Making Social Media an Instrument of Democracy

Thomas Wischmeyer. European Law Journal. Volume 25, Issue 2, March 2019.

In recent years, the responsibility of social media platforms towards their users and society at large has become a major political issue. However, the regulatory responses to the crisis of social media are still mostly considered unsatisfactory, as demonstrated by the widespread criticism of the German Network Enforcement Act of 2017. This article compares the current constitutional discourse on social media regulation with the debates that accompanied the last major transformation of the media landscape: the rise of broadcasting. While we certainly do not find a roadmap for social media regulation in the past, the key concept of the broadcasting discourse—the idea of media as a sphere of ‘institutional freedom’—can be applied to the challenges of today and can be used to strengthen the democratic function of social media.

Over recent years, two major narratives have developed in response to the question: What kind of challenge do digital technologies pose for society? The first, which understands digitalisation as a fundamental discontinuity in the very fabric of society, takes its arguments largely from the post‐war literature on technocracy. In the 1950s and 1960s, philosophers and sociologists argued that in the industrial age, technology had lost its emancipatory potential and that instrumental rationality, as it is embedded in our devices and artefacts, had infected human minds. Writers across the ideological spectrum including Martin Heidegger, Norbert Wiener and Helmuth Schelsky described why capitalism slowly transforms human power over nature, materialised in technology, into the control of technology over men and society. And they concluded that technological progress impoverishes human lives and ultimately enables a system of total domination. Herbert Marcuse, probably the most prominent voice of this movement, put it as follows: For the concept of technical reason is itself perhaps ideology. Not merely its application, but technique itself is domination … The aims and interests of domination are not ‘additional’ or dictated to technique from above—they enter into the construction of the technical apparatus itself. For technique is a social and historical project: into it is projected what a society and its ruling interests decide to make of man and things. The aims of domination are ‘substantive’, and belong to the form of technical reason itself. H. Marcuse, ‘Industrialization and Capitalism’ (1965) New Left Review. For an instructive analysis of Marcuse’s theory, see Habermas’ article ‘Technology and Science as “Ideology”‘, reprinted in J. Habermas, Toward a Rational Society: Student Protest, Science, and Politics (Beacon Press, 1970).

Today, more and more scholars are applying this narrative of a societal transformation to the rise of digital technologies. And indeed, historically the development of many digital ‘apparatuses’ and methods, including the Internet or artificial intelligence (AI), was initiated and financed by state actors for governance or military purposes. The dynamics of digital capitalism show some striking similarities to the so‐called ‘late capitalism’ of the industrial society, too. We are witnessing again an accelerating process of economic concentration, the rise of financially and politically powerful multinational corporations, and the emergence of oligopolistic market structures, which are stabilised by government intervention in the form of national planning strategies or by close cooperation between private IT companies and state actors, most excessively in China.

Moreover, the adherents of the transformation narrative claim that digital technologies have betrayed the original ideals of the digital revolution and have morphed from tools, which we employ to serve our ends, to express our thoughts and to communicate with others freely, into artefacts, which control how we act, think and communicate. Once again, ‘the construction of the technical apparatus’ seems to determine how we conceptualise ourselves and our society. Smart devices and applications change our self‐perception and our most private as well as our public communications—and reduce them to marketable items. ‘Algorithms’ are described as autonomous entities, which are woven so deeply into the fabric of our daily life that, unbeknown to us, they have started to shape our preferences, to create and curate our infospheres, and to nudge our decisions. Over time, even the rule of law might be replaced by the rule of algorithms.

If this diagnosis were true, digitalisation would strike to the very core of liberal democracies and a radical response would be necessary in order to preserve the modalities of our lifeworld. However, political platforms resembling the social movements, which developed in the past to counter the exigencies of the industrial age, do not exist (yet). Moreover, it is far from clear how the regulatory instruments we could employ to protect the normative status quo would look. The various proposals for new charters of fundamental rights in the digital age have so far mostly failed to offer a convincing perspective and to create a political momentum.

