The ability of a concentrated few to shape the speech environment has waxed and waned throughout history. The printing press exploded elite control over ideas, leading to revolution, conflict, and an unprecedented sustained increase in living standards (sometimes called “the Great Enrichment”). More recently, powerful publishers and broadcasters lost their gatekeeper role in the eruption of communications facilitated by internet platforms. As a result, the average person today possesses an ability and freedom to speak to a broad audience that is unparalleled in human history.

This freedom can and has been abused, and thus platforms face political and market pressure to control the speech on their services. People across the political spectrum accuse platforms of fostering hate, censoring speech, and harming journalism. Based on largely instinctual assessments that platforms face little competitive pressure on how they govern speech, some have sought to expand antitrust enforcers’ mandate beyond competition values.

Many question whether and to what degree large platforms are guilty of the accused harms. We touch on this debate. But our primary method is to assume such harms are serious and examine the proposed solutions.

Our conclusion: If private platform power over the speech environment is a problem, antitrust is the wrong solution. Competition can promote consumers’ free speech preferences. However, increasing the number of competitors will not help because current free speech concerns are not caused by a lack of competition. In fact, unleashing antitrust regulators to pursue non-competition-related goals would threaten free speech values. Removing key constraints on antitrust’s powerful tools—tools with a history of abusive and arbitrary use—would weaken antitrust’s ability to protect the competitive process and increase the risk that governments and others will abuse such tools to interfere with speech.

Everybody Hates Mark Zuckerberg—But Can’t Agree Why

People of diverse political views worry that popular internet platforms for user-contributed content such as Facebook, Twitter, and YouTube are harming free speech and even disrupting democracy in the United States and around the world. Many, especially on the right, fear that these companies exhibit biases in choosing what content to suppress or promote. On the left, many believe that these platforms facilitate hate and normalize extremism. Both fear that these online platforms hold outsized power over public discourse because so much speech flows through them.

The Supreme Court has expanded protection of free speech against government interference over the last several decades. But despite greater free speech protections in the traditional (government-owned) public square, some fear that the increasing prevalence of social media platforms will actually threaten free expression because these private companies, unlike the government, are not subject to the constraints of the First Amendment. Indeed, social media platforms—like any private person—have First Amendment rights against government interference with their content decisions. Efforts by Prager University and others to try to impose the restraints of the First Amendment on YouTube, Twitter, and other platforms in their relationships with their users have thus far failed. But in the absence of any clear legal standards on the free speech rights of social media platform users vis-à-vis the platforms’ own content decisions, debates about online speech are nevertheless shaped by notions of viewpoint neutrality and speaker discrimination in First Amendment law.

Many pundits, politicians, and individuals generally aligned with the Republican Party believe that online platforms are suppressing conservative speech online. Republican senators Ted Cruz and Josh Hawley, for example, assert that online-platform companies prevent conservatives from expressing and spreading their views. In a petition to the Federal Trade Commission (FTC), Cruz and Hawley complained that “never before in this country have so few people controlled so much speech” and demanded that the agency should investigate tech companies’ business and moderation practices. Hawley’s fears have motivated him to legislative action: He introduced the Ending Support for Internet Censorship Act with the intent of removing liability protections from online platforms that moderate content with non-neutral standards.

Other prominent Republicans, including President Donald Trump, media host Laura Ingraham, and campus antisocialism provocateur Charlie Kirk, contend that platform company employees and policies are biased against conservative viewpoints. They point to incidents like the suspension of Senate Majority Leader Mitch McConnell’s Twitter account or leaked internal emails at Google that suggest Google employees have strong liberal—or rather, anticonservative—leanings. Pew Research Center surveys also indicate that individuals aligned with conservative ideologies believe that social media companies favor liberal ideas over conservative views.

The companies are well aware of the right-of-center concerns. After the Susan B. Anthony List, a pro-life group, complained that Facebook took down several of its videos about abortion, the company eventually apologized and restored the videos to their site. Conservative groups and individuals also criticize online platforms for relying on left-leaning organizations to help them set policies for their services. For example, Amazon relies on the Southern Poverty Law Center (SPLC), a progressive public-interest advocacy and legal organization, to prevent charities from participating in AmazonSmile, its charitable contribution program. When the SPLC labeled the Alliance Defending Freedom (ADF), a religious liberty law firm, a “hate group,” Amazon did not allow individuals to donate to ADF through its online platform. Facebook recently released a detailed report compiled on its behalf by a law firm that catalogs the bias concerns of 133 conservative organizations, individuals, and lawmakers. And as Mark Zuckerberg himself noted in a Washington Post op-ed, “Lawmakers often tell me we have too much power over speech, and I agree.”

But concerns about online speech are a rare opportunity for bipartisanship today. Left-leaning organizations and individuals also worry about platforms’ roles in speech. It’s only a bit oversimplified to summarize right-of-center concerns as “platforms take down too much of our content” and left-of-center concerns as “platforms don’t take down enough right-of-center content.” To be sure, some left-of-center groups argue that the platforms favor and protect controversial right-wing views while deleting or disfavoring controversial left-wing views. They compare takedowns of conservative content versus their own or aligned content and accuse the platforms of uneven moderation actions. Some claim that content moderation by online platforms disproportionately affects racial and ethnic minorities. For example, some note that Facebook allowed a congressman to advocate the murder of “extremist Muslims,” yet it deactivated a Black Lives Matter activist’s account for calling white people racist. Other groups, including Palestinian media outlets and activist organizations working in conflict regions, have received moderation warnings or account deactivations based on Facebook’s content policies. These examples of moderation decisions against left-of-center voices are not well known to right-of-center audiences, exacerbating the perception that all moderation decisions are against right-of-center voices.

