Digital punishment: social media exclusion and the constitutionalising role of national courts

ABSTRACT

Today social media represent an essential instrument for exercising a broad range of fundamental rights. The phenomenon of social media exclusion, arising when a user is prevented from using specific social media websites or parts of them, therefore has profound fundamental rights implications. Based on the analysis of recent case law from the US and Germany, the article outlines the essential characteristics of social media exclusion. Preventing individuals from accessing social media does not nullify their constitutional rights. However, without social media one could not enjoy her fundamental rights to such an enhanced level as has become standard in recent years. This article argues that curtailing the possibility of accessing social media should be subject to minimal constitutional safeguards, and examines which measures have been put in place by national courts in the US and Germany to this end. Finally, this article reflects on the role played by national courts from a general perspective, contending that judges represent a main catalyst of the process of constitutionalisation of the social media environment. National courts articulate traditional constitutional principles in the context of social media, in this way solving constitutional collisions between the state constitutional dimension and that of private platforms.

KEYWORDS:

• Digital punishment
• social media exclusion
• constitutionalisation

[P]unishments are just in proportion, as the liberty, preserved by the sovereign, is sacred and valuable. […] [B]y justice I understand nothing more than that bond which is necessary to keep the interest of individuals united, without which men would return to their original state of barbarity. All punishments which exceed the necessity of preserving this bond are in their nature unjust.

— Cesare Beccaria, On Crimes and Punishments, ch 2 (1764)

1. Introduction

Over the past few years, recent technological developments have contributed to binding physical and virtual existence together in a way that, today, for many, these two are complementary and inseparable. The American computer scientist Mark Weiser argued that ‘the most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it’ (Weiser 1991). The recent outbreak of the Covid-19 pandemic has highlighted the crucial role that digital technology can play in the extreme circumstance of a global health crisis, where our freedom of movement is limited and physical social interaction discouraged. We have attempted to compensate for such restrictions by heavily relying on the possibilities offered by digital technologies. The versatility of these tools has provided us with affordable and easy to use ways to work and live an alternative social life.

In this context, social media have played a crucial role, by providing us with an invaluable source of information and allowing us to maintain in contact with family and friends. However, even in more normal times, the use of social media has become an integral part of our daily life (see Karppi 2018; Dowek 2017; cf. Turkle 1995; Markham 1998). Reading news, communicating, searching for a job, professing one's own political or religious faith are all examples of activities that many individuals habitually perform through social media websites.

From a legal perspective, this means that today social media represent a central instrument for exercising a broad range of fundamental rights that are generally protected in most national and supranational constitutional texts, such as the right to freedom of expression, to seek information, to peacefully assemble and to protest. What if one were therefore precluded from accessing social media? What are the consequences for an individual if the state, a social media platform or even another user banned her from a social media website or part of it? What are the fundamental rights implications of this form of digital punishment? Could an individual be said to fully enjoy the right to express herself, seek information, peacefully assemble or protest without having access to social media? These questions are neither academic nor abstract. The phenomenon of social media exclusion, arising when a user is prevented from using specific social media websites or parts of them, is a tangible reality. Over the past few years, various national courts in the US and Germany have attempted to address this issue, focusing in particular on its fundamental rights implications.

This article aims to analyse the consequences of social media exclusion from a fundamental rights perspective, in particular looking at the role played by national courts to ensure minimal constitutional guarantees in this context. Section 2 outlines the essential characteristics of social media exclusion, proposing a categorisation based on the analysis of recent case law emerging on the topic in the US and Germany. Section 3 illustrates the fundamental rights implications of social media exclusion. Strictly speaking, preventing individuals from accessing social media does not nullify their constitutional rights. However, at the same time, without social media one could not enjoy her fundamental rights to such an enhanced level as has become standard in recent years. Consequently, curtailing the possibility of accessing such tools should be subject to minimal constitutional safeguards. Section 4 examines which measures have been put in place by US and German courts to ensure that a balancing of fundamental rights is guaranteed in cases of social media exclusion. Section 5 finally reflects on the role that national courts in general may play in this context. The article contends that judges represent a main catalyst of the process of constitutionalisation of the social media environment. They translate and articulate traditional constitutional principles in the context of social media and, in this way, solve constitutional collisions between the state constitutional dimension and that of online private platforms.

2. Defining social media exclusion

Scholars from various social science disciplines have dealt with the issue of exclusion in the digital world since well before the advent of social media platforms. The concept of ‘digital exclusion’ traditionally denotes the problem of ‘disparities in access’ to information and communication technologies (Wong et al. 2009, 59; cf. Karavas 2010). This notion translates the idea of ‘social exclusion’ in the context of the digital society (see Byrne 2005). Digital exclusion would be the result of an individual's ‘deprivation’ from information and communication technology, the latter intended as one of the ‘necessities of modern life’ (Helsper 2012). Indeed, someone is socially excluded if she cannot access goods or services that are considered as essential in a specific societal context (see Hills, Grand, and Piachaud 2002). Digital technologies, rather than as an essential good per se, are regarded as a necessary instrument to access knowledge stored on the web and to interact with other individuals online. In other words, digital technologies allow individuals to participate in the so-called ‘information society’, the social reality as transformed by the digital revolution (see Wong et al. 2009).

The scholarship has explained the phenomenon of digital exclusion as the result of a multiplicity of interplaying factors. First of all, dynamics of economic nature: such as the ‘lack of affordability’ of information and communication technologies (see Wong et al. 2009), and issues related to the insufficiency of connectivity infrastructures in remote, rural, or simply underdeveloped areas, i.e. the so-called digital divide (see Norris 2001). But also other forms of cultural, social and political disadvantages, like a scarce degree of media literacy, disability, ethnicity and language (Helsper 2012; see also 2017).

This article analyses from a legal perspective a phenomenon that can be regarded as a species of digital exclusion: social media exclusion. The latter occurs when an individual is prevented from accessing social media platforms or parts of them. Social media exclusion represents a narrower concept when compared to digital exclusion. It does not imply that individuals are prevented from accessing digital technology in general, but only refers to cases where people cannot access social media platforms, a specific social media website or single portions of it. In order to make this discussion more concrete, the following sections will present three examples of social media exclusion directly derived from an analysis of existing case law, and will subsequently reflect on the essential characteristics of social media exclusion from a legal perspective.

