In current debates over the Internet’s impact on global democracy, the prospect of state regulation of social media has been proffered as a solution to problems like fake news, hate speech, conspiracy-mongering, and similar ills. For example, US Senator Mark Warner has proposed a bill that would enhance privacy protections required of internet platforms, create rules for labeling bot accounts, and change the legal terms of the platforms’ legal relationship with their users. In Europe, regulation has already been enacted in the form of the European Union’s General Data Protection Regulation (GDPR) and new laws like the Network Enforcement Law (NetzDG). This chapter will survey this rapidly developing field, putting current efforts of liberal democracies to regulate internet content in the broader perspective of legacy media regulation. As we will see, there are very different national approaches to this issue among contemporary liberal democracies, and in many respects the new internet regulations, actual and proposed, are extensions of existing practices. We conclude that, in the US case, content regulation will be very difficult to achieve politically and that antitrust should be considered as an alternative.
Media regulation is a sensitive and controversial topic in all liberal democracies. The US Constitution’s First Amendment protects freedom of speech, while media freedom is guaranteed in various legal instruments governing the European Union and the Council of Europe, as well as in the European Convention on Human Rights. Freedom of speech is normatively regarded as critical to the proper functioning of a liberal democracy, and the Internet was seen in its early days as a great boon to democratic discourse. Nonetheless, all modern democracies regulate media in various ways; even in the United States, where the First Amendment is regarded with great reverence, the state has over the centuries seen fit to draw boundaries around what can be said or shown on various media platforms.
In addition to normative hostility to restrictions on speech, many observers have maintained that it is not technologically possible to regulate media in the age of the Internet. The explosion of bandwidth for communications of all sorts that has occurred since the 1980s has made state control vastly more difficult than in the days when citizens relied on a handful of local and national newspapers and two or three broadcasting channels operating over finite, government-allocated radio spectrum. Regulating content on the Internet was said to be like “nailing jello to a wall” (Reference Allen-EbrahimianAllen-Ebrahimian 2016) because rapid technological change would quickly outpace any government mechanisms for censorship. The sentiment that normative and technical constraints undermine state control of online content is evident in John Perry Barlow’s “Declaration of the Independence of Cyberspace,” in which governments were told “You have no sovereignty where we gather” (Reference SegalSegal 2018).
China may ultimately prove that Barlow was wrong and that a high-capacity authoritarian government is perfectly capable of controlling speech on the Internet. Indeed, China envisions applying AI/big data techniques to the monitoring of citizen behavior under a “social credit” system. Short of this level of control, however, governments in modern liberal democracies that want to protect freedom of speech nonetheless use various mechanisms to regulate legacy media, mechanisms that have been carried forward into the digital era.
There are, however, not only large systematic differences in approach between Europe and the United States with regard to the role of the state but important differences as well among European countries. We focus here on France and Germany. These differences did not begin with internet regulation but are evident in the long prior experience of dealing with legacy media.
Modern liberal democracies do not typically regulate speech via censorship directed by the state, as in authoritarian regimes, that is, by having a government official approve or alter the content carried by a media channel. Rather, the mechanisms tend to be indirect and include controlling access to media channels through licensing; managing sources of revenue to support private media; setting broad guidance as to what type of content is deemed acceptable or not; promoting certain content through public broadcasting and other mechanisms; and establishing more or less permissive regimes for private citizens, including politicians and other public figures, to press defamation and invasion of privacy claims. In addition, states have sought to encourage self-regulation by content providers. These are the same techniques that have been carried forward into the Internet Age. An alternative approach to media regulation is antitrust. The logic here is that, in the marketplace of ideas, fair competition among ideas depends in part on meaningful competition among the platforms that carry ideas. Media pluralism, therefore, is a source of resiliency against propaganda, demagoguery, and extremism coming to dominate democratic discourse. Thus, in most liberal democracies, print media have not been subject to extensive content regulation because these markets are usually decentralized and competitive. In contrast, broadcast media, and in particular TV, have been much more highly regulated everywhere because of the formerly oligopolistic or monopolistic position of broadcasters. Today, one could argue that internet platforms like Google and Facebook occupy a position similar to that of legacy television networks: Because of their scale and reach, decisions that they take with regard to content moderation are far more consequential than for print media. An alternative to state regulation of content would therefore be antitrust actions designed to increase the number of platforms and reduce the reach of the current internet giants.
Media Governance in France and Germany
France and Germany are the two largest media markets in continental Europe. Though media governance in France and Germany have much in common when contrasted against other European countries and the United States, noteworthy distinctions emerge when the two countries are contrasted with each other. These distinctions are interesting in their own right and as illustrations of how the two largest media markets on the European continent approach media governance. These distinctions are also important contextual factors to consider as France, Germany, and other actors within the European Union grapple with whether and how to pursue national and/or European approaches to addressing disinformation.
Of particular relevance to the present discussion is how France and Germany differ in their respective conceptions of media pluralism and especially in the role envisioned for the state in safeguarding or promoting media pluralism. As we shall see, the two countries’ initial forays into regulating problematic content on online platforms already show signs of these differences, which emerge as important historical themes in how the two countries have regulated legacy media. Reference Hallin and ManciniHallin and Mancini (2004) present a framework for comparing media systems among the Western liberal democracies of Europe and North America based on four dimensions: the developmental arc of a mass circulation press; the nature and extent of linkages between the media and the broader political system, or political parallelism; norms and practices associated with journalistic professionalism; and the role of the state in shaping the media system. Using this framework, the authors describe three archetypical models for the relationship between a political system and its media (Reference Hallin and ManciniHallin and Mancini 2004, pp. 73–75). These models – Polarized Pluralist, Democratic Corporatist, and Liberal – provide a useful analytic framework for assessing recent developments in the media landscape in Europe and the United States. The media systems in these countries existed well before the Internet became a global force, and key concepts, traditions, and assumptions from these legacy governance frameworks help frame debates about whether and how to regulate internet platforms and the content they carry.