Others appear less concerned with the specific outcomes of such decisions than with the fact that the power to make those decisions sits in private hands. While many on the political left want social media companies to regulate the types of messages shared on social media platforms, they also share conservatives’ concerns about the influence and power of large social media platforms over users’ ability to express themselves. Chris Hughes, a Facebook co-founder, protests thus: “The most problematic aspect of Facebook’s power is Mark [Zuckerberg]’s unilateral control over speech. There is no precedent for his ability to monitor, organize and even censor the conversations of two billion people.” Zephyr Teachout and Lina Khan claim that corporations can exert greater political influence due to their large size and permissive reforms in campaign finance laws. Separately, Teachout has argued that online platforms, especially social media companies, reduce the number of local newspapers and “really good news” sources and enable the spread of propaganda. Larry Kramer, a constitutional law scholar, likewise frets that social media platforms make it too easy for “the really awful information” to reach the general public.

Senator Elizabeth Warren has attacked social media companies for wielding too much influence due to their large size and has retrospectively criticized mergers between online platforms like Instagram and Facebook as harmful to competition and society. She has argued that the large size of social media companies enables them to censor speech and spread lies and misinformation. After Facebook removed several of her campaign ads last March, Warren tweeted:

Curious why I think FB has too much power? Let’s start with their ability to shut down a debate over whether FB has too much power. ... I want a social media marketplace that isn’t dominated by a single censor.

Demonstrating the bipartisan nature of these concerns, Senator Ted Cruz retweeted Warren, adding “she’s right—Big Tech has way too much power to silence Free Speech.”

Yet It Has Never Been Easier for an Individual to Speak to a Broad Audience

Ironically, the very platforms accused of stifling speech are the ones making it easier today than ever before for individuals to speak to a large, even global audience. Facebook, Twitter, Google, and others in the crosshairs of the antitrust crusaders have weakened the ability of a concentrated few to control ideas. Steve Yelvington argues that the internet has shifted the ability to create and distribute information away from institutions such as traditional media companies and toward the general public. Large publishers and broadcasters have seen their gatekeeper role undermined with the development of internet-powered communications. New technologies and innovations allow individuals to fill the roles traditionally held by publishers and journalists as they can instantaneously film, photograph, and describe events with mobile devices and online platforms. Today, individuals can and have built international followings, and many make a living speaking online—whether through Instagram accounts, gaming streams, monetized YouTube videos, podcasts, or subscription email lists. It has never been easier for one person, anywhere, to test his or her ideas in the global marketplace.

In countries with authoritarian regimes, social media and new types of information technology threaten these governments’ long-standing control over information and their societies. Citizens living under repressive regimes, including those in Iran, China, and Egypt, have used the internet and online platforms to organize protests against oppressive policies and unfair elections. Even more democratic and liberal governments show an aversion to allowing new forms of communication when it threatens their domestic industries and cultural traditions. Farhad Manjoo, a columnist for The New York Times, argues that European governments fear the new outlets for entertainment, news, and communication provided by American technology companies because of the challenges they face in controlling them.

The democratization of speech and the disruption of the traditional media gatekeeper role may also disproportionately undermine incumbent politicians. The “incumbency effect” is well understood, with the resources of incumbency and traditional media coverage of incumbents giving them a leg up against challengers. But research indicates that direct access of political challengers to masses of voters through social media platforms may reduce incumbents’ advantages. In nondemocratic states like Egypt, social media platforms may have played an even larger role in disrupting established political authorities.

U.S. conservatives have long complained that mainstream media sources are biased against them. If true, one might expect that the democratizing influence of social media platforms would give a greater voice to conservative speakers. And indeed, despite the litany of conservative complaints about platform bias, the current evidence suggests that conservative ideas have found their voice on online platforms. An NBC survey of the most popular authors on Facebook in February 2018 showed that out of the ten most popular online authors, six held and published conservative viewpoints. Donald Trump himself stated that “the fact that I have such power in terms of numbers with Facebook, Twitter, Instagram, etc., I think it helped me win all of these races where they’re spending much more money than I spent.” And some conservative outlets have had outsized influence on social media compared to major media outlets. For example, in the last quarter of 2019, Breitbart’s Facebook page, despite having one-eighth the following of CNN, “racked up more likes, comments, and shares ... (57.8 million) than The New York Times, Washington Post, Wall Street Journal, and USA Today combined (42.6 million)” and “outpaced each of the broadcast news networks, MSNBC, and CNN.”

But it is not just conservative voices that have found an outlet online. Social media has fostered political movements ranging from the nascent to the national to the international. The Black Lives Matter movement, #MeToo, and the Parkland kids’ gun control movement all originated online and grew through social media. Other examples of social media-driven movements include the Arab Spring uprisings as well as the recent Hong Kong protests.

Social media also empowers more than just political change. Social media platforms increase access to information in critical situations, allow people to raise money for local causes, and empower entrepreneurs. They also help us stay connected to extended family and friends, find help, and do our jobs. Social media is not just cat videos and memes.

If it is so easy for individuals to speak online today, why are people accusing the platforms of stifling speech? It is worth considering the motives of the disrupted incumbent gatekeepers. Antitrust scholar Ramsi Woodcock has argued that writers and reporters see certain big tech companies as a competitive threat, and their reporting is “colored by writers’ sense of professional vulnerability to the tech giants.” No surprise, then, that over “the first seven months of 2019, the [New York] Times published more than 300 articles mentioning Google, Facebook, or Amazon and antitrust, including an Op-Ed by a Facebook founder calling for breakup, an article discussing legal changes required to ‘take down big tech,’ and another musing on what Amazon will do once its ‘domination is complete.’” In this environment, politicians, including senators and state attorneys general, have benefited greatly from attacking big tech because, as Woodcock notes, “while [the press] may not be as well financed as Amazon, Google, or Facebook, writers can offer their friends something more valuable than money: publicity.” And these criticisms of big tech are not limited to the news pages. The News Media Alliance, which represents a large coalition of newspapers, has called for legislative changes to Section 230 of the Communications Decency Act which would impose significant costs on online platforms. Meanwhile, the same group is also seeking government help to compete with big tech companies—and gaining traction. House Judiciary Committee Chairman David Cicilline, who is leading a congressional antitrust investigation into big tech, has co-sponsored legislation favored by the News Media Alliance that would create an antitrust exemption permitting news outlets to form a cartel to negotiate with online advertising companies like Google and Facebook.

