Private Agents of Democracy

The first model assumes continued private ownership of social media, but calls for legislation that turns the platforms—at least to some extent—into agents of democracy. The underlying assumption is that ethical self-regulation of platforms—favored, for example, by Nicolas SuzorFootnote ¹⁹—will not be sufficient to achieve the required change.Footnote ²⁰ What is needed, according to this view, is stronger government regulation and increased oversight by public agencies. As the European Parliament put it regarding the Digital Services Act and the Digital Markets Act: “we need laws, not platform guidelines.”Footnote ²¹ Especially in the context of the European Union (EU), but also at the theoretical level, this model has connections to the idea of digital sovereignty, according to which states should more proactively assert authority over the internet. In terms of the content of new social media laws, it has been argued that Germany’s post-war regulation of public and private broadcasting could serve as an example, as it was supposedly “ambitious and comprehensive” in its “commitment to the principles of democracy, pluralism and diversity.”Footnote ²² One direction this could take is giving citizens greater control over the data that is a by-product of their social media use. Roberta Fischli envisages a “data-owning democracy” where user data can be pooled for collective purposes, for example to improve public services.Footnote ²³ In this regard, the right to data portability in the EU’s General Data Protection Regulation can be seen as a first step.

Public Service Social Media

The second model takes inspiration from the example of traditional public broadcasting, aiming at state-funded social media to complement the realm of private platforms. Public service social media could be designed with the interests of democratic citizens at heart because they would be independent from market imperatives.Footnote ²⁴ One such proposal is Hélène Landemore’s Citizenbook, which is to be imagined as a variant of Facebook run by a non-profit foundation and “repurposed as a deliberative platform for democracy.”Footnote ²⁵ All citizens would be automatically registered at birth and be provided with spaces and mechanisms for online deliberation, including digital mini-publics connected to decision-making processes. New democratic jobs would be created to maintain the platform and to facilitate its use, for example by providing information on topics to be deliberated.Footnote ²⁶ As tentative beginnings of such a structure one can interpret public platforms such as Decidim and vTaiwan, which enable citizens to participate in local and national policy-making.Footnote ²⁷ Other proposals aim at a less top-down and more decentralized ecosystem of “alternative social media platforms designed to meet the needs of local geographic communities and communities of interest,” complemented by a web interface that ensures a single sign-on and interoperability of the different sites.Footnote ²⁸

Platform Socialism

The third model, which is the most radical vision, calls for social ownership of digital platforms. While this often involves ideas for public service social media,Footnote ²⁹ the proponents of platform socialism do not merely seek to place new players in the public sphere but aim for a more general transformation of the digital economy, which includes “the breakup of capitalist monopolies in the communication, media, and digital sector.”Footnote ³⁰ The goal is to reorganize social media as digital commons that are collectively owned and controlled, in order to empower both the platforms’ employees and their users. The institutional arrangements envisaged for this purpose range from “top-down nationalization schemes” to “direct forms of workers’ control” from below.Footnote ³¹ Depending on the proposal, the notion of digital commons can go as far to include “common control of the mines where natural resources are extracted that form the physical foundations of digital technologies.”Footnote ³² Generally, social media (and other platforms) are meant to be characterized by social ownership, workplace democracy, and participatory rights that allow users to co-determine the development of digital infrastructures.Footnote ³³ While this model is the one most distanced from current realities, alternative social media such as Mastodon can give an idea of how the envisaged platform co-operatives might be structured.

The question of which of these models Habermas should adopt—as the most promising approach for maintaining a media structure conducive to democratic opinion and will formation—is not easy to answer. Democratic societies aiming to regulate social media need to make choices at two levels. At a first level, there is the decision between the three organizational models, i.e., between private agents of democracy, public service social media, and platform socialism. At a second level, within each model, there is the choice between various possible ways of implementation, i.e., decisions on the concrete content of new rules for existing platforms or, depending on the model, on the goals, rules, and structure of new platforms. The right choices depend on what we expect social media to contribute to a democratic society. The proponents of the different models take the second step before the first—they advocate structural reforms of social media without having determined their adequate function(s). They suggest that the platforms are to play a constructive role in democracy and, in doing so, apparently assume that we already know what we want from them. But, as I argue in the next section, this is hardly the case.Footnote ³⁴

