No, Florida Can’t Regulate Online Speech

Republican Governor Ron DeSantis has promised that Florida will soon enact “the most ambitious reforms yet proposed” for “holding ‘Big Tech’ accountable.” The bill would force large “social media platforms”—entities that enable users to access “a computer server, including an Internet platform and/or a social media site”—to apply their content moderation standards in a “consistent manner,” to change those standards no more than once a month, and to let users turn off algorithmic promotion or post sorting. It would also block websites from moderating content posted by politicians during an election. “We’re going to take aim at those companies,” DeSantis says, “and pull back the veil and make sure these guys don’t continue to find loopholes and gray areas to live above the law.”

Although DeSantis poses as a champion of free speech, his bill would trample on private companies’ First Amendment right to exercise editorial discretion. Private actors cannot be compelled either to host certain speakers, or to privilege some forms of speech over others. This is even more true of political speech, which, contrary to DeSantis’s claims, the bill is likely to suppress. Although DeSantis frames some parts of it as a campaign finance measure, the bill does not even regulate campaign contributions. And although he frames other parts as a consumer protection measure, the bill in fact targets entities precisely because of what they choose to say (or not say). The bill is, in essence, nothing more than an attempt to impose a new Fairness Doctrine on the internet.

This content-based regulation would compel social media platforms to carry government-backed speakers and speech. This is unconstitutional, as can be seen from what happened to another bad Florida law. That one, passed in 1913, gave political candidates a right to reply to critics, free of charge, in the paper that published the criticism. In Miami Herald Publishing Co. v. Tornillo (1974), the Supreme Court unanimously struck the law down. “The choice of material to go into a newspaper,” the court wrote, “constitute[s] the exercise of editorial control and judgment.”

Only once has the Supreme Court upheld a “fairness” or “equal time” mandate on privately owned media. But that was a special case. In 1969, Red Lion Broadcasting Co. v. FCC upheld the Federal Communication Commission’s Fairness Doctrine only because broadcast frequencies are scarce, they are owned by the public, and the government licenses their use—clear “state action.”

The court has repeatedly held that digital media enjoy the same First Amendment protection as traditional media. DeSantis counters that Big Tech companies are monopolistic. The plaintiff in Miami Herald made a similar argument—and, indeed, many local markets really did have only a single newspaper. Yet the court ruled that no degree of monopoly power could diminish the First Amendment’s protection of newspapers’ editorial discretion.

Many critics of Big Tech—not only DeSantis but also politicians in other states, such as Texas, that are considering bills similar to Florida’s—have used terms like “town square” and “public forum” in arguing that the First Amendment constrains, rather than protects, the editorial discretion of large websites. But social media platforms, even big ones, do not qualify as “public fora,” in the technical legal sense, because they don’t do anything traditionally and exclusively done by the government—like running literal town squares. “[M]erely hosting speech by others is not a traditional, exclusive public function,” Justice Brett Kavanaugh wrote for the court in 2019, “and does not alone transform private entities into state actors subject to First Amendment constraints.” In February 2020, the U.S. Court of Appeals for the Ninth Circuit affirmed that YouTube is not a public forum under this definition.

No one, not even a political candidate, has a First Amendment right to force a private actor that is not a public forum (or some other form of de facto state actor) to provide a platform for speech. On the contrary, tech companies have a First Amendment right to free speech and free association—and may therefore decide whom they will and will not host.

DeSantis’s bill would let users opt out of “post prioritization” (the placement of content “ahead of, below, or in a more or less prominent position than others in a newsfeed, feed, view, or search results”) and “shadow banning” (measures, including ones “not readily apparent to a user,” that “limit or eliminate the exposure” of a user or her content “to other users of the social media platform”). Yet these are matters of editorial discretion, as Miami Herald makes clear. Likewise, Hurley v. Irish-American Gay, Lesbian, and Bisexual Group of Boston barred the city of Boston from dictating which signs or messages a private organization had to allow at a St. Patrick’s Day parade. The First Amendment rules that protect newspapers and parade organizers also protect digital media. Websites cannot be told whom to host or what to say.

