life in pixels

Did Facebook Cause Riots in France?

Photo: Etienne De Malglaive/Getty Images

In the days since protesters wearing yellow vests (gilets jaunes) began taking to the streets of France in the hundreds of thousands to voice their opposition to a proposed gas tax hike — and, more broadly, to the deeply unpopular administration of President Macron — English-language media has begun circling around a particular story of causation: This is Facebook’s fault. In Bloomberg, Leonid Bershidsky writes that “Street riots in Paris are less about a tiny fuel tax hike than the power of social networks to radicalize their users”; on Medium, Frederic Filloux argues that Facebook is “fueling the French populist rage.” Most widely circulated is a lengthy and detailed Buzzfeed article headlined “The ‘Yellow Jackets’ Riots In France Are What Happens When Facebook Gets Involved With Local News.”

The general story goes something like this: Earlier this year, Facebook changed the way it sorts its news feed in the hopes of reducing partisan squabbles and links to “fake news,” and began promoting posts from friends and family and semi-private Facebook groups over posts that link off the website. Unwittingly, however, the company was promoting mass unrest, by pushing into users’ feeds, memes, and rants from the semi-political populist groups that would become the core organizational structure of the gilets jaunes.

It’s a compellingly dystopian way of thinking about the riots, in which hundreds of people have been injured, especially if you’re a Facebook critic or skeptic. Look at what Facebook brings to stable democracies! Look at how Facebook leads good citizens astray! The problem is that there’s very little evidence being provided for this particular narrative. We know that Facebook has adjusted its News Feed sorting in various ways over the last year. We know that some of the protesters have used Facebook to organize themselves. But it seems like a big stretch to go from there to calling the movement “a beast born almost entirely from Facebook,” as Buzzfeed does.

That’s not to say that Facebook was irrelevant to the protests. There seems to be consensus that the social network is the organizational platform of choice for the gilets jaunes. But the idea that popular outrage is more about “the power of social networks” than actual French politics, as Bershidsky argues, seems very wrong, and more than a little irresponsible. “Some in Paris have suggested all gilets jaunes are driven by fake-news and conspiracy theories on Facebook, & are somehow uneducated,” Guardian Paris bureau chief Angelique Chrisafis tweeted on Friday morning. “That was not what I found and it would be a mistake to think that … ” At one barricade, Chrisafis spoke with a wide range of citizens “united in fury at Macron’s way of running France — what they called his top-down approach cut off from ordinary people’s experiences. Everyone could angrily quote examples of Macron’s ‘arrogance.’” This sounds like real grievance, not inauthentically promoted “fake news.”

Facebook is obviously an important component to the gilets jaunes protests, but some care is needed about how we attribute causality or we run the risk of misunderstanding not just the situation in France, but the relationship between social media and protest movements. Studies of the “Arab Spring” indicated that while social media was an important structural factor in enabling the fast pace and organizational capacity of the protests; its use almost always followed, rather than led, the core political activity — that is, it was an accompaniment to broad political dynamics rather than their driver. Twitter or Facebook might have accelerated or heightened those dynamics, or become a field for their interplay, but they were not primary causes.

In discussing the role of Facebook in the gilets jaunes protests, we might want to keep in mind the “first principle” of the internet’s relationship to protest movements, articulated by political scientists Gadi Wolfsfeld, Elad Segev, and Tamir Sheafer in their paper “Social Media and the Arab Spring: Politics Comes First”: “One cannot understand the role of social media in collective action without first taking into account the political environment in which they operate.” One aspect of the political environment of France right now is the president’s deep unpopularity: is that a more important a causal factor than Facebook? Given what we know about the protestors’ anger and Macron’s popularity, the riots would likely have happened even in Facebook’s absence, just, perhaps, in a different way, at a different speed, or in different places. It seems insulting to France’s long national heritage of rioting to imply that the gilets jaunes are only gathering in the streets thanks to an American company.

But we should also start moving away from theoretical discussions that imply the possibility of some counter-historical “control world” without Facebook, which has now helped shape global politics (and been shaped by global politics) for a decade. Facebook is less a “cause” than a condition, as sociologist Zeynep Tufekci implied when she distinguished between different kinds of causality to answer the question of whether or not the Egyptian or Tunisian uprisings were “social media revolutions”:

In this schema, material causes are the substrate of things. Does metal cause cars? In some sense, cars as we know them wouldn’t exist without metals so it meets a “but for” definition of causality. So, in some sense cars are caused by metals in that no metals, on cars — at least in their current form. However, in everyday usage, most of us tend to use the other two definitions, the efficient cause, i.e. cars are there because someone manufactured them; or the final cause, i.e. cares are there to take us from place A to place B in a speedy (but polluting!) manner.


So, I think most of the people using the term “social media revolution” are using it in the sense of a material cause. As I asked on Twitter during the debate, would we call the French Revolution a printing press revolution? Surely, the invention of the press is a strong antecedent of that revolution. But also surely, that revolution was made by people, through political action. So, the printing press just defines the milieu in which the revolution took place; it is an inseparable part of the French revolution even though it is not the efficient (political uprising) or the final (establishing a republic) cause of the French revolution. But you cannot really imagine a French Revolution, of the kind that happened, without the printing press.

It’s always tempting to lay blame for political change at the feet of a company that has consistently shown itself to be clumsy and insensitive about politics. But to the extent we’re able to separate out and measure Facebook’s effect on society, it seems increasingly clear that its influence is more important on a macro, structural level — in the way it shifts and opens up an entire media-political ecosystem — than on an immediate, individual behavioral one. Studies of the 2016 U.S. election have shown, for example, that Fox News was a more important immediate vector for spreading conspiracy theories than Facebook was, but that Facebook’s dominance has indirectly created an environment — a hyperpartisan “attention backbone” — that encouraged Fox to spread those conspiracy theories. And even this demonstrates how difficult it is to isolate Facebook as a variable, rather than (as Tufecki puts it) the milieu in which politics takes place. Would we say that Donald Trump’s election was more about Facebook than about American politics? Or that his presidency was a beast born almost entirely of Facebook? To say the same about the gilets jaunes risks missing the forest for the trees.