Still, while the mainstream press may be less than objective on the topic of internet platforms, it is obvious that this expansion of speech online is not problem-free. Online, as in real life, people do not always exercise freedom responsibly. Because social media provides easy access to large audiences, individuals with extreme and even hateful and dangerous views can find each other online and reinforce each other’s viewpoints. Online anonymity and pseudo-anonymity protect some from the real-life consequences of their uncivil or abusive behavior. In past mass media environments, gatekeeper publishers and broadcasters limited the ability of fringe perspectives to break through to an audience. Those gatekeepers screened material so that it would appeal to a broad audience. This meant that innovative new ideas faced barriers to reaching a national audience, but so too did ideas that large numbers of people found offensive or abhorrent. The internet has routed around those gatekeepers and added novel methods to continue humanity’s long streak of some people being terrible to others—only now, horrible behavior can go viral in ways that many others can see.

In short, the internet is a very large mirror reflecting a wide swath of human behavior, and we do not like everything we see.

The Problem Isn’t a Lack of Competition

So, people disagree greatly on exactly how speech problems manifest on platforms. Bias on the left or on the right? Too much moderation or too little? Yet no matter what speech concerns people have, a growing number argue that antitrust enforcement would help. They disagree over the nature of the problem but seem to agree that platforms lack competition and believe that increasing competition would help address the free speech problems they worry about. But while competition can promote free expression, the specific problems people worry about are not caused by a lack of competition—therefore, increasing the number of competitors is unlikely to address these problems.

Competition Can Help Protect Non-Market Values Such as Free Expression—But Not Every Speech Problem Is a Competition Problem

The focus on competition as a remedy is based on a sound premise: Competition can indeed help protect free expression. Competition incentivizes companies to deliver what consumers want, including whatever environments for expression they prefer. But consumers do not get everything they want even under vigorous competition, so the fact that some consumers are dissatisfied does not mean that the market lacks competition.

People typically think of competition as delivering reduced prices. But price is one of many different characteristics that influence consumer choices. Other important characteristics include quality, brand, color, source (“Louis Vuitton” or “Made in USA”), and manufacturing process (“made from recycled materials”). Consumers also consider socially desirable attributes of a product, such as fair-trade coffee or conflict-free diamonds. Companies compete to make the product with the bundle of characteristics that most appeals to consumers while still generating a profit for the company.

Free expression, as reflected in moderation policies and practices, is one characteristic of platforms that many consumers certainly care about. As such, we would expect competition to drive companies to invest in and experiment with moderation models and techniques to satisfy these consumers. Indeed, we do see examples of platforms trying different moderation approaches and evolving their approaches over time.

But while competition can spur companies to deliver what consumers want, including platform practices that protect free expression, the absence of any “perfect” set of practices does not prove that the market lacks competition or that there is an antitrust violation. The fact that consumers do not get exactly what they say they want from the market with respect to free expression does not mean the market has failed or lacks competition. In fact, few consumers get exactly what they want in any context. Many of the characteristics that consumers typically desire conflict with each other: price and quality; quality and source; convenience and environmental impact, to give just three examples. Given these tradeoffs, consumers search for the product or service with the bundle of characteristics that best meets their preferences. Different consumers can have widely varying, even conflicting preferences. Companies compete to meet these preferences, but in a world of limited resources they can never fully satisfy all the preferences of every consumer. Consumers will often buy a slightly improved version of the existing product, even if the improvement is simply a lower price. Indeed, if a product or service existed that perfectly satisfied every consumer’s needs and wants, there would be no need for competition.

Thus, if the platform marketplace today lacks any certain set of platform moderation practices, this probably is not caused by a lack of competition. There are many other more plausible causes. It could be that creating that set of moderation practices imposes tradeoffs on price or convenience that no consumer would want. It could be that no one has developed the right technology or business model to effectively deliver that set of practices.

Or—and this seems extremely likely—it could be that the set of moderation practices that most appeals to one group of consumers completely alienates another group of consumers. Indeed, as discussed above, people on the left and the right have diametrically opposed complaints about the platforms. It appears impossible for any platform to adopt a moderation policy (or even a policy of no moderation) that would fully satisfy all groups. It would be like trying to make a single carbonated beverage that satisfies both diehard Coke and diehard Pepsi fans—impossible, because they simply want different things. We would not claim that a market lacks competition because it has not produced the universally appealing soda. Likewise, we should not blame a lack of competition for not producing a platform with ideal content moderation practices, whatever those may be.

The market quest for the perfect platform is further complicated because many consumers demand not only a platform with their preferred content moderation policies, but also that this platform have broad appeal. Loosely moderated platforms like Gab exist. But many of those seeking greater free speech protection online want their freedom of expression to come with a mainstream platform with access to a large and diverse userbase. They want Coke instead of Pepsi, but they want Pepsi drinkers to drink Coke too.

Breaking Up Companies Won’t Mitigate Extremism or Reduce Bias

Because our current content moderation problems are not caused by a lack of competition, increasing competition through antitrust enforcement or other means is unlikely to improve the situation. Indeed, forcibly increasing the number of competitors may not address free speech problems and could exacerbate them.