The constitutional limits on compelling political speech are even stricter than those on compelling speech in general. Yet DeSantis wants to give special privileges to political candidates. He proposes to exempt them from social media websites’ content moderation policies, and he wants to impose fines on websites that deplatform candidates during an election.

In short, DeSantis wants tech companies to host certain speakers and viewpoints against their will. This is unconstitutional. Consider Washington Post v. McManus, a 2019 U.S. Court of Appeals for the Fourth Circuit opinion by J. Harvie Wilkinson III, a prominent Reagan appointee. A Maryland law required large websites to publish and retain lists identifying who bought political ads, and stating how much they paid. A group of news websites sued.

The Maryland law raised several constitutional alarms. It was content based and, therefore, presumptively unconstitutional. Further, the law singled out specific content: political speech. Because the free flow of political ideas is crucial to self-governance, McManus notes, content-based regulations of such speech are especially suspect. Finally, the law compelled speech—it “force[d] elements of civil society to speak when they otherwise would have refrained.” The court had no trouble finding the law unconstitutional as applied to the plaintiffs.

The constitutional problems are even bigger with DeSantis’s proposed bill. The Florida proposal singles out campaign-related speech, it does so precisely because the speech is political, and it compels platforms to speak.

McManus makes two especially important points. First, when the government favors certain topics, the result is likely to be less speech. “Faced with th[e] headache” of regulation, wrote Wilkinson, “there is good reason to suspect many platforms would simply conclude: Why bother?” McManus came before the court “against a backdrop where platforms are not exactly eager to host political advertising.” Twitter, Wilkinson observed, had recently banned political ads—they were just not worth the trouble. The Maryland law would have burdened such speech even further.

Likewise, here, a law that seems to promote speech, by forcing websites to give politicians certain privileges, could easily suppress speech, by convincing those websites to ban politicians altogether.

Second, McManus explains that the Maryland law created “a constitutional infirmity distinct from garden-variety campaign finance regulations.” Placing speech burdens on intermediaries (for example, newspapers that publish ads) differs from placing them on political actors themselves. Burdening intermediaries is more likely to deter speech. Unlike political actors, “third-party platforms” are likely to “view political ads no differently than any other.” “When the onus [of regulation] is placed on platforms,” McManus concludes, “we hazard giving government the ability to accomplish indirectly via market manipulation what it cannot do through direct regulation—control the available channels for political discussion.”

Notably, nothing McManus says hinges on an entity’s “neutrality”—an important point given the repeated, though dubious, conservative claims that platforms discriminate against users on the political right. Although McManus refers to “neutral third-party platforms,” this plainly means platforms independent of political actors themselves. “For sure,” McManus clarifies, “platforms are obviously attentive to what their advertisers are saying; the Boston Red Sox are unlikely to accept ads from a group extolling the virtues of the New York Yankees.” Likewise, Fox News is unlikely to accept ads from MSNBC (or Democrats) accusing Fox News of bias.

The point can be inverted: An entity remains “neutral,” in the sense used in McManus, even though it bans content that affects advertisers. The Red Sox website need not host Yankees trolls whose comments antagonize other users, diminish the site’s popularity and alienate sponsors. What sets third-party newspapers and platforms apart from political actors, in short, is not that they have preferences about what content they host, but that they host the content “predominant[ly] ... to raise revenue.”

True, McManus was brought by a group of newspapers, and the court declined to “expound upon the wide world of social media.” It’s not just the traditional media, however, that enjoys protection from content-based regulations and compelled speech. As the Supreme Court has noted repeatedly, First Amendment protections “do not vary when a new and different medium for communication appears.” McManus concurs. “Indeed,” it says, “when a private entity, let alone a newspaper, decides to host political speech, its First Amendment protections are at their apex” (emphasis added).

