, was telegraph companies’ refusal to transmit certain kinds of messages, a refusal that Congress decided to forbid.41
The same is true in other contexts. PruneYard illustrated the rules in some states that require large shopping malls to allow public access.42 Some jurisdictions limit the power of places of public accommodation — which in some places include a wide variety of services — to reject or eject customers based on their political affiliation, beliefs, or speech.43 And these rules aren’t limited to situations where there is “corporate consolidation that had left control . . . in a few hands”;44 though Douek and Lakier rightly point out that such consolidation creates special risks to public debate, various rights of access apply even in the absence of such consolidation.45
Likewise, in New Jersey and Pennsylvania, private universities can’t exclude certain speakers from outdoor spaces they have generally opened up to public access.46 California generally limits private colleges’ and private high schools’ ability to restrict speech by students, including on-campus speech.47
Employers may also have to allow some employee speech on their property.48 Wacko Hurley had the right to exclude pro–gay rights floats from the parade he ran, because requiring him to include the floats would affect the content of what the Court viewed as his own speech.49 But typical California employers, for instance, wouldn’t be allowed to fire employees because of their pro–gay rights (or anti–gay rights) speech,50 likely including on-the-job speech, at least in contexts where other employee speech is allowed.51 The speech may be on the employer’s property (like the speech in PruneYard was on the mall’s property), but requiring employers to allow such speech usually wouldn’t disrupt the employer’s own speech.
And indeed, if we think governmental restrictions on speech are dangerous because they risk undermining democratic self-government, the marketplace of ideas, the search for truth, self-expression, or autonomy, we should likewise be concerned about private restrictions.52 That is especially so when the restrictions are backed by threat of real economic penalty, as with threat of firing or expulsion from college.53 Restrictions on access to the channels of communications — telegraph and telephone, email, or social media — can also substantially restrict public debate.54
Some negative theories of the First Amendment, under which “justifications for the freedom of speech focus not on any special value of free speech, but on the special dangers presented by government regulation of that right,” might not apply to private speech restrictions.55 The Moody majority opinion may take that view when it comes to restrictions on the curated feeds:
On the spectrum of dangers to free expression, there are few greater than allowing the government to change the speech of private actors in order to achieve its own conception of speech nirvana. That is why we have said in so many contexts that the government may not “restrict the speech of some elements of our society in order to enhance the relative voice of others.”56
But even under such negative theories, one might conclude that restrictions imposed by certain kinds of private entities likewise present their own “special dangers.” And one might conclude that the First Amendment allows restrictions that do not aim “to stop private actors from speaking as they wish and preferring some views over others,”57 but aim only to stop private actors from restricting others’ speech, at least when “the mandated access” wouldn’t materially “alter or disrupt” the private actor’s “own expressive activity.”58
Of course, the U.S. Constitution by its terms protects only against government restrictions. But legislatures have considerable authority to forbid private restrictions — much as legislatures have done in the equal protection context, for instance, by implementing statutory antidiscrimination rules that in some measure adapt Equal Protection Clause principles to private entities.