Now, for the proponents of the second narrative, this failure indicates that the idea of a transformation might be interesting from an intellectual point of view, but that the approach towards digitalisation in the realm of practical politics needs to be more pragmatic. For them, digitalisation represents an important development, the consequences of which we are only beginning to understand. However, rather than stressing the disruptive power of the digital transformation for society’s normative foundations, they emphasise the normalcy of technological progress. According to their ‘reconfiguration narrative’, digitalisation ultimately means the rise of a new set of devices and methods, which—like any new technology—brings opportunities and creates risks. The main challenge then consists in carefully evaluating the facts and in finding governance mechanisms that increase the benefits and reduce the costs for individuals and society. While the reconfiguration narrative does not blind itself to potential negative effects of digitalisation, its main concern is to make the rules and principles that were developed for the offline world fit for the digital age. If this turns out to be difficult for reasons such as the ‘un‐territoriality’ of data, functional equivalents need to be found.

The two narratives are not incompatible with each other. Rather, they operate on different theoretical, temporal and normative levels. Whether or not digital technologies à la longue will transform the structure of our communications is largely beyond the interest of the followers of the reconfiguration narrative. For them, digitalisation raises important questions today such as the re‐distribution of responsibility between programmers and consumers or the design of effective incentives for the creators of sustainable AI, which can be answered within the conventional legal and political framework. The transformation narrative, on the other hand, clearly aims at the constitutional plane. Nevertheless, the narratives do not simply coexist independently of one another. Its proponents represent different interests, they compete for attention and vie for political influence.

One field, in which this contest is currently very much ongoing and on which the remainder of this paper will concentrate, is the digital transformation of media or, more specifically, the rise of social media. Only a couple of years ago, the emancipatory function of social media was widely hailed and the proliferation of communication enabled by social media was considered to be a stimulus for democracy. However, the experiences over the past years have been sobering. Not only did social media fail to reinvigorate a meaningful public conversation, the technology is now used regularly to spread incitement to ethnic cleansing, to recruit terrorists, to distribute child pornography and ‘hate speech’ and to contribute to the proliferation of what is often called ‘fake news’. Even social media companies themselves now recognise that their product may endanger social stability.

All of this obviously plays into the hands of those advocating stronger regulation. And indeed, the responsibilities of online platforms ‘towards their users and society at large’ have become a major theme in politics. Nevertheless, the laws that have been implemented in response to the crisis over the past years still follow the political logic of incremental change when they try to apply offline rules to the online world. Technically, they focus on the protection of individual rights through the state or the platforms in order to minimise the normative costs for individuals and society without endangering the alleged benefits of the new technologies. But what if digitalisation indeed transforms the ways we think, act and communicate about private and political matters and thus changes the structure of public debate? What if the toxic culture of social media is not the product of a small number of Internet trolls, which we need to exclude from our conversation, but an in‐built feature of social media? In this case, the reconfiguration script and its emphasis on the protection of individual rights might at some point turn out to be insufficient. But what would an alternative look like? This is the question I shall try to answer in this paper. Because the effects of online media differ considerably across countries, I limit my analysis to the situation in Germany.

The paper is organised as follows: First, in Section , in order to assess the plausibility of either the transformation or the reconfiguration narrative, the rise of social media will be located within the theoretical framework associated with Habermas’s account of the public sphere. I will then show in Section that the most recent attempt to address the crisis of social media in Germany, the Netzwerkdurchsetzungsgesetz (Network Enforcement Act 2017—NEA), is still firmly committed to the logic of reconfiguration—and at the same time demonstrates the limits of this approach. However, while the competing transformation narrative works well in the realm of critical theory, its proponents have so far failed to provide workable concepts for a structural reform of social media. Against this backdrop, in Section I propose to look at the constitutional debates of the 1950s and 1960s, when the last major transformation of the public sphere happened—the rise of broadcasting. Obviously, we do not find a detailed roadmap for social media regulation in the past. But, so I argue, the key concept of the discourse on broadcasting—the idea of media as a sphere of ‘institutional freedom’—can be helpful also today. The paper concludes with some preliminary thoughts on how this concept can be operationalised for social media in future.

Social Media and the Public Sphere

The public sphere operates as an ‘intermediary structure between the political system, on the one hand, and the private sectors of the lifeworld and functional systems, on the other’. Here, social and political events are observed, selected and commented on according to their social relevance. The idea that a well‐functioning public sphere is a condition of possibility for a robust democratic polity is most frequently associated with the work of Jürgen Habermas. The nexus between mass media, the free press and professional journalism, on the one hand, and democratic accountability and liberal constitutionalism, on the other, is indeed a major theme of Habermas’s magnum opus Faktizität und Geltung (‘Between Facts and Norms’).