2.1. Three examples

From an analysis of existing case law, it is possible to distinguish three types of social media exclusion on the basis of the actors and, as will be illustrated in this section, specific underlying factors determining the users’ exclusion: in the first case, social media users were excluded by other social media users; in the second case, by the social media platform itself; and in the third case, by the law.

2.1.1. Users vs users

A first example of social media exclusion that one can find in existing case law occurs when social media users are prevented from accessing specific portions of social media platforms by other users. In the US, this typology of social media exclusion has materialised in cases where critical citizens were prevented from directly interacting on social media with their elected representatives. The most notorious case involved the President of the United States Donald Trump.

Trump is a prolific Twitter user (see Fuchs 2018). The use of Twitter by US presidents is not a novelty. Trump's predecessor, Barack Obama, was considered the first ‘social media president’(Acker and Kriesberg 2017). He tweeted for the first time back in 2007 and, in 2015, he created @POTUS, the official account of the President of the United States (Acker and Kriesberg 2017). Although @POTUS's credentials were handed over to President Trump after his election, the latter continued to use his personal account @RealDonaldTrump to communicate with the general public about his administration. Tweets posted using the @RealDonaldTrump account include not only personal comments, but also statements made by President Trump in his official capacity (Swartzwelder 2018). On multiple occasions, President Trump used his personal Twitter account to convey official news, even before they were officially published or communicated through traditional press channels. Every tweet posted by President Trump is usually retweeted, liked and commented on by thousands of users. Consequently, in 2017, US democratic representative Mike Quigley introduced the COVFEFE Act – intentionally echoing one of Trump's most famous typos, but actually being the acronym for the ‘Communications Over Various Feeds Electronically for Engagement’ Act – a bill proposing to amend the Presidential Records Act 1978 in order to require the US National Archives to store all presidential tweets: therefore not only those published on the official @POTUS account, but also those posted on President Trump's personal account @RealDonaldTrump (Quigley 2017).

Since his inauguration, several individuals, after critically commenting on Trump's tweets, have been ‘blocked’ by the President. ‘Blocking’ is a Twitter feature that allows a user to prevent another user from seeing and commenting on her own tweets. Users who have been blocked by President Trump can still read the President's tweets if they log in on Twitter using a different account or, in a simpler way, without logging in and simply accessing via their web browser, since one's tweets are in principle visible to anyone on a web-browser. Blocked users also remain able to read all the comments to President Trump's tweets and can in turn reply to those comments.

In 2017, seven individuals who had been blocked by President Trump on Twitter, together with the Knight First Amendment Institute at Columbia University brought a claim against Donald Trump, his social media manager and the White House Press Secretary (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018). The first lamented to have been prevented from accessing and commenting on the President's tweets, and the latter not to have had the possibility to read the potential comments that the blocked users would have posted if they had not been prohibited from doing so.

2.1.2. Platforms vs users

A second type of social media exclusion arises where social media users are prevented from accessing a social media website by the platform itself. These cases are common in German case law, where users who verbally abused other users were banned by specific social media providers. In a series of recent German cases, some users saw their account disabled by Facebook for having posted comments which allegedly infringed the terms of use of the platform, and in particular the prohibition of hate speech.

According to Facebook's Community Standards, hate speech falls within the category of ‘Objectionable Content’ (Facebook. ‘Community Standards’. https://www.facebook.com/communitystandards/objectionable_content). It is defined as ‘a direct attack on people based on what we call protected characteristics – race, ethnicity, national origin, religious affiliation, sexual orientation, caste, sex, gender, gender identity, and serious disease or disability’ (Facebook n.d.). Just to provide some examples, in the examined case law, Facebook considered hate speech not only a user's post invoking the use of water cannons and the army against asylum seekers (LG Frankfurt/Main, 10.09.2018 - 2–03 O 310/18 2018), but also posts in which users quoted a statement by the Hungarian Prime Minister Viktor Orbán defining Syrian refugees as Muslim ‘invaders’ (OLG München, 17.07.2018 - 18 W 858/18, n.d.), or classifying a newspaper as ‘pseudo-lefty’, ‘first-class warmonger’ and ‘denigratory rag’ (LG Frankfurt/Main, 14.05.2018 - 2–03 O 182/18 2018).

Similar expressions are not ‘allowed’, and were therefore removed by Facebook, because they generated ‘an environment of intimidation and exclusion and in some cases may promote real-world violence’ (Facebook n.d.). Moreover, in the cases presented above, it was not the first time that the authors of these expressions published similar content. For this reason, their accounts were disabled by Facebook.

2.1.3. States vs users

The third and last example of social media exclusion occurs where national law directly prevents specific individuals from accessing (some) social media platforms. Until 2017, in North Carolina, a national statute prevented registered sex-offenders who committed abuse against minors from using, once released, a broad range of social media platforms (Tolon 2016). The norm aimed to hamper access to a powerful means to identify and contact potential victims for individuals in this category.

In 2010, the local police identified a post on Facebook in which Mr Packingham, a registered sex-offender, manifested his joy after the dismissal of a traffic ticket. For breaching his obligation not to use social media, Mr Packingham was convicted in the first instance, but subsequently succeeded in appeal, arguing that the North Carolina's statute unduly restricted his right to free speech protected under the First Amendment. The case eventually reached the US Supreme Court, where North Carolina's statute was held unconstitutional (Packingham v. North Carolina 2017).

2.2. Essential characters of social media exclusion

The causes of the phenomenon of social media exclusion are different from those identified by social scientists as representing the basis of digital exclusion. Social media exclusion appears to be distinct from the form of digital exclusion deriving from the absence of conditions of net neutrality. Net neutrality is the principle whereby Internet service providers should not discriminate against online content (Nunziato 2009; Marsden 2010; Belli and Filippi 2016; Mueller et al. 2007). In this case, the central issue is represented by operators’ policies which charge more, slow down or even block specific Internet traffic. This behaviour can favour wealthier social groups who can pay for access to selected content and, in this way, exclude those individuals who cannot afford the cost of similar services. Conversely, in the case of social media exclusion, it is the behaviour of the excluded user, and not her economic conditions, that plays a central role. Social media exclusion is uniquely the result of the prior conduct of the excluded user. In the examples just presented, users publicly criticised other users, engaged in hateful comments, or committed crimes (e.g. sexual abuse) that could be facilitated by the use of certain social media.

