The debate over banning TikTok for its addictive effects on young people (not to mention its links to China) continues to roil Congress and the nation. These issues are, of course, important, but the threat posed to democracy by TikTok, as well as other media platforms such as X and Facebook, goes well beyond whether they are owned by a Chinese company. The problem is that social-media platforms’ business model undermines the social foundation of democratic participation.
That is, social-media platforms purposely seek to undo local communities to maximize profits. The reason is that social-media companies profit from our digital relationships, but not from our “physical” relationships, those that take place in parks, coffee shops, book clubs, and the like. As a result, social media platforms see the physical relationships between people as competing with the digital relations they offer.
But these physical relations are the foundation of our communal life, and as Tocqueville taught, without local communities, democratic engagement drops. This decrease in engagement will leave democratic institutions at the mercy of vested interests, undermining their performance. The declining effectiveness of the democratic process will sway many young voters to dismiss democracy as non-essential. In a way, profit-maximizing social media are succeeding where the totalitarian ideologies of the 20th century failed: redesigning society and its physical networks to achieve their profit maximizing goals and, in so doing, killing democracy.
A network of relationships is an asset, like a building or a machine, which needs to be “maintained”; otherwise, it depreciates. Physical relationships, like friends and family, are maintained by the voluntary efforts of their members, coaxed into it by existing social norms. Hence, the endless cycle of reciprocity in society, from our presence at the milestones of life (weddings, graduations, funerals, and the like) to the institutions that allow us to keep in touch (the occasional phone call or text message, Christmas cards, and so on).
Maintaining relationships is costly, and it is easy to free-ride on others to bear these costs (your friend calls you more often than you do, they attend your events more often than you do theirs, and so on). For this reason, physical networks aren’t well maintained unless there is a member of the network who has an ulterior motive to maintain them. In the obscure parlance of the economist, we say that there is an externality at work.
For example, in the United States, alumni fundraising is an essential component of college budgets. As a result, colleges heavily invest in maintaining the alumni network, providing a valuable service to all the former students who enjoy the network without paying for its maintenance (or paying through monetary donations). In Europe, government-funded universities don’t have any incentive to maintain alumni networks. Left to the voluntary effort of its members, alumni networks of European universities languish because no member wants to bear all the cost of organizing periodic reunions (from contacting all the alumni to securing the venue and caterer), which benefit all the alumni. All the alumni suffer as a result.
“We don’t own our followers—X does.”
Alumni networks aside, however, most physical networks emerge from spontaneous order: They are decentralized. In contrast, digital ones are designed and regulated by the platform owner: They are centralized. The reason is that digital relationships are built and maintained in a space that is controlled by a platform, and this control confers the platform with the ownership of the relationship. We can’t take our X (formerly Twitter) followers with us to a different platform, nor can we access our followers without X’s permission, which can be withdrawn at any time without cause. In contrast, X can advertise to our followers and connect them with other followers. We don’t own our followers—X does.
Not only are most digital networks centralized, but they also have another motive to create and maintain the network: to monetize it. Their economic motives fuel their interest in undermining physical networks. All else being equal, customers are indifferent between a digital relationship and a physical one, but platforms aren’t. They profit from digital relationships but not physical ones. In their effort to subsidize digital relationships and undermine physical ones, these platforms are even willing to pay us to remain glued a little longer to the smartphone and not go to the bar with our friends.
This distortion would be irrelevant if digital and physical relationships were perfectly interchangeable, but they aren’t. Physical relationships tend to be local. Even in a world of supersonic jets, traveling time and expenses represent an important impediment to long-distance physical relationships. In contrast, digital ones tend to be long-distance by definition: Not only is this their comparative advantage, but also precisely what platforms promote to avoid disintermediation. If we can meet a digital friend at the bar, we will reduce the platform’s profit. To avoid that, platforms promote long-distance relationships.