2. Political History. — As Douek and Lakier have pointed out, these concerns about private power have historically been associated with the political and judicial Left, though today they are more commonly heard on the political and judicial Right.59
Human nature being what it is, this likely partly reflects whose ox is being gored. When corporate power was associated with conservatives, the main concern about it stemmed from liberals and progressives. But now that many important employers and media platforms have embraced some causes of the Left, conservatives have unaccustomedly shifted away from their earlier “laissez-faire” position while “progressives, traditional skeptics of corporate power,” have unaccustomedly embraced it.60 And though those progressives may well view the power here “as a necessary means of protecting the public against lies and deception, and vindicating the rights of women and minorities online,”61 human nature suggests that people more commonly see “lies and deception” in statements that oppose their ideological worldview than in ones that fit it. (Consider, for example, the calls — largely endorsed on the Left — to suppress the story about Hunter Biden’s laptop as being supposed Russian disinformation, though the laptop ultimately proved authentic.62 Or consider the suppression of claims that COVID-19 may have stemmed from a Chinese lab leak, claims that are now generally believed to be at least plausible.63)
But perhaps part of the reason for the political shift is that both sides have learned valuable lessons over time. “[A] conservative is a liberal who has been mugged”;64 “[a] liberal is a conservative who has been arrested.”65 The experience of seeing conservative speech suppressed may have opened conservatives’ eyes to the perils that liberals had seen all along. The experience of seeing big businesses protecting values that liberals found important, and that the government was unable to protect, may have opened liberals’ eyes to the value of private ordering. And of course there have long been both liberals skeptical of governmental attempts to restrict private power over speech and conservatives open to some such restrictions.66
C. Function-by-Function Analysis and Permissible Nondiscrimination Rules
Sections I.A and I.B above also reflect the Court’s recognition that, in the words of Justice Jackson:
The question is not whether an entire category of corporations (like social media companies) or a particular entity (like Facebook) is generally engaged in expression. . . . [C]ourts must make sure they carefully parse not only what entities are regulated, but how the regulated activities actually function before deciding if the activity in question constitutes expression and therefore comes within the First Amendment’s ambit.67
The same focus was fully embraced by the Moody majority. The majority noted “that some platforms, in at least some functions, are indeed engaged in expression.”68 It noted that lower courts must ask “as to every covered platform or function, whether there is an intrusion on protected editorial discretion.”69 And it acknowledged, applying a pre-social-media precedent:
Of course, an entity engaged in expressive activity when performing one function may not be when carrying out another. That is one lesson of FAIR. The Court ruled as it did because the law schools’ recruiting services were not engaged in expression. The case could not have been resolved on that ground if the regulation had affected what happened in law school classes instead.70
1. Email and Direct Messaging. — In particular, when the Court raised the need to ask, “as to every covered . . . function,” whether the content moderation restrictions “intru[de] on protected editorial discretion,”71 it noted that:
[T]he answers might differ as between regulation of Facebook’s News Feed (considered in the courts below) and, say, its direct messaging service (not so considered). Curating a feed and transmitting direct messages, one might think, involve different levels of editorial choice, so that the one creates an expressive product and the other does not. If so, regulation of those diverse activities could well fall on different sides of the constitutional line.72
Justice Barrett’s concurrence echoed this, reasoning that courts need to consider “whether and how specific functions — like feeds versus direct messaging — are inherently expressive.”73
Other institutions have previously suggested that direct messaging functions, including email and text, should generally not be subject to broad content controls by the service providers. For instance, New York’s high court remarked, in rejecting any duty on the part of email systems to block defamatory messages, that an email system’s “role in transmitting e-mail is akin to that of a telephone company, which one neither wants nor expects to superintend the content of its subscribers’ conversations.”74 Likewise, though the FCC has held phone companies are not common carriers as to text messaging, the rationale for that decision was the need to block unwanted robotexting — and as to messages among willing customers, a concurring opinion assured readers that “[t]omorrow, like today, our text messages will go through.”75 State laws, whether the existing Texas and Florida laws or potential future laws, that seek to act on these concerns may well be allowed to do so.
Nor would the interest in restricting viewpoint discrimination as to direct messaging be the forbidden interest “in improving, or better balancing, the marketplace of ideas.”76 Rather, it would be the interest in allowing people to better communicate with each other, free from certain kinds of discrimination imposed by platforms — an interest similar to the one that has historically justified the government’s common carrier mandates imposed on telegraph and telephone communications, or antitrust access rules to important resources such as wire reports.77
2. Removal of Posts. — The Court’s decision also didn’t resolve whether state law may ban viewpoint-discriminatory removals of posts or of entire user accounts — removals that keep the posts or accounts from being visible even individually, by people who want to go to that Facebook page, X feed, WordPress blog, and the like.78 To be sure, the Court mentioned the platforms’ “removing” posts,79 as part of its description of what platforms do. But the Court’s specific holding, and the heart of its First Amendment analysis, turned on platforms’ ability to control their curated feeds, not to remove individual posts.80 The Court also expressly reserved the question whether platforms can exclude people based on their off-platform views:
In addition to barring “censor[ship]” of “expression,” the law bars “censor[ship]” of people. More specifically, it prohibits taking the designated “censor[ial]” actions against any “user” based on his “viewpoint,” regardless of whether that “viewpoint is expressed on a social media platform.” Because the Fifth Circuit did not focus on that provision, instead confining its analysis to the law’s ban on “censor[ing]” a “user’s expression” on the platform, we do the same.81
A law that leaves platforms free to exclude posts from curated feeds, but forbids them from removing users or even posts outright, might thus remain constitutional. Indeed, the rule upheld in Rumsfeld seems consistent with that: Congress was allowed to insist that universities not remove or otherwise exclude military recruiters, but presumably, universities would have remained free to choose whether to include military recruiters in any “Recommended Recruiters” list that they would choose to compile.82
To be sure, the Court’s discussion of Turner Broadcasting System, Inc. v. FCC83 may suggest broad platform authority to decide whom to exclude. The Court cited Turner as presumptively protecting cable systems’ right to “exercis[e] editorial discretion over which stations or programs to include in [their] repertoire,”84 free of government attempts to “forc[e] the operators to carry stations they would not otherwise have chosen.”85 “A private party’s collection of third-party content into a single speech product (the operators’ ‘repertoire’ of programming) is itself expressive, and intrusion into that activity must be specially justified under the First Amendment.”86
Yet this should be read, I think, in light of the Turner Court’s conclusion that platforms’ interest in “editorial control” was insufficient to invalidate the must-carry mandate, precisely because the mandate wouldn’t affect the content of the cable systems’ speech: The Court reasoned that there was no unconstitutional speech compulsion because there was “no content-based penalty” on speech or evidence “that must-carry will force cable operators to alter their own messages to respond to the broadcast programming they are required to carry”87 (though the Court also relied on cable systems’ “bottleneck . . . control” over what TV channels can come into the home).88
Likewise, in Hurley v. Irish-American Gay, Lesbian & Bisexual Group of Boston, Inc.,89 the Court distinguished parades from cable systems on the grounds that:
Unlike the programming offered on various channels by a cable network, the parade does not consist of individual, unrelated segments that happen to be transmitted together for individual selection by members of the audience.90
When a user seeks to directly access a Facebook page, WordPress blog, or X account — as opposed to seeking out a platform-curated feed — it is even more clear that the user is seeking “individual, unrelated” items “that happen to be transmitted together for individual selection by members of the audience.”91 And even if a cable system is seen as providing a “repertoire” of a few dozen channels (common in 1994), it’s not clear that the millions of pages, blogs, or accounts hosted on Facebook, WordPress, or X are those platforms’ “repertoire.”
3. Net Neutrality. — Moody also seems to bear on another important form of regulation, which goes beyond social media platforms: net neutrality rules that limit internet service providers’ (ISPs’) ability to favor or disfavor certain kinds of content traffic on their systems.92 Justice Kavanaugh, when he was a D.C. Circuit judge, argued that net neutrality rules violated the compelled speech doctrine, analogizing the rules to impermissible restrictions on the “editorial discretion” of newspapers, bookstores, newsstands, and cable operators.93 “The First Amendment,” Judge Kavanaugh argued, “affords editors and speakers the right not to speak and not to carry or favor unwanted speech of others, at least absent sufficient governmental justification for infringing on that right,” and suggested that ISPs necessarily qualify as such “editors and speakers.”94
But under Moody, compulsions to carry speech are unconstitutional “if, though only if, the regulated party is engaged in its own expressive activity, which the mandated access would alter or disrupt.”95 And Moody appears to reaffirm PruneYard and Rumsfeld,96 which make clear that sometimes entities may indeed be compelled to carry the unwanted speech of others — including when those entities (such as universities) engage in a great deal of speech in their various other functions.
To be sure, Justice Kavanaugh’s joining the Moody majority might mean he sees the Moody rule as consistent with his views about net neutrality97 — perhaps on the theory that net neutrality rules do “alter or disrupt” service providers’ “own expressive activity,” in a way that the rules in PruneYard and Rumsfeld didn’t alter or disrupt the property owners’ expressive activities. For instance, some might argue that ISPs’ favoring certain kinds of traffic (such as Netflix video streaming) over other traffic is indeed the ISPs’ “own expressive activity,” and thus that prohibiting such favoring would alter or disrupt that expressive activity. But following Moody, a court needs to be persuaded that providing faster transmission speeds to some business partners is indeed the ISP’s “own expressive activity.”98
D. No Decision About Speaker- and Content-Based Protections
The Florida law in Moody didn’t ban all removals by platforms: It banned only removal of materials posted by political candidates and journalistic organizations, as well as material about political candidates.99 The Texas law covered speakers and topics generally, but banned only viewpoint discrimination, and not viewpoint-neutral content discrimination,100 and excluded expression that “is the subject of a . . . request from an organization with the purpose of . . . protecting survivors of sexual abuse from ongoing harassment.”101 Likewise, even bans on viewpoint discrimination in direct messaging would presumably need to have some exclusions for spam, and determining what is spam may sometimes involve content judgments.