If it passes, the Florida bill will likely be challenged in court, and then promptly struck down under Miami Herald, Hurley and McManus. But some conservatives have turned to campaign finance law as an alternative route by which to constrain the content moderation decisions of tech platforms. Rep. Matt Gaetz raised this idea in May 2020. He filed a complaint with the Federal Election Commission (FEC), asserting that Twitter had made an “in-kind contribution to President Trump's political opponents” by attaching fact-check labels to Trump’s tweets. The GOP later filed a similar complaint against Twitter, alleging that it made a contribution by suppressing a New York Post story about Hunter Biden. The FEC has not responded, though this is unsurprising for an agency that usually deadlocks 3-3 because of its unique partisan balance or simply delays such decisions indefinitely.

Federal law regulates the giving of “anything of value”—including “in-kind contributions”—to a candidate, by “any person for the purpose of influencing any election for Federal office.” But a contribution isn’t simply what a service provides on the same terms to everyone. The FEC was once asked, for instance, whether a nonpartisan politics website contributed to candidates by letting them upload videos about their policy positions. The website gave each candidate an equal opportunity to upload content; it subjected each candidate to the same content restrictions; and it sought commercial gain, rather than electoral influence. Under these conditions, the FEC advised, the website was not contributing to candidates by distributing their videos.

Like that politics site, social media sites have uniform terms of service. Twitter, for instance, bars hate speech, harassment, election interference and doxxing—rules Trump violated routinely. Twitter’s decision to label Trump’s falsehood-ridden election tweets was, if anything, special treatment: Twitter could have justified expelling Trump long before it did so. Likewise, the New York Post’s Hunter Biden story—which accused Joe Biden’s son of arranging a meeting between Biden and a Ukrainian energy executive—violated general restrictions imposed by social media sites regarding content shared on their platforms: Twitter, for its part, objected that the article contained personal contact information and hacked material. An edge case, concededly, but a violation. Republicans are complaining, at bottom, about the platforms’ having terms of use. Even biased enforcement decisions, under those terms, are protected by the First Amendment and by campaign finance law itself. But so long as the platforms seek commercial gain, offer an equal chance to use the service, and subject users to the same restrictions—as all credible evidence suggests they do—campaign finance law never enters the picture.

Even if Gaetz’s charges of bias were true, campaign finance law would not apply here. Under federal law, a “news story, commentary, or editorial” distributed through a “broadcasting station, newspaper, magazine, or other periodical publication” is not a contribution under campaign finance law. This “media exemption,” the FEC has made clear, applies to entities “that cover or carry news stories, commentary, and editorials on the Internet,” as well as to entities that, consistent with “the advent of the Internet,” post content “on a frequent, but perhaps not fixed, schedule.”

To qualify for the media exemption, an entity need not be neutral. Recall that in McManus, the websites’ intermediary status strengthened their right to avoid speech regulation. Here, similarly, the only “neutrality” that the media exemption requires is that the media entity not be “owned or controlled by a political party, political committee, or candidate.” The FEC has opined, in fact, that an entity “would not lose its eligibility” for the exemption “merely because of a lack of objectivity in a news story, commentary, or editorial, even if the news story, commentary, or editorial expressly advocate[d] the election or defeat of a clearly identified candidate for Federal office.”

Were it otherwise, CNN might have to satisfy campaign finance requirements whenever it praised or criticized a candidate. Some Democratic FEC Commissioners have indeed pushed to impose an unconstitutional “objectivity” requirement of this sort, at least for any “debate” that’s “staged” between candidates. Gaetz adopts this Democratic talking point in his complaint. But an “objectivity” requirement cannot be squared with a media exemption that encourages “commentary.” The exemption rightly ensures that outlets—from MSNBC to talk radio to bloggers—can give their commentary any slant they like. Indeed, a publisher can promote a book by a candidate herself, and yet be covered under the exemption.