Habermas’s earlier work, however, is less teleological. In Strukturwandel der Öffentlichkeit (‘The Structural Transformation of the Public Sphere’), which is still indebted to the alienation theory of Adorno and Horkheimer, Habermas describes how developments in media business and media technology have transformed the liberal culture of coffeehouses and civic societies of the enlightenment age into the mediatised public sphere of the early twentieth century characterised by mass media, in particular newspapers and broadcasting. The main points of this analysis are, first, that the structure of the public sphere is not a given, but malleable and, in large parts, a product of the media we use to communicate and, second, that every transformation of the public sphere has important political ramifications. Not only is it possible to use the new media for new types of political campaigning, etc., but, according to Habermas, the invention of a new type of media also changes the very structure of public debate and enables a new type of politics. In Strukturwandel, Habermas depicts the change induced by the rise of mass media as profoundly ambivalent: newspapers, radio and broadcasting not only enabled modern mass democracies but also fascism.

Many scholars have applied Habermas’s theoretical framework to the digital age and have argued that the rise of social media once again induces a structural transformation of the public sphere or at least contributes to the ongoing fragmentation of the public sphere(s). Numerous studies dissect the politics of social media and attempt to show how social media, due to their technological architecture and the financial interests behind them, reward certain types of speech and silence others, create and divide audiences, and thus slowly transform the way we communicate not only on private, but also on public matters.

The exact political effects of social media, however, are rather difficult to pinpoint as shown, for example, by the controversy among social scientists about the phenomenon and the impact of so‐called ‘filter bubbles’. Again, there is the obvious case that social media are employed for political advertising, for electoral battles, for direct communication between politicians and citizens, for government PR or for election meddling. Some speak of social media in this context as an ‘information weapon’. However, the concept of the structural transformation is interested less in these overt or direct forms of influence and more in the powerful, but diffuse ways, in which a new type of media changes the style and structure of the public conversation, which ultimately translates into a different conduct of politics. As Habermas has shown, the direction of this change is not fully determined by the design of the new medium, but also depends on the context of its use. And indeed, if we look at case‐studies on the political impact of social media, which were conducted across the globe over the past years, it is very difficult to generalise the findings. In stable democracies social media are employed for different purposes and affect politics differently compared to fragile societies. Against this backdrop, we probably need to speak less of ‘the’ digitalisation and rather recognise—following the literature on capitalism —that there exist ‘varieties of digitalisation,’ especially in the media sector.

However, the lack of a general theory is not an e contrario proof of the reconfiguration narrative. Rather, it demonstrates that it is probably impossible to decide the debate on the impact of digitalisation on the public sphere in global terms. In order to make a strong case for the structural transformation thesis, a thick description of a specific society à la Habermas would be necessary. This is beyond the scope of this article. Here, I will limit myself to highlight three key features of social media, which are widely agreed on and which we can plausibly assume to have a major impact on the traditional structure of public discourse in Germany.

First, social media are not good at moderation; they amplify intent and magnify opinions, moods or agendas. With the rise of social media, the balanced press statement is replaced by the provocative tweet. The speed of social media reduces the time for the discussion of a topic massively. In contrast, public debate in post‐war Germany was dominated by the desire to achieve balance, to accept the power of argument and to strengthen the newly established institutions. Instead of either/or, the practice of the sowohl/als auch (‘not only/but also’) became the communicative ideal. Moderation was key, in politics, in constitutional law, but also in the media. While the rise of social media has the potential to reinvigorate public debate, it certainly leaves less room for a nuanced public conversation. Or, as put in The Economist in 2017: ‘Because people are sucked into a maelstrom of pettiness, scandal, and outrage, they lose sight of what matters to the society they share. This tends to discredit the compromises and subtleties of liberal democracy and to boost the politicians who feed off conspiracy and nativism’.

Secondly, in Germany, as elsewhere, democratic politics builds on a certain level of conformity. In the realm of media, the public broadcasting system was designed as a great leveller, whose function was not only to inform, but to ‘shape’ public discussion and to ‘filter and distribute’ information, as the German Federal Constitutional Court once put it. The German sociologist Andreas Reckwitz has shown how deeply this moment of collectivity was embedded in post‐war Germany and other Western societies until very recently—and how radically late modernity, with its strive for individuality and its emphasis on singularity, breaks with this approach. Now, in stark contrast to traditional media, social media create curated infospheres and thus require and reward singularisation. In political terms, this puts a premium on radical positions.