Information

A first way to look at social media is as a new information technology. A functioning public sphere depends on a media system that provides citizens with political news, analysis, and commentary in a way that enables them to engage in political debates. While not originally created for this purpose, social media “function as transmitters of important facts about what is going on in the world.”Footnote ³⁷ In contrast to traditional media, they do not produce information but provide platforms for third-party content. This implies specific democratic affordances. In particular, social media offer new possibilities to access and circulate political news. For many citizens, social media have become a key source of information,Footnote ³⁸ providing a sphere in which they are presented with a wide range of potentially relevant content, from private blogs to quality newspapers. At the same time, the networked form of communication characterizing social media puts all users in a position to curate content, for example by promoting political information they deem important among their contacts. In some places, social media also serve as an important substitute for traditional sources of information, which are increasingly disappearing. For example, in areas where local newspapers close down—or at least lose significance—social media often “replace or augment local information, as people connect directly with their children’s schools or with community organizations.”Footnote ³⁹

Deliberation

There is a long-standing tradition of seeing the democratic value of the internet, and, with the rise of the so-called Web 2.0, of social media, in creating spaces for deliberation.Footnote ⁴⁰ This approach comes in two versions. According to the first, digital platforms offer new spaces for the kind of “anarchic” political communication that Habermas traditionally locates in the arena of civil society. Social media are seen as having the potential to approximate the function of eighteenth-century coffee houses in Western Europe, that is, enable citizens to gather informally to debate political issues.Footnote ⁴¹ The second view assumes that, in order to actually promote rational discourse, social media need to provide more formal spaces of deliberation than they do at present, such as digital mini-publics established to discuss specific issues. The idea is that “guided” political communication could “break the silos, filter-bubbles, and echo-chambers in which individuals currently prefer and are in fact encouraged to segregate themselves by platform designs created to maximize ad revenue rather than quality deliberation.”Footnote ⁴² Both versions assume that social media offer opportunities to mitigate problems of limited space and time that face-to-face deliberation in large societies is always confronted with. Digital platforms are more easily accessible than offline forums and, if they combine synchronous and asynchronous communication, can potentially include larger numbers of participants in deliberation.

Activism

Here the emphasis is on the platform’s potential to provide a stage for disruptive practices. Proponents of this view assume that “dissent plays an important role for initiating and sustaining social progress and learning processes” in democracies—and point out how social media provide new opportunities to engage in political contestation.Footnote ⁴³ Digital platforms allow for a new type of large-scale “connective action” that differs from traditional social movements in that it is characterized by loose coordination among (often) physically dispersed actors rather than classical organizational structures including leadership.Footnote ⁴⁴ Prominent examples from recent years are hashtag campaigns such as #MeToo and #BlackLivesMatter. As these cases indicate, it is important not to reduce social media to an online stage, but to see that they can play a critical role in initiating, coordinating, and sustaining offline protests and movements, as could be observed in the Arab Spring.Footnote ⁴⁵ If one emphasizes these affordances, the democratic value of social media primarily lies in its potential to enable citizens—especially marginalized groups and even non-citizens—to form counter-publics, to effectively voice dissent, to mobilize and coordinate collective action, and to challenge existing institutions. Social media appear as instrumental for political activism that can, as Robin Celikates puts it, “function as a corrective to democratic deficits that seem to be a structural part of actually existing liberal states.”Footnote ⁴⁶