Would such rules themselves be viewed as impermissibly content-based? The Court left this question unsettled. The Court did not, for instance, resolve whether the laws should be judged under “strict or intermediate scrutiny,” because it concluded that any restrictions on the platforms’ newsfeeds — their “curated compilation[s]” — would fail even intermediate scrutiny.102
Nor do the Court’s precedents resolve the issue. To be sure, laws that restrict speech are generally subject to strict scrutiny if they are content-based, even if they are viewpoint-neutral.103 But the Court has never decided how courts should review laws that selectively protect speech from private restrictions, as many laws do.
Laws protecting employee speech, for instance, often focus just on “political” speech, or even just on speech about political campaigns.104 Federal labor law limits employers’ ability to restrict employees’ speech related to unionization or labor conditions, and limits unions’ ability to punish their members for such speech.105 State and federal whistleblower and retaliation laws also limit employers’ ability to punish employees for their complaints about certain kinds of illegal or civilly actionable conduct.106
Likewise, laws that protect speech in places of public accommodation often protect only speech about certain broad categories of political matters.107 Some state legal rules that provide some right of access to shopping malls (often inferred from state constitutions) protect only signature gathering for ballot measures, not speech on other topics.108 Some telephone company common carrier laws don’t require the regulated businesses to provide access to dial-a-porn.109
To be sure, perhaps all these protections against private power, including protections against Big Tech private power, should be subject to strict scrutiny unless they protect all speech equally with no content classifications (except perhaps for exclusion of speech that fits within a First Amendment exception). But nothing in Moody so mandates.
I’m inclined to think that these sorts of content-based but viewpoint-neutral speech protections are indeed constitutionally permissible. In a sense, they are like the sort of benefit program discussed by the Court in Iancu v. Brunetti110 and by Justice Barrett’s concurrence in Vidal v. Elster111: The government is offering something to speakers — here, protection from retaliation or exclusion — as a means of promoting a diversity of speech. While the benefit must be viewpoint-neutral, it may be limited based on content.112 So if the legal rules discussed in the preceding paragraphs aren’t rendered unconstitutional by their preference for some categories of speech, then social media platform access requirements shouldn’t be rendered unconstitutional by the viewpoint-neutral but content-based protection rules they may draw.
E. Uncertainty About Disclosure Rules
Sometimes, private power can be constrained by transparency requirements rather than outright hosting mandates. The Moody Court suggested that such requirements, as applied to social media moderation policies and decisions, must be judged under Zauderer v. Office of Disciplinary Counsel,113 a case that dealt with compelled disclosures in lawyer advertising.114 In this regard, the Court echoed the Fifth Circuit’s decision in NetChoice, LLC v. Paxton,115 which said that the “control[ling]” Zauderer precedent “established that States may require commercial enterprises to disclose ‘purely factual and uncontroversial information’ about their services.”116
Yet the quoted material in Zauderer comes from a sentence that relies on the disclosure being within “commercial advertising”117 — and the Moody disclosure rules weren’t limited to compelling disclosures in social media platform ads, or even more broadly compelling disclosures as part of commercial transactions that the platforms entered into with their users.118 They would apply even in the typical relationship between social media platforms and users, where no money changes hands.
To be sure, the relationship does involve services by a profit-seeking business (albeit one that chiefly profits through advertising rather than direct payment). And indeed, in National Institute of Family and Life Advocates v. Becerra,119 the dissenters would have applied Zauderer to compelled speech by entirely noncommercial entities120 — the clinics in that case were generally nonprofits that provided free services to their clients, for ideological reasons rather than for profitmaking.121 It thus appears that many Justices are open to applying Zauderer to at least some mandated disclosures in relationships between service providers and clients, including disclosures (1) in speech that does not “propose a commercial transaction” (the commercial speech definition used in Zauderer122), (2) in relationships in which the client does not pay, and (3) in relationships in which the speaker is not commercially motivated.