Thirdly, social media are based on immediacy and break with the idea of mediation. Individuals can reach their followers directly and do not need to rely on traditional gatekeepers, such as the press or broadcasters. When political parties set up their own newsroom to address the general public, traditional media are simply bypassed. For society, this comes at a price, because the decline of traditional media means that the substantive and procedural standards that they adhered to are becoming less relevant, too. Especially for public broadcasters, these ethical and legal standards included the requirement to involve so‐called socially relevant groups (churches, trade unions, NGOs, etc.) in decision‐making processes and to ensure that their programme represents the plurality of opinions represented in society. None of this exists (so far) for social media companies. But without representation, there is no mediation and, again, less moderation.

The Logic and Limits of the Reconfiguration Narrative

It is remarkable that the long‐term political implications of social media have been ignored until very recently. While traditional media are, inter alia, subject to various internal duties of care exactly because of their influence on society, social media were not even treated as media, but instead subsumed under the favourable regime for online intermediaries. This means that while the responsible editor of a newspaper can be personally liable if criminal content is published in her paper, social networks are largely exempted from liability for all illegal activities committed through their services. Even if the differences between traditional and social media are, in many respects, greater than their similarities, the privileges of social media are increasingly difficult to justify in light of their political impact and of the ongoing process of media convergence.

It is therefore no surprise that lawmakers have started recently to consider a tougher stance on social networks. This turn from limited liability to ‘responsibility’, which includes demands for transparency and for stricter liability, has already received widespread attention in the literature. One of the first laws which fully commits itself to the new responsibility paradigm is the German Netzwerkdurchsetzungsgesetz of 2017 (Network Enforcement Act—NEA). According to the explanatory memorandum to the bill: [S]ocial network providers have a responsibility for the culture of debate in society that they have to live up to. In view of the fact that the existing instruments and the voluntary commitments of social network providers are not working sufficiently well and that there are considerable problems in enforcing existing laws, it is necessary to introduce compliance mechanisms for social networks that are subject to fines in order to incentivise them to take effective and immediate action against hate crimes and other criminal content on the Internet. Deutscher Bundestag, ‘Begründung zum Entwurf eines Gesetzes zur Verbesserung der Rechtsdurchsetzung in sozialen Netzwerken’ [Explanatory Memorandum to the Network Enforcement Act], BT‐Drs. 18/12356, 16 May 2017, 11.

Despite its commonsensical aim, the NEA has been widely criticised by legal scholars as being incompatible with Directive 2000/31/EC (E‐Commerce Directive) and Article 5 of the German Basic Law, the constitutional right to free speech. The details of this debate are mostly beyond the scope of this article. However, while many scholars argue that the law goes too far in its attempt to regulate online speech, the real problem with the NEA is that the law goes hardly far enough. Upon a closer look, the NEA is a paradigm case of what I have called the reconfiguration narrative, and its attempt to apply offline rules to the online world fails to address the challenges social media pose for society.

Already the title of the law makes clear that the purpose of the Network Enforcement Act is anything but revolutionary. Rather than introducing new rules for so‐called ‘hate speech’ or ‘fake news’, the NEA only aims to improve the enforcement of the existing laws on illegal speech in social networks. And even in terms of enforcement, the NEA does not take a progressive position. Despite EU safe harbour laws, online intermediaries operating in Germany, including social media companies, have actually never been completely shielded from liability. Rather, they always had to act promptly on complaints of third parties whose rights had been infringed by their users. If they failed to comply, they could be liable for the infringement as if they themselves had posted the content. The NEA now affirms this principle. While the original draft bill had taken a more radical approach and had intended to oblige social networks to ‘take effective measures against new uploads of illegal content’, this provision was removed during parliamentary debate, because critics had pointed out that such an obligation would most likely be incompatible with Article 15 of Directive 2000/31/EC.

The only really new thing about the NEA is that the Act introduces additional procedural and organisational obligations for intermediaries. Section NEA obliges social network providers to ‘produce and publish half‐yearly German‐language reports on the handling of complaints about unlawful content on their platforms’. And section NEA requires providers to ‘maintain an effective and transparent procedure for handling complaints about unlawful content’. Section (ii) NEA specifies that content which is manifestly unlawful must be removed or blocked within 24 hours of receiving the complaint; however, the social network can try to reach an agreement with the competent law enforcement authority to extend this period. According to section (iii) NEA (merely) unlawful content must be removed or blocked ‘immediately’, which means generally within seven days of receiving the complaint.

