Did Echo Chambers and Fake News Really Tip the Election?
In the aftermath of the 2016 election, fake news has taken center stage. Fake news, it is surmised, pushed the election away from Clinton and towards Trump, enabled by personalization on Facebook, Twitter, and Google. Yet few have dealt critically with the research on the echo chamber effect, fake news, or the psychology behind political opinions. While companies should be applauded for cutting spam, the fake news narrative lacks serious rigor and support, which weakens the simplistic argument that it tipped the election.
The story that has recently emerged with fake news seems to fit into a coherent narrative that has been building since the mid-2000s. Sites serving individualized content have segregated users into like-minded groups. In this election especially, fake news filled into spaces that were already politically split. The result limited understanding between political opposites and exacerbated the already partisan political conversation.
Yet, the research cited throughout this piece contradicts this simple story. Social media isn’t as ideologically split as many believe. Seeing diversified viewpoints doesn’t lead to moderated political views. And social media hasn’t led to a dramatically divided political life. The challenge for critics and commentators alike is to parse out already existing social tendencies from the structure of these new mediums and then review specific decisions made by network operators within this context. Taken together, we should temper our views on the power of fake news.
While it is true that people tend to cluster on Facebook and Twitter with like-minded groups, they also cluster when they aren’t online. Because our online presence is an extension of our social selves, this should be expected. However, research consistently finds that social media sites like Facebook and Twitter allow us to maintain more extensive social networks and build social capital. No longer are cousins, high school friends, and casual acquaintances seen just around the holidays. Online social networks allow more consistent contact. These weak tie relationships, as they are called, introduce us to new perspectives and to novel information.
In other words, social media should increase the range of experiences and news. Scholars working with Facebook data found that weak ties actually drive the majority of information on social networks. Others have corroborated this finding, instead suggesting that the site actually sustains ideologically diverse communication.
Herein lies a central paradox of diverse viewpoints. Exposure to others doesn’t often result in a changed mind, quite the opposite. We all rely on mental shortcuts that generate systemic biases. So instead of updating our views when presented with contradictory evidence, everyone tends to discount that new information and cling onto their basic beliefs even more. This tendency is called motivated reasoning and it finds support in countless surveys across numerous disciplines.
Political views are especially durable to change. Negative information about a candidate doesn’t shake support, but actually leads to its rejection. In one seminal experiment on this effect, people who watched a presidential debate and heard their preferred candidate support views contrary to their own didn’t waver in their support of their nominee but updated their opinions to conform to his political ideology.
When political scientist Brendan Nyhan ran an updated version of this experiment, he and his coauthors found similar effects. Trump’s convention speech in July served as a baseline, since he suggested that crime was increasing in that talk, even though the long-term trend has been towards less crime. After providing participants with corrective stories, both Clinton and Trump supporters did change their view about crime but didn’t change their view about the candidates. Moreover, those supporting Trump upped their tendency to say that crime statistics are “not very accurate” or “not at all accurate” from 16 percent to 43 percent.
If the shoe were on the other foot, we should expect Clinton supporters to decry the statistics. As has been consistently shown, partisans of all stripes tend to think the media is against them even if information is neutrally presented. In other words, media outlets are seen as being hostile to all partisan views.
The hostile media effect often blinds us to the reality of both online and offline news. Most of what people consume is fairly moderate. Even as news consumption has been shifting towards online sources, which are slightly more ideological than offline sources, a dramatic shift towards partisan bubbles is hard to find. Rather, most people still learn political news from mainstream, relatively centrist sources. Both sides of the aisle also tend to read the same outlets. Yet how those articles are received by readers are informed by prior political views.
Only extreme partisans buck the trend of moderate news and consume more overtly ideological news. But partly this is because both consistent liberals and conservatives discuss government and politics more than four times as often as moderates in some cases. If we assume that these groups seek out information to confirm their prior beliefs, as research shows, then it is a logical conclusion that they would consume more partisan news because they are far more active in political discussion. Yet Pew found that only about 13 percent of all people talk about politics in any format every day, whereas the vast majority, around 58 percent of all people, discuss politics just a few times a month or less. Altogether, partisans skew the data.
A perception about increased political partisanship dovetails into a feeling that outlets are becoming more ideologically insulated. But the reality here is also more complex. While those in Congress have shown more consistency in voting and beliefs, leading to sharper divisions among the parties, there is little evidence that Americans have done the same. Rather, the ideological sort in Congress since the 1970s has given rise to more consistent voting preferences. No longer do voters split the ticket. Even self-identified moderates and independents vote with consistent loyalty for one party or another.
Fake news must be understood within these dynamics, which have been building for some years. Although small, extreme partisans on both sides are vocal, since most people don’t often discuss politics, a few very vocal persons can change the perception even though it is rare overall.
For those that care to dig for the truth in the fake news story, more questions should exist than answers. How much did active and consistent liberals or conservatives contribute to the sharing of fake news? Did the presence of fake news actually change voting behavior or did it merely reflect consistent but covert tendencies, as the literature on moderates suggest? And even if the erroroneous reports were corrected, would these people have voted differently?
While Facebook, Twitter, and Google should be given praise for looking more closely at stories rife with misinformation, in discussing the fake news stories, commentators should do the same. Far too often, critics of media rely on comfortable thinking, but the complexity and importance of this subject demands we give it respect and rely on research.
Much like the news we consume, criticism of fake news needs proper citations.