For those concerned with platforms’ role in distributing harmful content, a mere increase in the number of platforms is unlikely to reduce the total amount of harmful content. If new platform competitors arise or are created through antitrust enforcement, individuals who want to say vile things will have more options of places to say them.

Some might argue that new competitors would adopt different moderation policies and create a greater variety of “safe” places. But is the goal to create spaces where it is safe to say awful things? It seems unlikely that many of those concerned with objectionable content online would be satisfied if that content were left online but was simply less visible to them. Furthermore, creating smaller platforms that appeal to narrow groups could further push the most hateful and vile to specialized platforms that would reinforce their tendencies rather than temper them. Services like 8chan demonstrate the potential toxicity of small platforms. For those who want platforms to adopt their preferred moderation policies, wouldn’t it be easier to convince or coerce a few large platforms rather than many smaller platforms? And given the difficulty and cost involved in moderating user-generated content, larger platforms may be more able to effectively curb harmful content.

Breaking up social media services could also reduce their total value. Each additional user on a social media site increases the value of the site, certainly to advertisers but also to users who seek to maximize the potential audience for their speech. This phenomenon, known as a network effect, means that many smaller platforms would not deliver the same total value as a few larger platforms.

Additionally, the largest platforms are now American-based, attentive to some degree to American concerns about and understanding of free expression. But breaking up these platforms could facilitate the rise of platforms based in China (like TikTok), which would raise all sorts of policy concerns, not least ones about free expression.

Nor is breaking up companies likely to address concerns about platform bias. Even if more vigorous or novel forms of antitrust enforcement increased the number of competing speech platforms, it may not increase the vibrance or variety of speech online. Indeed, it could force us into siloes of common opinion, diminishing viewpoint diversity on any one platform. As anyone who has spent time on modern social media platforms can attest, competition between ideas—often exemplified by interaction and debate—can happen within a single platform. This isn’t always the case: Filter bubbles can form, and it is possible to choose friends or follows so that one shields oneself from any difference of opinion. But, in general, larger platforms provide more opportunity for exposure to and dialogue with a greater range of viewpoints. In fact, “[c]ontrary to popular belief, we now hear more diverse voices than ever before—studies suggest that most people do not live in Facebook or Twitter echo chambers and ‘filter bubbles.’”

Some argue that the threat of consumers leaving and going to a different platform would help temper platforms’ content practices, and that this threat is more plausible when there are more competing platforms. But the literature on network effects (defined above) suggests that breaking up platforms would have little effect on platform bias. Markets with network effects tend to be characterized by competition for the market, rather than competition in the market. The platform that already has the most users will attract the greatest number of new users, creating a snowballing “winner takes all” dynamic. As such, in the fight to compete for the market, platforms face competitive pressure to adopt moderation policies that appeal to (or at least do not offend) the most potential users. Thus, even if antitrust enforcers forced more platform choices into the marketplace, those new choices will all face incentives to appeal as widely to their market as possible, which could result in similar moderation policies across competing platforms—and no increased variety in speech.

There are several other reasons that inflating the number of competitors through antitrust may not improve the robustness of speech online. Creating competitors won’t change the desire of certain people to speak to the largest possible audience. More specialized, smaller platforms are less attractive to attention-seekers who want to communicate a message to the broadest possible audience. Network effects mean that such attention-seekers will generally prefer the largest platforms irrespective of their moderation policies. And breaking up companies is unlikely to satiate those who want platforms to change their moderation policies. Even if platforms were smaller than they are today, the biggest platforms will almost certainly continue to face pressure about their moderation policies from diverse interest groups.

Current Antitrust Law Does Not Permit Breaking Up Companies Based on Free Speech Concerns

Even though increasing the number of competitors is unlikely to fix online bias or extremism, that hasn’t stopped people from trying. Indeed, many of those who are concerned about free speech online want to use antitrust law to address these problems. There are two rough categories of proposals to use antitrust to address speech problems. In this section we discuss proposals to use current antitrust law to break up tech companies and thereby create more competition. In the next section we discuss proposals to change current antitrust law to address speech problems. We also briefly mention a third, non-antitrust approach that would use regulation to try to generate competition.

Many propose using existing antitrust law to “break up” companies, which they believe would create more competition and thereby benefit freedom of speech and online expression. These types of proposals tend to be the vaguest of the three categories. They focus on the size of the largest platform companies and their market shares but do not articulate which specific company practices are anticompetitive under the law or how antitrust enforcement would improve freedom of expression.

This type of proposal is particularly prevalent among Republican politicians and regulators. For example, Senator Cruz argues that Facebook’s large size and its alleged censorship of political speech makes it a potential target for antitrust regulators. In a recent speech, Makan Delrahim, the assistant attorney general responsible for overseeing the antitrust division of the Department of Justice, indicated that a lack of competition in the digital economy could restrict free speech online. Indeed, a DOJ task force is investigating “the widespread concerns that consumers, businesses, and entrepreneurs have expressed about search, social media, and some retail services online.”

Democrats have similar proposals. Typical of these proposals is Chris Hughes’s assertion that Facebook faces “no competition” and therefore is held unaccountable for its bad behavior. He argues that government officials should use antitrust to break up large technology companies and then empower regulators to set standards and rules for social media speech. Robert Reich, U.S. Secretary of Labor in the Clinton administration, has said, “Antitrust should be used against Facebook and Twitter. They should be broken up” so that “the public will have more diverse sources of information.”

Despite the bipartisan urge to use antitrust to address free speech concerns, current antitrust law likely does not support breaking up tech platforms based solely on the free speech concerns discussed above. Supporters of these proposals informally frame the antitrust violations as monopolization or attempted monopolization cases. But being a large company with a high market share—or even being a monopoly—does not violate antitrust law. Antitrust law has rejected the simplistic conclusion that “big” is inherently “bad” because that approach was harmful to consumers and arbitrary in application.

