Has the internet spelled the end of democracy?
When most people ask this question, they are thinking about what the internet does to the politics of governments: the Cambridge Analytica scandal and QAnon, the app-driven election campaigns of populist strongmen like Jair Bolsonaro in Brazil and Rodrigo Duterte in the Philippines, #fakenews, #deepstate and so forth. There are lots of good reasons to worry that the answer might be yes.
But what if we ask the inverse question? What is the fate of democracy on the internet itself? That is, do we experience democracy online? Are the services you use or the communities you’re part of or the terms of service you sign meaningfully democratic? What has the internet taught us about the governance of everyday life? Just as the French diplomat Alexis de Tocqueville saw the “art of association” in civic life as the engine of democratic government in the young United States, how we associate on apps and screens has consequences for what we expect from the halls of power.
The optimists and the pessimists on internet democracy share a basic assumption: The internet is ushering in something radically new and different about how people relate to each other. Either it is warping users’ brains or it is realizing the long-elusive “global village.” Either it is supercharging protest movements or reinventing authoritarianism — or, for subtler observers, both at the same time.
Online, it’s rare to find what we might expect from a well-run civic association — officers, elections, meeting minutes and more. Instead, we tend to see unelected admins, moderators and so-called “benevolent dictators” running online communities, from Facebook groups to subreddits. These practices have a long history that sheds light on politics in the internet age.
There are much older forms of collective rule than the trappings of democracy most of us see today — the competitive elections, the party platforms, the term limits. A great many societies on multiple continents over several millennia practiced a form of rule in which leaders are held accountable not at the ballot box but, instead, by the fact that if they sought to rule alone, and not jointly with their people, they could never get anything done.
We can call this type of governance “early democracy.” It closely parallels what is happening in many corners of the internet today. It is a form of governance where rulers — like online benevolent dictators — were not necessarily chosen by their people, but they faced significant constraints on their rule. Most importantly of all, rulers were in a weak position with respect to society.
The core feature of early democracies was that rulers relied on their people to provide information about production and to aid with governance, whether it be raising revenues or engaging in external defense. They did this in the form of councils and assemblies and by delegating authority. It resembled an online platform where those at the center are always dependent on outside developers for content.
The other core feature of early democracies was the threat of exit: Instead of voting at the ballot box, people could vote with their feet. Here again, we see a parallel in many, though certainly not all, online contexts. It is the threat of moving to another community that keeps would-be dictators benevolent.
Council and assembly governance has been prevalent in so many societies that it appears early democracy was something that came naturally to humans — it wasn’t just the Greeks who invented it. Europeans saw this when they began to conquer other peoples on other continents, even if they often failed to appreciate the importance of what they witnessed. In the 17th century, French Jesuit missionaries in North America discovered that the people they called the Huron (they called themselves the Wendats) had a sophisticated version of council governance that ranged from village, to tribe, to a larger confederacy. The Jesuits referred to this as the “estates general” of the Huron.
Further south in Mesoamerica, Hernán Cortés discovered in 1519 what would come to be known as the republic of Tlaxcala: a society with a governing council but no single central ruler. Cortés thought this resembled the form of governance that he knew to exist in Venice, Genoa or Pisa. Several millennia earlier, there had also been the Mesopotamian kingdom of Mari, which, like the Huron and the Tlaxcalans, relied heavily on council governance.
Without governing jointly with their people, rulers would have accomplished little. Early democracy prevailed whenever rulers were in a weak position because they lacked information and a state bureaucracy that could allow them to rule on their own. When people’s allegiance could be compelled rather than earned, it was autocracy that ensued, and this happened in places like China, Sumeria and ancient Egypt. When we think of some of today’s largest internet titans, and the massive resources that they have at their disposal, we can see a similar situation.
If we want to understand what is happening on the internet today, then it makes sense to focus on early democracy — as well as early autocracy — rather than modern democracy, where popular participation in governance occurs more or less only at the ballot box. Modern democracy is a system where those who govern are elected by universal suffrage. But modern does not necessarily mean better.
It is striking, when you stop to consider it, how few aspects of modern democracy have made their way into online communities. A typical union or parent-teacher association elects its leaders, but Facebook groups usually do not; the company provides no tools designed to enable such an election. Online communities do not tend to establish juries, elected boards, bylaws, referendums or other procedures of self-governance. Cases where such structures do exist, like Wikipedia, are outliers.
Meanwhile, moderators expend tremendous amounts of lonely time and energy on community upkeep, generally without compensation. Maintainers of open-source software projects struggle to onboard partners and successors. It seems like more regimented forms of democracy could come in handy.