Apart from the first type of social media exclusion, where users exclude other users, in the second and third example, social media exclusion is a legal – when imposed by law – or quasi-legal effect – when we think of the hate speech cases forbidden by Facebook's Community Standards. Social media exclusion is not a phenomenon as large as digital exclusion, but is nevertheless the product of factors of general applicability. In our first example, Twitter's terms of use allowed President Trump as well as any other user to limit the possibility of interacting on the platforms of other individuals. In the German cases, Facebook banned a series of users, acting on the basis of its general policy. In the Packingham case, a national law imposed a series of restrictions on sex offenders.

In light of these considerations, it is also possible to distinguish the phenomenon of social media exclusion from what the scholarship has called ‘cyberostracism’ (Williams, Cheung, and Choi 2000). The latter denotes situations where individuals are ‘ignored or excluded by peers or groups’, especially in the context of social media (Schneider et al. 2017, 385; cf. Sobczak 2017). In such cases, users are not prevented from accessing social media or parts of them, but only de facto secluded from certain social interactions occurring on platforms. Cyberostracism is a phenomenon that reflects an unpleasant, though widely common, form of behaviour of the offline world. It has certainly social consequences, but does not present serious legal implications. Conversely, the phenomenon of social media exclusion analysed in this article has significant legal repercussions on our fundamental rights.

3. Fundamental rights implications

3.1. Social media as a new tool to exercise fundamental rights

Man is a social animal: from Aristotle to Einstein, many thinkers argued that this is a quintessential characteristic of human beings (Aristotle n.d.; Einstein 1954). Certainly, this circumstance partially explains why in the twenty-first century social media have such a huge success. At its origins, the Web largely reflected a traditional distinction of the offline world: there were, on the one hand, content creators, such as journalists, commentators, writers, etc., and, on the other hand, a wide number of readers-listeners-viewers. The advent of the Internet was revolutionary: it allowed anyone to become a content creator, and an unprecedented amount of people across the world to access such resources. Nevertheless, the original distinction between producers and consumers of content essentially subsisted.

What the advent of social media changed, leading in this way to the so-called Web 2.0, was precisely such a clear separation of roles (Fuchs 2010; Obar and Wildman 2015). For the first time, the readers-listeners-users could also easily become creators of web content. There was no longer a dividing line between producers and consumers, but a combination of the two: many ‘prosumers’ creating and, at the same time, consuming their own content (Ritzer and Jurgenson 2010). However, social media are not merely the results of ‘uroboric’ practices of their consumers, thousands of users who publish and at the same time read web content in an unending circle, such as snakes swallowing their own tails. The real richness of social media that is now contributing to the development of what is generally called Web 3.0 lies in their being social spaces, technology-based networks that through the interaction and cooperation of users create new knowledge (Fuchs 2010; Barassi and Treré 2012; Obar and Wildman 2015). Content published online is not merely the tail of the user, but the result of a relationship, the produce of an exchange among individuals. This is probably the most ‘social’ aspect of these media: they are not only webpages created by ordinary individuals, but also social spaces of virtual interaction.

Regarding this phenomenon from a legal perspective, it is possible to affirm that social media represent an extraordinary new tool allowing people to exercise a broad range of fundamental rights rotating around the exchange of information. Reading news, communicating, publishing content, disseminating religious or political beliefs, organising online assemblies or protests are only a few examples that show how today social media can be used to exercise our right to information, freedom of expression, freedom of religion and opinion, freedom of assembly and right to strike.

3.2. Essentiality of social media

However, today social media are not only one of the many tools through which we can exercise our fundamental rights, but have progressively become a crucial instrument for fully enjoying our essential freedoms in the digital age. Today, the boundaries between physical and virtual life are no longer clearly demarked. Physical life expands into the virtual realm, so that a physical existence cannot be considered as such without its virtual expansion. Social media have become an integral component of our social life (see Karppi 2018; Dowek 2017). We could not imagine exercising some of our fundamental rights, including our freedom of expression and information, and right to protest, assembly, business and religion without relying on the benefits of easily accessible social media platforms populated by millions of individuals. As this section will illustrate, recent case law has recognised that exercising a broad range of fundamental rights through social media has risen the standard to which individuals can exercise their basic freedoms. Strictly speaking, preventing individuals from accessing social media does not nullify their constitutional rights. One can object that there is always an ‘analogue’ way to exercise them. However, at the same time, without social media one could not enjoy fundamental rights to such an enhanced level as has become standard in recent years.

In the US, the Supreme Court's ruling in Packingham v. North Carolina explicitly recognised this new central role played by social media. Justice Kennedy, echoing the seminal ruling in Reno v. American Civil Liberties Union decided exactly twenty years earlier, not only affirmed that ‘social media offer relatively unlimited, low-cost capacity of communication of all kinds’ (Packingham v. North Carolina 2017, 5), and allow anyone with an Internet connection to become a modern ‘town crier’ (Packingham v. North Carolina 2017, 8), but also represent

what for many are the principal sources for knowing current events, checking ads for employment, speaking and listening in the modern public square, and otherwise exploring the vast realms of human thought and knowledge (Packingham v. North Carolina 2017, 8, emphasis added)

as well as ‘the most powerful mechanisms available to a private citizen to make his or her voice heard’ (Packingham v. North Carolina 2017, 8, emphasis added). Social media represent the ‘parks’ and ‘streets’ of the digital society, the public fora par excellence where every individual can freely express their opinions. Twenty years after Reno, the decision that for the first time identified the quintessential tenets of the relationship between the Internet and First Amendment's rights, the Supreme Court incontrovertibly concluded that:

While in the past there may have been difficulty in identifying the most important places (in a spatial sense) for the exchange of views, today the answer is clear. It is cyberspace – the “vast democratic forums of the Internet” in general, […] and social media in particular (Packingham v. North Carolina 2017, 4, emphasis added).