Size and market share can indicate market power (the ability to increase price or reduce quality without losing market share), but they can also signal a competitor who succeeded by better meeting consumer demands. Prior to the 1980s courts primarily focused on the size of companies in antitrust cases, and as a result they frequently punished efficient companies that grew large by outcompeting their competitors to give consumers what they wanted. As economists have demonstrated (and experience has confirmed), larger companies can generate significant benefits, such as efficiencies of scale that enable lower prices. Antitrust enforcement focused on size alone often punished companies that served consumers better.

Although size or market share would seem to be a relatively objective test, antitrust enforcement before the 1980s was also surprisingly arbitrary. Sometimes a company’s purportedly threatening size was puzzlingly small. For example, in the famous Brown Shoe v. United States case, the defendant company held approximately 7 percent of the national shoe market. Aggrieved competitors also frequently weaponized antitrust to protect themselves from competition. The focus on size rather than conduct produced a body of cases so arbitrary that Justice Potter Stewart said the only consistency he could see was that “the Government always wins.”

Because of these flaws, courts in the late 1970s and into the 1980s turned away from a “big is bad” approach and toward more sophisticated analyses to identify harmful conduct. Courts adopted many procedural and substantive tools that determine whether a business practice is anticompetitive by its ultimate effects on consumers. This so-called “consumer welfare standard” recognized that consumers often can benefit from large companies.

As a result, market power—even monopoly power—by itself is not a violation of the antitrust laws. To succeed on a monopoly claim today, one must first define the relevant product and geographic markets and demonstrate that the accused company has power in those markets. But even if the accused company has a 100 percent share of the defined market, that alone is not sufficient to violate the antitrust law. The company must also engage in some specific exclusionary conduct that is helping the company obtain or maintain its monopoly. Exclusionary conduct is that which impairs the opportunities of rivals while not furthering competition on the merits. Examples include inducing a boycott of competitors, refusing to make certain deals, and engaging in predatory pricing.

Those who would use existing antitrust law to protect free expression have not defined the relevant product and geographical markets that apply to platforms. Nor have they shown that any accused company has significant market power in such a defined market.

But most importantly, they have not demonstrated that companies have engaged in exclusionary conduct to obtain or maintain that market power. Content moderation practices do not constitute exclusionary conduct sufficient to establish a monopolization claim. Such practices are internal to the company and not directed at competitors. Moderation practices do not make it more difficult for competitors to enter the market. Given the historical lack of transparency around moderation practices, it would be hard to argue that any of the major platforms gained or maintain their current market share through their choice of moderation practices. Selecting any specific moderation practice is unlikely to disadvantage a competitor. Furthermore, even if certain moderation practices harm some consumers, they often benefit many others. Those who would bring an antitrust case against the platforms will not be able to point to moderation practices as exclusionary conduct.

Platforms may be engaged in conduct other than content moderation practices that exclude competition, harming rivals and consumers. If so, it may be appropriate to bring a monopolization case under current antitrust law to address that conduct. But those who allege that platforms’ activities are anticompetitive will have to base their case on something other than content moderation practices—at least under existing antitrust law.

Changing Antitrust Law’s Reliance on the Consumer Welfare Standard Could Undermine Free Speech Values

A second category of proposals tacitly acknowledges that current antitrust law is not well suited to tackle free expression concerns. These proposals argue that the law’s emphasis on consumer welfare overly constrains the scope of antitrust. They would revise the law to empower antitrust to address broad concerns, including concerns about free expression. Such proposals range from changing the way antitrust law measures consumer welfare to expanding antitrust enforcement to prioritize political considerations over consumer welfare.

The Consumer Welfare Standard Is the Core of Modern Antitrust

Antitrust’s consumer welfare standard is often misunderstood and misrepresented. Examining some critics of existing antitrust enforcement, one could get the impression that they want to repeal the consumer welfare standard (CWS) as if it is a regulation or statutory text. But as one recent comment to the FTC explained,

[T]here is confusion about what the CWS is and what it is not. It is not a particular legal doctrine that is to be applied strictly in every situation and which, in itself, leads to particular results. It is a bundle of analytical tools and doctrines—subject to revision as new circumstances develop—that all point toward a common goal: increasing the welfare of consumers. These tools include standing/injury requirements, conduct requirements, effects analysis, burdens of proof, market definition requirements, and, ... presumptions.

Thus, the consumer welfare standard is the name for the courts’ decision that, when deciding whether business conduct is anticompetitive, the arguments will focus on whether the result of the conduct is good or bad for consumers. Antitrust expert Herbert Hovenkamp describes the consumer welfare standard as seeking to “encourage markets to produce output as high as is consistent with sustainable competition, and prices that are accordingly as low,” a goal that aligns with consumer interests while running contrary to the interests of cartels or less competitive firms that need high prices to survive.

Antitrust did not always have this singular goal. Judges, plaintiffs, practitioners, and academics developed it over decades of experience and learning. As both supporters and critics of the consumer welfare standard note, early antitrust regulators often sought to promote competition but frequently had an overt political intent to rein in powerful companies or to pick winners and losers in the marketplace. Terse, vague statutes gave very broad authority to enforcers, but courts struggled with how to judge what were “unfair methods of competition” or how to make sense of a statute that prohibited “every contract ... in restraint of trade,” for example. Courts began to grapple with why certain conduct was harmful but other conduct was not. In some early cases courts appeared to protect the competitive process. In others, it’s more accurate to say they protected certain companies from competition.