Perhaps this is sound. Perhaps there should be more room for compelled disclosure in such relationships, so long as the compulsion doesn’t change the speaker’s other speech123 (as it would have in, say, Hurley or Miami Herald Publishing Co. v. Tornillo124 or Riley v. National Federation of the Blind of North Carolina, Inc.125). But it does appear to be a shift from an earlier understanding of Zauderer, and it seems likely that many lower courts may not notice the shift — consider, for instance, the recent NetChoice, LLC v. Bonta,126 which held that Zauderer didn’t apply to certain platform reporting requirements because they “regulate[] far more than mere commercial speech”127 — or perhaps notice it but push back against it.
II. Restraints on Private Power Following Murthy
The law also sometimes constrains governments’ use of private power, especially when such government action is aimed at accomplishing goals that are forbidden to the government itself. Even when there’s nothing unconstitutional about a private entity doing something on its own, the government may be constrained in its ability to partner up with the entity to accomplish the government’s goals.
Thus, for instance, say that you rent out a house to a tenant. The lease might allow you to access the property, with sufficient notice, and state law may recognize that right.128 If you observe evidence of a crime when you are lawfully accessing the property — or even when illegally exceeding your rights to access the property — and report it to the police, the police can use that information without violating the Fourth Amendment: The information came from your private search, not a government search.129
But say the police call you up and say, “We know you have the right to inspect your tenant’s apartment; might you exercise that right, please, and in the process see if you can spot any evidence of crime? No pressure: If you say no, we won’t retaliate against you. But we hope you’ll be a good citizen and help us out.” And say you do inspect the apartment, find evidence of crime, and turn it over to the police. That evidence will be treated as the fruit of a Fourth Amendment violation, because the government prompted the private search (even if it didn’t coerce the private search).130
The same is likely so if the government urges a private person to interrogate someone, or to discriminate based on race.131 The government has immense power, but is limited by constitutional constraints. Landlords, roommates, employers, and others also have immense power in the aggregate — in part precisely because they aren’t limited by those constitutional constraints — but are limited by the fact that, as private entities, they can’t do some of the things the government can do. Courts have resisted allowing the government to combine those two sets of powers.
The question in Murthy was whether this principle also applies to government requests that private entities limit users’ speech.132 The answer, after Murthy, is that we don’t know.
Murthy said even less about the First Amendment than Moody did, because the Court decided that the challengers in Murthy lacked standing to bring their claims: They didn’t sufficiently show that the government’s requests to platforms — whether viewed as coercion or just persuasion — affected how the platforms treated these particular plaintiffs, and (more importantly for injunction purposes) how the platforms would treat them in the future.133 Former U.S Secretary of Labor Robert Reich wrote that:
[In Murthy], the US supreme court said federal agencies may pressure social media platforms to take down misinformation — a technical win for the public good (technical because the court based its ruling on the plaintiff’s lack of standing to sue).134
But the Court in Murthy of course did not say that it’s permissible for federal agencies to pressure social media platforms this way.
And, returning to standing, presumably someone was likely affected by the government’s requests: Presumably government officials put in the time and effort to make the requests in order to have some effect. To be sure, it’s theoretically possible that every time the government asked platforms to remove certain material, the platforms would have done so in any event on their own, even without such a request. But it just doesn’t seem likely. Mark Zuckerberg’s letter to Representative Jim Jordan acknowledging the “pressure[]” from the government accepted that Meta was ultimately responsible for its removal decisions (“Ultimately, it was our decision whether or not to take content down, and we own our decisions”).135 Yet the letter suggests that the pressure did indeed affect Meta’s “choices,”136 including ones that Zuckerberg wishes he hadn’t made.137
Some people might thus be able to show actual injury from federal pressure on platforms, and perhaps even the likelihood of future injury, especially if the federal government decides to make such pressure part of its normal arsenal.138 And if state governments do something similar, they could be sued under 42 U.S.C. § 1983 for actual damages, which wouldn’t require a showing of likely future injury.139
If such action is coercive, then courts could indeed find a First Amendment violation. Indeed, that’s what the Court did in this past Term’s National Rifle Ass’n of America v. Vullo140 decision: The state government official defendants there allegedly coerced banks and insurance companies to stop doing business with the NRA, rather than allegedly coercing internet platforms, but the First Amendment coercion analysis should be much the same in either situation.141
Yet if the action merely involves persuasion rather than coercion — or perhaps the government systematically working together with social media platforms — then the Court would have to confront the question that it temporarily avoided in Murthy: Is the government free to encourage platforms to restrict user speech, so long as it does so noncoercively? Or, as with the Fourth Amendment, are there limits on the government trying to accomplish, using private power, things that it isn’t allowed to accomplish by itself?