The recent German case law seems to be in line with the position expressed by the US Supreme Court. German judges too explicitly recognised that Facebook today provides a ‘public space for information and the exchange of views’ (OLG München, 17.07.2018 - 18 W 858/18, n.d.; OLG München, 24.08.2018 - 18 W 1294/18 2018; LG Frankfurt/Main, 10.09.2018 - 2–03 O 310/18 2018).

3.3. Constitutional relevance of social media exclusion

Nowadays, social media have become constitutionally relevant because they are essential to the exercise of a series of fundamental rights. They represent a vital, integral component of the contemporary public space (Heldt 2020). Notwithstanding their private nature, social media are the square, the public garden, the streets of the twenty-first century (cf. Fuchs 2014). They embody the space for social interaction of billions of individuals, as never a physical space could do. Such a prominent position that social media have acquired vis-à-vis fundamental rights has a series of legal, and more precisely, constitutional implications.

Firstly, theoretically speaking, one could infer that, by recognising social media as an essential instrument of exercising fundamental rights in the digital society, a right to social media should be implicitly recognised (Celeste 2018). If today social media have risen the standard to which individuals can exercise their basic freedoms, one should have the right to enjoy their rights at such an improved level. A very similar debate emerged in relation to the possibility of recognising Internet access as a fundamental right (Tully 2014; Frosini 2013; Hert and Kloza 2012). As in that case, the most extreme corollary of the recognition of a right to social media would imply the creation of a social right, meaning a positive right to obtain from the state all necessary technical, financial and educative resources to access social media, and at the same time a positive obligation imposed on social media platforms to admit any users (cf. Hartmann 2013).

However, this solution was not considered in the analysed case law. US and German courts never recognised a social right to social media. Not only from a substantive perspective, because of the apparent political, social and economic implications of mandating their respective governments and online social media platforms to ensure access to social media. But also due to the constitutional architecture of these jurisdictions: indeed, both in the US and in Germany, federal courts limit themselves to monitor legislative action rather than interpreting their respective constitutions to dictate the legislative agenda of their own states (Tushnet 2009; Badura 2015; Hartmann 2013).

In the US and Germany, the essentiality of social media platforms in the contemporary architecture supporting the enjoyment of fundamental rights did not lead to imposing positive obligations on public and private actors. Nevertheless, national courts explicitly recognised the constitutional relevance of social media exclusion. Preventing individuals from using social media platforms or part of them restricts their capability to fully enjoy their fundamental rights. For this reason, the possibilities of imposing social media exclusion should be clearly limited according to constitutional principles. Over the past few years, national courts have progressively affirmed a series of basic guarantees, constitutionalising a space mostly subject to the internal rules of private platforms.

4. Constitutionalising social media exclusion

4.1. Obligations of special users

Recent US case law has imposed a series of obligations where phenomena of social media exclusion involved accounts of special categories of users, in particular politicians. The importance of introducing constitutional safeguards in these circumstances is apparent in light of the implications for other civil and political rights. Excluding individuals from accessing and interacting with content published by politicians on social media significantly restricts the enhanced possibility that online platforms offer citizens for engaging in the democratic life of their country.

In the Trump case, where the US President blocked a series of opposers on Twitter, the type of social media exclusion was relatively limited. Blocked users were prevented from directly reading and engaging with the President's account using their own Twitter profiles. The legal question at the core of this case, as summarised by US District Judge Naomi Reice Buchwald at the outset of her decision delivered on 23th May 2018, was ‘whether a public official may, consistent with the First Amendment, “block” a person from his Twitter account in response to the political views that person has expressed […]’ (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018, 1). After finding that the blocked individuals wanted to engage in political speech, a form of expression protected by the First Amendment, and after demonstrating that Twitter's interactive space could be considered as a ‘public forum’, the district judge concluded that President Trump was not entitled to block any users without violating their constitutional right to free speech (see Roberts 2019; Berry 2018; cf. Beausoleil n.d.; see also West 2018).

A central part of the judgment was devoted to ascertaining whether President Trump's Twitter account represents a public space susceptible to being controlled by the government and, as such, triggers the protection of the First Amendment (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018, 39 ff). Such a provision prohibits the government from restricting citizens’ freedom of speech. To verify the applicability of the First Amendment, US courts have developed the so-called ‘forum doctrine’, which aims to verify whether the speech under consideration took place in a space owned or controlled by the government (see Berry 2018; Briggs 2018). In Knight First Amendment Institute v. Trump, the district judge had to apply this test in the new context of social media platforms (see Briggs 2018; Berry 2018; cf. Beausoleil n.d.). Although Twitter is privately owned, Judge Buchwald recognised that the account @RealDonaldTrump is de facto controlled by the President of the United States and his social media staff. The district judge did not take the whole Twitter platform into consideration, although recalling that the Supreme Court had affirmed that social media represent one of the most important spaces for the exchange of opinions in Packingham (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018, 60; see also Packingham v. North Carolina 2017). The users blocked by President Trump were only banned from accessing a tiny part of such a social media platform, namely the President's tweets. Furthermore, President Trump's tweets manifestly represented a form of ‘government speech’, which is explicitly excluded from the scope of the First Amendment. The district judge eventually managed to isolate the virtual space in which users can interact with President Trump's tweets as the public forum to take into account. Such space had been designated as a forum for public interaction by the President himself (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018, 55 ff; cf. Smolla 2018). Judge Buchwald was not able to identify any ‘compelling state interest’ justifying a restriction of free speech rights in the interactive space of the account @RealDonaldTrump, and concluded that blocking users amounted to ‘viewpoint discrimination’ which is not susceptible to justification in any case (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018, 63).

Trump's counsel objected that in this way the President would see his First Amendment rights to select his interlocutors curtailed (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018, 64). The district judge held that President Trump does not have a duty to answer under the First Amendment, but conversely maintains, in his official capacity too, a ‘right to ignore’ (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018, 65 ff). The latter, however, would not be exercised by blocking users, but simply by ‘muting’ them. Muting is another Twitter feature enabling a user not to see another user's tweets or notifications. Judge Buchwald cleverly observed that by replying to a Tweet, a user does not only address the author of the original Tweet, but also the other users who have previously interacted with that Tweet. Muting users would be the only solution for a public official who wants to ignore critical replies without infringing third parties’ First Amendment rights (Knight First Amendment Institute at Columbia University et al. v. Donald J. Trump et al. 2018, 66).

