You Like this and I Like that: is Facebook just an echo chamber, or is something more complicated going on?

Does Facebook create ‘echo chambers’ for its users, or is something more complicated going on? Nick Anspach (York College of Pennsylvania) created mock News Feeds and found that seeing friends endorse and comment on political items made participants 25% more likely to select political news from sources with which they agreed – and 40% more likely to pick news from counter-attitudinal sources. Recommendations from friends clearly help to expose people to a wider range of political views. But the flipside of this finding is that when they encounter arguments that challenge their existing attitudes, politically-sophisticated audiences regularly reject them, causing a backlash effect.

checking phones new york

Photo: Jim Pennucci via a CC-BY 2.0 licence

In today’s fragmented media environment, it is common to hear talk of echo chambers, filter bubbles, or selective exposure. No matter the term used, the concept is generally the same: those interested in politics choose from like-minded media sources, surrounding themselves with opinions that reinforce their partisan sentiments. A Democrat, for example, is more likely to get her news from MSNBC or the Huffington Post than from Fox News or Breitbart. Cable television and the internet has led to a proliferation of news options, so it is now easier than ever to find niche sources that share our attitudes.

Indeed, these technological advances have also expanded our entertainment options, giving audiences a chance to opt out of news and ignore politics altogether. This expansion of choice has created a traditional media environment where most people avoid partisan media, and those that don’t increasingly find themselves in echo chambers.

And yet, despite this new ability to tailor our media diet as we see fit, the rise of social media has further changed how people encounter political news. In traditional contexts, exposure to news is usually the result of a purposive act; I decide to pick up a newspaper, visit a website, or to tune into a news broadcast. But on sites like Facebook, audiences often encounter partisan news inadvertently. Users may log on for a few moments’ diversion, but encounter political news in the same space as their friends’ posts or pictures. In fact, a recent Pew report finds that 78% of those who receive news from Facebook do so when they are logged on for other reasons. Social media thus funnel partisan media to millions of individuals who may have otherwise avoided such information.

But if most Americans ignore partisan news in favour of entertainment options in traditional contexts, why wouldn’t they do the same on social networking sites? In an article recently published in Political Communication, I examine why we pay more attention to political news when it is shared through sites like Facebook. I find evidence that the social aspect of social media drives users towards news that they may normally avoid. Social media features such as Twitter’s retweets and Facebook’s Likes and comments make it easier than ever to share, endorse, and discuss content, facilitating a level of interaction not possible with TV broadcasts or newspapers. The result is that someone who may not be particularly interested in politics may nevertheless decide to read a partisan post if it has generated an interesting discussion among her friends.

To test the influence these social endorsement features collectively have on our selection of news, I created a survey experiment in which participants selected articles from a series of mock Facebook News Feeds. In one mock feed, none of the included posts received Likes or comments. In that feed, a majority of participants chose an entertainment post over a political post. Of those that did select a political article, 66% chose the article that shared their ideological commitment. This indicates that absent social media’s endorsement features, people behave on Facebook as they do in traditional media environments: most tend to avoid politics, and those that don’t tend to select from like-minded sources.

However, once I manipulated the mock News Feeds to include social endorsements, interesting patterns emerged. Likes and comments attributed to fictional individuals did nothing to change participants’ selection behaviours, but attributing endorsement activity to the participants’ actual friends and family members (whose names and endorsement activity I collected in an earlier portion of the study) caused participants to select those endorsed posts. With friends’ Likes and comments included, participants were 25% more likely to select political news from sources with which they agreed, and perhaps most interestingly, 40% more likely to select news from counter-attitudinal sources. Social media endorsements not only compel users to read political news, they also draw audiences out of their echo chambers. That I only found an endorsement effect for friends and family members, and not for fictional individuals, speaks to the importance of the social aspect of social media. Social networking sites’ ability to share and endorse content easily sets them apart from traditional types of media, and this distinctive feature increases political news exposure for thousands, if not millions, of previously uninterested individuals.

But what effect does this exposure to political news and partisan discussions have on these new audiences? On one hand, social media may facilitate good democratic discourse, informing each side of the merits of the other side’s ideas. But on the other hand, political debates on social networking sites have a reputation for being overly aggressive or ill-informed. If true, witnessing such discussions may turn people off from politics. Additionally, exposure to partisan media may also polarise social media audiences. When exposed to arguments that challenge their existing attitudes, politically-sophisticated audiences regularly reject those arguments, causing a backlash effect that further polarises those most interested in politics.

And though there is reason to suspect that uninterested or uninformed individuals will be the least likely to dig in when their existing attitudes are challenged, research currently under peer-review finds that politically-unsophisticated Facebook audiences also reject information with which they disagree. Such behaviours exacerbate, rather than mitigate, mass polarisation, suggesting that though the potential benefits of social media are many, we should be equally cognisant of the platform’s detrimental effects on our democracy.

This post represents the views of the author and not those of Democratic Audit.

Nick Anspach is an Assistant Professor in Political Science at York College of Pennsylvania. His research focuses on political psychology and political communication, specifically social media, political interest and learning, and psychophysiological responses to political stimuli.