Legitimacy Beyond the State: The Case of Social Media

Political philosophers have begun to clarify, qualify and defend some of the moral concerns about large digital platforms like Meta, YouTube, TikTok and X that are often voiced in public debate, such as that they manipulate and exploit us, violate our privacy, or amplify dangerous speech. (Benn and Lazar 2021; Howard forthcoming)

But the problem with platform power is not just its abuse but its mere existence. This distinction should be familiar to political philosophers; it parallels the often-made observation that although a law made by a benevolent dictator could be justified at the bar of justice, its enforcement might be impermissible nevertheless. What gives them the moral right to govern us as they do?

This is the classic problem of legitimacy - for some the fundamental question of political philosophy. In my DPhil thesis, I will mount a legitimacy critique of platform power, and consider the sorts of checks, such as democratic authorisation and procedural rights, that could render it permissible.

The problem of legitimacy arises from exercises of power. Suzor distinguishes between three different 'faces' of platform power: content moderation, content curation, and platform architecture. (Suzor 2019) Are documents how opaque algorithmic moderation systems rob precarious content creators of their livelihoods (Are 2022); Wu argues that curation systems shape the distribution of attention online, differentially empowering users to influence one another (Wu 2017), and Lessig popularised the idea that platform architecture regulates users 'in the sense that bars on a prison regulate movement of a prisoner', allowing platforms to exercise pre-emptive power. (Lessig 1999)

While social scientists have offered an impressive diagnosis of platform power, their normative evaluation is elliptical at best. Often, such authors seem to think that simply naming platform power is sufficient to show that users have a moral right to participate in platform governance. (Lehdonvirta 2023)

To plug the gaps in their arguments, we need a normative theory of legitimacy, which explains why exercises of platform power are wrongful unless subject to forms of democratic legitimation.

One might hope that the extant philosophical literature on legitimacy would give us some guidance here. One is liable to be disappointed, however. For most existing theories of legitimacy are formed in the shadow of one institution: the state. This leaves them ill-equipped to evaluate platforms, for two reasons.

First, such theories focus exclusively on coercive power. This narrowing is not only unjustified (Quong 2014; Kolodny 2023), but also excludes the non-coercive faces of platform power (such as content curation and platform architecture) from the domain of a legitimacy critique from the outset. Second, state-centric theories characteristically recommend regulations that are ill-suited to check SMCs.

How can transparency requirements be applied to black-box algorithms that are inscrutable to both users and developers? And how can standards of consistency govern recommendation systems tailored to individual preferences?

Thus, I will need to excavate the deeper values underlying state-centric conceptions of legitimacy, and give an account that is general enough to encompass the power of SMCs.

Finally, I will consider how the particular case of digital platforms sheds light on some first-order questions in political theory. First, it could turn out that the reach of principles of democratic legitimacy is significantly wider than is often thought, encompassing firms (Anderson 2017), the relationship between states and prospective immigrants (Abizadeh 2008), international organisations and religious institutions (Cordelli 2017).

Second, my account will shed light on the relationship between legitimacy and political obligation. Considering a very new form of power should help to cast these old questions in a new light.