Still, even under the autocratic defaults of most platforms, there is considerable space for democratic practice. Community moderators feel pressure to behave well because their members could fairly easily leave the community and form a new one. Maintainers of open-source software projects can drive away talented contributors if they behave badly. The threat of exit, as in early democracies long ago, carries considerable force. People with power online have reputations to protect, and they can face a barrage of public shaming if they’re not careful.
Even Mark Zuckerberg appears to grasp this. As part of a post-mortem publicity tour after the 2016 election, the outcome of which many people blamed him for, he published an essay titled “Building Global Community.” It outlined plans — since implemented — to shift the platform’s focus from its news feed to groups, to “meaningful communities.” And it suggested that, “as the largest global community, Facebook can explore examples of how community governance might work at scale.”
Zuckerberg seemed to recognize that neither he nor his employees could oversee the norms and needs of a global network alone; better community-scale governance would be needed. He even spoke of users as “citizens.” And since then, the platform has introduced a quasi-independent oversight board whose initial members the company appointed.
But still: no ballots for community members, no modern democracy. And Zuckerberg remains one of the most unaccountable people in the world.
What if he went further? Consider the example of Python, an open-source project that produces one of the most popular programming languages in the world. For decades it was led by a self-described “benevolent dictator” — founder Guido van Rossum — until he announced his intention to step down in 2018.
The community entered into a period of deliberation and soul-searching, exploring everything from more dictatorship to open anarchy, only to emerge the following year with a new structure: a community-elected board, one probably not so different from what governed the civic associations de Tocqueville saw during the early days of modern democracy in the U.S.
If the pattern of Python spreads to other online communities, we may see the internet retrace offline history’s steps toward a reinvention of modern democracy.
The history of offline democracy, however, offers cautionary lessons. Modern democracy is far from the inevitable culmination of participatory politics. As political scientists have long observed in America, most people participate politically only when they vote. The risk with modern democracy, therefore, is citizen disengagement and distrust of government.
The trappings of modern democracy have not yet come to the internet, and it is possible they won’t — or if they did, they might not work very well. Would people really feel they held power if all they could do was vote for members of a governing board every few years? Perhaps making the internet more democratic can be better done by starting with the practices of early democracy and working from there toward possibilities that modern democracy hasn’t tried yet.
Democracy can mean many things and be practiced in many ways. As premodern societies attest, it does not always mean elections. The ancient Athenians, for instance, relied heavily on the selection of officials by random lot. Among the Huron, chiefs inherited their positions as part of a lineage, but they could be formally deposed if people were unhappy, and councils constrained their power.
With accountability crises cascading across our modern networked world, it is time to get serious about the next evolution of democracy online — or risk jettisoning any semblance of democracy altogether.
Historically, democracy has tended to evolve and flourish when power couldn’t become too concentrated. Thus, proposals for more active antitrust enforcement of online platforms resonate back through the centuries; similar limits on concentration could apply to communities within platforms as well.
Also relevant are more technical proposals for requiring platforms to be more interoperable with their data and functionality; if users can move from platform to platform more easily, carrying their data with them, they can put greater pressure on platforms to be accountable, just like in early democracies.
Making public investments in free, open-source software, as European governments are increasingly doing, also creates a kind of “public option” that offers exit alternatives and sets common standards for private platforms. Adjusting the logic of Elizabeth Warren’s Accountable Capitalism Act, platforms of a certain size might be required to include user representation on independent oversight bodies or corporate boards.
Regulation need not come only from above. Also important for evolving platforms’ democracy is enabling new forms of counterpower. This can include building associations among groups of users, such as sellers on Amazon and drivers on Uber, or trusts that hold the data of social media users, so that ordinary people can use collective leverage to make their voices heard. Some platforms could be reconstituted as cooperatives or trusts. Counterpower can force platforms to develop creative strategies for ensuring that people’s voices are heard before an outright confrontation occurs.
Advanced online democracies may end up looking very different from modern democracy — which is, after all, largely a product of 18th-century ideas. Governance hackers are experimenting with opinion polling through artificial intelligence, cryptographic juries for resolving disputes and the dynamic selection of delegates in real-time, issue by issue. Rather than simply replicating modern democracy, we have an opportunity to develop previously untried forms of democracy.
If online governance is to evolve out of its current crises, toward something that is more rather than less democratic, we will need to approach the task with some humility. Rather than imagining that our admins and moderators and super-CEOs have this mostly figured out, we need to recognize where we are, historically speaking. In the language of the techies, online governance is early-stage.