Shortly after hearing about the massacre in San Bernardino, California, I retweeted something that I soon regretted. In theory, it was benign—a link to a widely shared article about how 2015 had seen more mass shootings than days in the year. I added a single word of commentary: “whoa.”

Days later, however, Mother Jones editor Mark Follman wrote an op-ed for The New York Times explaining that that statistic, while technically true, is deeply misleading. Yes, hundreds of shootings in the United States have killed or injured at least four people this year, Follman wrote. But this year, the country has seen just four of the random, mass shootings we all live in fear of—the type we saw in San Bernardino and in Oregon, the type I thought I was referring to in that tweet. The original statistic still points to hard truths about gun violence in America, but not, necessarily, the random acts of violence we refer to as “mass shootings.”

I read Follman’s piece and thought back to how quickly and thoughtlessly I had shared the earlier article. I hadn’t taken a minute to consider the nuance that might lie behind those numbers. I wanted to contribute to the conversation, but the only contribution I made was to the glut of misinformation and fear-mongering that clouds both sides of the gun control debate—misinformation instantaneously spread far and wide on social media.

I deleted the tweet, but knew I already had become part of the problem. And if you are one of the millions of Americans who take to Twitter and Facebook to share one-liners about gun control, Donald Trump, or any of the other partisan issues tearing at the civic fabric of the US, then, well, you may be part of the problem, too.

The Echo Chamber Effect

It’s become a reflex of the Internet age: A mass shooting happens, or Trump says something bigoted, and people flock to social media to share their two cents. It can sometimes feel like activism, like partaking in meaningful debate about some of the most important issues facing the country. But it’s not helping. Not for the most part, at least.

The problem is, true political debate is rare on social media. Instead, most of us preach to the converted, reaffirming our own beliefs and seldom paying attention to the other side. Social scientists call it the echo chamber effect, a now well-documented social media phenomenon.

One Pew Research Center study published last year illustrated in great detail just how polarized those echo chambers can be. Researchers analyzed thousands of tweets containing the hashtag #My2k, which the White House launched to start a conversation about a $2,000 tax increase that was pending in Congress. The study showed that less than 1 percent of the tweets and follows exchanged using the hashtag occurred between liberals and conservatives, which defeated the point of the hashtag. The study also found that conservative users linking to a completely different news sources than liberal users.

Pew characterized these so-called “polarized crowds” this way: “Polarized crowds on Twitter are not arguing. They are ignoring one another while pointing to different web resources and using different hashtags.”

Two Vocal Extremes

But echo chambers in and of themselves aren’t the dangerous part. What’s dangerous is the fact that they rarely, if ever, reflect the views of moderate Americans. Instead, research shows, it’s people on either end of the ideological spectrum who have the most input. Unfortunately, research also shows that people on either end of the spectrum are the least likely to see what the other side has to say, and, therefore, the least likely to ever see eye-to-eye.

“You have very vocal minorities on two sides of the spectrum,” says Dominique Brossard, who studies the way controversial scientific ideas are discussed on social media at the University of Wisconsin-Madison, “and then the vast majority in the middle that are pretty silent about it.”

Which is why, immediately after a mass shooting, there often are two, and only two, lines of criticism on social media: Gun control would prevent future tragedies, or more guns could help people protect themselves from such attacks. Typically, there is no middle ground. There is no nuance. It’s no surprise, then, that tweet by tweet, post by post, the middle ground on which so many Americans still stand is rapidly disappearing from public discussion.

This, some researchers say, helps to explain not only the stalemate on the issue of gun control, but also the rise of Trump as a legitimate contender for the Republican presidential nomination. “I wonder, without social media, to what extent Donald Trump would be as successful as he is right now,” says Pablo Barberá, a research fellow at New York University’s Center for Data Science, who researches social media and politics.