Countries described by the Polarized Pluralist model (France, Greece, Italy, Portugal, and Spain) feature an elite-oriented print media with relatively small circulation and a comparatively more popular broadcast media. Freedom of the press and the rise of commercial media industries developed relatively late in these countries. Linkages between the media and politics tend to be durable and mutually reinforcing and characterized by advocacy journalism, political parallelism, and the instrumental use of the media by political and commercial actors through regulation and/or ownership to advance their broader political and economic interests. The state tends to play a large role in regulating, owning, and/or financing media, and public broadcasting tends to be politicized.
In contrast to the Polarized Pluralist countries, the Democratic Corporatist countries (Austria, Belgium, Denmark, Finland, Germany, the Netherlands, Norway, Sweden, and Switzerland) experienced early development of press freedom and a robust commercial media industry, especially in print media where readership and circulation levels are still among the highest in Europe. Newspapers affiliated with political parties and interest groups have been common throughout their recent history, however, and political parallelism is relatively high, though declining. Advocacy journalism remains an important feature of the media landscape but coexists alongside neutral, information-oriented reporting. The media is regarded as an important social institution worthy of state support and protection, and the profession is marked by high degrees of professionalism and formal organization. Democratic Corporatist countries tend to have long traditions of liberal democracy (with Germany standing out as a notable exception, as we will see) and well-organized social groups coexisting within a corporatist framework that emphasizes consensus and rational-legal authority.
The third model is the Liberal model, which Hallin and Mancini most closely associate with Canada, Ireland, the United Kingdom, and the United States. Like the Democratic Corporatist countries, the Liberal countries also feature strong traditions of press freedom, a commercial mass-circulation press, and early development of liberal institutions. Unlike the Democratic Corporatists, however, the role of the state is generally more limited, and the media is shaped largely by market forces as opposed to partisan, ideological, or other instrumental purposes.
In both France and Germany, achieving and sustaining media pluralism is a fundamental goal of policy because it is viewed as essential to a vibrant, competitive marketplace for ideas and thus vital to democratic discourse. In France, this ideal can be traced back to the 1881 Law on the Freedom of the Press, which established a liberal regime for the press that reflected the view of French elites in the Third Republic that the primary threat to pluralism was excessive state control of the press and that safeguards against excessive state interference were needed. The consensus among the ruling former French Resistance elites after World War II, however, was that the pendulum had swung too far in the direction of liberalization during the interwar period, with the result that press barons had accumulated too much control over the media. As Raymond Reference KuhnKuhn (2011) explains, “[w]hereas in 1881 the concern of the legislators had been to promote the liberty of the press by protecting it from political control of the state, in 1944 the emphasis was placed on removing economic threats to press freedom from capitalist owners” (p. 12). Thus, in 1944 France put in place a number of measures to shield the press from market pressures, including a framework for provision of financial aid from the state to print media – a practice that continues today.
For France, however, media pluralism means more than just a competitive media market; for broadcast media especially, it means a pluralism of distinctly French media that upholds and delivers French linguistic and other politico-cultural values to French citizens, in an increasingly competitive global media marketplace dominated by American popular culture (Reference KuhnKuhn 2011; Reference EkoEko 2013). Though France has on occasion engaged in outright censorship – perhaps most notably in the 1950s and early 1960s during the Algerian war of independence – it has more typically directed policy and state resources to promoting preferred content, through subsidies and content requirements, over outright censorship of disfavored content. The French approach of promoting preferred content can be seen in the political competition for influence over public broadcasting, where such competition resulted in a changing series of regulators, the Haute Autorité de la communication audiovisuelles (HACA), the Commission nationale de la communication et des libertés (CNCL), and finally the Conseil supérieur de l’audiovisuel (CSA) in 1989.
For Germany, media pluralism is viewed as a bulwark against the return of totalitarianism. The structure of German public broadcasting, with its devolution of broadcasting governance to the Länder, reflects this orientation. Instead of establishing one or a small number of national-level public broadcasting outlets, as the British and French did, German public broadcasting is organized and managed at the German state, or Länder, level. This decentralized structure was imposed on Germany by the Allied Powers following World War II as a way of guarding against the monopolization of media power by any single political actor (Reference Hallin and ManciniHallin and Mancini 2004). Indeed, concern about the monopolization of the German media by illiberal forces – and the concomitant emphasis in Germany on media pluralism – is a recurring theme in Germany’s governance frameworks for media, which generally seek to “limit state power in order to avoid the recurrence of totalitarianism” (Reference Hallin and ManciniHallin and Mancini 2004, p. 161).
The public broadcasting ideal in Europe has been codified by the Council of Europe, and all of the Council’s members are required to establish public broadcasters conforming to it (Reference Mutu and CorralMutu and Corral 2013). For example, German public broadcasters in the Länder are governed by independent boards comprised of representatives from political parties on an apportioned basis and members of civil society, such as trade unions and professional associations – a typically corporatist approach to governance. A recommendation adopted by the Council of Ministers in December 2000 (Council of Europe 2000) calls on member states to guarantee the independence of broadcasting regulators, referring in turn to Article 10 of the European Convention on Human Rights and decisions of the European Court of Human Rights that define freedom of information as a fundamental right. Control over content and determination of what constitutes the public interest is not to be set directly by governments but rather delegated to a professional body whose standards, in theory, reflect a belief in impartiality, nonpartisanship, and a broad sense of public interest.