In January 2019, the US Court of Appeals for the Fourth Circuit decided the case Davison v. Randall, where Mr Randall, the chair of a county school board, removed a series of critical comments and allegations of corruption, and banned Mr Davison for twelve hours from his official Facebook page (Davison v. Randall 2019). Mr Randall was found to have acted unconstitutionally by operating viewpoint discrimination on a public forum. Indeed, Randall's official Facebook page was considered to manifestly bear all main characters of a public forum, having Randall himself expressly invited his constituents to engage with it for any queries or comments. The Court of Appeals referred to the District Court decisions in Knight First Amendment Institute v. Trump and in Morgan v. Bevin to highlight that the case law did not agree as to whether social media pages of politicians should be considered as public fora. In the latter decision (Morgan v. Bevin 2018), a number of users lamented of having being blocked on Twitter by the Governor of Kentucky, Matt Bevin, and banned from his Facebook page. District judge Gregory Van Tatenhove argued that Bevin's social accounts did not amount to public fora, being written in his personal capacity (Morgan v. Bevin 2018, 1011). Policymakers would not have any obligation to listen to their citizens as citizens would not have any right to be heard (Morgan v. Bevin 2018, 1011). Moreover, from a subjective point of view, Judge Van Tatenhove observed that Bevin never wanted to set up a public forum, a space for discussion through his social media account. The Governor of Kentucky conversely intended to disseminate his views on some online private platforms, and was fully entitled to ban or block other social media users by the same rules of these platforms (Morgan v. Bevin 2018, 1011). Remarkably, the District Judge also argued that, if the contrary were true, Governor Bevin would not have the possibility of offering the public image he prefers on social media (Morgan v. Bevin 2018, 1012). On this point, it is interesting to note that the District Court for the Eastern Virginia District in Davison v. Randall did not infer that a public official should be prevented from moderating content and users on her social media page (Davison v. Loudoun County Board of Supervisors et al. 2017). Conversely, the District Court stressed that ‘a degree of moderation is necessary to preserve social media websites as useful forums for the exchange of ideas’, and that, consequently, ‘neutral, comprehensive social media policies […] may provide vital guidance for public officials and commenters alike in navigating the First Amendment pitfalls of this “protean” and “revolution[ary],” Packingham, 137 S. Ct. at 1736, forum for speech’ (Davison v. Loudoun County Board of Supervisors et al. 2017, 31).

In July 2019, the decision of the District Court in Knight First Amendment Institute v. Trump was upheld by the Court of Appeals for the Second Circuit (Knight First Amendment Institute v. Trump 2019). In August 2020, President Trump requested a writ of certiorari from the Supreme Court (Trump v. Knight First Amendment Institute (Petition for a writ of certiorari) 2020). In the meantime, in July 2020 the Knight First Amendment Institute at Columbia University sued the US President again, lamenting that @RealDonaldTrump continues blocking individuals who were blocked before Trump's inauguration as US president and those who cannot demonstrate that they were blocked on the basis of their political views (Knight First Amendment Institute at Columbia University v. Trump (complaint for declaratory and injunctive relief) 2020). In this way, President Trump offers an unreasonably restrictive interpretation of the District Court decision, affirming that the President's blocking of users amounts to viewpoint discrimination prohibited by the First Amendment.

4.2. Constitutional guarantees of digital penalties

In the German case law, national courts dealt with the blocking or disabling of users’ accounts operated by Facebook. This form of social media exclusion has more extensive consequences in comparison with the case of blocking operated by another platform user, and usually represents an ancillary penalty imposed by social media in case of recidivism in the publication of contents violating the platform's policy. Although banning and content removal can be regarded as two forms of censorship and are often analysed together by courts (see Heins 2014; Gillespie 2018; Langvardt 2018), this section will exclusively focus on the issue of platforms banning users as a form of social media exclusion.

The German case law affirmed that social media have, in principle, a right to ban users from their platforms by virtue of their ‘virtuelles Hausrecht’, their right as ‘digital householder’ to decide who enters their property and who has to leave. This right was first recognised for the owner of a chat forum in a case decided by the Regional Court of Bonn in 1999 (LG Bonn, 16.11.1999 - 10 O 457/99 1999; OLG Köln, 25.08.2000 - 19 U 2/00, n.d.; see Ladeur 2008; see also Schwenke 2012). German courts acknowledged this prerogative of the platform both in cases of a perpetual ban – a sort of social media ‘death penalty’ (see LG München I, 25.10.2006 - 30 O 11973/05 2006; AG Hamburg, 11.09.2012 - 18b C 389/11 2012; AG Kerpen, 10.04.2017 - 102 C 297/16 2017) – and in cases of temporary suspension of all, or some, of the user's account functions (see LG Frankfurt/Main, 14.05.2018 - 2–03 O 182/18 2018; OLG München, 17.07.2018 - 18 W 858/18, n.d.; OLG Dresden, 08.08.2018 - 4 W 577/18 2018; OLG München, 24.08.2018 - 18 W 1294/18 2018; LG Frankfurt/Main, 10.09.2018 - 2–03 O 310/18 2018). German judges held that the right of the digital householder can be exercised not only when a user posts abusive content which is not constitutionally protected or illegal content according to the German criminal code – as demanded, for instance, by national legislation on online content moderation which is currently in force in the country, the so called German Network Enforcement Act 2017 (see Schulz 2018; Spindler 2017) – but also when the user's post infringes the platform's terms of use (see OLG Dresden, 08.08.2018 - 4 W 577/18 2018; LG Frankfurt/Main, 14.05.2018 - 2–03 O 182/18 2018; LG Frankfurt/Main, 10.09.2018 - 2–03 O 310/18 2018). In this way, social media platforms are entitled to ban users on the basis of standards regulating freedom of expression that could be stricter than those established by state law. However, such a right of the virtual householder to ban users is not absolute, being subject to a series of limitations which have been progressively established by the German case law.

