Is social media destroying democracy? It’s a question that animates discussions around the world. But it’s striking that the debates so often focus on the “social media” part of this question and seldom interrogate the other part — democracy. What actually is democracy, and what parts specifically might social media be undermining? These are not academic issues but ones that strike at the heart of how governments, platforms and civil society conceive of this problem.
Even more striking is where this question is asked. In early June 2021, more than 100 prominent scholars signed a statement of concern about American democracy. Similarly, an enormous amount of ink was spilled over the Facebook Oversight Board ruling on former US president Donald Trump’s suspension from the social media platform as well as over Facebook’s subsequent decision to lift the special designation for politicians, a move again based on Trump’s use of the platform. Yet 96 percent of the world’s population lives outside the United States. One statement of concern from 35 global civil society organizations noted that “a knee-jerk response for a few select markets might be easier and cheaper for the companies, but not appropriate for defending democracy and human rights.” Whose democracy counts when global rules are made on the basis of one US politician’s behaviour?
Instead of talking about democracy as an abstract, ill-defined ideal, we can examine different definitions of democracy and consider the role of social media in each. By breaking down the components of democracy, we can piece together a more complete picture of how platform governance might address each issue. Perhaps even more crucially, we can follow a global thread to see how solutions that might seem to work in one democracy may prove harmful or extremely complicated in another.
Scholarship on democracy is, of course, voluminous, and the concept itself debated. What democracy meant in the nineteenth century — before, for example, women could vote — is obviously very different from what democracy means in the twenty-first century. Indeed, Steven Levitsky and Daniel Ziblatt’s book How Democracies Die drew on historical examples from around the world to understand how democracies can fade away. So, too, democracy differs depending on where we are — a point that seems less clear in the realm of social media, where companies often seek global solutions.
In a report for the European Union, Stephan Lewandowsky and his colleagues defined democracy using the three essential principles of “equality, representation and participation.” Jennifer Forestal has reached back to Hannah Arendt to posit that democracy online needs to foster “both equality and difference.” Forestal suggests that online spaces such as Facebook should be judged on their ability to bring together users around shared interests while simultaneously enabling them to disagree and put forth diverse ideas. Other definitions of democracy discuss notions of popular sovereignty or responsiveness to public demands.
Yet each one of these definitions can be used to justify both democratic and undemocratic solutions. One example of this problem lies in ideas around sovereignty. In February 2020, Ursula von der Leyen, president of the European Commission and former German defence minister, described the European Union’s new digital strategy as gaining “tech sovereignty” from US-based platforms and retailers such as Amazon, Google and Facebook. Such debates have underlain the drafts for the Digital Services Act and the Digital Markets Act currently under debate in the European Union.
In early June 2021, more than 100 prominent scholars signed a statement of concern about American democracy. Similarly, an enormous amount of ink was spilled over the Facebook Oversight Board ruling on former US president Donald Trump’s suspension from the social media platform as well as over Facebook’s subsequent decision to lift the special designation for politicians.
At the same time, few seem prepared to confront how such discourse can unintentionally sound remarkably similar to efforts in authoritarian regimes such as Russia. In early 2019, for instance, Russia passed a “sovereign internet” bill to “ensure the reliable operation of the Russian segment of the internet in the event of disconnection from the global infrastructure of the World Wide Web.” The Russian bill clearly differs substantially from European efforts. For one thing, Russia is aiming to create an autonomous system that can easily be closed off from the outside world. At the same time, it shows how sovereignty talk can swiftly slip from democratic to authoritarian realms.
Other solutions, too, can seem welcome in one democracy but worrisome elsewhere. For some observers, the German Network Enforcement Act (Netzwerkdurchsetzungsgesetz, or NetzDG) of 2018 showed that states could push social media companies to adhere to national law. The NetzDG law required social media companies to respond within 24 hours to reports of speech that violated any of the 22 statutes of German speech law or else face steep fines. While this law has vastly increased the number of content moderators working in the German language for Facebook, for example, it also raised concerns about serving as a model for less democratically minded countries that sought to suppress speech, as Jacob Mchangama and Joelle Fiss pointed out in a report for the Copenhagen-based think tank Justitia.