Another way in which this orientation manifests is print media, by way of contrast with France. On the one hand, both France and Germany have traditionally had vibrant local and regional newspaper markets. As recently as 2013, for example, half of all newspapers sold in Germany were regional papers (Reference Stelzig, Schwarzer and SpitzerStelzig 2015, p. 71). The French display an even more marked preference for regional papers over the national press (Reference KuhnKuhn 2011, p. 38). Overall readership of newspapers, however, is declining in both countries (Reference KuhnKuhn 2011, p. 38; Reference Kolo, Weichart and MurschetzKolo and Weichart 2013; Reference Lardeau, Le Floch and MurschetzLardeau and Le Floch 2013). Matthieu Lardeau and Patrick Le Floch highlight the comparatively high cover prices of French newspapers, due to declining advertising revenues and long-standing production and distribution inefficiencies in the French print industry, as a major factor turning readers away. Kuhn agrees, and identifies changing consumer preferences and competition from free newspapers as additional relevant factors (Reference KuhnKuhn 2011, pp. 40–42). In the case of Germany, Castulus Kolo and Stephan Weichart point to changing consumer preferences and competition from internet platforms for advertising dollars as critical factors impacting the German newspaper industry. French newspapers benefit from an elaborate system of direct and indirect subsidies that keep many otherwise nonviable newspapers alive. Germany, however, has no tradition of direct subsidies, limiting its state support instead to indirect measures, such as preferential tax treatment. As Reference Kolo, Weichart and MurschetzKolo and Weichart (2013) explain, “[t]raditionally, Germany’s postwar governments have decided not to directly support their press as fears of intervention into editorial affairs of newspapers remained widespread” (p. 215). How to support the newspaper industry while steering clear of French-style direct support remains, as of this writing, a matter of considerable debate in Germany (Reference Kolo, Weichart and MurschetzKolo and Weichart 2013).
The examples of public broadcasting and print media highlight how France and Germany differ with respect to the appropriate role of the state in promoting media pluralism. France has tended to pursue a more direct, interventionist role for the state in promoting a particular brand of media pluralism that emphasizes French nationalist and cultural values, from subsidizing French newspapers to mandating that broadcasters carry French cultural content. Germany, however, has purposefully avoided comparable forms of direct intervention, out of concern that the state might abuse these powers to undermine pluralism in pursuit of illiberal ends.
When we turn to digital media, we find that recent developments in both countries aimed at addressing misinformation and other problematic content reflect the experience with legacy media. On November 20, 2018, French lawmakers, with French President Macron’s backing, passed a “law on the fight against the manipulation of information”. The law establishes an expedited judicial procedure for adjudicating complaints by individuals and organizations, including the French government, about alleged “fake news” and its proliferation in the period leading up to elections. It also imposes heightened transparency obligations on platforms during these periods and requires them to police their platforms for fake content between elections. The French broadcasting regulator, the Higher Audiovisual Council (CSA), is empowered under the law to strip broadcasters of their licenses if the CSA determines that the broadcaster is spreading offending content on behalf of a foreign government.
Germany has also enacted a new law to address problematic content, the “Act to improve the enforcement of the law in social networks,” also known as NetzDG. Rather than endow the federal German state with new authorities for initiating and adjudicating complaints about content, however, NetzDG puts that responsibility in the first instance on the platforms. One of the “Frequently Asked Questions” that the German Federal Ministry of Justice and Consumer Protection (BMJV) poses to itself about NetzDG is, “Why should social networks decide what is legal and what is not? Is that not the job of law enforcement agencies?” The BMJV goes on to answer:
This obligation is already set out in section 10 of the Telemedia Act. According to this Act, service providers are obliged to delete unlawful content as soon as they become aware of it. It is therefore up to service providers themselves to decide whether content is unlawful when such content is reported. Those who operate services and/or infrastructure and make these available to third parties have a duty to limit their own role – whenever still possible – in any abuse of the infrastructure and/or services they provide. Operators themselves are responsible for doing this after receiving specific indications that such abuse is taking place. There is no general judicial-scrutiny provision in German civil law which would imply that any issues surrounding disputed content could, or indeed must, first be resolved by the courts.Footnote 1
This approach, according to the BMJV, helps “guarantee a free, open and democratic culture of communication.”
To be sure, both governments have maintained close political ties with media that add important nuance to the relationships between the state and media in each country. According to Reference Hallin and ManciniHallin and Mancini (2004, p. 168), for example, Germany’s corporatist governance of public broadcasting is criticized by some as falling short of the corporatist ideal of pluralistic multi-stakeholder governance, since the stakeholders that comprise the governance boards often have partisan or ideological leanings that are consistent with the political representatives on the board, such that a board may in fact be comprised mainly of like-minded partisans. Media policy making in France, on the other hand, is political in a different way – it is usually conducted by the president in a deliberative context dominated by elites, including the media organizations themselves (Reference KuhnKuhn 2011). Indeed, French presidents have routinely sought to, and in most cases succeeded in, at least partially refashion elements of French media governance to advance one or more political objectives (e.g., Reference KuhnKuhn 2011, p. 21). In this respect, Macron’s Paris Call in November 2018 and his initiatives on press subsidies are in step with a line of media-related measures initiated by his predecessors. In addition, the media in both countries have in the past adhered to norms that have had the practical effect of shielding politicians and incumbent parties from political threats. In France, for example, journalistic norms about not reporting on politicians’ private lives have protected political figures, perhaps most notably former French president François Mitterrand, from scrutiny of their character (Reference KuhnKuhn 2011). In Germany, mainstream media have for decades refused to give much media exposure to the far right, which Antonis Ellinas (Reference Ellinas2010) argues is an important reason for why the German far right has lagged behind its ideological brethren in Austria, where the far right has experienced greater electoral success.
