Vol. 33, No. 06 (August 2023) pp. 69-74.

SOCIAL MEDIA, FREEDOM OF SPEECH AND THE FUTURE OF OUR DEMOCRACY, Lee Bollinger and Geoffrey Stone (eds.). Oxford: Oxford University Press, 2022. pp. 404. Cloth $99.99. ISBN-13: 978-0197621080. Paper $24.95. ISBN-13: ‎978-0197621097.

Reviewed by Cary Federman, Department of Justice Studies. Email:

This is the fifth collaboration by Lee Bollinger, the president of Columbia University and a law professor, and Geoffrey Stone, a professor of law at the University of Chicago Law School. Including this volume, they have edited three books on speech, one on the press and national security, and co-authored a book on affirmative action. This one is the timeliest, as it concerns the freedom of speech, the internet, and democracy.

The purpose of the collection is to explore the main issue confronting the freedom of speech today: “what to do about ‘bad’ speech on the internet?” (p. xv). According to the various contributors, “bad speech” includes: “abusive speech,” “extreme speech,” “genocidal speech,” “harmful speech,” “hate speech,” and “hateful speech.”

The book is divided into four parts. Prior to Part One, Bollinger and Stone write an “Opening Statement.” Four lawyers from Debevoise and Plimpton follow with an essay on the regulation of harmful speech. Part One provides an overview of the problem of the internet and speech. Part Two addresses Section 230 of the Communications Decency Act, the cornerstone of the problem of the internet, speech, and democracy. Part Three looks at the problem of algorithms and content moderation. Lastly, Part Four addresses possible reforms. Prior to the editors’ conclusion is an essay by “the Commission,” a group convened by the editors and tasked with making “a set of specific recommendations about what we should and should not do” (p. xviii) regarding the internet, speech, and democracy. (The Commission members are: Katherine Adams, of Apple; Jelani Cobb, of the New Yorker, Martin Baron, formerly the Executive Editor of the Washington Post; Russ Feingold, the former Democratic Senator from Wisconsin and the current President of the American Constitution Society; Lee Bollinger; Christina Paxson, the president of Brown University; Hillary Clinton; and Geoffrey Stone.)

Most of the contributors make the case for internet speech regulation. But what’s missing from this collection is a clear explanation of how and why internet law developed toward libertarianism, and what effect that has had on democracy. It is obvious that there is a new orthodoxy emerging regarding speech and democracy. The old orthodoxy comes from Oliver Wendell Holmes’s dissent in Abrams v. U.S. (1919). Holmes claimed that the marketplace of ideas must allow for all manner of speech, on the idea that the truth will win out eventually. In that spirit, the century that followed legalized pornography, obscenity, and seditious speech. It effectively decriminalized libel of public officials and figures. That century’s jurisprudence was indifferent (if not, at times, hostile) to the claims of local communities that did not want pornography or obscenity in their neighborhoods or schools, or to have a communist for a High School teacher. A half century into the free speech century, the Warren Court told us that a latitudinarian understanding of speech was beneficial for democracy, even if it prevented the deliberate sense of the community from restricting speech to preserve local democracy. Now, however, the contributors to this collection tell us that the “most powerful communications technology magnifies…harms exponentially beyond anything we have encountered before” (p. xv). But no essay in this collection explains how this came about, or why this seemingly new situation requires accommodating a hate speech exception into First Amendment law.

Larry Kramer provides a basic overview on the need to rethink speech, as do the editors, but Kramer, like several contributors, looks to the European Union (EU) and international human rights law to support greater regulation of speech. The EU, for example, requires that “video-sharing platforms” “prohibit users from uploading hate speech,” anything connected to terrorism, and “child sex abuse, or other content harmful for minors” (p. xxxix). There is, also, a non-binding code of conduct among the European Commission and various social media platforms that polices “hate speech.” The First Amendment, of course, has no hate speech exception.

But this seems not to deter the contributors. As Jamal Greene writes in a perceptive essay on speech and human rights law, “[a]n unregulated public platform [in the US] might even violate various human rights norms by dint of its lack of regulation” (p. 169). Greene is willing to restrict hate speech, rape jokes, and ethnic slurs on a hypothetical government-run platform because it could come into conflict with international human rights law, even though the Supreme Court frowns upon viewpoint discrimination from government entities. Rejecting “viewpoint discrimination” as an “unhelpful” rule for content moderation, Green writes that “[f]ronting a public platform loaded with ethnic slurs, rape jokes, and swastikas would seem to sit in serious tension” (p. 174) with a responsibility to international human rights law.