First of all, German courts affirmed that the decision of a platform to ban a user should be objectively justified and not arbitrary (LG Bonn, 16.11.1999 - 10 O 457/99 1999; AG Hamburg, 11.09.2012 - 18b C 389/11 2012; cf. LG München I, 25.10.2006 - 30 O 11973/05 2006; OLG München, 17.07.2018 - 18 W 858/18, n.d.; OLG München, 24.08.2018 - 18 W 1294/18 2018). The case law not only stressed the importance of the presence of rules, and their transparency, but also emphasised that platforms’ policies on banning users should be based on objective criteria (OLG Dresden, 08.08.2018 - 4 W 577/18 2018). In two decisions, the Higher Regional Court of Munich held, for example, more specifically in relation to the removal of content, that Clause 5.2 of Facebook's ‘Statement of Rights and Responsibilities’, the platform's terms of service until the entry into force of the General Data Protection Regulation, was ineffective because it disproportionately disadvantaged users by entrusting to Facebook a unilateral, and almost unrestricted, right to remove their posts (OLG München, 17.07.2018 - 18 W 858/18, n.d.; OLG München, 24.08.2018 - 18 W 1294/18 2018). Clause 5.2 indeed read:

We can remove any content or information you post on Facebook if we believe that it violates this Statement or our policies.

The Munich court argued that this provision granted Facebook too wide a discretion in determining whether a post should be removed. Conversely, a subsequent decision by the Higher Regional Court of Dresden recognised that Facebook's new Terms of Service, as updated after the entry into force of the General Data Protection Regulation, were no longer characterised by such an arbitrary nature as they unambiguously conditioned the removal of content on the infringement of the rules on prohibited content set out in the terms (OLG Dresden, 08.08.2018 - 4 W 577/18 2018).

Secondly, German courts established the criteria of interpretation that a platform should adopt to assess the conformity of a post with its own policies and, consequently the need to adopt further measures, such as the suspension or deactivation of a user's account (see OLG München, 24.08.2018 - 18 W 1294/18 2018). When interpreting a user's post, platforms should take into account the general and objective meaning that an unbiased, informed, but at the same time average, reader would give in that specific context (OLG München, 24.08.2018 - 18 W 1294/18 2018, para. 48 ff).

Thirdly, German courts established that a user's ban should not be automatic, as it could be if it automatically followed the removal of a post due to an ‘unchecked complaint of another user’ (OLG Dresden, 08.08.2018 - 4 W 577/18 2018). The Regional Court of Frankfurt am Main, however, was still hesitant in determining whether a user should have a right to be heard or warned before being banned (LG Frankfurt/Main, 14.05.2018 - 2–03 O 182/18 2018).

Fourthly, the German case law held that platforms’ ‘sanctions’, including banning the users, should be proportionate, and in particular should take into account the ‘seriousness of the offence and the past conduct of the person’ (OLG Dresden, 08.08.2018 - 4 W 577/18 2018). The High Regional Court of Dresden expressed a positive view on Clause 3.2 (now 4.2) of Facebook's new Terms of Service which establishes a series of gradual sanctions for the publication of prohibited content by a user, ranging from a warning to the partial/temporary or total/definitive disabling of the account (OLG Dresden, 08.08.2018 - 4 W 577/18 2018).

Fifthly, German courts therefore seem to share the view that the definitive ban of the user from a platform is not merely justified by a single violation of the platform's code of conduct, but is admissible in cases of repeated infringements (LG München I, 25.10.2006 - 30 O 11973/05 2006; LG Frankfurt/Main, 10.09.2018 - 2–03 O 310/18 2018; cf. LG Frankfurt/Main, 14.05.2018 - 2–03 O 182/18 2018).

4.3 Proportionality of national law

In the first instance, Mr Packingham was condemned for breaching his obligation not to use social media, as required by the law of North Carolina (Packingham v. North Carolina 2017). He subsequently succeeded in appeal, arguing that the North Carolina's statute unduly restricted his right to free speech protected under the First Amendment. The case was eventually brought before the US Supreme Court, after that the North Carolina Supreme Court reversed the decision of the Court of Appeal. In 2017, the US highest federal court unanimously held that North Carolina's statute was unconstitutional (see Butler 2017; Miller 2018; Burnette-McGrath 2019). Even if the protection of minors is considered a valid governmental interest, the ‘unprecedented’ restriction of the scope of free speech engendered by the prohibition of using social media was deemed unacceptable (Packingham v. North Carolina 2017, 8). North Carolina's statute was not sufficiently ‘narrowly tailored to serve a significant governmental interest’ (Packingham v. North Carolina 2017, 6).

In Packingham v. North Carolina, the US Supreme Court did not recognise a new right to social media as some journalists wrote (see Kravets 2017; Lapowski 2017). However, the Court acknowledged the central role of social media in exercising fundamental rights by requiring that any restriction of the use of these platforms be proportionate to the aim pursued. Imposing forms of social media exclusion is therefore admissible in principle, especially when justified by the protection of another fundamental right, such as the protection of minors. However, the respect of the constitutional principle of proportionality is essential when assessing the legitimacy of social media exclusion imposed by national law.

5. The role of national courts

The analysis of the case law on social media exclusion shows that national courts are seeking to protect fundamental rights in the novel context of social media. Constitutional rights rotating around the freedom of information and expression as well as principles, such as proportionality, the rule of law, and procedural guarantees are reinterpreted and articulated so as to fit the social media environment. In this sense, one can talk of a process of constitutionalisation of the social media environment (Celeste 2019b, 2019a). The next two sections will clarify to what extent national courts play a role in this context and will investigate the tools which so far have been utilised by judges in their decisions.

5.1. Bridging constitutional dimensions

A comprehensive view of the analysed case law on social media exclusion shows that national courts are instilling a series of traditional constitutional guarantees in the social media environment. The Packingham case presents the most traditional dynamic in terms of constitutional adjudication (cf. Prakash and Yoo 2003). The US Supreme Court had to assess the validity of a national statute vis-à-vis a series of parameters established by the US Constitution and case law, such as the right to free speech and the legitimate interest of the state to protect minors. Here constitutional principles collided with a balancing of values reflected in national legislation.