Missouri v. Biden,142 the Fifth Circuit decision reversed on standing grounds in Murthy, concluded that the government indeed went too far in encouraging private entities to restrict speech (even in the absence of coercion). For one thing, the court concluded, the interactions between the government and the platforms were systemic and not just occasional: “consistent and consequential,” “repeated[],” “persistent[],” and “relentless.”143 And, likely more importantly, the interactions involved “press[ure] for outright change to the platforms’ moderation policies,”144 which caused “a lasting influence on the platforms’ moderation decisions without the need for any further input.”145
The result was that the government “was entangled in the platforms’ decision-making processes,” with the government “becoming intimately involved in the various platforms’ day-to-day moderation decisions” in a way that produced “an extensive relationship with the platforms.”146 Because of this, “the resulting content moderation, ‘while not compelled by the state, was so significantly encouraged, both overtly and covertly’ by CDC officials that those decisions ‘must in law be deemed to be that of the state.’”147
This may have been the Fifth Circuit’s attempt to distinguish the government’s actions from the routine individual interactions between government officials and the media that Justices Kagan and Kavanaugh discussed during oral argument, referring to their own experiences in the White House.148 Perhaps those kinds of interactions wouldn’t be as “consistent,” “repeated[],” “persistent[],” and “relentless,” at least by the same government office (even if they happen often when one views the government as a whole). And, likely more importantly, such interactions wouldn’t involve changes to media policies or government “entangle[ment]” or “intimate[] involve[ment]” in the media’s decisionmaking.
To be sure, this may be an odd line to draw. When it comes to coercion, after all, even one-time coercive demands to intermediaries (for example, “cut off financial services to Backpage.com, or we’ll prosecute you”149) are unconstitutional. Likewise, even a one-time request by the police asking a private property owner to engage in a search would be state action subject to Fourth Amendment constraints.150 Yet some line might be necessary, to constrain the union of government power with private sector power.151 The Fifth Circuit’s opinion is of course no longer binding precedent, since it was reversed on standing grounds; but it remains to be seen whether its analysis will remain persuasive to future judges.
III. Restraints on Government Officials Moderating Online Content Following Lindke
Lindke, like Murthy, asked the Court to draw a line between public power, which is constrained by the First Amendment, and private power, which is not. But in Lindke, a government official was indubitably acting: The question was whether the official was acting as an official or as a private citizen.152
Here I think I disagree in some measure with Douek and Lakier’s analysis, though more as a matter of emphasis: I appreciate their critiques of the Lindke opinion, but think that on balance the opinion makes sense, because the “space of privacy for government officials when they use[] social media to publicize their work and promote their political agenda”153 is indeed worth preserving.
Before being anything else, after all, officials are people. They often act the way other people act, and are entitled to considerable latitude in doing so. Thus, for instance, even when the Establishment Clause was seen as generally restricting governmental endorsement of religion, individual officials remained free to go to church and avow their own faith, in their individual capacities — and even to include religious references in their public speeches.154 Indeed, attempts to protect the state from undue capture by the church were themselves struck down on First Amendment grounds when they interfered with officials’ ability to be both public servants and clergy members.155 Likewise, equal protection principles limit government officials’ power to discriminate based on sex when choosing government employees,156 but presumably not when choosing, say, roommates, where such discrimination may itself be constitutionally protected.157
Government officials running for reelection often give speeches that are understood as their own private expression. A candidate for initial election is free to organize an event that’s only open to those who support the campaign, or allow public attendance generally but exclude attendees who display signs or wear clothing that supports the other side.158 A candidate for reelection is free to do the same.159 An officeholder’s excluding people based on their political affiliation or political viewpoint may be forbidden when the officeholder is acting as officeholder, but permitted when acting as candidate.