These provisions hopefully incentivise social media companies to make their internal filtering systems both more transparent and more responsive to rights violations. Moreover, in terms of rule of law and democratic accountability, it is laudable that the NEA tries to bind social media companies and their community standards more tightly to the substantive rules for free speech developed by German lawmakers and the Bundesverfassungsgericht, the Federal Constitutional Court (FCC). However, one should not overlook the fact that the NEA effectively sanctions these internal standards, which, as Kate Klonick and others have analysed, are functioning today as powerful ‘systems of governance’ for online speech and which are routinely employed by social media companies to curate user content in order to make their product more attractive and relevant for some users, while excluding and policing others. While this fact alone does not make the NEA unconstitutional, as some have argued, it again demonstrates the limited ambition of the Act, which is entirely based on the reconfiguration logic of ‘what is illegal offline, should also be illegal online’.

This impression is further confirmed when looking at the actual impact of the law. Of course, the law is still rather young. The bill was passed on 30 June 2017 and went into effect on 1 October 2017, giving companies a grace period of three months until 1 January 2018. Therefore, one can only speculate about the long‐term effects of the NEA. However, the reporting requirement of section NEA helps to shed light on its short‐term effects. In July 2018 Google, Facebook and Twitter presented their first transparency reports.

In the first six months of 2018, YouTube received around 215,000 NEA complaints; 58,000 of them were considered to be justified (around 27%). Most cases in which content was deemed to be illegal in the sense of Article 1 NEA were also judged to be in violation of YouTube’s community standards and therefore not only blocked in Germany, but removed completely from the platform. Amongst the grounds for removal were ‘hate speech or political extremism’ and the ‘violation of personality rights or insults’, which were the most important categories. In the same time period, Twitter received 264,000 complaints of which only 29,000 were considered to be justified and blocked or deleted (around 11%). Facebook so far has not integrated the NEA into its existing reporting system, which might be the reason why the company only received 1,704 NEA complaints between January and July 2018, of which only 362 were deleted or blocked (around 21%), because they were illegal within the meaning of the NEA or violated the community standards or both.

When we compare these numbers to the amount of content blocked or deleted by social media because of a violation of internal community standards, we can see that these numbers are far from impressive: Between January and March 2018 YouTube alone deleted all in all 9,790,082 videos; most of them were flagged automatically before anyone could view them. In the same period, Facebook took down several million pieces of content as part of its community standards enforcement process. And it is probably safe to assume that most removals now based on the NEA could also have been based on a violation of community standards as well.

What is even more important is that the law, with its focus on the eradication of specific illegal speech acts, completely ignores the question whether the function of social media as ‘amplifiers’ and as ‘singularisers’ and the decline of traditional media with their commitment to ‘mediation’ trigger a new transformation of the public sphere. In other words, the law provides a highly individualised (and as such only partially effective) answer to a structural and collective problem. Moreover, the law misses the fact that mass media are starting to lose their ‘public service function’ to social media. Again, this does not necessarily make the law unconstitutional. However, it differs from the way transformatory processes of the media landscape were dealt with in the past.

The Transformation of the Public Sphere from A Constitutional Perspective

For some observers, the break of social media with the ideals of mediation, collectivity and moderation indicates a structural incompatibility of the technology with liberal democracy. Social media are then associated with the rise of populism and political polarisation. As the historian Barbara Stollberg‐Rilinger has recently remarked: ‘Supposedly direct affirmation in a virtual Volksgemeinschaft takes the place of argumentative discourse. A certain digital style of communication and a populist style of politics mutually reinforce each other’.

However, as analysed above, the politics of (social) media is more complex. Going back to Habermas’s Strukturwandel, one can read that at the beginning of the twentieth century, the rise of mass media was described in terms which have little to do with the idealised version of a free press which implicitly underlies today’s critique of social media. Mass media, Habermas writes, were feared, because they ‘created a new category of influence, namely a power that, when used manipulatively, robbed the principle of publicity of its innocence. The public sphere, which became both pre‐structured and dominated by the mass media, grew into a political arena, where topics and contributions were not only used to influence others, but also as disguised means to control communication and to change individual behavior’.

