Beyond Oversight. Advancing Societal Constitutionalism in the Age of Surveillance Capitalism

— Angelo Jr Golia, Senior Research Fellow, Max Planck Institute for Public Comparative Law and International Law, Heidelberg.

Facebook’s Oversight Board (OB) has sparked great interest in an already rich debate over the constitutionalisation digital spaces. The establishment of a non-state adjudicator with jurisdiction over the freedom of speech exercised on FB (2.8 billion monthly active users worldwide) seems to mark a step towards a more advanced digital constitutionalism, and the publication of the OB’s first decisions has only fostered this trend. Lorenzo Gradoni has recently highlighted the remarkable autonomy of the OB toward its own creator. Despite the relative silence of FB’s Community Standards and OB’s Bylaws, the OB gave relevance to human rights law (HRL). He refers to these decisions as the OB’s Marbury v. Madison, and recalls the theory of societal constitutionalism (SC) to argue that it performs a constitutional review ‘in full swing’. I would like to expand on this last point, and use SC’ framework to argue how the focus on content moderation decisions may distract from more fundamental potential harms that come about from FB.

Societal constitutionalism and Facebook: a love story?

SC has been often used to frame forms of normativity in the regulatory spaces opened by new technologies. FB’s processes have gone through a juridification (structural coupling between a social sub-system and its own law), and a stable hierarchy has emerged between primary and secondary norms à la Hart (Terms of Service and Community Standards v. established practices, decisions and codes implemented by both algorithms and human editors). A divide has emerged between organised/decisional and spontaneous arenas, together with its distinctive power struggles and contestation practices (FB’s corporate governance and moderators v. users and affected outsiders). The establishment of the OB ticks a further box of SC’s checklist, i.e. the emergence of a form of judicial review within FB’s private order.

Further, SC conceptualises democratisation as the institutionalisation of a system’s self-contestation. The never-ending series of scandals and abuses emerged in the last years – in which  the Cambridge Analytica and the Myanmar affaires are only the peak of the iceberg – sparked global backlash and contestation, boycott campaigns, legal sanctions, and of course also (threats of) public regulation, coming from a multitude of individuals, movements, as well as economic and political actors. FB’s constitutionalisation has so far taken the form of a reflexive response to massive, sustained, and ever-growing pressures from its own spontaneous arenas and social environment, rather than from actual, hard law obligations or the often timid sanctions by national bodies. By voluntarily (but not spontaneously) binding itself to the decisions of a relatively independent adjudicator, sensitive to external impulses (especially the pool of legitimacy represented by HRL) FB also channels social energies and internally stabilises the possibility of contestation. By these means, it aims to legitimise itself at the level of global society, while still preserving its autonomy. Here again, the constitutionalising potential of the interaction between “public” and “private” codes of conduct – emerging in the internalisation of Ruggie’s UNGP into the OB’s Rulebook – has been explored by SC.

The list could go on. What matters is that the trajectory followed by FB’s normative order fits with SC’s core concepts. Does this mean that FB has now a constitution properly so called?

Facebook: what needs to be limited?

According to SC constitutionalisation takes place if and insofar as the system’s own normativity performs both constitutive/foundational and limiting functions towards its specific communication medium (power, money, knowledge, etc.) and, more generally, the expansive dynamics that might endanger social environments and ultimately itself. SC thus captures in legal-theoretical terms the social processes that at any given moment take place in a system, to verify whether and to what extent they are limited by legal norms. The emergence of a normative hierarchy, of a (quasi-)judicial scrutiny, as well as of constitutional arenas and processes points to an ever more solid constitutive/foundational dimension. Such developments contribute to formalising/autonomising FB’s system by legal means, and to building its own symbolic narrative, enshrined in its (admittedly quite bombastic) mission statements.

What about the limiting function? The jurisdiction of the OB is limited to disputed takedowns of single posts or comments. To be sure, the relevance of filtering fake news, hate speech, cyber-bullying and -stalking, and thus deciding over the limits of free speech of almost three billion users cannot be underestimated. Still, this is only the most visible side of a subtler, broader, and to some extent de-personified societal process. Indeed, what comes to the surface are mostly decisions on the content of the speech, i.e. the what of communication. However, even more problematic and largely outside the OB’s radar are the negative externalities of social media-based communication, i.e. the how of communication.