In short, early antitrust had multiple, conflicting goals, including limiting political power, protecting smaller competitors, and increasing consumer welfare. The statutes offered no principled way to balance these conflicting goals. The result: arbitrary, unpredictable, and often politicized enforcement. The consumer welfare standard gave coherence to this shapeless assertion of government authority and clarified a goal that nearly everyone agrees is important, even if some think other goals are more important. Rather than try to balance various conflicting goals, courts could focus on one goal and hear arguments from each side about how that goal was or was not furthered.

The consumer welfare standard has other benefits, too. Consumer welfare is a more justiciable standard than political power, for example: While consumer welfare can be challenging to measure, it remains far easier to quantify and compare than political power, which depends not only on the company at issue but also on the political orientations of the current government leadership. The consumer welfare standard also avoids the complexity of balancing many different interests: As FTC Commissioner Christine S. Wilson has said, “If consumers are harmed by reduced output, decreased product quality, or higher prices resulting from the exercise of market power, then this result trumps any amount of offsetting gains to producers or others.” And, unlike protecting smaller competitors, the consumer welfare standard better incentivizes overall economic efficiency.

The consumer welfare standard clarified the goal of antitrust, but that does not mean the law is static. For decades the goal of increased consumer welfare has been established in law and practice by experts across the political spectrum. But there remains a healthy and productive debate about how to achieve that goal in any specific case. Antitrust practitioners, academics, and advocates frequently and vigorously debate how to apply the legal, economic, and policy tools of antitrust. And antitrust has been able to incorporate new economic evidence and arguments developed by researchers and academics.

Proposals to Expand the Definition of Consumer Welfare to Address Free Speech Values Are Underdeveloped

Some recent proposals would expand the definition of consumer welfare to include certain free speech values. These proposals are not fully developed and lack empirical evidence.

For example, the Electronic Frontier Foundation (EFF) argues for “an updated consumer welfare standard that recognizes censorship as a cognizable harm to consumer welfare, and therefore an indicator of anticompetitive effects of digital platform market power.” According to EFF, online platforms’ free-of-charge advertising-based business models mean that “price cannot be the sole metric for evaluating conduct by internet platforms.” Chris Hughes similarly criticizes antitrust for its “narrow reliance on whether or not consumers have experienced price gouging” and argues that enforcers should also evaluate other effects of mergers and business conduct, such as reduced innovation.

However, EFF and Hughes are attacking a strawman version of the consumer welfare standard. In antitrust cases, raising prices is a commonly alleged anticompetitive effect and can be very persuasive when demonstrated. But antitrust also considers non-price effects of business conduct. As the government’s Horizontal Merger Guidelines state, “Enhanced market power can also be manifested in non-price terms and conditions that adversely affect customers, including reduced product quality, reduced product variety, reduced service, or diminished innovation.” Even when prices are the core effect being alleged, they are often quality-adjusted prices. And some of the biggest antitrust cases—including the cases against AT&T, IBM, and Microsoft—alleged reduced innovation as a core part of each case. Indeed, “[b]etween 2004 and 2014, the FTC challenged 164 mergers and alleged harm to innovation in 54 of them.”

EFF nominally acknowledges that the consumer welfare standard “also recognizes harms to innovation and product quality” but still claims that “current antitrust doctrine does not capture the full range of consumer harms.” More specifically, it claims that consumers are harmed by the inability to switch to platforms with more favorable content moderation policies. However, as described earlier, there can be many reasons other than competitive dynamics for why the market lacks products or services with some desirable feature. The market absence of some specific suite of features cannot be grounds for a successful antitrust action.

EFF argues that the consumer welfare standard should be expanded to incorporate harms created by content moderation. But the harm they allege—diminished quality of moderation—could already be considered under the current consumer welfare standard. They admit that content moderation is common and may be motivated by business need but assert that “large platforms tend to do it poorly.” This sounds like an argument connecting market power to reduced quality. Indeed, they propose quantifying censorship by focusing on how censorship “reduce[s] the service’s value to users, limiting the amount of expressive, communicative, and informational content they can access and share with each other (or excluding some users entirely).” This framing fits the existing consumer welfare approach.

But their argument remains skeletal. As noted in the previous section, not every act of a large company that disadvantages some party is anticompetitive—nor should it be. It is insufficient to assert that poorly done content moderation reduces a service’s value. A persuasive case would have to demonstrate that a company exercised its market power to lower content moderation quality in a way that harmed consumers compared to moderation in a competitive market.

Proposals to Politicize Antitrust Law Will Harm Free Speech Values—And Cause Collateral Damage to the Economy

Other proposals would more radically change antitrust law to address speech and other issues. Teachout and Khan, for example, would address platform political power by adopting “a general spirit” of antitrust that would allow the government to limit a company’s size and regulate market concentration, even if doing so would actually reduce consumer welfare. Relying on nonlegal definitions of monopoly they argue that the structure of corporations and markets are political questions that should be answered through political means. Specifically, they urge regulators to use antitrust to break up companies or limit their size in order to constrain their effect on politics, including on political speech.

This urge to use antitrust as a broad political tool is a core theme of populist and neo-Brandeisian antitrust reformers. Neo-Brandeisians argue that antitrust enforcement ought to expand beyond its traditional focus on consumer welfare and competition to address a wide range of policy problems, such as labor, inequality, political power, and, yes, free speech. They seek to use the powerful tools of antitrust as “meta-legislation” to achieve policy objectives.

Yet changing antitrust to a general-purpose tool to achieve a wide range of political objectives is a recipe for disaster. History amply demonstrates that politicians use antitrust for personal and political gain. Increasing the discretionary nature of antitrust would exacerbate such abuses. And, because it is a broadly applicable area of law, changing antitrust would bring such abuses, both private and public, upon a wide swath of the economy, not just technology platforms.