Voting

Social media have the potential to facilitate voting processes. As Landemore has argued, they could be used to organize online referendums and to enable liquid democracy—vote delegation on a case-by-case basis.Footnote ⁴⁷ Social media can also be understood as a sphere of “reactive democracy,” where online crowds “vote” in informal plebiscites.Footnote ⁴⁸ Paolo Gerbaudo argues that political communication on social media follows a logic that is best understood as a plebeian public sphere—in contrast to the notion of a bourgeois public sphere outlined in Habermas’s early work.Footnote ⁴⁹ Due to the logic of reactions (likes, follows, etc.), social media enable the formation of “online crowds,” i.e. “non-organized collectives … that, while lacking the element of physical proximity characteristic of traditional crowds, nevertheless seem engaged in an experience of gathering or ‘crowding’.”Footnote ⁵⁰ This enables a plebeian public sphere whose protagonists follow an affective rather than cognitive and informational logic of intervention and which is generally characterized by “emotional mobilization rather than reasoned argumentation.”Footnote ⁵¹ While reactions on social media can have a qualitative character (e.g., comments), they are primarily quantitative in nature. As such, they can serve as a measure of popularity for political ideas, policy proposals, and so on. In Gerbaudo’s view, “reactions can be understood as a vote.”Footnote ⁵² Online crowds forming on social media in response to offers made by political elites resemble preference aggregation in plebiscites. From this perspective, digital platforms serve as “the stage of a permanent referendum in which people constantly judge everything that is on offer.”Footnote ⁵³

Representation

Yet another perspective suggests that the democratic value of social media could lie in enabling improved political representation based on techniques of web scraping. The practice of “demos scraping” has been described in an illuminating way by Lena Ulbricht—although she sees it rather critically.Footnote ⁵⁴ According to Ulbricht, demos scraping is a way of generating politically relevant knowledge through the automated analysis of trace data. It consists in a repurposing of data generated as a by-product of individuals’ (not necessarily political) online activities.Footnote ⁵⁵ “Data available on the Internet and in social networks” is meant to provide precise, comprehensive, and real-time aggregate information about what citizens want—and thus an effective way of tracking the “will of the people.”Footnote ⁵⁶ Demos scraping is meant to provide an alternative to opinion polls that is unaffected by the biases of survey situations. According to proponents, it “generates a superior form of knowing the demos that surpasses the insights gained from traditional disciplinary data about citizens, such as administrative data and censuses.”Footnote ⁵⁷ Demos scraping can take the form of sentiment analyses aiming to identify social media users’ attitudes towards particular policies—and thus an indication of which positions could win majorities. In these processes of data collection, in which online behavior is treated as “an expression of political preferences and intentions,” citizens “are unconscious of their role.”Footnote ⁵⁸ However, political actors collecting the relevant information can then use it to make “representative claims.”Footnote ⁵⁹ To what extent these are successful depends on how voters respond to the offers made, that is, what political choices they make. From the perspective of demos scraping, then, social media—or, more specifically, the data generated through user activity on the platforms—appear as a tool of governments and other political actors whose democratic value consists in improving representation. Web scraping enables political elites to gain more knowledge about “what the people want” than elections alone can provide and thus enables more responsive legislation.

Given that social media have such a variety of affordances, their adequate role in democracy can be interpreted in different ways. Not all of these options call for the same regulatory approach, nor can they necessarily all be promoted at once. For example, if social media are to provide citizens with an information technology, the model of private agents of democracy might be sufficient. Public regulation could require that platform companies take measures to ensure that information circulating on their services is reliable and that there is no arbitrary interference with content provision. By contrast, if social media are to provide civil society with an infrastructure for deliberation, a more ambitious restructuring appears necessary. To promote a deliberative function of social media, it needs to be made sure, for example, that the built environment promotes this particular mode of exchange and that the flows of communication are not distorted by algorithms designed to serve the platform companies’ economic interests.Footnote ⁶⁰ As this might be difficult to achieve through legislation without undermining the platforms’ business model, public service social media appear more promising. However, state-run platforms might in turn conflict with the goal of empowering activism. If the point is, for example, to enable citizens to push for institutional change, publicly managed social media are probably not the right approach. Instead, platform socialism, especially the idea of new platforms built “from below,” seems more adequate. The general point here is that we need to decide what should be social media’s contribution to democracy. As long as this is undetermined, reform debates are futile.