Media and technology have always had a deeply intimate, symbiotic relationship. Advances in communications technology inevitably impact the development, distribution, and consumption of media. At the same time, consumer demand for content and new media experiences can in turn drive the market for new communications technology. A salient feature of the French approach to media governance, in contrast to Germany and that country’s traditional export orientation, is France’s strong dirigiste streak. This streak is evident in France’s various efforts to support French media outlets in the face of international competition. According to Reference Lardeau, Le Floch and MurschetzLardeau (2013), “while French regulators in the 1980s claimed to be preoccupied with efforts to limit concentration and thwart the voracious appetite of domestic press barons, today’s emphasis has shifted to enabling French media empires to grow sufficiently large and prosperous to compete with international rivals” (p. 196). The dirigiste streak is also evident in its deployment of media policy in the service of wider French industrial policy initiatives. For example, Reference KuhnKuhn (2011, p. 16) suggests that one of the motivations behind France’s decision to extend the daily broadcast schedule for television in the 1960s was to help French television manufacturers sell more televisions.
A particularly compelling example of French industrial policy relating to the media is Minitel, the “[p]rofoundly French” internet platform (Reference Mailland and DriscollMailland and Driscoll 2017). Minitel was a modem-based videotex platform and network service developed and implemented in the 1980s by the French postal and telecommunications regulator PTT; Minitel remained in service until 2012, when it was retired. The catalyst behind the French government’s considerable investment in Minitel – which included giving a Minitel video device costing several hundred dollars to every household in France for free – was a report commissioned in 1976 by French President Valéry Giscard d’Estaing on telecommunications and computers from two leading French experts, Simon Nora and Alain Minc. Giscard was deeply concerned about the state of France’s telecommunications infrastructure, which lagged far behind other industrial countries, and about the country’s ability to compete with the United States in both the technological and the cultural spheres, as advances in telecommunications and computers created new opportunities – and risks – for content creation and distribution. Giscard initiated a major overhaul of France’s telecommunications infrastructure in 1975 and directed Nora and Minc to look ahead at how France should approach digitization. Nora and Minc’s 1978 book-length report, The Computerization Society, was a bestseller in France and was a call to arms for the French government to intervene decisively with industrial policy aimed at building a competitive “telematics” industry in France – Nora and Minc’s phrase for the increasing convergence of the telecommunications and computer industries (Reference Nora and MincNora and Minc 1981, p. 4). The authors predicted a “computer revolution” that “will alter the entire nervous system of social organization” and “affect the economic balance, modify power relationships, and increase the stakes of sovereignty” (Reference Nora and MincNora and Minc 1981, pp. 3–4). Especially in light of the “IBM challenge” from the United States, which threatened to “encroach upon a traditional sphere of government power, communications” (Reference Nora and MincNora and Minc 1981, p. 6), France needed “a deliberate policy of social change”:
It must support the [French] companies that provide computer-related services, a sector that is dynamic but fragmented; it must allow powerful public intervention in the field of research, provided incentives linked to the activities of the manufacturers for the component parts of the computer industry, and finally, once its strategy has been determined, it must allot a proper role to the national manufacturer of large computers.
President Giscard, a conservative, approved a videotex pilot program in 1978 (Reference Mailland and DriscollMailland and Driscoll 2017, p. 55), and his successor, François Mitterrand, a socialist, brought Minitel to market five years later.
Media Regulation in the United States
There is an extensive literature on American exceptionalism that documents the ways in which the United States has always been an outlier among other developed liberal democracies with regard to state regulation. Seymour Martin Lipset, among others, has noted that the American state modernized later than did the state in other advanced societies, was less extensive, and achieved a lower degree of professionalization (Reference LipsetLipset 1995). American political culture remains highly suspicious of concentrated political power and has created an expansive set of political institutions to protect citizens from government power. The private sector has a much more positive valence as the locus of entrepreneurship and individual freedom. In Europe, by contrast, the private sector is regarded with greater suspicion, while government is generally regarded more as a protector of public interest that is needed to safeguard citizens from excessive private power. How European governments exercise this protector role, however, is itself an important source of variation in how democracies regulate media.
These broad generalizations show up in a variety of domains related to technology regulation. European regulators, for example, in recent decades have primarily focused on protecting consumers from abuse of their privacy rights by private corporations, as exemplified by the EU GDPR, which establishes a comprehensive framework for consumer privacy. In the United States, there is no such comprehensive framework; instead, privacy protections are embedded in a variety of sector-specific laws and regulations, such as the Health Insurance Portability and Accountability Act (HIPAA) and the Fair Credit Reporting Act (FCRA), which govern health- and credit-related data, respectively. In addition, Europeans have always been more willing to regulate speech than Americans; many European countries enacted laws banning certain forms of hate speech (e.g., publication of Nazi propaganda and symbols) well before the rise of the Internet, in contrast to the United States where such speech is legally protected. Private speech is also more constrained in certain European countries as a result of tougher libel laws that facilitate civil litigation against private individuals.