The resort to EU and human rights law to regulate the internet is interesting, but not compelling. Surely the Warren Court, which liberalized speech in the name of furthering democracy, bears some responsibility for the libertarian attitudes the contributors are criticizing. Yet the Warren Court makes no appearance in these pages. Also missing is viewpoint diversity. Four Democratic Senators, two past and two present, contribute to this collection, but there are no Republican Senators. There is also no confirmed libertarian among the contributors. The general tenor is toward regulation, if not censorship, or what Cass Sunstein calls “speech-protective approaches” (p. 61) to regulate misinformation and disinformation.

Newton Minow, the former chairman of the Federal Communications Commission under President Kennedy, asks: “couldn’t there be a requirement that nobody can be on the internet if they use hate speech?” (p. 290). More troubling is his daughter Nell Minow’s response. She says that, although there is no agreed upon definition of hate speech, “extremist Christians are going to say it’s hate speech for people to make pro-LGBTIA comments” (p. 290). In other words, the problem with hate speech isn’t its undefined parameters. The problem is that non-approved groups will exploit hate speech for their own reasons. Both statements require more explanation than we are given.

There is, moreover, a general sense among the contributors that capitalism and other vast structures have played an outsized role in creating our current problem, thereby denying agency to Supreme Court Justices, Congress, and perhaps the American people themselves. Throughout this volume, we are told that social media sites operate for profit, as if profit-seeking companies are an anomaly in the United States. Rather than stressing the connection between a free market and free speech, Jack Balkin writes that the free market provides “incentives to act irresponsibly and amplify false and harmful content” (p. 233). Thus, he wants to reform “informational capitalism,” and is concerned about “surveillance capitalism” – the phrase comes from Shoshana Zuboff, who argues that capitalism takes human experiences and reduces them to raw data, an idea as old as capitalism. Although not every essay blames adjectival capitalism and the “great asymmetries of power and knowledge” that exist “between the digital businesses that collect data from end users and the end users themselves” (p. 249), for the current state of democracy, the internet, and speech, a Foucauldian postmodernism, focused on panopticism, surveillance, and power and knowledge, pervades the volume.

Balkin is the only essayist to address democracy and speech. He laments the fact that “a few large institutions effectively control what people see, hear, or read,” thus undermining the “health and vibrancy of the public sphere” (p. 241). He speaks, however, not of democracy as such, but of “cultural democracy” (p. 238), which he defines as creating “broad opportunities for cultural practices.” Although Balkin doesn’t explain and describe this in detail, he argues that cultural democracy should replace democracy as it has been practiced because there is a “new kind of digital public sphere that lacks the connective tissue of institutions and practices necessary to sustain the underlying values of freedom of speech” (p. 245).

But this begs the question: where were these underlying values when the Warren Court was striking down state laws against pornography and obscenity, and banning prayer in schools? Balkin, moreover, doesn’t say why cultural democracy is superior to procedural democracy, or why we need adjectives to describe democracy and capitalism. His first solution, however, is not unwise, i.e., greater antitrust enforcement. But his second solution creates more problems than he addresses. He calls for the promotion of “public-regarding institutions governed by professional norms for curating information and producing knowledge” (p. 241).

The lawyers at Debevoise & Plimpton in fact argue for a strengthened FCC or Federal Trade Commission, or for an entirely new agency that would fulfill this role. Balkin advocates a trade-off: social media companies that want immunity from lawsuits have to agree to be regulated “as information fiduciaries.” That is, these companies have to ensure “confidentiality, care, and loyalty to the people whose data they collect and use” (p. 249). He also wants social media companies to “allow government regulators to inspect their algorithms at regular intervals for purposes of enforcing competition law, privacy, and consumer protection obligations” (p. 253).

Because social media sites are lightly regulated, they have strong incentives to collect personal data, promote hate speech, and allow for misinformation and disinformation. Consequently, “we’re drowning in lies,” Emily Bazelon states (p. 42). One could, of course, say we have always been drowning in lies. Anyone who has read the Federalist and Jeffersonian press during the Alien and Sedition Acts debate knows this. But now the stakes are higher, for some reason. “‘Misinformation costs lives,’ the World Health Organization and other groups said in September 2020 about what they saw as a crisis,” Bazelon writes. But the real crisis of speech and democracy came one year later. The events of January 6, 2021, clearly exacerbated the collective problem of speech, the internet, and democracy. But it also revealed that internet misinformation and disinformation are problems of scale; they are not, of themselves, novel. Unfortunately, no contributor considers the historical problem of “fake news.”