The Center–Periphery Model

Habermas’s model of political communication in media society provides a “map” that outlines flows of communication in democratic systems.Footnote ⁶¹ The model draws on the political sociology of Bernhard Peters. According to Peters, processes of opinion and will formation in modern societies take place along an axis of center and periphery of the political system. The center is characterized by its competences to make collectively binding decisions, which the periphery lacks.Footnote ⁶² Habermas gives this sociological model a normative twist, arguing that democratic processes ideally take their starting point at the periphery of the political system, in civil society, where citizens publicly articulate their concerns and—through media-based mass communication, intermediaries such as political parties, and ultimately elections—feed them into the political system.Footnote ⁶³ The institutions of the center subject these issues to formal processes of deliberation and, where appropriate, follow up with collectively binding decisions. Crucially, this process is to work as a “feedback loop.”Footnote ⁶⁴ Political elites, especially strong publics, publics with decision powers, should process citizen inputs and formulate them as considered public opinions that can serve, along with relevant information and arguments, as the basis for further debate in weak publics, with possible rounds of repetition, until the process eventually culminates in decisions.Footnote ⁶⁵

According to Habermas, these feedback loops involve three levels of political communication, each constituting a different arena characterized by a specific mode of communication. The first arena is the political system, where state institutions such as the government, administration, parliaments, and courts are engaged in institutionalized discourses and negotiations. The second is the public sphere, which is populated by elite actors, for example politicians, lobbyists, experts, advocates of general interests, and intellectuals, who engage in media-based mass communication. The third arena is civil society, where everyday communication among citizens takes place in both arranged and informal relations. This is the realm of various kinds of networks, social movements, and the like.Footnote ⁶⁶ What Habermas outlines here is a deliberative system, that is, a set of interconnected elements that together enable “a talk-based approach to political conflict and problem-solving,”Footnote ⁶⁷ where “persuasion that raises relevant considerations” leads the way.Footnote ⁶⁸ Crucially, not every element of a deliberative system needs to operate in a deliberative mode, or to the same degree. The deliberative quality of political decisions is expected to be produced by the system as a whole. This leaves room, for example, for disruptive acts aiming to make previously excluded voices heard in deliberation.Footnote ⁶⁹ In Habermas’s words, there is a division of labor between the different arenas and only “across the full scope of the process of legitimation” can we expect deliberation to fish “the reasonable elements of opinion formation out of the murky streams of political communication.”Footnote ⁷⁰

For Habermas, the public sphere has a key role in this division of labor, as an intermediate system between civil society and the political system. Its task is to ensure that in the feedback loops of opinion and will formation “relevant issues and controversial answers, requisite information and appropriate arguments for and against will be mobilized.”Footnote ⁷¹ Elites in the public sphere need, first, to take up inputs from civil society that need to be incorporated and reflected upon in broader political discussions and, second, to monitor the processes in the political system and provide the audience with critical commentary. In this way, a functioning public sphere ensures that the deliberative system is “both ‘plugged in’ to the source of legitimate authority, the demos, and to the outlet of binding collective decisions and executive power,” as John Parkinson puts it.Footnote ⁷² Two conditions need to be met in order for the public sphere to live up to these expectations. First, there must be a media system sufficiently independent from its social environments, especially from the political institutions of the center but also from economic interests. This presupposes, among other things, a media constitution. Second, there is a need for civil society structures that empower citizens to participate in public discourse. The periphery needs to be able to maintain its own autonomous publics in order to act as a sensor for social problems that need addressing and as a basis for new political impulses.Footnote ⁷³

A striking point in Habermas’s recent work on the digital public sphere is that he analyzes the effects of social media without ever explicitly placing them in his model of the deliberative system. In fact, there is considerable ambiguity when it comes to their democratic function. On the one hand, Habermas emphasizes how social media, in contrast to traditional media, empower all users to become authors. Citizens can use the platforms “like blank slates for their own communicative content.”Footnote ⁷⁴ However, as he points out, social media do not perform the editorial function of classical journalism, which results in a lack of quality control. This perspective on social media suggests that the platforms should be treated as part of the media system—they play a (yet to be defined and regulated) role in the production and circulation of political news and information. On the other hand, Habermas emphasizes that social media have become spaces for conversations “that had previously been reserved for private correspondence.”Footnote ⁷⁵ As a result, when citizens engage in political discussions on social media, which they also do frequently, they often fail to recognize the need to follow standards of public rather than private autonomy. This take on social media as spaces of citizen-to-citizen communication suggests that the platforms need to be placed in civil society, as part of the structures that enable citizens to create autonomous publics. In short, Habermas’s discussion implies two options: social media might be a functional equivalent of quality newspapers or the coffee houses of the digital age.

