Meta-Analysis: Social Media Does Little to Inform Citizens

Justin Hendrix is CEO and Editor of Tech Policy Press.

In his Facebook manifesto, “Building Global Community,” the platform’s founder and CEO, Mark Zuckerberg, pondered how the site might contribute to a more “informed community.” The phrase appears six times, alongside other phrases such as “civic engagement” and “common understanding.” But has Facebook– or Twitter, or social media more generally– contributed to a more informed citizenry?

A meta-analysis of past science on the subject published in the Journal of Communication says not really.

Hebrew University of Jerusalem’s Eran Amsalem and University of Haifa’s Alon Zoizner looked at 76 studies that investigate “the relationship between social media use and political knowledge,” which combined represent data gathered from 442,136 people. While a growing number of people report that they get news through social media sites, the results of the review “reveal little evidence that people learn about politics on social media,” with an average effect of virtually zero across all the studies that is “consistent across social media platforms, types of political knowledge, samples, countries, and periods.”

“Our results show that despite the huge potential for learning on social media, in reality, citizens seem to learn very little about politics from using these platforms,” write Amsalem and Zoizner.

In order to determine the extent to which specific characteristics identified in the underlying studies may influence the effect of social media use on political knowledge, the authors looked at five categories: “general knowledge” of the “basic institutional arrangements” of politics; knowledge of “day-to-day politics and recent political events”; “issue-specific knowledge” of particular policy domains; “campaign-related knowledge” of candidates, parties, and their issues; and an “other” category that describes, for instance, the combination of one or more of these definitions into one variable. The meta-analysis found “no statistically significant influence on political knowledge” across any of the five types.

The authors say these findings– which are mostly based on studies of Facebook and Twitter– are important to the ongoing debates about “the consequences of social media on democratic citizenship,” and that they support scholars whose work suggests “that access to the digital information environment, and to social media in particular, has no effect or even has a negative impact on democratic citizenship.” This is particularly troubling, say the authors, since people are spending more and more time with social media and less time with traditional news media, which decades of research suggests do in fact “contribute substantially to citizens’ political knowledge,” a finding that “cannot be generalized to social media.”

The results hold across both election and routine periods, raising questions about whether the investments Facebook and other platforms have made in initiatives such as addressing misinformation and developing hubs for civic information have had any meaningful impact in the face of the negative effects of social media, including information overload, misinformation and other factors that may impede citizens from becoming more informed about politics.

Interestingly, the null effect is somewhat counter to the views of ordinary citizens, at least according to one recent poll. In 2022, the Pew Research Center fielded a survey on technology use and views of social media and its role in democracy in 19 countries. It found that most people believe they know more about foreign and domestic events thanks to social media.

Perhaps, like Mark Zuckerberg, they are more optimistic than they ought to be.