The United States has typically relied much more heavily on industry self-regulation than have European democracies, and this tradition has carried on into the digital age. Neither the Federal Communications Commission (FCC) nor any other federal regulators have sought to lay down formal rules as to the kinds of content that internet platforms can carry. Government agencies have expressed concern over certain forms of content like child pornography or terrorist training materials that might be deemed illegal per se – and not simply because they were posted on the Internet. Digital platforms have responded to such pressures by taking down such content to avoid criminal and reputational liability but not necessarily because the government was endowed with special powers through a law like the German NetzDG.
Indeed, the internet platforms were spared from the threat of private liability for content they hosted by Section 230 of the Communications Decency Act of 1996. This section is frequently misinterpreted as requiring that internet companies act like neutral platforms (i.e., by not curating content) if they are not to lose this protection from private lawsuits. In fact, the intention of the legislation was the opposite: Were such liability to exist, the platforms would respond by not seeking to moderate content at all for fear of being held liable for their editorial decisions. This protection was seen as an effort to promote the rapid growth of internet platforms and placed the burden of content curation squarely on the platforms themselves. The latter sometimes try to argue that they are simply neutral conveyors of other people’s content, but it is clear that they have been able to promote or demote various forms of content at their own discretion.
The nature of this self-regulation has shifted dramatically as a result of Russian involvement in the 2016 American election, the Cambridge Analytica scandal, and the realization that extremist views and polarization were being fostered by platform algorithms that promoted conspiracy theories and outlandish personal attacks over high-quality information. Politicians on both the left and the right began suggesting that platforms might need more overt forms of regulation to protect American democracy.Footnote 2 Facebook founder and CEO Mark Zuckerberg went so far as to suggest that his industry needed some form of government regulation. Facebook and Google began to tune their algorithms and started to make more overtly political kinds of decisions as to what constituted acceptable content. Google’s YouTube, for example, banned conspiracy theorist Alex Jones from using its services. Facebook, for its part, tweaked its algorithms and beefed up its human content moderation staff substantially beginning in 2017. It is also in the process, as of this writing, of establishing an independent appeals process for adjudicating complaints alleging improper content mediation decisions by Facebook.
There is plenty of precedent for private, for-profit companies making these sorts of prudential decisions. The US First Amendment does not protect the right of individuals to use privately owned platforms; indeed, the First Amendment protects the right of those platforms to carry whatever content they see fit. Only the government can be accused of censorship under its terms. Private corporations like those that own the New York Times or the Wall Street Journal make decisions about appropriate content all the time; that is what media companies do.
The problem with self-regulation by companies like Facebook and Google is not a legal one; the issue they raise is one of basic legitimacy, brought on by their scale. Traditional newspapers like the New York Times make decisions to carry only certain content and not others in a print media market that is still relatively decentralized and competitive. Consumers have the option of switching from the Times to a different print media outlet if they dislike its coverage.
The same is not true in the world of online platforms. Owing to economies of scale and scope, the large internet platforms (predominantly Facebook and Google) have grown to such a size that they effectively constitute the public square, not just in the United States but in dozens of other countries around the world. A takedown by YouTube is far more consequential than a decision by a legacy media company not to carry a particular writer or point of view, since there are few other channels for reaching so wide an audience. Facebook exercises government-like powers, even though it is not a government; it is a private, for-profit company largely controlled by a single individual, whose primary objective is not necessarily to serve the public interest of his political community. Facebook can try to acquire the trappings of a government, like its own internal Supreme Court or its own currency, but these efforts in the end do not make its behavior more democratically legitimate.
This is not a new or unprecedented situation. Democracies have faced the issue of monopolistic or oligopolistic media companies in the past, during the heyday of broadcast television. The universal response by liberal democracies was to regulate broadcast media, and many continue to do so to varying degrees. This power to regulate media exists despite the commitment of all modern liberal democracies to protect fundamental freedom of speech and is as true in the United States as it is in Europe or Asia.
In order to understand the possibilities for state regulation of internet platforms in the United States, it is therefore useful to look at the history of regulation of legacy media.
Regulation, Competition, and Privatization of Legacy Broadcast Media
Lack of competition is what initially induced democratic countries to regulate broadcast media, since the introduction of radio and then television offered new channels of mass communication with limited bandwidth. This was true in the United States as in other democracies. Congress began regulating radio broadcasting with the Radio Act of 1927, which created a temporary Federal Radio Commission whose primary job was to allocate scarce broadcast spectrum (Reference Head and SterlingHead and Sterling 1982, pp. 140–141; Reference SparSpar 2001). Federal authority over broadcasting was expanded by the passage of the Communications Act of 1934, which established the FCC as the principal national regulator, first of radio and then of television broadcasting.
From the beginning, there were constitutional challenges to the FCC’s ability to regulate broadcasting, not just from the perspective of the First Amendment but also regarding the national government’s authority over interstate commerce and its ability to deprive private actors of property by denying them a license under the Fifth Amendment’s takings clause. The latter two issues were ultimately decided in the government’s favor by the courts, but the first issue – the state’s right to regulate content – has remained controversial up to the present. The 1934 act also stated that broadcast media were not “common carriers” like railroads or trucking companies that could be forced to accept content from all comers – an issue that is potentially important in the context of present-day debates over net neutrality (Reference Head and SterlingHead and Sterling 1982, pp. 405–409).