Political-Sociological Indeterminacy

However, the issue is not simply whether social media should be understood as part of the media system or as spaces for informal debate. To bring out the multiple options of placing them in the center–periphery model, I introduce three categories, derived from recent debates in democratic theory about the conceptual building blocks of systems approaches—democratic practices, sites, and agents.Footnote ⁷⁶ My point is that the center–periphery model, and the feedback loops Habermas envisages, are in principle compatible with different ways of combining these elements and of organizing their relations. This leads to competing options of promoting (different) democratic affordances of social media. I define the three categories as follows.

• • Democratic practices include formal and informal political activities such as assembling, associating, campaigning, deliberating, going on strike, news reporting, protesting, representing, and voting.

• • Democratic sites are institutionalized and non-institutionalized arenas where such activities can take place, e.g., community centers, the media, mini-publics, parliaments, the public sphere, squares, and town hall meetings.

• • Democratic agents are individual and collective actors that engage in democratic practices and populate democratic sites, e.g., citizens, civil society organizations, elected officials, governments, journalists, lobby groups, political parties, and social movements.

These categories bring into view the multifaceted nature of social media. Depending on how they are designed, social media can enable different kinds of democratic practices such as deliberation or political protest. Moreover, while social media can be seen as a democratic site, different kinds of arenas can also be created within their realm (e.g., digital mini-publics). Finally, social media can fulfill different functions for different democratic agents. For example, governments and social movements are likely to use them for different purposes. Even if we follow Habermas in assuming that the point of social media regulation is to maintain a functioning public sphere, the platforms’ proper place in the center–periphery model is anything but self-evident. This becomes even clearer if we consider these categories in light of social media’s democratic affordances. Depending on which one we aim to promote, different agents will be enabled to engage in different practices; social media will take on a different character as sites of democratic processes.

Let me highlight some of the options, moving from the periphery to the center of the political system. If activism is seen as the main democratic affordance, social media are clearly placed in the periphery, meant to enable citizens to disrupt the center’s routines to push for greater responsiveness or even institutional change. Similarly, if one aims to shape social media as sites of non-institutionalized deliberation, they appear as an important infrastructure of civil society, the periphery, although now as spaces for political debate contributing to the formation of considered public opinions. However, if we see social media as an information technology, they take on the role of an intermediary between civil society and political institutions. In this scenario, social media allow various agents—across the center–periphery divide—to access, disseminate, and exchange political information. If we take yet another perspective and regard social media as sites for institutionalized deliberation, connected to decision-making processes (e.g., digital mini-publics), they move closer to the center—with the goal of strengthening the link between citizens and political institutions.Footnote ⁷⁷ Once we pursue the idea of (quasi-)voting in informal plebiscites, social media no longer serve citizens but become governance tools of the center. In this vision, governments, political parties, elected officials, etc. use digital platforms to test the popularity of their political projects among voters. Finally, if we see social media as a way of improving representation by gathering information about the preferences of citizens (who are basically unaware of this process), they become a part of the center, taking on a role similar to political parties.

A purely political-sociological perspective does not put us in a position to give a definitive answer to the question of social media’s place in democracy. Given social media’s various democratic affordances, they cut across the boundaries of Habermas’s categories of center and periphery. Social media can and de facto do play a (sometimes more, sometimes less significant) role at various points along this axis. Thus, if we are seeking guidance for democracy-enhancing social media regulation, we need to switch to a normative mode of reasoning and ask what the platforms’ contribution should be.

Rebooting the Extraordinary Mode of Problem-Solving

Social media do not have a “natural” place in democracy. Thus, their proper role is not just a diagnostic but also a normative question—and one that needs to be answered with a view to the challenges contemporary democracies are facing. I approach this question in a reconstructive manner, in line with Habermas’s view that the common distinction between empirical and normative theories is oversimplified.Footnote ⁷⁸ Following his method of rational reconstruction, normative considerations about democracy—for example, assessments of changes in democratic processes and institutions—need to take their cue from “the implicitly assumed normative contents of empirically established practices,” which need to be explicated “from the participant perspective, i.e., in a performative attitude.”Footnote ⁷⁹

