What the Research on Social Media’s Impact on Democracy and Daily Life Says (and Doesn’t Say)

A recent article by Jonathan Haidt in The Atlantic, titled “Why the Past 10 Years of American Life have Been Uniquely Stupid” has garnered a lot of attention. It is easy to see why just from that title. While Haidt certainly raises valid questions about social media’s impact, the preponderance of research tells a far more nuanced story and many of these questions lack clear answers.

For example, evidence simply does not support the idea that Facebook, or social media generally, is the primary cause of polarization. Research from Stanford last year looked in depth at trends in nine countries over 40 years, and found that in some countries polarization was on the rise before Facebook even existed, and in others it has been decreasing while internet and Facebook use increased. 

Research showed that polarization increased among demographics least likely to use the internet and social media, suggesting greater internet use is not associated with faster growth in political polarization among US demographic groups.

There are also studies showing that mainstream media plays a bigger role in disseminating disinformation than popularly accepted. A Harvard study published just ahead of the 2020 US election found that a campaign meant to cast doubt on the legitimacy of mail-in voting was driven more by elites and through mass media with the study finding that “social media played only a secondary and supportive role.” It went on to note that addressing this is “likely to require more aggressive policing by traditional professional media, the Associated Press, the television networks, and local TV news editors.” 

More and more research discredits the idea that social media algorithms create an echo chamber that causes polarization and political upheavals. The Reuters Institute noted that people who use online search and social media for their news are “significantly more likely to see sources they would not normally use.” And in Are Filter Bubbles Real?, author and professor Axel Bruns notes: “[M]ost claims about echo chambers and filter bubbles and their negative impacts on society are significantly overblown. These concepts are very suggestive metaphors, but ultimately they’re myths.”

Haidt’s article states, “Recent academic studies suggest that social media is indeed corrosive to trust in governments, news media, and people and institutions in general.” That is not actually what the majority of credible studies show. Nor is it consistent with the shared experiences of anyone who lived through the decades preceding social media’s emergence where plenty of angry voices were amplified on cable television, talk radio, and in newspapers.  

The strength of civic institutions has been declining long before social media was invented. One of the seminal works about this is Bowling Alone by Robert Putnam and was published in 2000. Trust in institutions has also been declining for decades, especially in America. Pew’s trust in government index, which goes back to the 1960s, shows the clear trend line. In fact, as the Pew study says, trust in government in the US has been declining since Watergate and shows no sign of accelerating since social media was invented. The World Values Index also shows how trust varies in different countries even as it is declining in the US. 

One of the most recent comprehensive sources of academic literature looking into the question of social media’s impact on democracy was published by the Digitization and Democracy working group. There was no clear consensus on social media’s role due to the variety of other social factors at play, as one of the members of that working group laid out in a compelling Twitter thread. 

None of this is to suggest that the concerns Haidt raised aren’t valid, especially as they relate to social media’s design contributing to the strident tone of some online discourse. However, the piece also assumes — without sufficient evidence — that the design of social media alone is the key driver of social changes such as a breakdown in critical thinking and the demise of bipartisanship that we clearly see in American political discourse.  

Social media has given a voice to billions of people around the world. It is a technology that is as empowering as it is disruptive. It even helps many people connect to content about the very issues raised in this piece. For example, many people found and read this piece in The Atlantic through social media. 

We need more academic research to better understand social media’s true impact, especially on democracy in America. That is one of the reasons why Meta is investing in open research and transparency including the Facebook and Instagram Election Study, which is a large-scale intensive collaboration between internal researchers and external academics to understand the impact of our products on key outcomes in the US 2020 election, including polarization. We are also supporting the URL shares dataset release and other efforts to study our impact on elections and democracy while still protecting the privacy of our users.

There’s more we can do to improve our own platform, based on what the research actually says, to help amplify the good and minimize the bad. And that work is continuing.