Mchangama and Fiss note that NetzDG incorporates democratic safeguards and that many authoritarian countries would likely have proceeded on this path anyway. But they suggest that NetzDG “seems to have provided several states with both the justification and the basic model for swift and decisive action.” Some policy makers have argued that authoritarian states will censor and limit speech regardless of how democracies act. Others believe that democracies should create solutions that cannot be easily copied for anti-democratic ends. Such critics worry that NetzDG’s very construction lent itself to adaptation by authoritarian states seeking to restrict legitimate critique. While many authoritarian countries were already pursuing such censorship measures, it is important to consider how democratic countries might have set a higher standard of democratic accountability that would be harder to co-opt for authoritarian purposes.
Another example is the idea of making a local representative legally liable for a company’s actions. This approach seems intuitively helpful in democracies such as the United Kingdom, which was unable to subpoena Facebook CEO Mark Zuckerberg to testify before the British Parliament in 2019. That problem helped to spark the creation of the International Grand Committee, a coalition of governments around the world seeking to hold social media companies to account. And, indeed, holding C-suite executives liable for their company’s actions could make a tremendous difference to their actions.
At the same time, Turkey’s new social media law of 2020 makes a similar-sounding requirement. Since Turkey has already severely constrained the freedom of newspapers, television and the judiciary, social media seemed to be the last bastion of freer political debate. Forcing Facebook, Twitter and other social media companies to choose between appointing a local legal representative or being blocked from the country could be “the nail in the coffin for online freedom of speech” in Turkey, according to Cathryn Grothe of Freedom House.
In the end, it did not matter too much to Canadians if they could not see ads served by Google when there were plenty of other places to learn about the election....what would happen if we ran a simulation of that Canadian policy in other democracies around the world?
Again, Turkey’s suggestion does differ from other democracies’ idea of how to hold executives liable. In Turkey, the representative will be made liable for content removals ordered by the court, which are supposed to happen within 24 hours. But this should remind democracies that no policy suggestion lives in a vacuum. Even more crucially, we should remain wary of solutions based on negotiations conducted solely between companies and governments. Civil society organizations warn about potential dangers to free speech for a reason: they have seen how regulations can be co-opted to suppress human rights around the world.
Given the global nature of platforms, therefore, it may behoove democracies to add global ramifications to their list of policy considerations. When Canada promulgated its Elections Modernization Act in time for the 2019 federal election, the act contained important transparency provisions for advertisements on social media. Facebook chose to comply and serve election ads in Canada; Google did not.
On the one hand, this seemed to show that democracies can still use regulatory power to ensure that their elections are run in a free, fair and transparent fashion. In the end, it did not matter too much to Canadians if they could not see ads served by Google when there were plenty of other places to learn about the election.
On the other hand, what would happen if we ran a simulation of that Canadian policy in other democracies around the world? Would the policy similarly make sense for these jurisdictions? Might such a test be as important a part of policy making as algorithmic impact assessments?
As David Murakami Wood recently warned, national solutions will only go so far: because platforms are “emerging alternatives to nation-states,” each local solution can only be “part of a process.” The other issue is that solutions have unintended consequences, often in places far beyond their original implementation. These are some of the many reasons why a global platform governance research agenda is so important. Sonja Solomun has laid out many more in describing a new platform governance research network that was launched with its first conference in March 2021 (I was one of the co-organizers). Policy makers who believe in democracy have a duty to think about and understand the other 96 percent of the world’s population.
Once we recognize that democracy itself is not a stable historical or geographical concept, we can ask more interesting questions, such as, “How is social media affecting different aspects of democracy in different places?” We can recognize that a regulation that seems to shore up democracy in one place may undermine another democratic value elsewhere.
One way to think about democracy is as a Venn diagram of values. For different democracies, these values overlap in different ways. And, indeed, no democracy really achieves all those ideals. As Astra Taylor has put it, democracies strive to be democratic but perhaps never wholly are. One crucial way for them to move forward is to recognize that policies made in one democracy can have unstable, unexpected and undemocratic consequences elsewhere.
There are no easy solutions to the dilemmas that I have raised. Democracies want and need to move forward with approaches to address the ecosystemic challenges of platform governance. But the cautionary tales outlined above mean that European and North American democracies cannot remain ignorant of the ramifications of their policies elsewhere.