The grounds for content regulation in American law were initially laid by the “public interest” standard written into the Radio Act of 1927 and carried forward by the Communications Act of 1934. This standard said that private broadcasters were expected to serve not just their own commercial interests but a broader public interest as well. The act did not define what that interest was but left it up to the FCC to establish guidelines. In practice, the FCC did not write highly detailed rules defining public interest; rather, there was a general expectation that broadcasters would include content relevant to their local communities and would provide balanced coverage of political issues. This standard was enforced by the threat that the FCC might not renew a broadcaster’s license if it did not deem it compliant with its general guidance (Reference Head and SterlingHead and Sterling 1982, pp. 410–412).
The argument for public broadcasting was made initially on the basis of spectrum scarcity, which evaporated with the development of cable television and the Internet. The oligopolistic or quasi-monopolistic positions of broadcasters began to give way in virtually every country to a much broader and more diverse media landscape. State regulatory authorities, beginning in the United States and Britain, began opening up their broadcast sectors to private actors. Competition began in Britain with the introduction of ITV in the 1950s and expanded in the 1980s with the growth of powerful private channels like Rupert Murdoch’s Sky TV. These private broadcasters began to compete vigorously with each other and with the incumbent public broadcasters. The latter lost significant market share and saw a large erosion of their revenue base as user fees declined and advertisers shifted to alternative platforms.
This shift was not just technological. The 1980s saw the rise of politicians like Ronald Reagan and Margaret Thatcher, who attacked the regulatory state as the source rather than the solution to public problems and who believed that free markets would ultimately produce a fairer distribution of resources than heavily regulated markets. They received intellectual support from an economics profession dominated by a rising form of market orthodoxy that argued that prices constituted accurate reflections of relative scarcities – the so-called Chicago School (Reference AppelbaumAppelbaum 2019).
This pro-market revolution in economic thought had major implications for media regulation. In the early days of the FCC, there was a presumption that the government was a guardian of public interest and that commercial interests on their own would not necessarily produce positive outcomes in terms of either efficiency or democratic control – hence the “public interest” mandate in the 1934 Communications Act. Even Herbert Hoover, long regarded as the epitome of a free-market enthusiast, argued that the broadcast sector was crying out for regulation (Reference Head and SterlingHead and Sterling 1982, pp. 139–140). By the 1980s, however, this view had changed to one that held that largely unregulated media markets would best serve public interest and that state regulation did not serve a neutral “public interest” but was politically driven.
The pro-market revolution of the 1980s had consequences for the other form of possible state intervention: antitrust. Led by Robert Bork, Aaron Director, George Stigler, and others, a number of conservative legal scholars and economists began arguing for a much narrower understanding of the grounds on which the government could launch antitrust actions. Bork argued that the original Sherman Act envisioned only one standard for public concern, which was consumer welfare as measured by prices and/or quality. Government antitrust suits against technology companies like IBM and Microsoft, which dragged on for years and ended up costing large amounts of money, were criticized for being wasteful and misdirected.
The growing hostility toward media regulation was best illustrated by the rise and fall of the FCC’s “Fairness Doctrine.” The latter had its origins in the “public interest” provision of the Communications Act and was strengthened in 1959 when Congress amended the act’s Section 315 which had exempted broadcasters from having to provide equal time coverage for certain kinds of news programs. The new law stated “Nothing … shall be construed as relieving broadcasters … from the obligation imposed upon them under this Act to operate in the public interest and to afford reasonable opportunity for the discussion of conflicting views on issues of public importance” (Reference Head and SterlingHead and Sterling 1982, pp. 476–477). The FCC interpreted this as a statutory endorsement of equal time coverage and an elaboration of the public interest provision of the original act. The FCC could, in other words, behave a bit like a European public broadcaster in forcing private media to provide what the agency regarded as balanced coverage of political events.
The FCC interpreted the Fairness Doctrine as permitting it to force broadcast stations to allow responses to personal attacks. The Supreme Court took up the question of whether this exercise of state power was compatible with the First Amendment in the 1969 Red Lion Broadcasting v. FCC case. In a preview of contemporary controversies over conspiracy theories and fake news, the FCC wanted to compel a Christian radio station to allow replies to a conservative commentator who alleged plots by the FBI and CIA in attacking a critic. The Court upheld the constitutionality of the FCC’s action on the grounds of spectrum scarcity, the uniqueness of broadcasting, public interest, and the state’s fiduciary responsibility (Reference Head and SterlingHead and Sterling 1982, pp. 477–478).
The Fairness Doctrine continued to be controversial, especially among conservatives. They believed that it was being used by the government to shut down conservative voices and that the FCC could never be truly impartial in its enforcement of the rule. By the 1980s, there was also a growing belief among economists that markets, left to their own devices, would be self-correcting. This view was used to critique antitrust actions, under the assumption that large scale implied efficiency rather than an undue use of market power (Reference WuWu 2018). Yet it was also applied to the market for ideas; in line with classic First Amendment thinking, it was argued that good ideas would eventually drive out bad ideas and that the government did not have a legitimate role in policing thought. Consider this passage from a 1985 law review article:
[W]ere there no fairness regulations, the most a broadcaster could hope to gain from misinforming or misleading its listeners is the allegiance of those already ideologically committed to the broadcaster’s point of view. That allegiance, probably depending on the issue addressed, may or may not counterbalance the loss of viewers who are not ideologues. But, in the absence of the doctrine, broadcasters would have almost no incentive to provide erroneous or one-sided information to those who do not want it or to refuse all coverage of issues that interest have to convince some many viewers or listeners. They do, after all, have to convince someone to turn on the set and tune in their frequency.