Historical analogies must be used carefully. However, the case of mass media demonstrates that each new medium comes with its own strengths and pathologies. Moreover, even if it is true that the structure of the public sphere is a product of the media we use, one should not forget that the media are, to a certain extent, also what we want them to be. This means that, eventually, a potentially destructive transformation process can be put on the right track through regulation. If mass media developed over the course of the twentieth century from disguised means to control the populace into a forum for democratic debate, why should a similar development not be possible with social media?

However, such a project cannot be accomplished solely through laws that follow the logic of ‘what is illegal offline, should also be illegal online’. How a transformatory media technology can be shaped in a way to reflect democratic values and to counteract radicalisation is illustrated by the regulatory response to the rise of broadcasting. In the case of Germany, on which I am concentrating again, this response was initiated in the realm of politics, then put into statutory law and, finally, elevated to the rank of constitutional law.

In post‐war Germany, there was broad political agreement that broadcasting needed to serve a democratic function. After the Volksempfänger had served as an extremely effective propaganda tool for Hitler, the totalitarian potential of the medium was obvious. Leading the way in articulating a vision of broadcasting as a tool for democracy was the military government in the American Zone of Occupation, which held, in an instruction from 1947: ‘It is basic US Military Government Policy that control over the instrumentalities of public opinion such as press and radio must be diffused and kept free from governmental domination’. This instruction reflected the Mayflower doctrine of the US Federal Communications Commission from 1941, which required private broadcasters to ‘provide full and equal opportunity for the presentation to the public all sides of public issues’. In the Mayflower decision, the US Federal Communications Commission had further argued that ‘[r]adio can serve as an instrument of democracy only when devoted to the communication of information and the exchange of ideas fairly and objectively presented. […] Freedom of speech on the radio must be broad enough to provide full and equal opportunity for the presentation to the public of all sides of public issues. The public interest—not the private—is paramount’.

When the re‐organised German Länder started to enact media regulations in the late 1940s, they acknowledged the democratic function of media, too. Article 3 of the Bavarian Press Act states explicitly: ‘The press serves the democratic idea’. This political vision was then translated into a set of statutory norms, including the obligation for journalists to report in an independent and balanced manner. Particularly important, however, were the rules on the internal organisation of public broadcasters. State broadcasting corporations were established with the right of self‐government and broadcasting councils were created in which different social groups were given a seat. The broadcasting council for the state of Hesse, for example, consisted of delegates from churches, workers and employers’ associations, teachers, universities and the state government. Through this form of pluralist representation, the ideals of moderation and mediation were injected into the organisational DNA of the new medium.

This regime was then constitutionalised in arguably one of the most important cases in the German Federal Constitutional Court’s history. The case started in 1960, when the federal government under Chancellor Konrad Adenauer tried to establish a nationwide television programme in Germany by means of a partnership agreement under private law between the Federal Government and the Länder. Because the negotiations with the Länder were unsuccessful, Adenauer’s government went ahead on its own and founded the DeutschlandFernsehenGmbH, a private limited liability company. The Länder challenged this act before the FCC. In its so‐called ‘first television ruling’ of 28 February 1961, the Federal Constitutional Court ruled in favour of the Länder.

The most interesting part of this judgment concerns the interpretation of Article 5 of the Basic Law, which guarantees the ‘freedom of speech’ as well as the ‘freedom of the press and freedom of reporting by means of broadcasts and films’. The Court starts by acknowledging the transformatory power of the new medium: ‘broadcasting is more than just a “medium” for the formation of public opinion; it is an imposing “factor” in the formation of public opinion’. But if broadcasting is an ‘indispensable means of modern mass communication’, an interpretation of the freedom to broadcast only as a negative right against the state seems insufficient. Rather, the Court argues, the freedom to broadcast must be interpreted in light of its function for society. This means that the right protects only such behaviour, which aims to promote public and private communication. Moreover, the right establishes a positive duty of the state to regulate its exercise in a specific rights‐enhancing manner. This Court derives this dialectical concept of a right by interpreting Article 5 of the Basic Law as a guarantee of an ‘institutional freedom’ (institutionelle Freiheit). Leaving aside its complex philosophical tradition, the concept of ‘institutional freedom’ can be traced back to Weimar, in particular to Rudolf Smend’s influential paper at the fourth meeting of the Vereinigung der deutschen Staatsrechtslehrer, in which he had emphasised the ‘social, group‐forming function’ of the freedom of expression. In the 1950s and 1960s, the concept was re‐discovered and re‐interpreted in democratic terms by various scholars, most prominently by Peter Häberle. According to Häberle, every meaningful concept of a right must take into account the social context in which the right is exercised: Freedom is worthless ‘without the factual conditions of its use’. Freedom is therefore never only freedom from, but always also freedom by the law and in the state. Applied to the freedom of speech, this approach—or frame —obviously radically differs from, e.g., the US constitutional tradition.