The local nature of physical relationships forces us to compromise. When the pool of potential friends is limited, we can’t be too picky; we must accept what we can find. Being stuck with the locals, our only option is to improve the quality of our relationships. In a digital sphere, where we can instantaneously connect with the entire world, why should we invest energy in making the current match work? By searching a bit longer, we can find a better match. The quality of any specific relationship becomes less important when there is a large pool of readily available alternatives. It is quantity over quality. As a result, digital platforms create shallower relationships among like-minded individuals that are easy to make and easy to drop.
The emphasis on quantity over quality is exacerbated by the gamification embedded in the platform’s design, carefully curated to maximize engagement and the platform’s profits. The number of followers and likes is prominently featured. Thus, users end up maximizing the number of relationships, not their quality. As many others have noted, in a boundless virtual community, the best way to get attention is to be extreme, to radicalize your position, and even to attack others. In a physical community where people constantly interact face-to-face, it is difficult to dehumanize adversaries. Not so in virtual communities. Online people are just avatars like the ones we are used to shooting in video games. Online, we don’t live with the consequences of our actions when we offend others because we do not observe the pain we inflict on them. We only live with the benefits: more retweets, likes, and followers. This is why otherwise calm and lovely people transform into aggressive beasts online.
The substitution of physical networks with digital ones dramatically alters how we relate to self-government. Democracy was born at the local level. The first versions in ancient Greece and Asia Minor, restricted as they were, were based on a tight-knit community of free men. The genius of modern American democracy was to retain some of these local roots as it became delegated across an expanding distance.
Local communities help democracy work through the production and diffusion of the information needed to monitor local officials. At bars, churches, and parks, the information on local elected officials is aggregated and then circulated with the help of local newspapers. As Tocqueville wrote, “associations make newspapers, and newspapers make associations.” The death of local newspapers has much to do with losing advertising revenues to digital platforms. Yet reduced interest in local administration has also killed the demand for local news.
“A captured democracy functions poorly.”
Without information, political participation dwindles, and vested interests have an easy time capturing local administrations because most people don’t vote or don’t vote in an informed way. A captured democracy functions poorly, reducing incentives to participate and the trust in the democratic process. If, as Tocqueville hypothesized, local communities are the gymnasium of democracy, then these trends do not affect just local administrations but also national ones.
Overlapping local networks also foster cooperation because they amplify the benefits of repeated interactions on cooperation. Engaged local communities, where members participate in different overlapping networks, can cooperate more effectively, thereby encouraging individuals to cooperate in other settings.
Democracy at the local level is made of small interests—fixing potholes, say— rather than big principles. When small local needs are involved, it is much easier to find a compromise. When big ideals are involved, any compromise seems like a betrayal. At the national level, intransigence wins the day. Thus, without a healthy experience of local politics, the political debate tends to radicalize.
All this is lost when physical networks give way to digital ones. No wonder democracy is being undermined among the cohort most likely to be plugged into digital networks and plugged out of physical ones.
We can’t turn back time. Digital platforms are here to stay; in any case, they bring many benefits we want to enjoy. So what are we to do? There are only two paths forward. The more conservative (lower-c) one takes the current form of democracy as given and tries to minimize the damage inflicted by the rise of digital networks. The more progressive (lower-p) approach tries to envision how representation should be changed to make democracy thrive in a digital world.
Let’s start with the conservative approach. Social and private returns of physical interactions are always misaligned, as we have demonstrated, and the advent of digital platforms makes that misalignment worse, with adverse effects on the workings of democracy. Could the kind of minimal intervention preferred by economists—fixing externalities and realigning private with social returns, such as through regulation or taxes—get the job done?
In most countries, television is heavily regulated or government-owned. While the idea of government-owned social media seems unappealing, the possibility of regulation isn’t. In China, TikTok warns its customers when they spend too much time in front of the screen. It doesn’t do it out of concern for the negative effect that excessive screen time has on its customers, but out of pressure from the Chinese government. Thus, it is possible to limit the adverse effects of social media, but there must be political will.