In the past, Barberá says, Trump’s more radical ideas—like his proposal to ban Muslim people from traveling to the United States or to deport 11 million undocumented immigrants—would have been covered by traditional media, “but they would be contained.” In the limitless world of social media, not only is there more conversation about these extreme ideas, but the people talking about them are themselves more extreme. Social media conversations are also increasingly informing the traditional news cycle, particularly this election season, as media outlets like CNN and CBS News turn to Facebook and Twitter data to drive their coverage.

“For those at the very extreme end of the distribution, this reinforces their extremism, and, in a way, it could legitimize this type of opinion,” Barberá says.

A Petri Dish of Fear and Misinformation

But there were other problems with my initial tweet about mass shootings. By inflating the apparent frequency of random mass shootings, the tweet preyed on people’s sense of fear. And, as one recent study out of the University of Buffalo demonstrated, fear is a key driver of viral content on Twitter. The study analyzed images shared on Twitter associated with the #guncontrol hashtag and found images that accentuated either fear or humor were the most likely to be retweeted.

A paper published in the journal Energy Policy, meanwhile, explored how Americans perceive the risks associated with nuclear energy and found that the more media coverage people consume, the more fearful they become. “Heavy consumers of media, rather than becoming more informed and uniform in their opinions, become more divided as disasters and their related issues are covered in newspapers, television and the Internet,” says Brossard, an author of the study. The researchers refer to this trend as the “social amplification of risk.”

This phenomenon was apparent during last year’s Ebola crisis, which researchers say the US media covered in a way wholly disproportionate to the risk Ebola posed in the United States. One study found that news stories about Ebola in the US that originated in traditional media spread like contagion on Twitter, with each story yielding tens of thousands of tweets.

It’s all made worse, Brossard says, by the fact that so much of what we’re afraid of on social media is misinformation. It’s no wonder, then, that one of Trump’s most egregious tweets this election season—featuring an inaccurate infographic of crime statistics by race—received 11,389 likes and 8,626 retweets at last count.

“As a species, we are fearful animals, and it’s easy to seed fear,” Brossard says. “It’s much easier to seed fear than to seed hope, and the rhetoric of hate tends to spread.”

There Is Hope

And yet, even as the most extreme Americans find audiences on social media, researchers see a few signs that social media may have a moderating effect on people in the middle of the spectrum, who tend to be much less vocal. That’s because, says Barberá, social media does have the potential, at least, to expose users to more new ideas than they would experience merely chatting with friends and family. Though it’s true that social media users mostly encounter views that align with their own, Barberá says, they still at least occasionally run into ideologies different from their own. For those who aren’t radical to begin with, that interaction can have a moderating effect over time.

These findings come from a working paper in which Barberá analyzed the Twitter profiles of millions of Twitter users and developed a model that determined whether they were liberal or conservative based on who they follow. Barberá cross-checked a sample of these users against their publicly available voter registration records and found that who people follow tends to be an accurate indicator of political affiliation.

Barberá then tracked how these follower networks changed over the course of 18 months. He found that when moderately conservative users were exposed to just a small percentage of content from liberal users, they tended to follow more and more liberal users as time went on, and vice versa.

For Barberá, that’s a sign that exposure to conflicting ideas online can make people more moderate over time. And that’s good news, Barberá says, because, his research shows that the vast majority of people are seeing enough conflicting ideas to experience that moderating effect. The bad news, of course, is the research also showed that the more extreme people’s views are to begin with, the less conflicting information they tend to see and that, in turn, can make them more extreme over time.

Still, these findings offer a little bit of hope in contrast to what is otherwise bleak picture. They suggest that when people choose to listen and not just talk on social media, they’re actually finding more common ground than they might otherwise have thought possible. And that’s a crucial discovery. As social media becomes more central to our lives, news outlets and politicians alike tend to listen to the people who have the loudest voices online. In reality, it’s the loudest who should be doing the listening.

Excerpt from: 

Social Media Is Making the Debate On Guns—And Trump—Worse