In its television decision, the FCC then used the concept of an institutional freedom in order to highlight the importance of a pluralist organisation of broadcasters for a democratic society. The core paragraph of the judgment reads as follows: Article 5 of the Basic Law at least requires that this modern instrument for the formation of public opinion is not put at the mercy of the State or one particular social group. Broadcasters of programmes must thus be organized in such a way that all conceivable groups are able to have influence in their organs and have their say in overall programming and that for the contents of overall programming, guiding principles are binding that guarantee a minimum of substantive balance, objectivity and mutual respect. This can only be ensured when these organizational and substantive principles are generally made binding by law. Article 5 of the Basic Law therefore calls for the enactment of such laws. BVerfGE 12, 205, 262.

Through the functional analysis of the fundamental right and the concept of an ‘institutional freedom’, the Court thus elevated the statutory rules on the internal organisation of broadcasters to the rank of constitutional law. Moreover, it bridged the gap between individual rights and democratic accountability. And—this is especially important for the current debate on social media—it did so not only for public broadcasters. Rather, the Court explicitly recognised the possibility of private broadcasting. However, it held that such private enterprises would also have to be organised in a way that ‘offers adequate assurance that all socially relevant forces have their say in a manner similar to that in public corporations and that freedom of reporting remains uninfringed’.

Outlook

There are obvious differences between broadcasting and social media that prevent a simple transfer of regulatory arrangements, which have been developed for the world of radio and television with its linear and hierarchical structure, to the polycentric and dynamic social media. Therefore, the detailed rules that the FCC has derived from the nucleus of ‘institutional freedom’ since 1961 offer limited guidance for those interested in making social media an instrument of democracy.

The point of comparing broadcasting and social media is, rather, that the fundamental change of the media landscape which we are currently experiencing with the rise of social media is neither unprecedented, nor is it advisable to leave this process to the ‘vagaries of societal forces’, as the FCC has put it in one of its later decisions on broadcasting. Rather, without comprehensive regulation that pluralises their internal organisation and sets substantive standards that guarantee balance, objectivity and mutual respect, social media companies will never develop into an ‘instrument of democracy’.

However, while the potentially corrosive effects of social media have been addressed more aggressively in recent politics, statutory law and constitutional doctrine still do not adequately reflect the democratic function of social media. So far, the hands of EU Member States are tied, because EU law, in particular the E‐Commerce Directive, protects social media platforms. In order to enable stricter regulation, the relevant Articles in the Directive need to be amended. Even more important, however, is the constitutional discourse which currently is almost exclusively concerned with the negative rights of users and platform operators against government intervention. In the case of the NEA, critics were quick to accuse the NEA of censorship and predicted that the German FCC would nullify the law because it gave social networks an incentive to ‘over‐block’ user content, which would constitute a violation of the freedom of expression and communication of those users whose content was blocked or deleted. Recently, scholars have started to extend this argument by applying the doctrine of horizontal effect claiming that social media companies as the operators of privatised public spaces (‘public fora’) are ‘indirectly’ bound by fundamental rights.

But while the protection against the state and powerful private actors remains of utmost importance, fundamental rights in the realm of media must not be understood ‘solely as a defense against state influence’, to paraphrase again the FCC; according to the theory of ‘institutional freedom’ we also need a ‘positive system’ which protects against any misuse of a new medium, be it through the state, other social groups or individuals, and which further ensures that media ‘reflect the diversity of subjects and opinions which play a role in society as a whole’.

Looking back to the regulation of broadcasting, one is impressed by the attention that regulators devoted in the post‐war period to institutional, organisational and procedural questions, and by their commitment to the principles of democracy, pluralism and diversity in designing their laws. Society would certainly benefit from social media regulation being equally ambitious and comprehensive today.