Alternatively, we can also turn to subsidies (or taxes) to support physical, local interactions. After all, online sales were tax-exempt at the dawn of the internet. It was a form of subsidy to promote a new sector. Thus, nothing prevents subsidizing physical stores and venues for the interaction benefits they produce. That is, we can make it cheaper to remain engaged with our physical surroundings by giving a price advantage to the economic activities taking place in them.
It is unlikely, however, that such approaches will work in this context. The optimal design of regulation or taxation requires stability and predictability, and the digital economy is too voluble at this stage to lend itself to this approach. Furthermore, this would be too indirect an approach. The concern is with democracy. Regulation and taxes affect incentives at the margin, in the hope that individuals might devote more time to their physical lives, which in turn could improve the democratic process. This seems a bit of a stretch. Shouldn’t we try to attack the problem more directly?
The more progressive approach would start by questioning how democratic representation has been implemented since the American Revolution. In the United States, the pool of candidates for a jury isn’t elected, but is randomly drawn. Jurors aren’t elected, because the only voters interested in that election would be the parties involved in the trial, all but ensuring a biased verdict. In the same way, in a digital world where citizens don’t care about their physical locations, local elections can’t guarantee fair representation; only a random draw would.
“We should take inspiration from the idea of the citizen assemblies.”
For this reason, we should take inspiration from the idea of citizen assemblies to administer the local provision of public goods. This is particularly important when there are strong vested interests. At the national level, these vested interests are generally large corporations, and at the local level, developers and public-employees unions. Take, for instance, the local police. Its union is a crucial ally in mayoral elections, and it can easily jeopardize the re-election of an incumbent mayor by de-policing in protest of a mayoral decision. The result is that it isn’t the elected mayor who controls the police but the police who control the elected mayor. The same is true for the teachers’ union.
A randomly drawn assembly would have much more control over the police. It wouldn’t owe its election to the police, nor would it cave in for fear of not being re-elected. Most randomly selected representatives, knowing they will return to be subjected to the police after their brief mandate, would want to ensure that the police are effective, but not abusive. Thus, citizens’ assemblies can reduce the distortions produced by representative democracies and increase the trust in the democratic process.
The beauty is that this experiment could be launched on a small scale. We can carve out some public provisions of local public goods (parks, policing, or street cleaning) and entrust citizens’ assemblies to manage them, re-energizing local democracy and rediscovering Tocqueville, so to speak. We can then study the success of this new form of democratic representation and its costs. If the experiments are successful on a small scale, we can scale them up. Thus, the cost of the trial is limited, and the potential benefit is very large.
It is here that the technology of digital networks could even help revitalize our democracy. Our citizen assemblies would be geographically local and would have a strong physical component. But there is no reason why they couldn’t harness the power of digital networks, from the virtual meeting to the sharing of important documents, data, subcommittee virtual rooms, and libraries, and so on. Technological advancement, indeed—but at the service of local democracy.
Digital networks live outside society, so to speak; they propose another society, and they aim to substitute for the array of physical networks we live in. And Big Tech platforms have the upper hand, because they have a governance advantage: They are owned, whereas physical society is mostly not. They are in an antagonistic relationship with physical society, and they are much more powerful, akin to having a grand chess master compete against an average player. Thus, Big Tech might well lead to the demise of our traditional society—and, with it, democracy.
Digital platforms didn’t set out to destroy democracy, to be sure, at least not initially. Rather, democracy’s demise is the collateral damage involved in a business model that, to be as profitable as possible, requires the destruction of the physical networks of society. Society can, of course, fight back and pursue a variety of strategies to reassert itself against the full digitalization of life. However, this is a risk for digital platforms, which may see, in the rebirth of local democracy, a threat to their business model.
The question is then whether, realizing this risk, digital platforms may turn against democracy itself. Should we fail to weigh the gravity of the question, we will find our chance to shape our future in a democratic way slipping out of our hands.