Seen through this reconstructive lens, a key problem of current politics is that democratic feedback loops increasingly fail to work as intended. The normative expectation that citizens be able to programme the content of collectively binding decisions is frequently frustrated. Habermas himself distinguishes between a (self-referential) routine mode and a (responsive) extraordinary mode of the political system.Footnote ⁸⁰ According to this distinction, the periphery is not always in the driver’s seat. Rather, only in exceptional cases, when certain problems attract special public attention and trigger an intensified search for solutions, are citizens able to determine the direction of decision-making in the center.Footnote ⁸¹ However, there is a growing sense that citizens face difficulties to force democracy into the extraordinary mode at all, not just in the day-to-day business, but even in the face of major crises—as the institutional inertia with regard to climate change and the emergence of protest movements such as Extinction Rebellion illustrates. Democratic systems are increasingly disconnected from their demoi.Footnote ⁸²

Using Habermas’s metaphors, the question today is how to reboot the extraordinary mode of problem-solving. In light of this, it would be misguided to treat social media as a technology whose democratic affordances should primarily be shaped towards the interests of agents at the “top” of the deliberative system. It may well be the case that governments (and other political actors of the center) can use techniques such as web scraping and informal (online) plebiscites to map citizens’ preferences. And it may be that such uses can reasonably be described as democratic—as attempts to improve representation and interest aggregation. However, given that contemporary democracies are plagued by a growing asymmetry between weak and strong publics, social media should not be turned into instruments of political institutions. The priority should be for citizens to regain the capacity to initiate feedback loops. This democratic agency depends at least in part on critical infrastructures that allow civil society to engage in political exchange, coordination, mobilization, and contestation—for which social media today play a crucial role. The goal of platform regulation should be to promote citizens’ capacities to interrupt the center’s routines and to influence the content of political decisions. If social media are to make a constructive contribution to democratic opinion and will formation, they need to strengthen the periphery.

This view finds support in citizens’ normative expectations as users of social media. These come to the surface, for example, when interference with certain practices is scandalized as an encroachment on democracy. Elon Musk’s purchase of Twitter (later renamed X) in 2022—and his subsequent interventions in the functioning and culture of the platform, including banning a number of journalists and activists—is an instructive case. The “public reaction to the acquisition goes beyond discontent”—in fact, “it seems to imply a perceived loss,” as X is apparently “understood by many to play … a key role in the public sphere and in many ‘subaltern counterpublics,’ and thus in democracy.”Footnote ⁸³ How people have responded, the critique and worries they have articulated, indicate what users presuppose about X’s role in democracy. The general assumption underlying the public reaction is that, despite the fact that X is a privately owned company, it should be treated as a collective good, a shared space of citizens. Musk’s intention to redesign X to make it more profitable is considered a threat to political equality: “[I]f Musk’s plan is to monetise everything, then ultimately, that will mean that access to that public square will be based on ability to pay.”Footnote ⁸⁴ What is being claimed is that the platform must be maintained as a “domain of mutual trust in which citizens feel confident that they can debate and criticize freely on the basis of a shared consensus about reality”—in particular, in order to able to “speak truth to power.”Footnote ⁸⁵

These reactions reflect the significance that X and other social media have for the periphery. In the eyes of many citizens, the added value of social media primarily lies in offering a space for the formation of autonomous publics and a realm from where political elites can be challenged. Citizens operate on the idealizing assumption that digital platforms play a key role in the empowerment of civil society. In his analysis of the digital public sphere, Habermas argues that democracy can only function, and be maintained, as long as there is not too large a discrepancy between citizens’ presuppositions and political reality:

Habermas’s focus is on how these conditions are undermined by online mis- and disinformation and the emergence of semi-publics that refuse to engage with each other. Here, the rationalizing power of the public sphere is called into question. However, considering the structural failure of feedback loops, the more serious gap between normative expectations and political realities today seems to be that social media do not live up to the role citizens ascribe to them. What is at stake is the power of citizens to engage in consequential (politically effective) debates. What citizens see threatened in the case of the former Twitter, and what is implied in the public reactions to Musk’s acquisition of the platform, is social media’s potential to provide a basis for bottom-up politics. This should be the main concern of platform regulation.