History demonstrates that unconstrained antitrust is abused by government and private parties, including to stifle speech

Empowering antitrust authorities to incorporate goals other than promoting the competitive process gives government a very powerful tool to use at their discretion. Government could abuse such a tool to directly or indirectly coerce limits on speech to favor a certain view or to silence criticism of government itself. In the past, the U.S. government has demonstrated a propensity to do just this.

Critics of the consumer welfare standard see an increase in arbitrary action as a feature of their proposals. Teachout and Khan see returning to politicized antitrust as a positive. But in the area of the First Amendment, courts are often wary of laws that the government can enforce for any or no reason, because such laws place a pretext of legitimacy on otherwise illegitimate government action. That is, laws that can be enforced whenever there is the political will to do so can be used to punish disfavored speech or speakers.

Would regulators seek to use antitrust law to coerce the treatment of social media content? As shown by examples discussed in this paper, many politicians calling to “break up big tech” hope that antitrust enforcement will change platforms’ content moderation practices. Given that many of those demanding antitrust enforcement want to change how online platforms govern users’ expression, we should expect that regulators would likewise seek to use antitrust tools to shape online speech to match the views of the president, other politicians, or the regulators themselves. At the very least, the newly empowered regulators of social media platforms would face political pressure to impose the changes in content moderation sought by those supporting these new powers.

In the past, political pressures have influenced antitrust enforcement in the United States. Since the formation of the antitrust laws, political actors have used antitrust regulators to shape investigations and enforcement actions to suit their own personal or political purposes.

Theodore Roosevelt, one of the earliest proponents of “trust-busting,” authorized a lawsuit against J.P. Morgan’s Northern Securities Trust, which controlled several railroads. However, Roosevelt left Morgan’s trusts that owned industrial assets untouched because he considered them “good trusts” that would cooperate with the government. Truman and Eisenhower took the opposite approach to antitrust enforcement in the oil industry, arguing that national security concerns negated any potential competition concerns and, in some cases, actively encouraged coordination between major oil firms. President Franklin D. Roosevelt reportedly used the threat of a full-scale monopolization investigation of the motion picture industry to cajole industry leaders to meet with him in private, where he hoped they could be “straightened out.”

The worst presidential abuses of antitrust authority, however, were by Presidents Johnson and Nixon. Each used antitrust enforcement to target media companies that threatened their political influence. President Johnson went after a newspaper, the Houston Chronicle, which had long criticized the Kennedy and Johnson administrations and had supported Nixon’s presidential candidacy in 1960. After Johnson won reelection in 1964, he allegedly contacted the Chronicle’s publisher, who also happened to be the president of a bank with a merger before the Justice Department. Johnson told the publisher that merger approval would only be granted if the Chronicle agreed to support Johnson throughout his presidency.

Nixon similarly used the threat of antitrust enforcement to intimidate broadcast network executives at ABC, NBC, and CBS, the only three privately owned television networks broadcasting nationally at the time, into providing better coverage of him and his administration. In fact, he preferred the threat of action over actually bringing a lawsuit, noting, “If the threat of screwing them is going to help us more with their programming than doing it, then keep the threat. ... Don’t screw them now. [Otherwise] they'll figure that we’re done.” He asked the attorney general to not file an antitrust case against the networks but instead to “hold it for a while, because I’m trying to get something out of the networks.” One of his aides noted that “keeping this case in a pending status gives us one hell of a club ... something of a sword of Damocles.” Indeed, Nixon’s staff bragged that the threat of an antitrust action coerced NBC into airing a wedding special on the president’s daughter and CBS into listening to White House input on its stories.

The Nixon administration also settled an antitrust case against the International Telephone and Telegraph Company (a case Nixon had strongly opposed) after the company had allegedly donated $400,000 to Nixon’s reelection campaign.

Abuse of antitrust laws has occurred even in administrations that sought to scale back antitrust enforcement. The Reagan administration staffed the antitrust division of the Department of Justice and the FTC with officials so supportive of Chicago School economics that the result was characterized as “an essentially laissez-faire approach in practically all areas of antitrust.” Yet even that administration interfered with antitrust enforcement to suit the administration’s political desires. When the Justice Department began investigating potential collusion between airplane manufacturers and airlines, the Reagan administration ordered the agency to stop due to concerns that an investigation would make it more difficult to negotiate transatlantic airline fares with the United Kingdom. Reagan also personally provided input on antitrust cases involving telecommunications and steel companies, shaping decisions to the preferences of a politician rather than to the requirements of the law.

However, as antitrust law became more grounded in economic evidence and argument, the ability of enforcers to apply it arbitrarily to achieve other political or speech-related goals became constrained. As a United States filing with the Organisation for Economic Cooperation and Development explains,

The U.S. Federal Trade Commission ... and the Antitrust Division of the Department of Justice (“DOJ”) [] do not consider employment or other non-competition factors in their antitrust analysis. The antitrust agencies have learned that, while such considerations “may be appropriate policy objectives and worthy goals overall ... integrating their consideration into a competition analysis ... can lead to poor outcomes to the detriment of both businesses and consumers.” Instead, the antitrust agencies focus on ensuring robust competition that benefits consumers and leave other policies such as employment to other parts of government that may be specifically charged with or better placed to consider such objectives.

In other words, the tools of antitrust are not the correct tools to implement even worthy noncompetition policies.

Still, current politicians remain inclined to use antitrust to address personal grievances with certain companies. As a presidential candidate, Donald Trump criticized CNN over its coverage of his 2016 presidential campaign and vowed to block the then-proposed merger between the outlet’s parent company and AT&T. After winning the election, President Trump continued to criticize the merger and the Department of Justice’s antitrust division filed a lawsuit to block the merger. Numerous individuals, including the CEO of AT&T, questioned whether the DOJ’s opposition to the merger was politically motivated. And as noted earlier, several sitting senators advocated for increased antitrust regulation of online platform companies because, they alleged, these firms discriminate against individuals with conservative beliefs.

