Over the past year, it felt like almost every public question was transformed into a social media question. The Black Lives Matter protests in the summer of 2020 quickly led to a debate about whether Trump’s tweet “When the looting starts, the shooting starts” was an incitement to violence. Discussions about the proper response to the COVID-19 pandemic bled into meta-discussion about whether social media platforms should allow the spread of patently false vaccine-related disinformation. The real and the online worlds overlapped and have at times even seemed to merge: On Jan. 6, protesters stormed the U.S. Capitol, brandishing memes as well as weapons — an extremely online incursion into the physical representation of American democracy. As the controversies piled up, the public turned to social media companies for decisive action, for “moral clarity.”
Such moral clarity has been so hard to find because we have all been muddled about what social media actually is. Social media companies often argue that they are simply neutral platforms, like postal services or telephone lines. This neutrality means that they have only a limited legal and moral responsibility to moderate their content. Critics, on the other hand, argue that social media platforms are more like newspapers or magazines. Because social media companies effectively curate content by promoting certain kinds of posts above others, the argument goes that they ought to have an editorial commitment to public-spiritedness: to fact-check popular posts, to restrict hate speech and to limit medical disinformation.
Among the Silicon Valley tech workers themselves, a third metaphor dominates: that of engineering and design thinking. By solving hard, technical problems over and over again, the brilliant engineers of Silicon Valley create technical solutions that, in aggregate, will build a new and improved world, the same way that engineers tinkering with steam engines in the 18th century incrementally created the technology of the Industrial Revolution.
But none of these metaphors have helped provide the moral clarity or legitimacy we have been looking for. This failure has led to the prominence of a fourth metaphor: that social media companies are governments. Like governments, they should be run by expert, transparent, fair, universal and equitable procedures. But this metaphor is still incomplete because it fails to account for the novel form social media governments have taken.
It helps to go back to early modern Europe (roughly 1500-1800), when some of the most aggressive, innovative and powerful organizations were very similar to social media companies. I’m thinking about the company-state. Like companies, company-states sought profit, were owned by shareholders and employed staff. But like states, company-states held territory and, within this territory, regulated economic, religious and civic life, conducted diplomacy and even waged war.
These company-state territories were in very different shapes than conventional states: They were very far away from their home countries and were often scattered in loosely connected networks of forts, ports, concessions and other enclaves. The most powerful company-state was arguably the British East India Company, which, after humble beginnings, came to control much of the Indian subcontinent for nearly a century. The historical analogy of the company-state example can help clarify our conceptual muddle over what social media platforms really are, and in so doing, help us think about possibilities for social media’s future.
“Content moderation, after all, is at its root a public problem — a question about what the community can say and do and thus about what the community is.”
State Power In A Different Shape
Social media companies don’t control territories, they don’t fly flags, they don’t collect taxes and they don’t command armies. But nevertheless, they, like the company-states before them, are hybrids blending aspects of the company and the state.
Two interlocking powers give the major social media companies state-like authority. Firstly, they have secured an effective monopoly over civil society and the public sphere. Political action now happens online. Since at least 2010, every social movement of any energy has been organized, in large part, on social media. Political speech also happens online. If you don’t want to go on Facebook, Twitter or YouTube (and to a lesser extent, Reddit, Instagram and LinkedIn), you simply cannot have a voice in the public conversation. Competition does not challenge companies’ effective monopoly because of the field’s powerful network effects. New social media companies like TikTok or Parler can only secure niche audiences, not the feeling of everyone’s here commanded by Facebook in particular.
Secondly, this effective monopoly over the tools of collective action and the spaces of public discussion gives social media companies enormous power of surveillance and control. So far, this power seems to only have been used to sell us things — to serve us creepily serendipitous ads for that product we were talking about a minute ago. But the power could be used more aggressively: to shift the direction of democratic politics, perhaps, or to start a riot, or to shift state policy. It’s hard not to suspect that this power is already being used against us in some subtle way.
“Social media companies have failed to legitimate their public power because they have responded to such public problems like engineers.”
This power over the public has been difficult to grapple with because — for lack of a better word — it has a radically new shape. The contemporary nation-state can be understood as a 2-D polygon: You can draw a nation on a map and point to say that this colored area here is America. Social media states, by contrast, use public power in networks that indiscriminately cross both state borders and cultural boundaries, making them very hard to point to. They seem to be everywhere, in that any given interaction on social media links people in dozens of countries, using physical infrastructure like undersea fiber optic cables that cross-continental shelves. Yet, at the same time, these interactions are also nowhere, the actual connections being so temporary and fragmented.