Section 230 of the Communications Decency Act is the heart of the problem, and there are three essays that discuss this issue. Section 230 “shields internet companies from liability for speech on their platforms” (p. xvii). Mary Anne Franks takes a hard line against Section 230, for it “curtails the rights of the vulnerable in favor of the privileges of the powerful” (p. 67). She offers a model revision of Section 230, restricting a fair amount of what passes for protected speech and diminishing social media sites’ immunity, to allow for those injured by online speech “to have their day in court” (p. 79). Erwin and Alex Chemerinsky’s essay is the most protective of free speech in the volume. They note that the federal courts have not looked favorably upon modifications of Section 230. Consequently, the Chemerinskys’ call for prudence when considering changes to something as new as internet speech. Finally, Senator Sheldon Whitehouse does not believe that Congress – the deliberative body chosen by the founders to make law – should have anything to do with internet law. He wants to repeal Section 230 and let the “courts sort it out” (p. 104). Later, he reconsiders and comes up with ways that Congress – prohibited by the First Amendment from restricting speech – can encourage social media companies to take down disinformation. In particular, he wants “Congress to require a ‘notice-and-takedown’ system removing Section 230 protections when a company willfully refuses to remove unlawful content” (p. 115). He wants greater transparency regarding algorithms, and to hold social media companies liable for their “profit-making decisions” (p. 117) that Section 230 currently immunizes as free speech.

Nathaniel Persily says that it is algorithms that “make ‘decisions’ about speakers and content quality” (p. 203). For that reason, Jack Balkin writes that social media sites should be treated differently than “broadband companies and telephone companies” because they are “curators who organize and moderate content” (p. 239). Mary Anne Franks argues that online speech is not uncensored. It is, rather, “filtered, arranged, promoted, altered, and labeled in accordance with corporate interests” (p. 66). To be sure, the algorithms are not spontaneous creatures. They are, Kate Starbird writes, “shaped by our actions” (p. 224), although she admits that “it’s hard to know how [algorithms] work.” For those perplexed by what algorithms do, RenĂ©e DiResta’s clearly informed essay on “algorithmic literacy” is recommended.

The Commission’s proposal to reform internet law is the heart of the volume’s argument. The Commission is concerned about the “real-world consequences” of the “spread of false or misleading information” (p. 315). The Commission mentions that social media sites have not, on their own, restricted hate speech or speech that promotes “violence or self-harm, defamation, harassment, and invasion of privacy” (p. 315). Although the Commission acknowledges First Amendment problems with “hate speech,” it never defines it, and yet it seems to know hate speech when it sees it (a view shared by nearly all the contributors).

Recognizing limits on regulation or censorship under First Amendment law, the Commission nevertheless recommends that courts render distinct the “use of algorithms or artificial intelligence to replicate and amplify speech…from the content of speech” (p. 316). They also recommend that Congress create an agency or give an already existing agency the power to “oversee issues relating to regulation of social media” (p. 325). And they want greater involvement by “civil society groups” (p. 317). The Commission does not fully explain this – they use the example of Major League Baseball moving the 2022 All-Star Game out of Atlanta because of Georgia’s voting rights law – but given the potential danger to the freedom of speech this creates, the Commission needed to say more about this. Finally, noting the global reach of the internet, and consistent with the theme of the volume to look to EU and human rights law, they argue that “US policymakers and regulators should work with allies and international bodies to ensure that any reforms have as much global reach as possible” (p. 318).

The essays in this volume are all well-written and clearly argued. That there is a bias in favor of regulation, in some cases bordering on censorship, from both foreign and domestic sources, probably has something to do with the slant one finds in elite law schools and other elite institutions these days. The volume, however, makes an important contribution to the literature on free speech, particularly because the next one hundred years will not mirror the previous century’s libertarianism. For this reason, this volume belongs in anyone’s library interested in the freedom of speech.


Abrams v. U.S., 250 U.S. 616 (1919)

© Copyright 2023 by author, Cary Federman.