If Packingham still reflects a traditional scheme of constitutional adjudication – constitutional law vs national law, the Trump cases as well as the German case law interestingly show a novel pattern: national courts had to bridge two constitutional dimensions. State constitutional principles faced internal constitutional principles of private platforms. In this context, national constitutional law is no longer articulated to assess the validity of a statute. National constitutional law collides and is balanced against the internal constitutions of social media, as established in their terms of use (see Teubner 2012; Celeste 2019a). Courts recognise the validity of these private constitutions in so far as they do not affect basic fundamental rights (Teubner 2018). If this happens, courts generalise and re-specify traditional constitutional principles in the context of social media (see Celeste 2019b). Constitutionalisation by national courts in this context does not merely represent a transplant of updated constitutional values in a space which is completely deprived of them. They solve a series of constitutional collisions (see Teubner 2012; Celeste 2019a). National courts operate a balance between the legitimate interest of private platforms to dictate their internal constitutional rules and the necessity to ensure that traditional constitutional values are not completely sacrificed in the biggest, privately-owned public square of contemporary society.

The essential public role played by online platforms vis-à-vis fundamental rights represents the core argument justifying the constitutionalisation of the social media environment. In principle, private spaces should not offer the same level of constitutional guarantees as public spaces (see Clapham 2006; Cherednychenko 2007). They are regulated by instruments of private law, such as contracts. Contexts such as families, private clubs and associations may well impose internal rules restricting fundamental rights (see Shelley v. Kraemer 1948). However, those private spaces do not exercise any dominant, essential role on our existence. Everyone is in principle free to leave her family or to join another club or association. Conversely, social media today represent a significant component of – not to say they monopolise – our space of public interaction online. Some German decisions referred to the ‘dominant’ and ‘quasi-monopolistic’ position held by Facebook as public communication space (OLG München, 17.07.2018 - 18 W 858/18, n.d.; OLG Dresden, 08.08.2018 - 4 W 577/18 2018; LG Frankfurt/Main, 10.09.2018 - 2–03 O 310/18 2018). In light of this circumstance, the Higher Regional Court of Dresden rejected the opinion of the lower court which had claimed that banned users would have multiple alternatives to exercise their freedom of expression, and instead observed that Facebook, notwithstanding its private nature, acts as a platform for public communication and performs ‘functions previously assigned to the state as a public service’ (OLG Dresden, 08.08.2018 - 4 W 577/18 2018; cf. OLG München, 24.08.2018 - 18 W 1294/18 2018).

Hence the need for national courts to interfere with the internal rules of these private platforms. National judges correct or direct the development of platform law in light of long-established fundamental rights. One could argue that they represent one of the main catalysers of the process of constitutionalisation of the social media environment. Judges generally anticipate the intervention of slow legislators (see Pollicino 2019; The White House 2020), and trigger or simply reinforce, in a ‘maieutic way’ (see Karavas 2010), instances of internal self-reflection and transformation at the level of platforms themselves, such as the recent creation by Facebook of an Oversight Board, which should act as an in-house adjudicating body responsible for complex cases of online content moderation (see Klonick 2019; Douek 2019). Nevertheless, one should not underestimate the obstacles that national courts encounter in this task. Mechanisms of constitutional adjudication are often not made for assessing the validity of internal rules of private actors. Depending on the national legal system, judges have adopted different tools to bypass such hurdles.

5.2. Instruments of constitutionalisation

Since the Middle Ages, legal scholars have fuelled the idea, allegedly derived from Roman law, that ‘dominium est ius utendi et abutendi re sua’ (Piccinelli 1980). The right of the owner (dominium) is not limited to use her property (ius utendi), but even to misuse and abuse it (ius abutendi). This conception of property is reflected across a plurality of legal systems. In Common law, ‘a man's house is his castle and each man's home is his safest refuge’ (Sir Edward Coke, The Third Part of the Institutes of the Laws of England (1628) ch. 73; in Ratcliffe 2018). In the previous section, we have seen how German courts have even translated the idea of Hausrecht in the right of the online platform owner to ban users.

This conception of property as a ‘terrible right’ (Rodotà 2013) apparently represents an insurmountable barrier to constitutionalisation. Nevertheless, national courts have successfully used new and traditional tools in order to instil fundamental constitutional principles in a space that would otherwise be subject to the capability of platforms to rule their private fiefs arbitrarily.

5.2.1. Constitutional law

In the US, the relevance of the right to property intended as an exclusive prerogative of the owner is reflected at constitutional level in the state action doctrine. The US Constitution merely binds state actors, and cannot interfere with the actions of private individuals and entities (see Kay 1993; Berman 2000).

The Packingham case reflects a traditional scheme of constitutional adjudication. Social media exclusion was imposed by means of national statute. Consequently, falling within the category of state action, this measure could be directly assessed against the constitutional parameters. The Trump case conversely presents an example of constitutional collision with the rules of a privately-owned platform. As seen, in this case, US courts managed to bypass the state action doctrine by underlying the public forum function of the interactive space below President Trump's tweets. The legitimate rules of a private entity, Twitter, were bent to satisfy a fuller enjoyment of the principle of free speech, enshrined in the First Amendment of the US Constitution. To do so, another mechanism developed by constitutional law, the public forum doctrine, was put in motion (see Briggs 2018; Heldt 2020). As highlighted by the US Court of Appeal for the Fourth Circuit in Davison v. Randall, the public forum doctrine applies regardless of the private nature of the forum (Davison v. Randall 2019). Even private property, if ‘dedicated to public use’, can be considered a public forum (Cornelius v. NAACP 1985, 801). And such public use is evident when a public official maintains ‘substantial control’ over a social media account (Davison v. Randall 2019, 25).

In Knight First Amendment Institute v. Trump and in Davison v. Randall, therefore, the constitutional doctrine of public forum was sufficient to bring a privately-owned online platform and its rules within the scope of constitutional adjudication. In the US, arguments based on private law considerations were only tangentially mentioned in Davison v. Randall. Circuit Judge Wynn wrote that, although Facebook's platform is the property of the homonymous incorporated company, according to its Terms of Service its users own the content they produce. Following this argument, one may contend that pages created by politicians represent intellectual property generated by public officials in the course of their employment, which are therefore owned by the government by virtue of their employment contract (Davison v. Randall 2019, 24). Nevertheless, Judge Wynn himself explicitly recognised that for the purpose of applying the public forum doctrine it is not necessary to definitively ascertain the ownership of these property rights.