Because of the networked nature of social media’s power, it’s not clear how it can be controlled. We can’t rely on nation-states’ judiciaries to regulate these evanescent multinational networks, because their networks cut across national borders. Nor can we call back to some set of agreed-upon traditions or norms to hold people to account for what they do or say, because online interactions are all so rootless and immaterial. The only bodies that seem capable of understanding and regulating what happens on social media are the social media companies themselves. Thus, we have looked to them for moral clarity.
Why Company Tools Cannot Solve Public Problems
The company and state parts of the company-state were historically often in tension. Tellingly, early modern company-states frequently exploited the territories that came under their control at great human cost. After the East India Company gained control of the taxation of the Indian provinces of Bengal, Bihar and Orissa in 1765, the company pushed to maximize its tax revenue so it could enrich its shareholders. Under this mismanagement, in 1769, a massive famine began, killing between 1 and 10 million people — perhaps one-third of the region’s population. Social media states’ hybrid nature means that they face a similar tension. Their company ethos — to maximize profit, to deal with problems bureaucratically — undermines their legitimacy, especially when faced with critical public issues.
Take the problem of content moderation. Fifteen years ago, social media’s content moderation practices were based on a combination of gut feeling and libertarian free speech norms. But this mentality ran into two problems. First, gut feeling was not scalable. You could rely on American college-educated tech employees to moderate American college-educated users, but once the moderators and users became truly global, impartial bureaucratic standards were needed to ensure fair and predictable policies in cultures as different as Sweden and Pakistan.
Second, libertarian free speech norms could allow quite toxic behavior, which discouraged new users and thus hampered growth. Most platforms pursued a similar solution to these problems of moderation, a company solution: the iterative development of procedures to more consistently, rationally and effectively determine whether individual pieces of content should be allowed to remain online. The goal is for any employee anywhere, from Hyderabad to Dublin, armed with the right trainings and looking at the right flowchart, to be able to decide whether any piece of content should be allowed to be shared on the platform. These efforts made social media spaces more welcoming, increased engagement and thus ultimately improved revenue — but they were only a success from the perspective of the company.
“There are hints that a similar process of regulation, break up and state co-option might happen to social media states today.”
These procedural solutions did not solve the public problem of how to make content moderation decisions be seen as legitimate. Content moderation, after all, is at its root a public problem — a question about what the community can say and do and thus about what the community is. Such problems are difficult to resolve because they often concern scarce resources, whose distribution will always make winners and losers: resources like status, recognition, truth, power and money. Institutions are legitimate to the extent that they can secure the acceptance of these fraught decisions, even among people who are on the losing side.
Social media companies have failed to legitimate their public power because they have responded to such public problems like engineers, looking to smart experts to craft maximally impactful solutions that can be scaled across a variety of different markets. Great for product. But bad for politics. As a result, social media companies have been vulnerable to public controversy, and with each new insult, each new misstep, the legitimacy of social media companies’ public power declines. This has left a yawning gap at the center of the online public sphere. It seems like there are no adults in the room. A new response is needed, or the liberatory promise of social media’s power will fade away.
Regulation, Exit, Democracy
As early modern company-states’ power increased, they also stumbled into public controversies that sapped their legitimacy. Even the East India Company was eventually cowed. After numerous scandals, including the Bengal famine, the British Parliament came to exert more and more oversight and control over the EIC, and eventually, the company withered to a mere state appendage. Company rule was formally ended in 1858. Most of the other company-states either failed or were reabsorbed back into their parent states. For the next century, company-state hybrids came to be seen as entirely illegitimate, especially when it came to the use of force.
There are hints that a similar process of regulation, break up and state co-option might happen to social media states today. In Europe especially, social media is coming under more stringent government regulation in the name of protecting privacy, which might limit their state-like powers of surveillance and social control. Furthermore, there are calls to break up social media companies, an action that would deprive individual companies of their effective monopoly on civil society.
China’s approach has solved the problem of the social media company-state in an entirely different way, by rendering social media an organ of the state itself. Through a combination of technological solutions, censorship and state-sponsored posting, Chinese social media has become largely coterminous with the boundaries of the nation-state, which allows the state to control social media just as it does real life.
These solutions are less promising than they seem at first because they clarify social media’s company-state hybridity, but they do not by themselves confront the public challenges of social media’s take-over of real life. If the example of the break-up of early modern company-states is any indication, such state-regulated and state-dominated social media may be more coercive, more restrictive and more manipulative than the private social media states we complain about today. The imperial dreams of the early modern company-state were at least hemmed in by the search for profit. The nation-state had no such compunction. It sought glory, gain and security, not profit, and so it satisfied its imperial ambitions recklessly.
“If the example of the break-up of early modern company-states is any indication, such state-regulated and state-dominated social media may be more coercive, more restrictive and more manipulative than the private social media states we complain about today.”