It was also argued that the Fairness Doctrine would stifle rather than encourage free speech, by deterring broadcasters from airing controversial content in the first place for fear of provoking an FCC action. The Fairness Doctrine was eventually rescinded by the FCC itself under Republican chairman Denis Patrick in 1987, in a 4–0 vote. The Democrats in Congress have subsequently tried unsuccessfully to reinstate the doctrine on several occasions but faced strong opposition from Republican administrations, including a veto by President Reagan and a threatened veto by President George H. W. Bush.
In contrast to its European counterparts, then, the United States by the early twenty-first century had taken a much more relaxed position than its European counterparts toward the regulation of legacy media. This was driven both by the explosion of bandwidth provided by cable and internet technology and by an ideological shift toward greater distrust of government regulation. The Europeans, facing similar pressures, have themselves backed away from pervasive state control over legacy media and permitted greater market competition, though their public broadcasters remain powerful players in many European countries. It is not the case, however, that the United States never took a more European-style approach to media regulation. The Supreme Court has upheld the constitutionality of the Fairness Doctrine in the Red Lion case; its abolition was the result of an administrative decision on the part of the FCC.
Conclusions
Regulation of legacy media may seem to be irrelevant to contemporary discussions of whether the Internet should be regulated in the interests of protecting democracy, in light of the huge differences between the technologies involved. Yet many of the older controversies remain the same and may provide legal precedents for future action.
As we have seen, the governments of modern liberal democracies have reconciled their efforts to regulate legacy media with their commitment to freedom of the press on the grounds of spectrum scarcity. As technology evolved during the 1980s and 1990s, this concern seemed increasingly outdated. The efforts by liberals to reinstate the Fairness Doctrine were driven in large measure by their unhappiness with the growth of Fox News, AM talk radio, and a host of new conservative media outlets that challenged the mainstream media’s dominance of the political and social narrative. Conservatives answered these criticisms by arguing that the new channels simply provided political balance in a media market that was diverse and highly competitive. If Fox News was attracting more viewers than CNN, that was the result of individual consumer choice and not a mandate by the government. In this they were correct; the legacy media market is not dominated by a single political point of view, however much one may object to biases by one outlet or another. If one does not like Fox News, one is always free to switch the dial to MSNBC.
The Internet at the outset promised to be as diverse and competitive as legacy media when it first came into existence as a public platform in the 1990s. Since then, however, it has moved in the direction of the broadcast media of the 1950s. Two very large platforms, Facebook and Google, and a third somewhat smaller one, Twitter, now serve as the primary channels of communication for hundreds of millions of people, not just in the United States but worldwide. It can be argued that the platforms have put themselves in a position comparable to the three main broadcast networks in the United States back in the heyday of over-the-air television. The platforms may not be monopoly providers of information, but neither were the broadcast networks; what they have in common is extraordinary influence over what millions of people see and hear through their decisions to prioritize certain content over others.
None of the internet platforms are necessarily promoting a single point of view. Yet, through their terms of service, they have maintained the right to police content, a right that is further protected by Section 230 of the Communications Decency Act, which shields them from liability for what they decide to carry. In addition, as private actors they enjoy free speech rights protected by the First Amendment. This has led to the paradoxical outcome wherein the platform’s ability to restrict speech on behalf of a foreign authoritarian government is actually protected constitutionally from US government regulation (Reference KellerKeller 2018).
In the early days of the Internet, content moderation centered around relatively uncontroversial issues like terrorist incitement, child pornography, cyberbullying, and the like. Yet, with the weaponization of social media by Russia and a host of other political actors, they have come under increasing pressure to restrict fake news, conspiracy theories, hate speech, and other toxic content flying around the Internet. They are, in other words, being asked to perform the same kinds of editorial functions that legacy media organizations like newspapers and TV channels have traditionally undertaken.
The question today, however, is whether the platforms’ near-monopoly or oligopoly position puts them in a similar position to the broadcast networks back in the early days of television. Because the market for social media is less competitive, a decision to remove certain content is much more consequential than the decision of, say, USA Today not to carry it.
As the platforms have grown more active in policing more political forms of speech in recent years, complaints have grown that they are showing political bias (Reference Wakabayashi and KangWakabayashi and Kang 2018). This in turn has led to a number of lawsuits against them, in some cases demanding that they be required to carry opposing political viewpoints. These “must carry” demands replicate the logic of the old Fairness Doctrine, that is, that the government could override the free speech rights of a private actor and compel it to carry certain material for the sake of political balance. To date, the courts have not supported such suits, but in light of precedents like Red Lion and subsequent Supreme Court cases, it is not clear in what ways contemporary internet platforms are different from legacy broadcast media.
There are two basic approaches to solving the problem of platform dominance. The first is to accept that dominance as an inevitable fact and to try to regulate platforms in the manner of legacy broadcasting. This has been one leg of the European approach to date. It is very unclear, however, what sorts of regulation would uphold public interest. Forcing the internet platforms to carry certain kinds of content in the name of political balance might have similar effects to the Fairness Doctrine, discouraging them from carrying controversial content in the first place. It is also not at all clear what a proportionate and legitimate approach to internet regulation along these lines might look like, in light of political polarization within societies and, given the global scope of service provision by internet platforms, different conceptions across countries of the appropriate role for the state in intervening in media markets. A number of European countries continue to enjoy a degree of political consensus that allows them to support public broadcasters that promote balanced coverage believed to be in the public interest. This would not seem to be possible in a highly polarized country like the United States currently. The boundaries of acceptable discourse have been challenged by the highest political authorities; expecting a federal agency like the FCC to police political debate is highly unrealistic.