In sum, a website that publishes regular commentary, of any bent, written by itself or by others, qualifies for the media exemption. Even a blog that publishes assorted writers, the FEC has held, is subject to the exemption. Twitter is simply a form of blog—called a “micro-blog,” because of its character limits. Gaetz objects that placing fact-check labels on tweets is not a “legitimate press function.” But this reads a “legitimacy” element into the media exemption that isn’t there. Even if you’re blogging in your garage—doing none of the due diligence that Twitter does—you’re as exempt as the Wall Street Journal.

Ultimately, it’s the First Amendment that constrains campaign finance law; the media exemption is merely an incomplete codification of First Amendment principles. Neither the FEC nor the states may use campaign finance law to force websites to grant special privileges to political candidates.

What DeSantis and Gaetz are really demanding is a reboot of the FCC’s old Fairness Doctrine—an attempt to mandate some form of “neutrality.” But Buckley v. Valeo (1976)—still the Court’s fundamental statement on campaign finance regulation—rejects a Fairness Doctrine-style argument for “equalizing the relative ability of individuals and groups to influence the outcome of elections.” “Legislative restrictions on advocacy of the election or defeat of political candidates,” Buckley declares, “are wholly at odds with the guarantees of the First Amendment.”

Finally, DeSantis’s bill would also attempt to regulate platforms’ terms of use. DeSantis wants more detailed content moderation standards; greater disclosure of those standards, and of a platform’s grounds for moderating a given user; and a ban on “frequent” changes to terms of use.

These may seem, at first blush, not to implicate the First Amendment. But laws that single out entities engaged in First Amendment activities, the Supreme Court has repeatedly said, are subject to First Amendment scrutiny. The Florida bill will have just such a targeted effect on social media websites.

Moreover, the bill directly regulates speech. The Supreme Court has applied a lower level of scrutiny to the regulation of commercial speech—that which “does no more than propose a commercial transaction.” To police traditional advertising claims or websites’ terms of service, a state need not satisfy strict scrutiny, as it does when it attempts to regulate noncommercial speech. DeSantis compares his platform rules to banking privacy and disclosure requirements, which are indeed commercial speech about business practices subject to government oversight.

But consumer protection law can’t be used to regulate noncommercial speech—and that’s what a platform’s community standards are: an inherently subjective set of limits (about, for example, what counts as “harassment”). Interfering with a website’s ability to set its takedown policies on disinformation and other abhorrent speech means forcing the site to leave such speech up, or to delay in taking it down. It means forcing the website to speak differently—a clear violation of the First Amendment.

Consider YouTube’s experience. In 2018 the site decided to modify its algorithms, so that viewers of extremist content would receive recommendations for more mainstream content. Before it could proceed, the site had to make difficult judgments about how to define extremism. The separation between tolerable falsehood and dangerous misinformation is subtle and subjective. Does a video on chemtrails cross the line? A video that argues the Sandy Hook shooting never occurred?

Next, YouTube had to make quick decisions about how to respond as extremist content evolved. In early 2019, seeking to better target and demote conspiracy videos, YouTube made at least 30 incremental changes to its algorithms. And the site has had to continue making changes as new forms of misinformation pop up—the “Plandemic” video, content advocating to “Stop the Steal,” and so on.

Although the Florida bill claims to seek consumer protection, it in fact strikes at free expression. To force a website to reveal more about how it moderates content, or to change its policies less often or more slowly, or to require that it be “consistent”—if such a thing is even possible—is to curtail that website’s editorial discretion, in violation of the First Amendment.

The underlying point is simple: The government cannot force a speaker to explain how it decides what to say. The government can no more compel Twitter to explain or justify its decision-making about which content to carry than it could compel Fox News to explain why it books some guests and not others. These are forms of noncommercial speech that turn not on facts, but on opinions. The government simply cannot compel such speech.