Another option is retreat — retreat from the openness of civil society into private spaces that are freer from controversy and surveillance. This happened to early modern British civil society. Starting in the 1670s, there was a brief flourishing of a novel public sphere in the London coffeehouse. Over the next forty years, men could walk into a coffeehouse and talk with other men across lines of class, creed and political interest about truly public matters that just a generation before had been considered the purview of the state. It was exciting and fertile, leading to developments in politics, finance, literature and science. But people found the experience of the coffeehouse quite uncomfortable, as it was noisy with bitter political arguments and constant misunderstandings.
After the 1760s, the coffeehouses became quiet and the conversations that once happened in them shifted to private members-only clubs. These preserved some of the openness of the old coffeehouse but in the comfort of private rooms, among like-minded people. Although this was a public sphere in some ways — people talked together, read newspapers and debated public issues — they were safe spaces where people would not find themselves unduly challenged by contrary viewpoints.
We are seeing a similar retreat from the public sphere today. The great energy of talk that once ran through Facebook in particular has, since 2016, been diverted to private Facebook groups and “group chats” like WhatsApp, WeChat and Telegram — spaces free from the surveillance of strangers. If this trend continues, it will break up social media companies’ monopoly on civil society. But it will also end that great potential of so many of us being on social media, talking together, in a common cacophonous project.
“Social media states should experiment with representative democracy.”
More appealingly, the problems of social media states could be solved not through regulation, nationalization or exit, but through a process of democratization. This is admittedly utopian, especially as contemporary liberal democracies are suffering a decline in their own legitimacy, but the utopian vision might be a useful provocation to future thinking. At its simplest, social media could be democratized through instituting votes on controversial decisions like whether to ban Donald Trump. One obvious benefit of such votes would be to legitimate the platforms’ most controversial decisions by giving everyone a voice in the process.
Such direct democracy has many challenges, not least of which is that actually-existing democracy is boring. Social media companies themselves attempted direct plebiscites in the past. After controversial changes to Facebook’s terms of service in 2009, for example, the company tried to put new company policies up for a series of votes. User participation was dreadfully low — the first referendum, on a controversial terms of service issue, only received 665,654 votes in total, about 0.3% of the platform’s userbase (then only 200 million people.) The direct democracy policy was finally buried in 2012.
Instead of direct democracy (which runs the risk of majoritarian online mobs), social media states should experiment with representative democracy. Instead of average users voting on individual policies that might be too complicated for them to dissect, they could vote on representatives, who could represent particular regions. People’s votes could be scaled by, say, their relative level of participation in those communities.
“Social media states could make democracy fun by gamifying online public life to encourage us to come together to do the hard work of politics.”
Early modern corporations provide an example of how to encourage the hard work of democratic action on the part of their members: They made their democracies fun by sugar-coating procedure with conviviality. The meetings at which important democratic business was done were annual feasts, filled with eating, drinking, toasts and songs, attended as much for their pageantry as they were for their democracy. They were also where different parties came together to argue about who got to benefit from the company-state’s riches.
In a similar way, social media states could make democracy fun by gamifying online public life to encourage us to come together to do the hard work of politics. Already, social media platforms manipulate their products to maximize engagement and social interaction. Our societies would be better if these tools instead rewarded more democratically healthy behaviors. Perhaps online discussion could generate agendas then considered by an elected body of representatives. Or the social media states could regularly channel some of their profits into real-world initiatives proposed online. With something more at stake than mere internet points, online civil society might become much deeper and perhaps more serious.
This conversation might lead to the formation of a new supranational democratic social media public sphere that would be capable of grappling with those intractable global problems that have proven too large for territorially-constituted polities to deal with on their own: problems like international tax avoidance, organized crime, wealth inequality and climate change. Competing political identities would inevitably contend with one another, but, hopefully, the ideological lines would be drawn in more expansive and productive ways than those that currently confine national democratic politics.
“These are decisions that cannot rightfully be palmed off to committees of experts or engineers — they must be determined by the public itself, a public that is larger, more inclusive and louder than any public ever before assembled.”
What is clear is that if the liberatory promise of social media is to be fulfilled, something political must be done. The prize is great: a truly global public sphere that can finally confront the immense public problems of the 21st century. The path is unclear, however. The solution will not be an East India Company 2.0. It will not be a Facebook Declaration of Independence. It will not be a social media General Electric. It will look as novel as the situation is. It is likely to be venal, crass, inequitable and partisan. But the opportunity of genuine political novelty should inspire us to think ambitiously about what this strange new hybrid company-state social media democracy might be and what tools could be developed to make it a reality.
The first step must be to acknowledge that when we discuss social media, we are no longer discussing the private actions of private companies. Instead, we are discussing the central problem of politics: who we are and what the common ground of our community is. These are decisions that cannot rightfully be palmed off to committees of experts or engineers — they must be determined by the public itself, a public that is larger, more inclusive and louder than any public ever before assembled.