Social media companies like FB normally have a business model – the “surveillance capitalism – centred around the monetisation of attention, as their profits mostly come from advertisers resorting in different ways to online platforms to increase the advertisement success rate. To maximise profits, such companies compulsively collect as much (meta-)data as possible. This model is oriented towards the maximisation of users’ engagement and, to that purpose, resorts to habit-forming digital products and persuasive technologies. The engagement growth compulsion pushes individuals and social actors (political movements; business entities; media, education, and research institutions, etc.) and ultimately FB itself to develop forms of addiction (yes, organisations and social processes may be addicted too).

Entities like FB, then, do not just shape individual behaviour but also drain other social sectors of their functional autonomy. Because of scale and pervasiveness they trigger vicious circles with potentially highly negative impacts on individuals and societal processes alike. The effects that addictive/addicted social networks can have on child, teenager, and adult well-being; on political engagement and politics in general; on human psychology and social fragmentation are by now well documented and have reached the general public. The phrase “human downgrading” has been coined to describe a system of mutually reinforcing harms: addiction, distraction, isolation, polarisation, fake news.

Social networks should not be demonised and have positive aspects: increased possibilities of connection, opening of new venues of participation, innovative modalities of socio-economic interaction, circulation of knowledge, de-marginalisation of communities and peripheral societies. Further, several online companies are not based on that business model. But here is the thing: the dangers posed by entities such as FB – the social dynamics to be limited by their constitution, if any – do not lie in the increased interconnection as such. Rather, it is the manipulation of individuals and the related social effects, deriving from a business model based on the maximisation of engagement.

Breaking the monetisation of attention as a normative agenda

Back to the OB, the dynamics triggered by the maximisation of engagement are not addressed, let alone limited by it. Put differently: even adopting SC’s theoretical framework, FB’s normative system is well behind in any process of constitutionalisation, because it does not perform any significant limiting function towards its most disruptive social dynamics. To make a comparison with state constitutionalism, it is like a constitution addressing the issue of abuses against slaves instead of slavery itself. The recent Digital Services Act (DSA) proposal of the EU Commission and several other instruments adopted in Europe in recent years, aiming to inject due process safeguards into the moderation activities, certainly go in the right direction, but their scope is still relatively narrow. Any attempt to establish an effective constitutionalisation in the digital sphere must target the vicious circle of monetisation of engagement and compulsion to social manipulation.

It is unlikely that such limits will be voluntarily established by FB. The recent blocking of a number of Australian news organisations’ accounts – as a reaction to a draft law which would make social media companies pay publishers for news content shared on their platforms – does not give much hope, despite the compromise finally reached with the Australian government. Further, the OB is not in the position of doing much, unless it adopts some spectacular and at the moment unforeseeable self-empowering decision, for example a ruling ordering FB to limit engagement maximisation techniques (e.g. by eliminating the “Like” or the “Comment” buttons).

A strong intervention by states and political actors is thus necessary. Indeed, even in SC’s pluralist framework state law remains central. As surveillance capitalism corporations often exercise de facto public functions in condition of (at least) oligopoly, classic nationalisation schemes could be an option in a classic welfare-state scenario. However, the multi-jurisdictional dimension of such entities and the economic competition among states make it unrealistic. Further, transferring the ownership to public actors would probably make it legally easier to enforce constitutional norms, but it does not address the problems of the attention-addicted digital economy.

Another and more realistic option would be resorting to anti-trust instruments. Many of the problems recalled depend on the sheer scale of the actors involved. The capacity to process an unprecedented quantity of data through ubiquitous technologies – and influence human behaviour based on them – could be addressed breaking corporations in smaller entities, which should not be allowed to match/pool their data. Here, after some measures taken by the EU Commission, it seems that the US has taken the lead.

However, even this option would not address the compulsion to engagement growth of the single entities. State law must find solutions, targeting the economic and financial structures that make possible and perpetuate surveillance capitalism. For example, introducing duties of funding diversification for digital start-ups and/or ceilings to revenues from direct or indirect advertisement, increasingly lower as number of users or data collected grows. Another option would be the separation and the limitation of data-sharing and -pooling among entities operating in different sectors of digital economy, e.g. those of social media and those operating in the “internet of things”.

These are, of course, only few rough ideas. The challenges raised by surveillance capitalism and human downgrading call for institutional and regulatory imagination. But they can only be addressed if constituent impulses are effectively internalised into the decision-making premises of the actors involved. SC may usefully contribute to such an agenda.

---

Suggested citation: Angelo Jr Golia, Beyond Oversight. Advancing Societal Constitutionalism in the Age of Surveillance Capitalism, Int’l J. Const. L. Blog, March 5, 2021, at: http://www.iconnectblog.com/2021/03/beyond-oversight-advancing-societal-constitutionalism-in-the-age-of-surveillance-capitalism/