And this remains true, I think, even when the campaign speech describes the officeholder’s accomplishments, or announces new plans. Officeholders acting as officeholders may announce their new plans as part of their jobs, or otherwise speak in a way that “facilitat[es] their performance of their [government] responsibilities” with “the purpose and effect of influencing the behavior of others.”160 But I don’t think this keeps them from doing so at campaign rallies, or requires them to admit adversaries to their rallies.
Nor is this a special feature of reelection campaigns (especially since many politicians are always acting with an eye toward the next election). Rather, it just reflects that government officials wear two hats, a governmental and an individual. When I was a UCLA professor, I was constrained by the First Amendment in my official decisions — I couldn’t just give lower grades based on a student’s party affiliation, I think — but it doesn’t follow that I was thus constrained when moderating an online discussion list that I had founded.161
And this is especially important given Moody’s recognition of the First Amendment value of content curation. Government officials, like private citizens, might want to curate the comments on their individual accounts, including based on viewpoint. Sometimes this might aim at promoting their political agenda, such as by blocking harsh criticism and perhaps thus helping them get reelected. Sometimes it might aim at promoting a more comfortable conversation for the great bulk of their constituents, for instance by blocking posts that they view as bigoted or just too nasty or personal.162 (Recall that governmental restrictions on “disparaging,” “scandalous,” or “immoral” speech are treated as viewpoint-based.163) Sometimes it might aim at both.
To be sure, government bodies, as noted above, might not be allowed to moderate comments based on such viewpoint-based criteria,164 since the comment threads are essentially viewed as “limited public fora,” and the government bodies are state actors.165 Likewise, government bodies that open up in-person public comment periods must be viewpoint-neutral in administering them.166 But some valuable curated speech products are lost as a result.
That loss might be inevitable given First Amendment doctrine as applied to government bodies. But it need not occur as to individual government officials’ social media pages. And ultimately I doubt that providing this sort of protection to public officials’ curation of the comments on their social media feeds — using the same tools that ordinary citizens use for their own feeds — will on balance “significantly undermine the public’s right to speak and to learn.”167 However much public debate may be burdened by speech restrictions imposed by behemoth platforms, the burden seems much smaller when created by individual officeholders controlling the comments posted to the accounts that they run. And even if elected officeholders’ speech interests are somewhat less than those of social media platforms (which is not clear168), those interests should suffice to justify protecting the officeholders’ editorial discretion as to the comments, even if that restricts in some measure the commenters’ ability to speak to fellow readers of that one social media page.
To be sure, this might just reflect that there was less at stake in Lindke than in the other cases. Individual officeholders’ power over comments on their own accounts is much less significant than social media platforms’ or government officials’ power over platform speech generally. Not much is either lost or gained to the marketplace of ideas when officials use that power (whether one views such use as valuable curation, harmful restraint on citizen feedback, or both). Still, I agree with Douek and Lakier that the public/private power question at the heart of Lindke is conceptually interesting, and Lindke may become an important precedent for future cases dealing with when officials are acting as government actors.
Conclusion
Douek and Lakier are right to recognize the central importance of private power as well as government power in free speech debates. Private entities are tremendously valuable — indeed, indispensable — in helping us as citizens sort truth from falsehood, and good ideas from bad. They need broad authority to do so, notwithstanding government attempts to constrain the entities’ editorial discretion. Yet they can unduly influence public debates, and unduly leverage their economic power into political power.169 This is why American law has long recognized that some constraint on private entities’ power over speech is constitutionally permissible.
And private entities offer tempting targets for governments to coopt — the larger and more influential, the more tempting. There need to be some limits on the government’s power to use private entities to restrict speech that the government may not restrict itself. That is especially well-settled when the government is coercing the private entities (something the government may be able to do even as to the richest and most powerful companies). But there may also be room for First Amendment doctrine that constrains some noncoercive interactions between the government and private entities as well.
* Thomas M. Siebel Senior Fellow, Hoover Institution, Stanford University; Gary T. Schwartz Distinguished Professor of Law Emeritus and Distinguished Research Professor, UCLA School of Law.