Relaxing the standards for antitrust action and broadening the scope of goals it seeks to achieve would roll back the clock in antitrust law, facilitating government abuse of antitrust law and providing a powerful tool for politicians, regulators, and enforcers who want to suppress disfavored speech or punish disfavored speakers.

Changing antitrust law would have collateral effects far beyond tech

Because antitrust law applies to nearly all industries, changing antitrust law to address online speech concerns would affect the entire economy, including many sectors that have nothing to do with free speech issues. Ditching the consumer welfare standard raises many concerns, as others have documented in detail. The mainstream consensus supports the consumer welfare standard as the right tool for antitrust, even as there is vigorous debate about how to apply it. Abandoning that standard for the entire economy based on free speech concerns in one sector is a drastic solution. Even if one is less skeptical of the negative effects, it would still be prudent to target the practices and industries one is concerned about rather than changing laws that affect every industry.

Relaxing the standards for antitrust enforcement would also fuel more private antitrust litigation. Plaintiffs who win private antitrust suits can receive treble damages. Such suits can be brought by competing firms or harmed consumers, but can also be done in bad faith. Even under the current consumer welfare standard, companies often use private antitrust litigation as a “tool for harassing, harming and extorting payments from other firms.” Economists Preston McAfee and Nicholas Vakkur have documented seven different purposes for which companies have abused private antitrust litigation:

  • to extort funds from a successful rival;
  • to change the terms of a contract;
  • to punish noncooperative behavior;
  • to respond to an existing lawsuit;
  • to prevent a hostile takeover;
  • to discourage the entry of a rival; and
  • to prevent a successful firm from competing vigorously.

Loosening the standards for alleging an antitrust violation or multiplying the types of harms that can be considered anticompetitive would affect the entire economy, giving politicians, companies, and individuals an even more powerful tool to abuse disfavored commercial rivals.

Using Competition Policy to Regulate Speech Online Would Be More Targeted Than Antitrust But More Indirect Than Other Forms of Regulation

We will quickly mention a third category of competitive policy proposals. Such proposals are not antitrust, which relies on ex post enforcement to correct anticompetitive actions. In contrast, competition policy relies on legislation to ex ante design an industry’s competitive structure. One example is Elizabeth Warren’s proposal for tech companies. She suggests classifying large technology companies as “platform utilities,” regulating company conduct, and requiring tech firms above a certain size threshold to divest themselves of certain businesses. She proposes doing so through new laws and regulations separate from antitrust law. Other proposals that fall in this category include the reportedly forthcoming White House executive order regarding political bias by social media platforms and the Digital Platform Act supported by certain advocacy groups.

Fully exploring the broad range of possible competition policies is outside the scope of this paper. However, two points are worth considering quickly. First, unlike changes to antitrust law, competition policy approaches can be more targeted. While problematic for many other reasons, such approaches can limit collateral effects better than sweeping changes to antitrust law. One key challenge for such proposals is to define boundaries between those intended to be subject to the laws and those not intended to be so affected in ways that cannot be gamed. Even so, they may create distortions between similarly situated companies, depending on which side of the line one falls.

Second, one problem with using competition policy to address free speech concerns is its indirectness. If one is already going to step outside the antitrust paradigm to pass laws and regulations, why not look for ways to directly regulate platforms to achieve the desired free speech goals? As discussed above, there are good reasons to think that increasing the number of competitors will not greatly increase the range of moderation practices. Setting specific rules could achieve the desired result.

The answer to why competition policy rather than direct regulation may be as simple as the First Amendment. The First Amendment protects private parties—including the content choices of private social media companies—against government interference. It would bar certain direct regulations of platform moderation practices. And even neutral competition policy approaches could run afoul of the First Amendment. The Supreme Court has “recognized that even a regulation neutral on its face may be content based if its manifest purpose is to regulate speech because of the message it conveys” or if it is discriminatory “in its practical operation.” Many of those who seek to employ competition policy openly seek to do so because of and to alter the content of online expression.

While competition policy can be more targeted than changes to antitrust law, it too can raise significant free speech concerns. These issues are worth further examination but will need to be done on a per-proposal basis.

Conclusion

Today we have more opportunities to speak to wider audiences and hear from more diverse voices than ever before. However, the same technological advances that have made that possible have also ignited new concerns about how those freedoms are used and who should regulate these new global speech forums.

Many have proposed using antitrust law to address these speech-related concerns. But while antitrust is a powerful tool, it is not the right tool for this job. Current antitrust law imposes significant legal challenges to breaking up internet platforms based solely on speech concerns. Even if such companies could be broken up, many of the speech concerns we have discussed will not be resolved—and may even be exacerbated—by increases in the number of competitors. The history of antitrust and First Amendment law in the United States warns of the dangers and downsides of arbitrary exercises of broad antitrust power. An antitrust law powerful and flexible enough to stop harmful speech and eliminate bias on platforms is an antitrust law so overbearing and arbitrary that it could be used to limit speech. We should look elsewhere to address our concerns about speech online.

 

Many thanks to Parker Kobayashi for his copious research assistance on this paper. Thanks also to Knight First Amendment Institute staff for careful and thoughtful feedback on this paper.

 

Printable PDF

 

© 2020, Neil Chilson and Casey Mattox. 

 

Cite as: Neil Chilson & Casey Mattox, [The] Breakup Speech: Can Antitrust Fix the Relationship Between Platforms and Free Speech Values?, 20-06 Knight First Amend. Inst. (Mar. 5, 2020), https://knightcolumbia.org/content/the-breakup-speech-can-antitrust-fix-the-relationship-between-platforms-and-free-speech-values [https://perma.cc/UNT7-SEK8].