5.2.2. Private law

German judges adopted two ways of justifying the platform's right to ban users in their capacity as digital householder, as well as the existence of a series of limitations to this right. In contrast to US judges, especially in the oldest decisions, German courts based such a right on private law principles (see LG Bonn, 16.11.1999 - 10 O 457/99 1999; Ladeur 2008). The owner of the platform is considered as a digital householder, who has in principle the right to regulate access to the platform. However, if the owner decides to open the platform to the general public, without individually vetting each user/entrant, then she may no longer arbitrarily prevent access or ban any users without infringing the prohibition of contradictory behaviour deriving from the principle of good faith established by the German Civil Code (§ 242 BGB, Leistung nach Treu und Glauben; see Karavas 2010).

According to German case law, the owner's opening of the platform would represent an offerta ad incertas personas, a binding offer to an indefinite group of people, and not merely an invitatio ad offerendum, an invitation to tender (see AG Kerpen, 10.04.2017 - 102 C 297/16 2017; cf. hiQ Labs, Inc. v. LinkedIn Corp., No. 17–16783 (9th Cir. 2019)). Consequently, when the user enters/accesses the virtual house, she concludes a contract, accepting the platform's offer including its explicit and implied terms. Only by abiding by these terms would the owner remain entitled to ban a user.

5.2.3. Third party effect

In Germany, more recent decisions, in addition to explanations based purely on private law, applied the doctrine of the indirect third-party effect of fundamental rights, the so-called mittelbare Drittwirkung (see Engle 2009; Karavas and Teubner 2005; cf. Hudson 2018). In Germany, fundamental rights not only bind public actors, but represent objective parameters capable of orienting all areas of law, including private law (see OLG Dresden, 08.08.2018 - 4 W 577/18 2018). Fundamental rights, therefore, must be taken into account when interpreting the effects of private platforms’ terms of service. This theory does not aim to minimise the level of interference of single rights, but rather to operate a balancing between the different rights involved (see OLG Dresden, 08.08.2018 - 4 W 577/18 2018).

In cases of platforms banning users, German courts recognised that the removed posts which triggered the suspension or disabling of the user's account, even if they allegedly breached the platform's terms of service, did not represent illegal speech, and were therefore in principle protected by the German constitution (see LG Frankfurt/Main, 14.05.2018 - 2–03 O 182/18 2018). This circumstance had to be reconciled with the right that the owner has as a virtual householder to decide if, who, when and how a user can access the platform. German courts concluded that these two rights can both be protected if the user cannot be arbitrarily banned from the platform, while at the same time the owner reserves the right to ban the user in specific circumstances in order to protect her legitimate interests (see OLG Dresden, 08.08.2018 - 4 W 577/18 2018; LG Frankfurt/Main, 10.09.2018 - 2–03 O 310/18 2018).

This approach of German courts, which is more sensitive to fundamental rights issues in the recent case law, can be explained by an increased understanding of today's centrality of social media in exercising our fundamental rights, and appears to be aligned with the position adopted by the US Supreme Court in Packingham.

6. Conclusion

There is a mobile and protean virtual border emerging in the social media environment, crossing both its outskirts and core. It is erected by different actors according to mutable criteria. It is not static or rigid, it creeps, it is invisible, but factually exists and is definitively tangible, concretely barring some individuals from engaging with a significant part of today's social life. Social media exclusion is a phenomenon with significant implications for our fundamental rights. Today social media allow individuals to exercise their essential freedoms in a way never experienced before. Preventing users from accessing specific social media websites or parts of them means restricting their possibility to fully exercise their fundamental rights. Social media exclusion is a form of digital punishment, which, although virtual, concretely affects individual freedoms.

The analysis of US and German case law on social media exclusion confirms a central role played by national courts in constitutionalising the social media environment. National judges are the joining ring between two constitutional dimensions: state constitutional law and the internal constitutions of online platforms. The private space of social media, apparently ‘lawless’ from a state perspective (Suzor 2019), becomes constitutionally relevant because of the social, public implications of its use. National courts have established that internal rules of social media platforms, such as those related to the possibility of blocking or banning users, should be balanced with freedom of expression and cognate rights, and should respect the minimal constitutional principles proper of criminal law, such as legality and proportionality. National courts are progressively recognising the constitutional relevance of social media in contemporary society. They consequently reinterpret and translate traditional constitutional principles in light of the mutated context of online social platforms (see Celeste 2019b).

National courts serve as a catalyser of the process of constitutionalisation of the social media environment. They do not act alone. Legislators, although slowly, are intervening in tackling issues such as disinformation and prohibited forms of speech. The platforms themselves are introducing constitutional principles and mechanisms, especially in the context of content moderation, as Facebook's Oversight Board demonstrates. Nevertheless, national courts play a ‘maieutic’ role in this context (Karavas 2010). They generate a virtuous circle by socratically stimulating and directing the process of constitutionalisation of the social media environment prompted by the other actors. Judges solve constitutional collisions between state constitutional law and the private constitutions of online platforms, in this way offering a path to follow in constitutionalising social media.

National courts did not infer a social right to social media from the essential role that these platforms today play in contemporary society. No positive obligations were imposed on states and online platforms to allow individuals to access social media websites. Nevertheless, the constitutional relevance of social media was recognised in the necessity of establishing precise boundaries for the phenomenon of social media exclusion. It is disproportionate for states to prevent individuals from accessing a plurality of social media. Public actors, and especially politicians or government representatives, cannot block or ban users from their social media accounts according to their political orientation, even if they are entitled to do so according to the rules of the platform. In cases of content moderation performed by both users and platforms, clear and proportionate rules should be in place to ensure that individual rights are not violated. Social media exclusion is a form of digital punishment with tangible consequences on our fundamental freedoms. Beccaria's quote about the justness and proportion of punishment is also, and especially, valid in the realms of private online platforms.

Acknowledgements

I presented an earlier version of this paper at the annual BILETA Conference in Belfast. I am thankful to all those who attended my presentation for their stimulating questions. In no particular order, I would like to thank my PhD supervisor Dr TJ McIntyre, my colleagues Clara Iglesias Keller, Amélie Heldt and Stefano Rossa, my research assistant Sorcha Montgomery, the anonymous reviewers of IRLCT, and the editors of this special issue for their ideas, helpful comments, constructive criticism and constant encouragement.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author gratefully acknowledges the support of the Irish Research Council Government of Ireland Postgraduate Scholarship.