The second, and not necessarily mutually exclusive, approach is to pursue policies aimed at increasing the degree of competition among internet platforms. As we have seen, liberal democracies have pursued a variety of mechanisms to support competition and pluralism, albeit with mixed degrees of success and based on different conceptions of pluralism and the role of the state. There is growing interest and focus on the competitive practices of large internet companies, but it is not immediately obvious how and whether alternative internet media platforms would affect the quality of civic discourse online and the incidence of disinformation (Reference Cremer, de Montjoye and SchweitzerCremer, de Montjoye, and Schweitzer 2019).
There is a growing consensus that antitrust law needs to be broken out of the framework established by the Chicago School back in the 1980s and modernized to meet the needs of the digital age. There are several prongs to this reassessment (Reference WuWu 2018; Reference KhanKhan 2018). The first is to broaden the courts’ understanding of potential harms arising from excessive concentration of power in the hands of a small number of private corporations. Robert Bork was insistent that the legislative history behind the Sherman Act encompassed only economic harms to consumer welfare and not potential political harms (Reference BorkBork 1966). It is not clear that he was correct in this assertion, since Senator Sherman himself articulated deep concerns about the political effects of concentrated private power on American democracy.
The chief political harm that contemporary platform power poses today is the one that we have just explored: The editorial decisions taken by Facebook and Google, while perhaps justified in themselves, have enormous power to affect overall political discourse in the countries in which they operate due to their sheer scale. They are not public institutions and have no legitimacy as, in effect, custodians of public interest. This problem would not be nearly as severe if there were greater competition among platforms, in the manner of legacy print media.
In terms of economic harms, it is clear that the existing consumer welfare standard measured by prices needs to be rethought. Many internet services are offered at zero price. The real costs to consumers arise elsewhere: foregone privacy when the platform sells their information to third parties or uses it to advance their own self-interest; foregone innovation when a large platform buys a start-up that is a potential competitor, and lost opportunities when platforms use their enormous data on consumer preferences to pursue exclusionary policies in adjacent markets (Reference KhanKhan 2017).
If one accepts the premise that large platforms create new categories of harms through reduced competition, the question then turns to potential remedies. Here, the very nature of digital markets poses severe constraints. Modern internet platforms enjoy enormous economies of scale and scope; the larger they are, the more potentially useful they are to advertisers and consumers. Competition exists not for market share but for the market itself, since scale and scope economies reward a single dominant player (Reference Cremer, de Montjoye and SchweitzerCremer et al., 2019). If a government were to try to break up Facebook, for example, in the way that it broke up AT&T, it is likely that one of the “baby Facebooks” would eventually come to assume the same dominant position that the original Facebook enjoyed.
It is beyond the scope of this chapter to review all of the possible antitrust remedies that have been suggested. One idea is to increase data portability, so as to make it easier for users to switch to alternative platforms. The idea of data portability has already been built into European regulations like GDPR; but privacy rules like GDPR themselves limit portability, since one’s network connections are part of one’s individual profile but also “belong” to those connections. For a group to move to another platform from Facebook would require consent by all the network members, something that is unlikely for a number of behavioral reasons.
A second idea is to restrict exclusionary behavior through the purchasing of potentially competitive start-ups. Facebook has already been subject to substantial criticism for its purchases of Instagram and WhatsApp and is busy seeking to integrate them with its existing services to make it harder for regulators to force a later disgorgement. This kind of scrutiny would require a redefinition of anticompetitive behavior away from the focus on total market share used in classic antitrust cases. In a Schumpeterian world of creative destruction, the most dangerous competitors are often not large rivals but small, nimble ones with good ideas.
Another suggestion for facilitating antitrust actions is to shift the standards of proof for potential harms from ones based on a strict rule of reason to structural remedies. One of the consequences of the Chicago School revolution in antitrust thinking was the establishment of a rule of reason standard that required antitrust enforcers to provide extensive empirical market analyses and projections of consequences for future markets of a given enforcement action. This increased the time and cost necessary to bring an antitrust action against a company and explains why the IBM and Microsoft cases dragged on for as many years as they did. A structural standard, by contrast, would define a certain market structure as per se anticompetitive, avoiding the need for such lengthy litigation. While both US and European competition law have been influenced by Chicago School thinking, there are more grounds for structural remedies in the latter case. The problem with this approach is knowing what structural standard to apply in a highly fluid technological world, one that is not arbitrary and likely to impede rather than promote competition.
The political climate surrounding internet regulation has shifted dramatically since 2016. The large platforms, and especially Facebook, have come under sustained criticism for their past behavior in facilitating Russian interference in the US election and for increasing domestic polarization by facilitating extremist speech, conspiracy theories, and the like. The Europeans have shifted in this direction earlier and more decisively than American regulators, imposing national-level laws like NetzDG or Europe-wide ones like GDPR and initiating antitrust actions against the large platforms. In the face of these shifting views, American regulators have been taking a fresh look at platform behavior. It is, nonetheless, not clear whether this will result in a consensus regarding either regulatory or antitrust remedies to the problems at hand. Regulation and antitrust are to some extent substitutes for one another; the problem of content moderation on the Internet could be dealt with under both approaches or the two in combination. Both approaches face substantial obstacles to implementation in practice, particularly in the United States where partisanship and polarization have reached new heights in recent years. Whether platform behavior will actually change in the face of these shifts in public opinion and state policies remains to be seen.
Open access