The myth of the online echo chamber
Contrary to popular belief, we now hear more diverse voices than ever before – studies suggest that most people do not live in Facebook or Twitter echo chambers and 'filter bubbles'. So why is global politics still so divided?
Back in the early 2000s many commentators were still marvelling at the freedom of the internet and its democratic potential when the US legal scholar Cass Sunstein offered a stark warning.
This virtual Wild West, he said, might allow us to overcome some of the social and geographical barriers between people, so that we establish a more balanced view of the world around us. But it was equally possible that we would simply erect new fences, as like-minded people siphon themselves into homogenous groups who all share the same viewpoints and gather their information from the same sources.
“Although millions of people are using the Internet to expand their horizons, many people are doing the opposite, creating a Daily Me that is specifically tailored to their own interests and prejudices,” he wrote. They would, in effect, live in ‘echo chambers’, leading to greater polarisation in a country’s politics.
You might also like:
• Is social media bad for you?
• How ‘collective narcissism’ is directing world politics
• What I learned by testing my stress levels on social media
Later commentators embraced the idea, while also pointing out that the technology platforms themselves may drive a further wedge between the different groups. Facebook and Twitter, for instance, may work out that you are more likely to click on shared stories from the New York Times rather than the Daily Mail, and so preferentially promote those stories on your feed.
“It’s done because there is so much information, because one person could not hope to consume it all,” says Elizabeth Dubois at the University of Ottawa. “And that’s a really helpful tool but it does mean that you end up in a bubble based on what that platform or company has decided is most likely to suit you and your purposes.”
Today, the risks of the echo chamber and the “filter bubble” are considered something of a truism, explaining the bitter divides in public opinion that often appear to toe strict party lines. Nearly 78% of Hillary Clinton voters support the Black Lives Matter movement, for instance, compared to just 31% of Trump voters.
But does this really arise from blinkered online behaviour? Or are more subtle dynamics at work?
Although there's little doubt that our reading habits shape our political opinions – and it's as-yet unclear the extent to which targeted advertising might sway voter behaviour – some striking recent studies suggest that the influence of echo chambers and filter bubbles may have been over-stated.
By its nature, social media will expose you to a number of other sources – Seth Flaxman
Consider a paper by Seth Flaxman and colleagues at Oxford University, which examined the browsing histories of 50,000 users based in the US. In line with the received wisdom, social media and search users tended to land upon more polarised news sources – Breitbart compared to Fox News, for instance – which might translate to a more extreme world view.
Crucially, however – and contrary to the concept of the online echo chamber and filter bubble – they were also more likely to visit sites expressing opposing viewpoints. Their media diet was more varied, overall. “It seems counterintuitive, but direct browsing often just consists of one or two sites that you regularly read – such as the BBC and CNN – while by its nature, social media will expose you to a number of other sources, increasing the diversity,” says Flaxman, who is now based at Imperial College London.
Flaxman emphasises that the study is based on data from 2013, and times may have changed. But a Pew survey around the 2016 US Presidential Election broadly agrees with his findings, with the majority of people reporting a range of opinions in their social media feeds. And the University of Ottawa’s Dubois has come to similar conclusions with her own studies.
Using a survey of 2,000 British adults, she found that the majority of people already reach outside their political comfort zone: they actively seek out additional sources that convey diverse views that do not match with their preconceptions. Indeed, just 8% of Dubois’s participants scored so low on her measures of media diversity that they could be considered at risk of living an echo chamber, visiting just one or two news services without other perspectives.
Dubois emphasises that even 8% of people living in an echo chamber is still a “worrying” number. But it is far less than most pundits would have anticipated. Most people should have a reasonable idea of what the other side are thinking on current debates.
Linguistic analyses showed that the Republican users began to use more emotive words on their feeds, as they were exposed to more liberal viewpoints
Indeed, there is now evidence that well-meaning attempts to counter the echo chamber and filter bubble, by reading more diverse news sources, may actually backfire – leading to more, and not less, political polarisation.
Along these lines, a team led by Christopher Bail at Duke University measured a group of more than 1,600 Twitter users’ political positions before paying them a small fee to follow a 'bot' that would retweet influencers from across the political divide.
About half of the participants took them up on the offer, but rather than developing a more moderate or nuanced stance on issues such as gay rights, most simply came to be more confident of their initial beliefs. (The effect was more pronounced for Republicans, who became significantly more conservative, whereas the Democrats retained roughly the same views.)
Various psychological theories might explain these findings.
One is “motivated reasoning”. Countless studies have shown that we are so attached to our political identities that we will devote extra cognitive resources to dismissing any evidence that disagrees with our initial point of view, so that we end up even more sure of our convictions.
Along these lines, linguistic analyses showed that the Republican users began to use more emotive words on their feeds as they were exposed to more liberal viewpoints. “Over time, we see increases in negative sentiment expressed towards opinion liberal leaders,” Bail said, “which we take this to be some evidence that a process like motivated reasoning may be underway, particularly because we saw the [negative feelings] increase over the course of the treatment.”
But an alternative possible explanation comes from the psychology of 'self-licensing' – the unconscious belief that once we have shown our open-mindedness in one situation, we have somehow earned the credentials to be more prejudiced later on. One study from 2008 found that people who had supported Barack Obama were subsequently more likely to express a potentially racist view, for instance. By reading a few dissenting voices on Facebook or Twitter, we may feel that we have already gained the right to be more dogmatic about our existing opinions. Anecdotally, at least, this seems to have been the case for a few of my own acquaintances following the UK’s referendum on Europe in 2016.
Bail offers a note of caution, however, pointing out that the retweets came from high-profile figures – people who may have been particularly alienating to the average Twitter user. “People really don’t like ‘elites’,” says Bail. So it remains to be seen if you would see the same backfire effect to messages from other, less provocative, representatives.
If you look at any measures of what people think about people on the other side, [they] have become vastly more hostile – Jonathan Haidt
All of which suggests that the problems of political discourse online appear not to come from the range of voices that we hear – as the idea of the echo chamber would suggest – but our unconscious reactions to them. As the social psychologist Jonathan Haidt puts it: “If you look at any measures of what people think about people on the other side, [they] have become vastly more hostile.”
The anonymity of our interactions online seem to make it particularly easy to dismiss other’s views, creating a more hostile environment for debate. (Read more about the reasons people are so hostile online).
“[The influence of] echo chambers in social media has been highly over-estimated. But that doesn’t mean that political polarisation isn’t being driven by other factors,” agrees Dubois.
Nor does any of this recent research deny the deliberate manipulation of social media and the influence that may have had on subsequent political events. A recent study in Science journal, for instance, confirmed that false news spreads much more rapidly than verified information from respectable sources. “It’s very possible that most people are not at risk of being stuck in an echo chamber, but they are still being targeted with specific ads based on their behaviour, or they are still being targeted with misinformation,” says Dubois.
Although there appears to be no easy solution to this online disharmony, experts such as Dubois hope that media literacy education at schools and university may help, teaching basic critical thinking skills and the ways to identify bias in an argument, allowing us to appraise news sources more thoughtfully. The World Wide Web may be approaching its 30th birthday, but we still have a lot to learn about the best ways to navigate the online environment.
David Robson is a freelance writer based in London, UK. He is d_a_robson on Twitter.